BOOK NUMBER
University of Ghana          http://ugspace.ug.edu.gh
ITRAPPING STUDIES ON GLOSSINA LONGIPENNIS CORTI AT 
NGURUMAN, SOUTH-WESTERN KENYA.
A THESIS SUBMITTED TO THE DEPARTMENT OF ZOOLOGY, 
UNIVERSITY OF GHANA, IN FULFILLMENT OF THE REQUIREMENT 
FOR THE AWARD OF A DOCTOR OF PHILOSOPHY DEGREE.
BY
CHARLES ANTIERME KYORKU, B.Sc.(UST, Ghana), M.Sc.(UG, 
Ghana)
FEBRUARY-, 198 9
University of Ghana          http://ugspace.ug.edu.gh
DECLARATION
I THE UNDERSIGNED DECLARE THAT THIS THESIS IS MY OWN 
ORIGINAL WORK WHICH HAS NOT BEEN SUBMITTED FOR ANY 
DEGREE IN ANY UNIVERSITY AND ALL SOURCES OF MATERIAL 
HAS BEEN DULY ACKNOWLEDGED.
SIGNATURE
University of Ghana          http://ugspace.ug.edu.gh
APPROVAL CERTIFICATE
THIS THESIS HAS BEEN APPROVED FOR SUBMISSION TO THE 
UNIVERSITY OF GHANA, LEGON,GHANA.
BY
MAJOR EXTERNAL SUPERVISOR ERNAL SUPERVISOR
University of Ghana          http://ugspace.ug.edu.gh
DEDICATION
I V
TO MY PARENTS, MY WIFE CHRISTY AND MY DAUGHTER ANATABA.
University of Ghana          http://ugspace.ug.edu.gh
VACKNOWLEDGEMENTS
My most sincere thanks go to my supervisor, Dr. R. D. 
Dransfield, whose keen interest and attention was the major 
force behind the completion of this work. His contribution 
ranged from suggestions on various fields of investigation to 
critical review of the work and statistical advice, not 
forgetting the inspiration drawn from his very meticulous 
approach to collecting ecological data.
I greatly appreciate the close collaboration I got from 
Mr. R. Brightwell throughout the study. Besides his major 
contribution on the trap development aspect of the work, most 
of the data were collated using computer programmes written by 
him.
I also thank the rest of the team members of the Nguruman 
project especially the Tsetse Programme technicians and the 
local field assistants without whom field work would have been 
impossible. I am particularly grateful to Mr. J. Kiilu for 
introducing me to tsetse dissection techniques, to Mr. J. 
Larinkoi for his help in marking in the field and extracting 
the data on mark-release-recapture experiments and to Messrs. 
D. Mungai and Z. Muriuki for regularly driving me at odd hours 
to run experiments. I remain indebted to the late Mr. J.
Mutel, one of the most dependable local field assistants, who 
recently passed away. I sincerely appreciate the review of 
the scripts by Dr. B. Williams. His red pen marks were a great 
contribution to the shape of this thesis. I gratefully
University of Ghana          http://ugspace.ug.edu.gh
acknowledge the use of his personal computer and a statistical 
programme written by him for analyzing some o£ the data and 
plotting some graphs. I am further indebted to Dr. Williams 
and Ms. Suzan Macmillan for their joint benevolent 
contribution to my upkeep in Nairobi and the moral support 
they gave me at the most trying period of my life.
Thanks are also due to Professor W. Z. Coker and Dr. S. 
Quartey of the University of Ghana for reading through some of 
the scripts and offering useful comments and to Dr. Nokoe for 
some statistical advice. I gratefully acknowledge useful 
discussions I have had with various ICIPE scientists including 
Drs. M. F. Chaudhury, S. A. Tarimo, R. Saini, S. Mihok and Mrs 
M. Owaga. I also appreciate the rewarding discussions with 
Drs. G. A. Vale, D. J. Rogers, and S. E. Randolph during their 
visits to the Nguruman project.
I am thankful to Messrs P. Lissamula for photographic 
work, to Messrs. R. Kruska, N. Muanga and F. Masika for 
graphic work and to Ms. K. Chaudury and Ms. E. Afandi for 
typing bits and pieces of the work. Many thanks are also due 
to the local Maasai community who provided a regular supply of 
cow urine for my experiments and to the keepers of the animal 
house at the Veterinary Laboratories, Kabete, for collecting 
buffalo urine.
My sincere thanks go to the Director of ICIPE, Prof. T.
R. Odiambo and the leader of the Tsetse Research Programme,
Dr. L. H. Otieno for the use of facilities in ICIPE in general 
and those of Tsetse Programme in particular. I sincerely
Vi
University of Ghana          http://ugspace.ug.edu.gh
VI t
appreciate the ready co-oporation of the ARPPJS academic co 
ordinator, Dr. M. E. Smalley, who did all that was possible tc 
make my task easier. The help of his secretary, Mrs A. Kumali 
and driver Mr. D. Isoso is also greatly appreciated.
I remain indebted to my friends Mr. P. Muange, Mrs. E. 
Sebitosi and Dr. M. Hassane for moral support and to members 
of the Ghanaian community in Nairobi, especially the families 
of Messrs. F. Tachie and B. Amoah, Dr. E. Suleman and Ms. A. 
Shika for contributing in diverse ways towards my stay in 
Nai robi.
Finally, I am grateful to the German Academic Exchange 
Programme (DAAD) for sponsoring my study under the African 
Regional Postgradute Programme in Insect Science (ARPPIS) 
programme and to the University of Ghana for granting me a 
study 1eave.
University of Ghana          http://ugspace.ug.edu.gh
v n j ;
ABSTRACT
Studies have been carried out at Nguruman, south-western 
Kenya, on Glossina longipennis Corti, a little known member of 
the fusca group of Glossina. The first objective was to 
develop an efficient trap suitable for both sampling and 
control purposes. Studies were then carried out on the 
population dynamics of G. 1ongipennis using the newly 
developed sampling methods. Lastly the trap/odour bait system 
was tested in a control situation.
Replicated Latin square design experiments were used to 
compare the performance of various trap designs and odour 
attractants. The Zimbabwe F3 trap proved more effective than 
the widely used biconical trap, especially for females. A new 
trap developed at Nguruman, called the NG2B, was also very 
effective and had the advantage of being cheap and easy to 
construct. Acetone and cow urine together increased the 
catches by 4-5X over unbaited traps, but when dispensed alone 
neither of them was effective. There was no significant 
difference between the attractancy of cow urine and buffalo 
urine. Trap catches were further increased when l-octen-3-ol 
was dispensed together with acetone and cow urine. A higher 
proportion of older flies was caught by the NG2B trap compared 
to the biconical but no significant difference was observed in 
the age structure of flies attracted by different odour baits. 
The effect of trap design on sample composition and the
University of Ghana          http://ugspace.ug.edu.gh
potential for using odour baited traps for sampling the fusca 
group of tsetse flies are discussed.
An electric screen adjacent to a baited target was used 
to determine the precise activity pattern of G. longipennis 
which is known to be crepuscular in behaviour. Morning 
activity started at about 15 minutes before sunrise at 0630 h, 
peaked at about 0615 h and ceased by 0700 h. The species was 
more active in the evenings, when activity began at about 30 
minutes before sunset at 1815 h, peaked at 1845 h and ended by 
1900 h. Males were regularly active before females. Light 
intensity was found to be the most important factor 
influencing activity. The relationship between activity 
pattern and cattle-fly contact is discussed.
Changes in the apparent densities of G. longipennis were 
monitored simultaneously using biconical and NG2B traps in two 
areas located 7 km apart. Both trap types showed similar 
trends in population changes but higher apparent densities 
were recorded with the NG2B trap than with the biconical trap. 
Apparent densities in both sexes were regularly observed to 
increase during the rainy seasons and decrease during dry 
seasons. Peak catches in one area were observed to precede 
those in the other area by one month. Flies spread out to more 
open areas during the cool wet seasons and concentrated in the 
thicker woodland during the dry seasons. The factors 
influencing changes in population densities, including 
movement between the two areas and between vegetation types, 
are discussed.
\*
University of Ghana          http://ugspace.ug.edu.gh
. X
Mortality rates estimated from ovarian age structure and 
from Moran curves were observed to be highest during the hot 
dry seasons and lower during the cool wet seasons. Adult 
mortality rates showed a significant positive correlation with 
maximum temperature and a negative correlation with minimum 
relative humidity. The effect of fly movement on mortality 
rate estimates and the reliability of the estimates by the two 
methods are discussed.
Dissections of female flies from NG2B traps showed that 
all non-teneral but nulliparous females and over 80% teneral 
females were inseminated. The average percentage distribution 
of the various pregnancy stages in trap samples were found to 
be very close to the values expected from the duration of the 
different stages, in contrast to the usual under­
representation of flies with third instar larvae for other 
tsetse species. The average abortion rate was 6% but ranged 
from 0% in the rainy seasons to 60% in the hot dry season. A 
significant negative correlation was observed between abortion 
rate and minimum relative humidity. A significant positive 
correlation was also found between fly size and minimum 
relative humidity of the previous month but one. A discussion 
is given of the immediate causes of abortions and their 
effects on population levels and of the factors influencing 
fly size.
University of Ghana          http://ugspace.ug.edu.gh
x l
The absolute population size of G. longipennis was 
estimated through mark-release-recapture experiments. The mean 
population size was estimated at 17,300 males (range 10,471 - 
25,703) and 16,900 females (range 14,125 - 20,892). The trend 
of changes in the absolute estimates corresponded with those 
in apparent estimates from trap catches. From the peaks in the 
recapture rate of marked flies, the feeding cycle of G. 
longipennis was found to be 2-3 days for males whilst for 
females the 9-10 day pregnancy cycle was the main factor 
affecting the recapture rate. There was a considerable amount 
of movement of marked flies between the two sampling areas but 
the movement was shown to be greater in one direction than the 
other.
A trial tsetse population suppression operation with 
baited NG2B traps was started during the course of the study. 
After 11 months of operation, the population levels of G.
1ongipennis were reduced by about 60% for males and about 90% 
for females. Much greater reduction levels were obtained for 
G. pallidipes. A discussion is given of the factors 
influencing the lesser impact on the population of G.
1ongipennis with suggestions on improving methods for the 
control of the species.
University of Ghana          http://ugspace.ug.edu.gh
TABLE OF CONTENTS Page
TITLE PAGE 1
DECLARATION 11
APPROVAL CERTIFICATE III
DEDICATION IV
ACKNOWLEDGEMENTS V
ABSTRACT VI If
LIST OF FIGURES XVII
LIST OF PLATES XXIII
LIST OF TABLES XXIV
CHAPTER ONE GENERAL INTRODUCTION 1
CHAPTER TWO LITERATURE REVIEW 11
2.1. INTRODUCTION 11
2.2. DISTRIBUTION OF GLOSSINA LONGIPENNIS AND 
OTHER FUSCA SPECIES
2.2.1. Geographical distribution 12
2.2.2. Ecological distribution (Habitats) 17
2.3. THE ECOLOGY AND BEHAVIOUR OF FUSCA FLIES 20
2.3.1. Breeding Sites. 21
2.3.2. Resting Sites 23
2.3.3. Activity 24
2.3.4. Host Preference 28
2.4. SAMPLING TECHNIQUES FOR FUSCA GROUP FLIES 29
2.4.1. The fly-round technique 30
2.4.2. The Stationary Bait Technique 33
2.4.3. Sampling the resting population 34
2.4.4. Trapping 3 6
University of Ghana          http://ugspace.ug.edu.gh
2.5. TRAP/ODOUR BAIT TECHNOLOGY 36
2.5.1. Colour stimulus 39
2.5.2. Trap design 44
2.5.3. Movement 46
2.5.4. Odour baits 47
2.4.5. Trap efficiency 51
2.5.6. The role of traps in tsetse control 53
CHAPTER THREE DESCRIPTION OF THE STUDY AREA 57
3.1. INTRODUCTION 57
3.2. LOCATION OF THE NGURUMAN AREA 60«
3.3. CLIMATE AND WEATHER 60
3.4. VEGETATION 62
3.5. MACROVERTEBRATES 65
CHAPTER FOUR DEVELOPMENT OF AN EFFECTIVE
TRAP/ODOUR BAIT SYSTEM 68
4.1. INTRODUCTION 68
4.2. MATERIALS AND METHODS 70
4.2.1. Traps tested. 70
a) The biconical trap. 70
b) The F2 and F3 traps 72
c) The NGU traps 7 5
The NQ1 M odel 75
The NG2 model 7 8
Thg__N_G3 . m odel 81
Xhs.. NG4 model. 83
4.2.2. Odours Baits 8 5
4.2.3. Exoerimpni-al Hoa-i™
University of Ghana          http://ugspace.ug.edu.gh
4.2.4. Details of experiments 88
4.2.5. Analysis of data 99
4.2.6. Test of trap efficiency 99
4.3. RESULTS 102
4.3.1. Efficacy of odours 102
4.3.2. Trap performance 105 
4 . 3.3.Efficacy of odours with the NG2B trap 113
4.3.4. Efficiency of the NG2B trap. 122
4.4 DISCUSSION 124
CHAPTER FIVE THE ACTIVITY PATTERN OF
G. LONGIPENNIS 13 6
5.1. INTRODUCTION. 136
5.2. MATERIALS AND METHODS 138
5.2.1 The electric screen 138
5.2.2. Meteorological data 139
5.2.3. Activity Period 141
5.2.3. Experiments 142
RESULTS 144
5.3.1. Fly numbers from various treatments 144
5.3.2. The daily activity pattern 146
5.3.3. Activity and physical factors. 151
5.4. DISCUSSION 154
University of Ghana          http://ugspace.ug.edu.gh
CHAPTER SIX POPULATION DYNAMICS OF GLOSSINA 
LON G I P E N N I S:I . APPARENT DENSITIES AND
MORTALITY RATES 160
6.1. INTRODUCTION 160
6.2. MATERIALS AND METHODS 163
6.2.1. Field Studies 163
6.2.2. Laboratory Studies 168
6.2.3. Analysis of Data 170
6.2.3.1. Fly numbers and sex ratio 170
6.2.3.2. Estimation of mortality
rates from ovarian age distribution 171
6.2.3.3. Estimation of mortality
rates from Moran curves 174
6.3. RESULTS 175
6.3.1. Apparent densities 175
6.3.2. Fly distribution in different 
vegetation types. 184
6.3.3. Age structure 189
6.3.4. Mortality rates from ovarian
age distribution. 195
6.3.5. Mortality rates from Moran Curves 199
6.3.5. Mortality rates and changes in the 
percentage of nulliparous flies. 204
6.4. DISCUSSION 207
University of Ghana          http://ugspace.ug.edu.gh
CHAPTER SEVEN POPULATION DYNAMICS OF GLOSSINA 
LONGIPENNIS: II. REPRODUCTION AND SIZE 215
7.1. INTRODUCTION 215
7.2. MATERIALS AND METHODS 217
7.3. RESULTS 218
7.3.1. Insemination rate 218
7.3.2. Abortion rate 218
7.3.3. Frequency distribution and size
of immature stages 220
7.3.4. Size of adult flies 226
7.4. DISCUSSION 230
CHAPTER EIGHT MARK RELEASE RECAPTURE STUDIES 236
8.1. INTRODUCTION 236
8.2. MATERIALS AND METHODS 239
8.2.1. Monthly mark-release-recapture 
experiments. 239
8.2.2. Seven-day continuous marking 
experiment 242
8.2.3. Analysis of data 243
8.3. RESULTS 249
8.3.1. Monthly marking experiments 249
8.3.2. Seven-day marking experiment. 258
8.3.3. Relationship between relative and 
absolute population estimates 264
8.3.4. The assessment of fly movement between 
Transect 1 and Transect 4 areas. 265#
University of Ghana          http://ugspace.ug.edu.gh
8.4. DISCUSSION 268
CHAPTER NINE SUPRESSION OF A G. LONGIPENNIS
POPULATION WITH BAITED NG2B TRAPS 276
9.1. INTRODUCTION 27 6
9.2. MATERIALS AND METHODS 277
9.3. RESULTS 280
9.4. DISCUSSION 286
CHAPTER TEN GENERAL DISCUSSION 290
SUMMARY 2 99
REFERENCES 303
University of Ghana          http://ugspace.ug.edu.gh
LIST OF FIGURES
Figure 2 
Figure 2 
Figure 3
Figure 3
Figure 3
Figure 4 
Figure 4 
Figure 4 
Figure 4 
Figure 4
Figure 4
.1: The distribution of the fusca group of Glossina 
in relation to the two zones of lowland rainforest
.2: The distribution of G. longipennis and G. 
brevipalpis in Kenya and near its borders.
. 1: Location of the main study area (boxed) in
relation to major geographical landmarks and other 
tsetse infested areas of the Nguruman tsetse fly 
belt.
.2: Profiles of monthly rainfall (histogram), mean 
monthly minimum and maximum temperature and mean 
monthly minimum and maximum relative humidity 
recorded at Nguruman over the study period.
.3: Map showing the different vegetation types in 
the main study area.
.1: Construction of the NG1 trap model.
.2: Construction of the NG2 trap model.
.3: Construction of the NG3 trap model.
.4: Construction of the NG4 trap model.
.5: Age distribution of female G. longipennis caught 
in the blue biconical trap baited with different 
chemicals.
.6: Age distribution of female G. longipennis caught 
in the different trap types all baited with 
similar doses of cow urine and acetone.
PAGE
14
16
59
61
63
77
79
82
84
120
121
University of Ghana          http://ugspace.ug.edu.gh
Figure 5
Figure 5
Figure 5
Figure 5
Figure 6
Figure 6
Figure 6
Figure 6
Figure 6
.1: Set-up of the electric screen and cloth target 
used in the study of activity pattern of G. 
longipennis.
.2: The relative levels of activity of G.
longipennis in the morning and in the evening.
.3: Activity patterns of male and female 
G. longipennis in the evening
.4: The relationship between activity and light 
intensity.
.1: Map of transect 1 area showing the different 
vegetation types and the transect along which 
biconical and NG2B traps were set.
.2: Map showing both transect 1 and 4 areas and the 
location of the River biconical traps along the 
01oibototo river in both areas.
.3: Monthly changes in the apparent densities of 
male (A) and female (B) G. longipennis in the 
biconical and NG2B traps set along TRl
.4: Monthly changes in the apparent densities of 
male (A) and female (B) G. longipennis in the 
river biconical and the transect 
biconical traps of transect 1 area.
.5: Monthly changes in the apparent densities of 
male (A) and female (B) G. longipennis in the 
biconical and NG2B traps set along TR4.
150
150
153
164
167
176
178
179
140
University of Ghana          http://ugspace.ug.edu.gh
181
183
183
185
186
188
190
191
193
193
196
6.6: Monthly changes in the apparent densities of 
male and female G. Iongipennis in 
the river biconical traps of transect 4 area.
6.7: Monthly changes in the percentage of females 
caught by the biconical traps.
6.8: Monthly changes in the mean percentage of 
females caught by the NG2B traps.
6.9: Seasonal changes in the relative distribution of 
male (A) and female (B) G. longipennis in 
different vegetation types along transect 1.
6.10: Seasonal changes in the relative distribution 
of male (A) and female (B) G. longipennis in 
different vegetation types along the river bed of 
transect 1 area.
6.11: Monthly changes in the relative distribution of 
male (A) and female (B) G. longipennis in 
the river traps of transect 4 area.
6.12: The age distribution of female G. longipennis 
caught by the NG2B traps of transect 1 area.
6.13: The age distribution of female G. longipennis 
caught by the NG2B traps of transect 4 area.
6.14: The relationship between wing-fray age
and ovarian age of female G. longipennis from 
transect 1 area.
6.15: The relationship between ovarian wing-fray age 
and ovarian age of female G. longipennis from 
transect 4 area.
6.16: Monthly changes in adult mortality rate 
estimated from ovarian age structure 
on transect 1.
University of Ghana          http://ugspace.ug.edu.gh
Figure
Figure 6
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
196
.17: Monthly changes in adult mortality rate 
estimated from ovarian age structure 
on transect 4.
.18: The relationship between adult mortality rate 
on transect 1 and the mean monthly maximum 
temperature. 198
.19: The relationship between adult mortality rate 
on transect 1 and the mean monthly minimum 
relative humidity. 198
>.20: The relationship between adult mortality rate 
on transect 4 and the mean monthly maximum 
temperature. 200
1.21: The relationship between adult mortality rate 
on transect 4 and the mean monthly minimum 
relative humidity. 200
i.22A: Scatterplot of male apparent densities
from transect 4 and the fitted Moran curve . 201
1.22B: Scatterplot of female apparent densities
from transect 4 and the fitted Moran curve. 202
1.23: Monthly changes in generation mortality rate
estimated from Moran curves. 203
>.24: Monthly changes in the percentage of
nulliparous flies in the transect 1 area. 205
>.25: Monthly changes in the percentage of
nulliparous flies in the transect 4 area. 205
M :  Monthly changes in the apparent abortion rate.
transects 1 and 4 225
1.2: The relationship between monthly abortion rate
and minimum relative humidity. 225
University of Ghana          http://ugspace.ug.edu.gh
227
229
22
252
252
256
256
257
261
261
7.3: Monthly changes in the wing vein length on
transects 1 and 4 in relation to changes in mean 
monthly minimum and maximum temperature and 
relative humidity.
7.4A: The relationship between the wing vein length 
in one month on transect 1 and the minimum 
relative humidity of the previous month but one.
7.4B: Linear relationship without data points for
months of suspected fly immigration to transect 1.
8.1: Changes in the recapture rate of male G. 
longipennis with days after marking.
(February 986- January 1987)
8.2: Changes in the recapture rate of female G. 
longipennis with days after marking.
(February 986- January 1987).
8.3: Changes in the recapture rate of male G. 
longipennis with days after marking.
(February 1987 - May 1987).
8.4: Changes in the recapture rate of male G. 
longipennis with days after marking. (June - 
October 1987 ) .
8.5: Changes in the recapture rate of female G. 
longipennis with days after marking.(February - 
October 1987).
8.6: Changes in the recapture rate of males with days 
after marking. (28th January 3rd February 1987).
8.7: Changes in the recapture rate of females with 
days after marking. (28th January 3rd February
1987).
University of Ghana          http://ugspace.ug.edu.gh
Figure
Figure 9
Figure 9
Figure 9
M :  Changes in apparent densities male (A) and 
female (B) G. longipennis in the monitoring NG2B 
traps except trap 18. 281
.2: Changes in apparent densities of male (A) and 
female (B) G. longipennis in 7 of the monitoring 
NG2B traps in the barrier. 283
.3: Bimonthly percentage reduction in the population 
levels of male and female G. longipennis estimated 
from apparent densities from the river biconical 
traps operating within and outside the suppression 
zone. 285
.4: The age distribution and mortality rate 
estimates of G. longipennis caught in the 
monitoring traps during population suppression 
operation. 287
University of Ghana          http://ugspace.ug.edu.gh
LIST OF PLATES PAGE
Plate 1: The Blue Biconical Trap 71
Plate2: The Zimbabwe F2 Trap 73
Plate3: The Zimbabwe F3 Trap 74
Plate 4: The ICIPE NG2B Trap 80
University of Ghana          http://ugspace.ug.edu.gh
Tabl e
Tabl e
Tabl e
Tabl e
Tabl e
Tabl e
Tabl e
Tabl e
PAGE
1.1A: Total trap catches, detransformed means,
and ANOVA F-ratios for G 1ongipennis using 
various odour baits. 10-
I.IB: Indices of increase for G. longipennis for
various odour Baits relative to an unbaited 
trap. 10 A
1.2A: Total trap catches, detransformed means,
and ANOVA F -ratios for G. longipennis using 
various trap designs. 10(
1.2B: Indices of increase for G. longipennis for
various trap types relative to a biconical 
trap. 10*/
1.3A: Total trap catches, detransformed means,and
ANOVA F-ratios for G 1ongipennis using the new 
NGU traps. 11C
1.3B: Indices of increase for G. longipennis for traps
of the NGU series.relative to the blue biconical 
trap. 112
1.4A: Total trap catches, detransformed means, ANOVA 
F-ratios for G. 1ongipennis using different 
odour baits. 11^
1.4B: Indices of increase for G. longipennis for
various experimental treatments relative to an 
unbaited NG2B trap. 11*
LIST OF TABLES
University of Ghana          http://ugspace.ug.edu.gh
Tabl e
Table
Tabl e
Table
Tabl e
Tabl e
Tabl e
Tabl e
. 5A: Total trap catches, detransformed means and
ANOVA F-ratios for G. 1ongipennis using 
various odour baits and modifications of the 
NG2B.
. 5B: Indices of increase for G. longipennis for 
various odour baits and trap modifications 
relative to a NG2B trap baited with low dose 
acetone and cow urine.
.6: Numbers of G. longipennis caught on the screens
and in the cage of the NG2B trap and the
estimated trap efficiency (E) for males and 
females.
.1: Numbers of G. longipennis caught at the electric
screens using various treatments (males and
females pooled).
.2: Catch per 15 minutes,total catch and the
percentage of female G. 1ongipennis caught at the 
electric screen during evening activity period.
.3: Catch per 15 minutes, total catch and the
percentage of female G. longipennis caught at the 
electric screen during morning activity period.
.1: Percentage distribution of uterine contents in 
monthly samples of G 1ongipennis with the NG2B 
traps on TRl.
.2: Percentage distribution of uterine contents in 
monthly samples of G longipennis with the NG2B 
traps on TR4.
117
118
123
145
148
149
221
University of Ghana          http://ugspace.ug.edu.gh
Table 7
Table 7
Table 7
Table 8
Table 8
Table 8
Table 8
3: Monthly means of uterine content lengths (mm)
for G .1ongipennis (Jan.-Sept. 1987)* 22s
4: Mean percentages and lengths of the four
stages in G. 1ongipennis pregnancy 22'
,5: Mean length of the largest developing follicule
at different pregnancy stages. 22<
1: Numbers of marked male G. 1ongipennis released 
each month, percentage recaptures within five 
days, numbers recaptured and total numbers caught 
on various days after marking in each month from 
February 1986 to January 1987. 25(
,2: Numbers of marked female G. longipennis released 
each month, percentage recaptures within five 
days, numbers recaptured and total numbers caught 
on various days after marking in each month from 
February 1986 to January 1987. 25.
3: Numbers of marked male G. longipennis released 
each month, percentage recaptures within 45 days, 
numbers recaptured and total numbers caught on 
various days after marking in each month from 
February 1987 to October 1987. 25
,4: Numbers of marked female G. longipennis released 
each month, percentage recaptures within 45 days, 
numbers recaptured and total numbers caught on 
various days after marking in each month from 
February 1987 to October 1987. 25
University of Ghana          http://ugspace.ug.edu.gh
Tabl e
Tabl e
Table
Table
Tabl e
Tabl e
Table
.5: Numbers of marked male G. longipennis released in 
successive marking occasions, numbers recaptured 
and total numbers caught on subsequent days after 
marking from 28th January to 3rd February, 1987. 
.6: Numbers of marked female G. longipennis released 
in successive marking occasions, numbers 
recaptured and total numbers caught on subsequent 
days after marking from 28th January to 3rd 
February, 1987- 
.7: Weekly recaptures of male G. longipennis marked 
from the 23rd January-3rd February 1987 
.8: Weekly recaptures of female G. longipennis 
marked from the 28th January to 3rd February.
.9: Relative and absolute population estimates for G. 
1ongipennis.
.10: The assessment of movement of G. longipennis 
between Transect 1 (TRl) and Transect 4 (TR4) 
from February 1987-September 1987.
.11: Recaptures of male G. longipennis marked and
released near Lake Jipe, Kenya ( Power, 1964).
University of Ghana          http://ugspace.ug.edu.gh
1CHAPTER ONE 
GENERAL INTRODUCTION
Long before the importance of trypanosomiasis was 
realized, a stable system involving tsetse, the trypanosomes 
and the African game animal, had apparently been established 
in the African environment. It is believed that the occurrence 
of trypanosomes and trypanosomiases in man and his domestic 
animals is a more recent association. The pathological 
response by man and domestic animals to the trypanosome is 
evidence that the parasites (trypanosomes) are still 
maladapted to these hosts (Duggan, 1970). Some aspects of the 
evolution and the ecology of tsetse and trypanosomiases in the 
prehistoric African environment are covered by Lambrecht 
(1964).
Control of trypanosomiasis was initiated by the 
colonialists who by 1908 saw that the sheer numbers of people 
involved and the obvious economic implications of the disease 
were a serious barrier to material progress. Various 
scientific committees were therefore dispatched to the field 
and some control measures, based on apparently ad hoc reports 
and recommendations, were embarked upon with the aim of 
eradicating the disease. Thus, tsetse and trypanosomiasis 
control started with very little scientific knowledge of the 
system that was involved in the prevalence of the disease.
This lack of sufficient basic knowledge is reflected in the
University of Ghana          http://ugspace.ug.edu.gh
2wide variety of control strategies that were tried. The main 
ones included the mass diagnosis and treatment of the disease, 
preventing people at risk from coming into contact with tsetse 
and the destruction of vegetation (tsetse habitats) and game 
animals (hosts). These methods were applied to varying degrees 
and in various combinations in different parts of Africa. The 
results obtained were equally varied, ranging from total 
failure in most places to eradication in a few. A 
comprehensive review of the various control strategies that 
were employed and their successes and failures, is given by 
Duggan (1970). He also pointed out that during all this time 
very little attention was being given to basic scientific 
research into a better understanding of the system involved.
By 1940 when it was realized that a more complex situation of 
several tsetse species and parasites was involved, it was 
concluded that eradication was not possible and that a better 
scientific understanding was necessary for more effective 
control.
Thus, by the 1950s, more attention was being directed 
towards basic research but the pace was slowed with the advent 
of DDT and other insecticides that rekindled the hopes of 
vector eradication. Allsopp (1984) gives a comprehensive 
review of the use of insecticides in tsetse control. Although 
eradication or successful control was achieved in some places 
with insecticides, reinfestation of control areas remained a 
major problem.
University of Ghana          http://ugspace.ug.edu.gh
3The limitations of insecticide application were apparent 
by the 1970's. Apart from the fact that no permanent control 
was being achieved in most cases, the high cost of 
insecticides and environmental pollution had to be taken into 
account. Following this realization, the search for 
alternative control measures was intensified. This started 
with the search for methods of refining the insecticidal 
control technique. Considerable research was carried out on 
the vector behaviour and ecology to provide information for 
discriminative and selective application of insecticides. 
Efforts were also put into the search for cheaper and less 
toxic insecticides and more economical and effective methods 
of insecticide application.
The use of insecticides still remains the most widespread 
method of tsetse control today despite its attendant problems. 
Owing to a shortage of foreign exchange, many African 
countries cannot afford the cost of insecticides to maintan 
the recurrent application which is necessary in most cases. 
Furthermore, the most effective methods of insecticide 
application require the use of aircrafts which again is high 
technology and out of the reach of most African countries. 
Hence control by these techniques is largely dependent on 
foreign aid.
Several lines of research on control strategies 
alternative to the sole use of insecticide were embarked upon 
by various workers. These include the search for a vaccine 
against trypanosomes, the breeding of trypanoto1erant animals,
University of Ghana          http://ugspace.ug.edu.gh
4the search for biological control agents for the vector, 
genetic control by the Sterile Insect Technique (SIT), 
physiological control and the development of effective baited 
traps and insecticide impregnated targets. There have been 
various degrees of progress in these fields of research.
Jordan (1986) gives an up to date appraisal of the various 
approaches and their future prospects. He identifies SIT and 
the use of baited traps and targets as the most promising 
approaches to vector control.
The SIT involves the release of large numbers of 
laboratory-bred sterile males into a wild population to 
compete with the wild males and produce sterile progeny. In 
theory this approach is most effecient for low density 
populations since the number of sterile males released should 
be about 10 times that of the wild males. Field trials of the 
SIT have therefore involved the initial reduction of the 
target tsetse population by insecticides followed by the 
release of the sterile males. By this approach, successful 
control was claimed in Tanzania against G. m. morsitans (Dame 
et a I., 1980) and in Burkina Faso against G. p. gambiensis 
(Cuisance et a l ., 1980) and G. p. palpalis (Politzar and 
Cuisance, 1982). However, the technique does not appear 
practicable on a large scale basis because of the high cost 
involved in rearing and sterilizing large numbers of male 
tsetse flies.
The research into developing effective odour baited traps 
and targets appears to be the most promising approach to
University of Ghana          http://ugspace.ug.edu.gh
5achieving long term tsetse control . Reseai. chers in. this field 
believe that effective control can only be achieved on a sound 
knowledge of the tsetse/trypanosomiasis system involved; the 
failure of past control attempts is attributed largely to the 
lack of adequate knowledge of the target systems. One vital 
requirement for understanding any tsetse/trypanosomiasis 
system is adequate knowledge of the population dynamics and 
behaviour of the tsetse species. This in turn requires 
effective sampling techniques.
For many years, most of the information gained on the 
distribution and ecology of tsetse was through the fly-round 
sampling technique. This placed limitations on the knowledge 
about those species that did not lend themselves to this 
sampling method. Therefore, research into trap development was 
initially aimed at producing effective sampling tools for 
population studies. The development of the biconical trap 
(Challier and Laveissiere,1973) was a success in this respect. 
The trap has been used for studying the distribution and 
ecology of important vector species in many parts of Africa. 
Better parasitological and epidemiological data have also been 
obtained from the samples taken by the improved sampling 
technique. More recent research has concentrated on the 
improvement of trap efficiency for control purposes. This 
involves the development of more effective trap designs and 
the search for effective odour baits.
It is along this line that a project was undertaken at 
Nguruman, by a team of ICIPE scientists, with the objective of
University of Ghana          http://ugspace.ug.edu.gh
developing new approaches to tsetse and disease control 
through a greater understanding of the population dynamics and 
disease epidemiology and more appropriate tsetse control 
strategies (Dransfield et al . , 1986a). To better achieve this, 
a multi-disciplinary approach was adopted to gather 
comprehensive information on all possible aspects of the 
tsetse/trypanosomiasis system in the area. Such information 
could then be used to build an epidemiological model on which 
any control strategy could be based.
The Nguruman area is in a semi-arid zone of Kenya in the 
Rift Valley. In a general survey on the distribution of tsetse 
in the Maasai reserves, Lewis (1934) reported the presence of 
both Glossina pallidipes Austen and Glossina longipennis Corti 
in the area. It is also known that animal trypanosomiasis is 
prevalent. The pastoral Maasai tribe depend mainly on cattle, 
goat and sheep for their livelihood and cattle especially are 
of great local economic and social importance. Therefore, the 
control of tsetse would be a great contribution to the needs 
of such a community. Research by the ICIPE team was 
concentrated in an area of about 100 square kilometres, 
occupied by one of the local Group Ranches.
The biconical trap (Challier et al., 1977) was initially 
adopted as a sampling tool for the tsetse population. This 
soon yielded much information on the G. pallidipes population 
(Dransfield et al., 1986a) but very little on G. longipennis. 
Through odour bait technology at Nguruman by Dransfield et a l . 
(1986b), the efficiency of the biconical was greatly increased
University of Ghana          http://ugspace.ug.edu.gh
7for G. pallidipes but less so for G. longipennis. However, 
there were adequate numbers of flies dissected to show that G. 
longipennis was also infected with both Trypanosoma congolense 
and T . vivax, the parasites responsible for trypanosomiasis in 
the cattle. Since the ultimate aim of the Nguruman project 
was to control the disease it would be desirable for a more 
complete operation to gather information on G. longipennis as 
well.
It was against this background that the study reported 
here on Glossina longipennis Corti was undertaken at Nguruman. 
Glossina 1ongipennis belongs to the fusca group of Glossina 
which include G. fusca Walker, G. nigrofusca Newstead, G. 
brevipalpis Newstead and eight others (see Chapter 2). This 
group of relatively large tsetse flies have been broadly 
described as forest dwelling species but G. longipennis and G. 
brevipalpis have become associated with drier climatic zones 
although the latter species still tends to be restricted to 
the most humid parts of such habitats. Like many fusca flies 
these two species have a crepuscular behaviour pattern. They 
are, however, known to be peculiar for feeding on large game 
and are limited in distribution to eastern Africa, thus 
forming the East African sub-group of the fusca flies. Here 
they may be found existing sympatrica11y or allopatrically 
with other species of Glossina including well-known vectors 
such as G. pallidipes Austen. G. longipennis has been recorded 
from all countries in East Africa especially Kenya, where it 
has a particularly wide distribution.
University of Ghana          http://ugspace.ug.edu.gh
8The first objective was to develop an adequate sampling 
technique for G. longipennis which would then be used for 
further studies on the species. The main sampling methods that 
have been independently developed for tsetse include the fly 
round technique, the use of stationary or mobile bait, 
trapping and the capture of resting tsetse. Trapping has 
however emerged as the most favoured technique owing to such 
advantages as the possibility of sampling populations at 
several places at the same time, the suppression of human 
factors and the use of standard materials (Challier, 1982). 
Several trap types have been designed and developed for 
different species of tsetse but owing to the general lack of 
attention on G. longipennis very few of these traps have been 
tested for this species (Owaga, 1981).
Trapping and odour bait technology was therefore chosen 
as the line along which studies should be directed in the 
search for a sampling tool. A successful trap if developed 
would then be used to study the dynamics of the resident G. 
longipennis population. Moreover, one of the objectives of the 
Nguruman project was to develop a technology to allow local 
Maasai to control rather than eradicate tsetse, probably 
utilizing odour baited traps or screens (Dransfield et al . , 
1986a). Thus if the efficiency of the developed trap could be 
increased to cost-effective levels through odour bait 
technology, it could be incorporated in such a control 
programme.
University of Ghana          http://ugspace.ug.edu.gh
9If tsetse control programmes are initiated in areas with 
both G. pallidipes and G .1ongipennis with a system that is 
more effective for G. pallidipes the results could be 
disappointing for two main reasons. First, although it is 
generally regarded as a minor vector possibly because of our 
inability to sample it, it could actually be important in 
disease transmission. Secondly, it is possible that G. 
longipennis could replace G. pallidipes as a vector if the 
latter species were successfully controlled. Evidence of such 
a possible 'take-over’ from one species by another has been 
reported from a number of control experiments carried out in 
Cote d'Ivoire (Laveissiere et a l ., 1988).
In one case Laveissiere and Couret (1983) observed that 
after 5 months of control operation carried out against a 
mixed population of G. palpalis and G. tachinoides, the total 
apparent density of the two species remained unchanged because 
the population of G. palpal is increased at the same rate as 
the population of G. tachinoides was being reduced. A similar 
observati on was made on G. pal 1icera replacing G. palpal is as 
the latter was being controlled (Laveissiere and Hervouet,
1988). In both situations the traps or targets used for the 
control operation were more effective for one species than for 
the other. According to Laveissiere et al., (1988), the more
affected species leaves an ecological opening for the less 
affected one to colonize up to the carrying capacity of the 
habitat. The disease situation may become more serious if the 
new dominant species is a better vector. He concludes that to
University of Ghana          http://ugspace.ug.edu.gh
10
avoid this risk, it is safer to carry out research into 
control strategies that are effective for all the tsetse 
species in a given system. The
work reported here on G. 1ongipennis seeks to achieve this 
objective for the G.pallidipes / G .1ongipennis system.
University of Ghana          http://ugspace.ug.edu.gh
11
CHAPTER TWO 
LITERATURE REVIEW
1. INTRODUCTION
Literature on tsetse and trypanosomiasis dates back 
to the turn of the century when various colonial 
governments began to take control measures against 
disease outbreaks in their colonies. Most of the early 
information, which concentrated mainly on the 
identification of disease foci and the vector tsetse 
species, was documented in government reports. A few 
workers carried out some basic research on the biology 
and ecology of the important vector species which were 
published in various scientific journals e.g. Nash (1930, 
1937) and Swynnerton (1921, 1936). Most of the early work 
was synthesized by Buxton (1955) featuring the general 
biology of tsetse flies. Subsequent work on various 
aspects including tsetse distribution, ecology and 
control strategies have been well reviewed by Glasgow 
(1963), Mulligan (1970) and Ford (1971).
Despite the vast amount of literature existing today 
on tsetse and trypanosomiasis, there is very little 
information on the fusca flies in general and G. 
longipennis in particular probably because they were 
never implicated in any disease outbreak. There is some 
information on the distribution of those fusca species
University of Ghana          http://ugspace.ug.edu.gh
12
that could be detected in the course of surveys but very 
little research work has been carried out on their 
bionomics. Since recent findings indicate that some of 
the fusca species could be important in disease 
transmission it is vital to gather more background 
information on the group. The present chapter therefore 
aims at bringing together all available information on 
the fusca group in general and G. longipennis in 
particular. In the various sections, the information on 
the fusca flies or G. 1ongipennis is presented in a 
background relative to other Glossina species.
2. DISTRIBUTION OF GLOSSINA LONGIPENNIS AND OTHER FUSCA 
SPECIES
2.1. Geographical distribution
Various schemes have been put forward by a number of 
glossinologists in an attempt to explain the present 
distribution of the various Glossina species (Evens,
1953; Machado, 1959; Glasgow, 1963). Synthesizing the 
different views, Ford (1970) described the distribution 
in the light of various factors including paleontology, 
feeding habits, climate and vegetation. He concluded that 
the dispersal of the genus leading to its separation into 
the phylogenetic groups fusca, morsitans and palpalis 
took place in the remote past. Climatic factors are 
considered most important in influencing the distribution 
of the various species. The humidity and temperature
University of Ghana          http://ugspace.ug.edu.gh
13
experienced by the puparia and adult are critical for a 
number of species. In a new analytical approach to tsetse 
population dynamics and distribution, Rogers (1979) 
showed how the distribution limits of some Glossina 
species can be defined by temperature and humidity 
ranges.
There are a number of publications on the 
distribution of various Glossina species in Africa 
including maps prepared by individual countries. The 
first edition of tsetse distribution maps was prepared by 
Potts (1953-54) and a second edition by Ford and Katondo 
(1977a). A revision by Ford and Katondo (1977b, 1979) 
provided separate maps for the three subgeneric groups 
using international colours and symbols for the various 
species as provided by OAU/STRC(1971). In the latest 
revision by Katondo (1984) it is pointed out that there 
is a lack in knowledge on the general distribution of the 
fusca group compared to the other two groups. This is 
partly because members of the fusca group are considered 
of less economic importance and partly because they are 
more difficult to detect.
Fig. 2.1 shows the general distribution of the fusca 
group according to Ford and Katondo (1977b). It is 
alleged to have dispersed from a central point around the 
Congo Basin with G. brevipalpis and G. longipennis 
evolving separately as they moved eastwards and were 
subsequently isolated by the East African Pleistocene
University of Ghana          http://ugspace.ug.edu.gh
Low land  ra in  forest
^  D is t r ibu tion  o f  species  o f  
\ \ \ \  the fusca g roup
Figure 2.1: The distribution of the fusca group of Glossina 
in relation to the two zones of lowland rainforest 
(Map taken from Jordan, 1986)
University of Ghana          http://ugspace.ug.edu.gh
15
forest (Ford, 1970) and/or the eastern Rift Valley 
(Evens, 1953). Thus, Machado (1959) recognizes two major 
subgroups each composed of relatively closely related 
species. One comprising G. fuses, G. fuscipleuris, G. 
haningtoni, G. nashi, G. schwetzi, G. tabaniformis and G. 
vanhoofi, generally has a west and central African 
distribution. The second subgroup comprising of G. 
brevipalpis and G. 1ongipennis has an east African 
distribution. The remaining three species, G. medicorum, 
G. nigrofusca and G. severini, appear to be intermediary 
between the two groups both in phylogeny and in 
distribution.
When G. longipennis was first described by Austen in 
1895 from records of the species in Kenya and Somalia, it 
was then speculated that it also occurred in Tanzania and 
Uganda. The distribution maps today show that the species 
generally occurs between latitudes 8° N and 5° S in 
Ethiopia, Kenya, Somalia, Sudan, Tanzania and Uganda 
(Ford, 1971; Ford and Katondo, 1977b). However, Kenya 
remains the country in which it is most widely 
distributed. G .1ongipennis occurs in the north, south­
west, east and the coastal hinterland of the country. G. 
brevipalpis on the other hand occurs in all the above 
mentioned countries but extends along the coast as far 
south as Mozambique. Fig. 2.2 shows the distributions of 
G .1ongipennis and G. brevipalpis mainly in Kenya and near 
its borders in neighbouring countries.
University of Ghana          http://ugspace.ug.edu.gh
THE  D ISTRIBUTION  O F G . L O N G I P E N N I S  AND  G -BR EV IP A LP I S  IN KENY/
AND  N E AR  I T S  BORDERS
16
G.  l o n g i p e n n i s  
G B r c  v i p a l p i s  
N a t i o n o l  b o n d a r i e s  
S t u d y  l o c o t i o n
E ldorei
• V-£trr>u
Figure 2.2: The distribution of G. longipennis and G.
brevipalpis in Kenya and near its borders.
University of Ghana          http://ugspace.ug.edu.gh
17
2.2. Ecological distribution (Habitats)
Generally, all tsetse flies in Africa are 
found in various parts of the two basic vegetation types 
on the continent; the large zones of lowland rain forests 
in west and central Africa, separated by areas of 
savannah and the lowland savannahs and isolated forests 
in east Africa. Within these main climatic zones, tsetse 
tend to inhabit certain vegetation types but not others. 
Tsetse infested areas of vegetation were termed 'fly 
belts' by pioneer workers (Buxton, 1955). Furthermore, 
different species appear to show preference for different 
types of vegetation cover, such that some species are 
referred to as woodland or game tsetse and others as 
forest or riverine species. Descriptions of some typical 
tsetse habitats by different workers have been reviewed 
by Buxton (1955) and Mulligan (1970). Ford (1970) gives 
an outline of the general habitats of the three 
subgeneric groups. Generally, the morsitans group is 
described as woodland inhabiting, the pal pal is group as 
riverine and the fusca group as forest dwelling. 
Swynnerton (1936) and Jackson (1955) provide detailed 
descriptions of the habitats of many East African species 
of tsetse. Various other workers have described the 
habitats of different species in different parts of 
Africa as reviewed by Challier (1982). The different 
habitats presumably offer the optimal ecological 
conditions to which the various species are adapted.
University of Ghana          http://ugspace.ug.edu.gh
18
However, as observed by various workers in Challier 
(1982) and Katondo (1984), some tsetse species can be 
encountered in habitats with which they would not 
normally be associated. On the other hand, certain 
vegetation cover may appear suitable for some species of 
tsetse but found to be without tsetse. Such anomalies may 
be due to changes in climatic factors or due to certain 
human activities which could force out resident tsetse 
populations or invite other species to an area.
Of the fusca group eleven of the twelve species 
inhabit various types of forest:
a) The lowland rainforest species, which are 
confined to the Congo Basin and the forests of Gabon, 
Cameroons and the Guinea Coast. These include G. 
haningtoni Newstead and Evans, G. nashi Potts, G. 
tabaniformis Westwood, G. vanhoofi Henrard, G. severini 
Newstead and in some areas G. nigrofusca Newstead.
b) G. fusca Walker, G. medicorum Austen, G. 
fuscipleuris Austen, G. schwetzi Newstead and Evans and 
in some 1ocalities G. nigrofusca Newstead occupy areas 
outside the lowland rainforest. They can be found along 
the edge of the forests, in the tongues of forest along 
water courses, and in the savannah in the forest islands 
far from the main forest block.
c)G. brevipalpis Newstead is found in the islands of 
forest often associated with water courses in East
Af rica.
University of Ghana          http://ugspace.ug.edu.gh
19
d) G. longipennis  Corti. occurs in arid habitats of 
East Africa.
Jordan (1962a) gave detailed descriptions of the local 
distribution of most of the west African fusca flies in 
relation to vegetation climate and seasons.
G. brevipalpis and G. longipennis are usually 
encountered together in the dry climatic zones of East 
Africa,. However, G. brevipalpis still seeks out the most 
humid and shady parts of such zones to inhabit.
Swynnerton (1921) remarked that G. brevipalpis is rarely 
found away from heavy shade. Moggridge (1949) described 
G. brevipalpis in the Kenyan coast as being confined to 
vegetation along water courses.
From various reports by different workers G. 
longipennis on the other hand seems to be well adapted to 
dry climate. Neave (1912) observed large numbers of G.
1ongipennis in dry semi-desert thorn bush country between 
rivers while on the river banks it was replaced by G. 
brevipalpis. He described G. longipennis as a 'desert
haunting species....... seen to be entirely independent
of water and indeed rather avoids it.' He thinks it is 
absent from the sea coast because the climate is too 
humid for it. Other descriptions of G. longipennis 
habitats as given by various workers include; relic 
secondary forest patches or thickets, dry grasslands with 
patches of bushes and trees, savannah woodlands and 
wooded steppes. The habitat of G. longipennis at the
University of Ghana          http://ugspace.ug.edu.gh
20
Kenyan coast has been described in Challier (1982) as 
semi-arid A c a c i a / Commiphora thorn bushland. Power (1964) 
also described the habitat of G. longipennis in his study 
area near lake Jipe, Kenya, as 'typical of the extensive 
dry Commiphora/Acacia bush which occupies almost half of 
Kenya and forms the principal habitat of G. longipennis. 
Owaga (1981) and Dransfield et al., (1986a) described the
Nguruman study area as belonging to a semi-arid zone with 
the vegetation consisting of open plains and Acacia 
dominated woodland.
Lewis (1934) gave a description of the Nguruman area 
and reported an unusual encounter with G. 1ongipennis in 
moist swampy flats and on river banks. Potts (1937) also 
reported such an atypical situation where G. 1ongipennis 
was inhabiting moist country with G. brevipalpis in 
Tanzania. He remarked that G. longipennis is sometimes 
forced to shelter in the riverine forest and thickets as 
it cannot always stand the dryness and heat of the semi- 
desert .
3. THE ECOLOGY AND BEHAVIOUR OF FUSCA FLIES
Most of the early works on tsetse ecology in 
general, were reviewed by Buxton (1955). Since then there 
have been updated reviews by Nash (1960), Mulligan 
(1970), Jordan (1974) and Challier (1982). Progress in 
ecological studies has undoubtedly been tied to the 
development of sampling techniques. Because of this, far
University of Ghana          http://ugspace.ug.edu.gh
21
less is known of the behaviour and ecology of the fusca 
than of the palpalis and the morsitans groups since the 
former are less amenable to the known sampling 
techniques.
3.1. Breeding Sites.
Because the puparia of tsetse require certain 
environmental conditions for development, pregnant 
females usually select the most conducive parts of the 
habitat to deposit larvae. Saunders and Phelps (1970) 
describe the characteristic breeding sites of the three 
subgeneric groups of Glossina based on puparia collected 
by various researchers. More recent work on breeding 
sites has been reviewed by Challier (1982). The selection 
of breeding sites by tsetse depends on the type and 
nature of the habitat and the prevailing climatic 
conditions.
Of the fusca flies, the forest dwelling species 
which live in a relatively uniform climate tend to 
scatter their puparia, such that no specific breeding 
sites have been recorded. The other species, being in 
less uniform and more open habitats, restrict larval 
deposition to sites characterized by maximum vegetation 
cover. Lewis (1939) recorded the puparia of G. 
fuscipl eutris under logs in dense riverine thickets. The 
puparia of G. fusca and G. medicorum were also found 
under logs but more in the dry season than in the wet 
season (Page, 1959). Lamborn (1921) and Swynnerton (1936)
University of Ghana          http://ugspace.ug.edu.gh
22
found large numbers of G. brevipalpis puparia under the 
shade of big trees in loose sandy soil with much humus. 
Although he did not find large numbers of puparia, 
Moggridge (1949) observed that G. brevipalpis appears to 
have preference for logs and the bases of thickets.
Lewis (1942) identified the most favoured breeding 
sites of G. longipennis as beneath logs, leaning tree 
trunks or stumps of felled trees near shrubby and woody 
thickets. Fewer puparia were found in the open country in 
the shade of isolated trees. He observed that puparia 
were usually deeply buried in friable soil of a mound but 
under logs were only one to two inches deep. Numerous 
sites were located throughout the Commiphora/Acacia 
savannah but puparia were most abundant in sites near 
dense bushes. In the Nguruman study area a few puparia of 
G. longipennis were recorded by Adabie (1987) in the 
riverine thickets which was found to be the best dry 
season breeding site for G. pallidipes in the same 
habi hat.
The influences of relative humidity and temperature 
on the duration, mortality and emergence of tsetse 
puparia have been discussed by several researchers 
including Nash (1931), Glasgow (1963), Challier (1973) 
and others reviewed by Challier (1982). The larger size 
of the G. 1ongipennis puparium is thought to contribute 
to its survival in drier conditions because it loses less 
water due to the smaller volume/area ratio than the
University of Ghana          http://ugspace.ug.edu.gh
23
smallei puparia of othet species.
3.2. Resting Sites
In a general review on tsetse sampling techniques, 
Muirhead-Thompson (1968) dealt with the early studies on 
resting sites. More recent studies have been reviewed by 
Hadaway (1977) and Challier (1982). The earliest detailed 
studies on resting sites were conducted by Nash and Davey 
(1950) on G. medicorum and G. fusca which like most fusca 
flies were not readily caught by the fly-round technique. 
Later on Nash (1952) showed through resting site studies 
that G. medicorum was a carrier of trypanosomes. Since 
then, very little work has been done on the resting sites 
of the fusca flies.
The search for resting flies by Nash and Davey(1950) 
in Nigeria yielded more G. medicorum and G. fusca flies 
than were obtained by any other sampling method. G. 
medicorum were found resting on the trunks of saplings, 
1-3 inches thick, in the high forest. Flies were usually 
resting head downwards on the shaded side of the trunk. 
They observed a daily variation in resting height ranging 
from one and half to 8 feet above ground. G. fusca on the 
other hand was not found resting in the high forest but 
in an adjacent teak plantation without undergrowth. Flies 
also rested head downwards 5-9ft above ground on the bark 
of the teak trees. Swynnerton (1921) showed in a cage 
experiment that G. brevipalpis rested on rough tree
University of Ghana          http://ugspace.ug.edu.gh
24
trunks, placing themselves in holes and grooves in the 
bark of the trunks that had colours close to that of the 
fly. He also observed that it was easier to locate males 
than females and pregnant females could not be located by 
searchers unless disturbed. He observed that flies could 
rest as high as 9ft in the cages. Jack (1941) observed 
large numbers of G. brevipalpis settling on roads and 
paths late in the evening. Moloo et a l ., (1980) recorded
G. brevipalpis resting on branches and trunks of trees 
and to some extent on rocks. There are no published 
reports on the resting habits of G. longipennis but 
Dransfield (pers. comm.) observed the species at Nguruman 
resting in the morning sun on rocks.
3.3. Activity
Before the idea of searching for resting tsetse was 
recognized, all tsetse sampling methods depended on flies 
coming to man, animal baits or traps. Thus, most of our 
knowledge on tsetse populations have been gathered on the 
active sections of the population. It was observed quite 
far back that the numbers of flies caught by these bait 
methods are dependent in part on their numbers in the 
area and in part on their activity at the time of 
sampling. Buxton (1955) gives a comprehensive review of 
some of the early researches into the factors influencing 
tsetse activity and pointed out the difficulties in 
separating the various factors involved.
University of Ghana          http://ugspace.ug.edu.gh
25
Activity in nature, which implies movement, is 
initiated for various life processes; the search for 
food, mates, and 1arviposition sites. In the past, 
several researchers have addressed the question of what 
controls the timing of tsetse activity. While recognizing 
the possibility of a complex interaction of climatic 
factors, different parameters have been emphasized to 
varying degrees by different investigators; Vanderplank 
(1948) on saturation deficit, Pilson and Pilson (1967) on 
temperature and Harley (1965) and Power (1964) on light 
intensity. According to Pilson and Pilson drastic 
depression in activity occurred above 32°C degrees but 
Barrass (1970) observed that high light intensities 
inhibit activity even below 32°C.
In more recent years it has been possible, through 
intensive laboratory studies by Brady (1970, 1972, 1973) 
and field studies by various workers reviewed by Challier 
(1982), to isolate the activity element i.e the readiness 
of the flies to move and come to baits. From the 
laboratory experiments, it was observed that tsetse flies 
exhibit spontaneous activity in relation to available 
energy reserves and that this spontaneous activity is 
controlled by an endogenous circadian clock (Brady, 1972, 
1973; Crump and Brady, 1979).
Very little work has been done on the fusca flies in 
general and G. longipennis in particular. Whereas most 
species of the morsitans and palpalis groups are active
University of Ghana          http://ugspace.ug.edu.gh
26
during the day, the few reports show that most fusca 
species are crepuscular. Nash (1952) reported on G. 
medicorum as being active in twilight hours. He concluded 
that the species was more active in the early morning 
than in the late evening. On the same species in Ghana, 
Chapman (1950) found that the largest number collected 
in a slow moving vehicle was taken at dawn (0600h). In 
one hourly catches made from bait bullock and buffalo, 
Kangwagye (1974) recorded more G. fuscipleuris in the 
evenings than in the mornings. Relatively low numbers 
were recorded throughout the daytime and the nighttime.
In a brief note on the activity pattern of G. 
brevipalpis Lamborn (1921) observed that flies were most 
active at dusk between 1800h and 1815h and no flies after 
then. Swynnerton (1921) also reported that G. brevipalpis 
remained quiet throughout the day but moved freely and 
buzzed at sunset, coming out to settle on unprotected 
vegetation. A detailed study on the activity pattern of 
G. brevipalpis was carried out by Harley (1965). Using a 
bait oxen he carried out hourly catches over 24 hours in 
different parts of the habitat in Lugala, Uganda. Peak 
activities were regularly recorded soon after sunset and 
before sunrise. He further observed that in shaded sites 
the morning peak was one hour later than in open sites, 
for both sexes. The evening peak occurred one hour 
earlier but only for males. By using hurricane lanterns, 
a few flies were also caught throughout the night. He
University of Ghana          http://ugspace.ug.edu.gh
27
was, however, not sure whether such flies were attracted 
to the light or to the bait animal. Among the various 
environmental factors that he measured during the study, 
he concluded that light intensity was the most important 
factor influencing activity.
On G. longipennis Neave (1912) observed that the 
species was inclined to feed early in the morning and 
late in the evening. Lewis (1942), using a screen-aided 
fly-round sampling method, recorded two flies during the 
day, more flies in the early morning and the largest 
catch in the evening. Peak catches were made at about 
1820h, when the sun was disappearing beneath the horizon. 
Attempts were made using powerful torches and lanterns in 
the dark but no flies were caught after dusk. Stationary 
lights were also set near the tents but no flies were 
attracted to them.
Power (1964) carried out a more detailed study on G. 
1ongipennis at the Kenyan coast near lake Jipe using the 
fly-round technique. Four catching teams each carrying a 
3 x 6ft brown screen were put on different straight paths 
(100yds long). The catchers walked back and forth along 
the paths stopping at 6 minutes intervals to catch flies 
with hand nets. Catching took place from dawn till about 
0900h and from 1745h till dark each day, for a fortnight. 
Measurements were made of temperature, relative humidity 
and light intensity. The results showed that the species 
was very active just after sunset, considerably less
University of Ghana          http://ugspace.ug.edu.gh
28
active at dawn and virtually inactive during the day. The 
pgFcentage catch on females was very low so the data were 
only analyzed for males. He also suggested that light 
intensity appeared the most important factor influencing 
activity with temperature playing a minor role.
3.4. Host Preference
The host preference of GIossina can be determined 
through the identification of blood meals from recently 
fed flies. The various methods available for blood meal 
identification has been reviewed by Weitz (1970). It has 
been shown through such studies that the feeding habits 
are very characteristic for different species of 
GIossina, although some variations may be observed at 
different times of the year under different climatic 
conditions. However the feeding habits of some species 
are similar to a large extent and can therefore be 
grouped according to the host most frequently fed on.
Five groups of tsetse have been created based on their 
feeding habits, details of which can be found in Weitz 
(1970) and Challier (1982).
According to this grouping scheme, G. 1ongipennis 
and G. brevipalpis belong to the group of flies that feed 
mainly on mammals other than suids and bovids. The 
earliest record on the host of G. longipennis came from 9 
blood meals identified as belonging to rhinoceros (Weitz 
and Glasgow, 1956). Subsequently, Weitz et al . (1958)
University of Ghana          http://ugspace.ug.edu.gh
29
collected 336 blood meals from resting G. longipennis at 
Kiboko, Kenya which were identified as follows: 74% were 
from rhinoceros (Diceros bicornis)* 16% from buffalo 
(Syncerus caffer), 10% from birds, elephant, pigs, cats 
and unidentified bovids. They suggested that ostrich 
(Struthio camel us) probably contributed the blood meal 
due to birds. Single feeds were found for man, dog, 
porcupine and aardvark. They concluded that since 
rhinoceros and buffalo were comparatively less available 
in the area than other animals like impala, hartebeest, 
Grant's gazelle the results showed a very strong 
dependence of G. longipennis on rhinoceros and buffalo.
4. SAMPLING TECHNIQUES FOR FUSCA GROUP FLIES
Sampling of tsetse started with reconnaissance 
studies, by workers involved in tsetse control in the 
early years of the century. Some of the early 
investigators made records of the numbers of flies caught 
on humans at certain spots over certain routes and kept 
notes on the type of species, pregnancy and hunger stages 
and age of flies caught (Buxton, 1955). These methods 
later developed into what is described as the 'fly-round' 
and the procedure for carrying it out was defined by 
Potts (1930). The technique was subsequently adopted as a 
sampling technique, first for assessing the efficiency of 
control operations and later on for general ecological 
studies (Smith and Renninson, 1961).
University of Ghana          http://ugspace.ug.edu.gh
30
As more emphasis was being laid on the need for a 
better understanding of tsetse ecology and behaviour to 
enhance better control strategies, other sampling methods 
were developed. Some of these later methods were aimed at 
sampling those tsetse species that did not respond to the 
fly-round sampling technique whilst others were meant to 
provide more information on tsetse populations than could 
be obtained by the fly- round technique. The main 
sampling methods that were developed included the use of 
stationary bait, capture of the resting population and 
trapping. All these methods including the fly-round 
technique have been developed modified and applied to 
varying extents in different geographical regions for 
different species of Glossina. Ford et aI. (1959),
Glasgow and Phelps (1970), Muirhead-Thompson (1968) and 
Potts (1970) have given general reviews on tsetse 
sampling techniques. More recent developments in sampling 
methods are reviewed in a section of Challier (1982).
4.1. The fly-round technique
Basically, the method involves using a group of fly- 
boys who act as combined bait and catchers to move along 
a path that is laid across a range of vegetation types in 
the habitat. They then stop at intervals to collect, 
using hand nets, the flies that land on them and on 
objects near them. Overall, the method has not been very 
effective for the fusca group of flies.
University of Ghana          http://ugspace.ug.edu.gh
31
Of the forest species in West Africa, Chapman 
(1950), Nash (1952, 1959) attempted sampling G. fusca and 
G. medicorum by this method. They made the common 
observation that these species were not attracted to man, 
even when dark screens were carried to increase the 
attractiveness of target. Very few G. fusca were recorded 
and G. medicorum was virtually undetected. Chapman (1950) 
was successful in sampling the species in Ghana only by 
using a slow moving vehicle. More extensive studies 
carried out by Jordan (1962a) on these species and the 
other west African fusca flies confirmed that the 
technique was not effective.
In East Africa, the method was also used in general 
surveys on the distribution of G. brevipalpis and G. 
longipennis. Lamborn (1921), Swynnerton (1921) and Jack 
(1941) made the common observation that G. brevipalpis 
did not land readily on man but flies could be caught 
with hand nets in the most active periods as they came 
out and landed freely on nearby vegetation and paths.
Jack (1941) observed that even when screens were carried, 
the attraction was poor. By using straight paths across 
different sections of the habitat, Moggridge (1949) was 
able to record adequate numbers of G. brevipalpis. He 
found that individual flies were taken in during the day, 
more flies at dawn and the most at dusk. All these 
workers observed that males formed the bulk of the 
catches and Jack (1941) found from fat content analysis
University of Ghana          http://ugspace.ug.edu.gh
32
that these were not hungry flies.
In general surveys on the distribution of G. 
longipennis, Neave (1912), Lewis (1934) and Potts (1937) 
all used the fly-round method. The numbers of flies 
caught were very low, single flies on some occasions, 
that only served to indicate the presence of the species 
in the area. Lewis (1942) however, incorporated a dark 
screen and was able to identify the most active period of 
the species from the catches made. A more intensive study 
was carried out by Power (1964) using four catching teams 
along straight paths. Each team carried a 3 x 6ft brown 
screen and walked back and forth stopping at 6 minutes 
intervals to make catches. The success of the technique 
was such that he was able to carry out a mark release 
recapture study to estimate the male fly population in 
the habitat.
The main drawback of the fly-round technique is the 
very low yield in female catch as has also been observed 
for non-fusca species. With the forest species, the main 
difficulty lies in penetrating the thick habitat. 
Generally, the numbers of flies caught depends on the 
catching efficiency of fly-boys. To get round the latter 
problem, Rogers and Smith (1977) developed a portable 
electric back-pack which can be worn by fly-boys so that 
flies coming to them get electrocuted.
University of Ghana          http://ugspace.ug.edu.gh
33
4.2. The Stationary Bait Technique
In this method large animals especially cattle are 
tethered and continuous catches carried out of the flies 
coming to them. This technique was developed with the aim 
of catching species that were not readily attracted to or 
are repelled by the presence of man. The method also has 
the added advantage that it is more feasible to extend 
the sampling periods longer than those used in the fly 
round method (Page, 1959).
The stationary bait technique has been found to be 
quite effective for detecting some fusca species which 
are not frequently attracted to man. Jordan (1962a) used 
it to sample G. fusca, G. haningtoni, G. nashi, G. 
tabaniformis and G. medicorum after failing to catch 
these species by the fly-round method. It was by this 
method of sampling that the vectorial status of G. fusca 
in West Africa was assessed by Nash (1965a). Among 
several animals that were tried, bait oxen were the most 
effective and sheep and domestic pig were least 
attractive. Rangwagye ( 1971, 1973, 1974) studied the 
activity pattern and ecology of G. fuscipleuns through 
catches made from bait bullock and buffalo. An intensive 
study was carried out by Harley (1965) to determine the 
activity pattern and vectorial capacity of G. brevipalpis 
using bait oxen (see section on activity for details). 
There are no published reports on the use of the animal
University of Ghana          http://ugspace.ug.edu.gh
34
bait method to sample G. longipennis but Dransfield et 
al. (unpublished) recorded fed flies in electric screens 
around a bait cow at Nguruman.
A number of factors have been observed to influence 
the catches made by the stationary bait technique. These 
include the variation in bait size, colour and scent 
(Saunders, 1964), the presence of alternative host in the 
neighbourhood and the visibility of the host depending on 
vegetation cover (Muirhead-Thompson, 1968). Vale (1974a) 
has showed that the presence of man can actually have a 
repellent effect on the catches of G. m. morsitans. He 
therefore used electrified grids of fine wire around the 
tethered oxen to electrocute flies attracted to the 
animals. This modification has the added advantage that 
the catch size no longer depends on the skill of the 
catcher.
4.3. Sampling the resting population
For many years, sampling methods were directed 
towards the active population. Nash and Davey (1950) 
initiated studies on sampling the resting populations.
The technique simply involves getting into the fly 
habitat and searching for resting tsetse.
Following the failure after several attempts to 
sample G. fusca and G. medicorum by the fly-round 
technique in the Olokemeji forest, Nigeria, Nash and 
Davey decided to search for resting flies. Observations
University of Ghana          http://ugspace.ug.edu.gh
35
were made over four days in an area of about 25 km2 . 
Search on the tree trunks in the high forest produced 
unprecedented catches of G. medicorum; 20 males and 17 
females (one blood-fed and one carrying a third instar 
larva). G. fusca was not found in the high forest but 
rather in a neighbouring teak plantation. They described 
the technique as being tedious, requiring a lot of 
patience and caution. One needs to move through the 
habitat pausing methodically to study every trunk; by 
first looking down the whole profile of the trunk on all 
sides then scrutinizing the bark. One should always 
advance into the sun as flies mostly rested on the shaded 
side of the trunk. Flies were caught with hand nets 
swooped downwards at a moderate speed. Fast swoops were 
observed to be ineffective. There are no reports of such 
intensive searches made on any other fusca species 
elsewhere. Chapman (1950) found it difficult to apply the 
method on G. fusca in Ghana because the vegetation was 
too thick to penetrate. For blood meal analysis to 
determine the preferred host of a number of GIossina 
species in Kiboko, Kenya, Weitz et a l . (1958) obtained
fed G. longipennis mainly by collecting resting flies 
from tree trunks.
The main problem with this method, in general, is 
the difficulty in locating motionless tsetse which often 
blend in colouration with that of their background 
(McDonald, 1960). The technique is also difficult to
University of Ghana          http://ugspace.ug.edu.gh
36
standardize because it depends very much on the skill of 
the searcher.
4.4. Trapping
The few attempts that have been made at using traps 
to sample some fusca species have shown traps to be least 
effective compared to other methods. Dransfield (1984) 
used the unbaited biconical trap to sample mixed 
populations of G. pallidipes, G. brevipalpis and G. 
austeni but recorded only low numbers of the latter two 
species. The biconical trap was also used by Gouteux 
(1983) to study the population dynamics of G. fusca and 
G. nigrofusca in the Ivory Coast. They remarked that the 
numbers recorded were very low. Kaminsky (1987) also 
reported relatively low catches of these two species in 
Liberia using biconical traps. A number of traps 
including Moloo's trap (Moloo, 1973), Langridge's trap 
(1968) and the biconical trap (Challier et al., 1977) 
were tried for G. longipennis at Nguruman with very 
little success (Owaga, 1981). Since one of the 
objectives of this study is to develop an effective trap 
for this species a more general literature review on 
trapping technology, is given below.
5. TRAP/ODOUR BAIT TECHNOLOGY
This technique involves the use of devices to catch 
the flies after they have been attracted with a stimulus
University of Ghana          http://ugspace.ug.edu.gh
37
of some kind. For various reasons, more attention has 
been directed at improving trapping as a sampling 
technique, than the other methods. Glasgow and Phelps 
(1970) think that trap catches are less biased than human 
catches and that a population can be conveniently sampled 
at different places at the same time using several traps 
and over long periods. An efficient trap can also be 
incorporated in a control campaign.
Challier (1977) reviewed trapping technology 
including the description and performance of various 
traps and screens, the use of odour attractants with 
traps and the behaviour of tsetse in relation to traps.
He classified traps into three categories based on the 
principle on which they function. These include; 
attracting screens, falling cages and artificial refuges 
and traps proper (’tridimensional1).
Attracting screens are simple visual attracting 
surfaces of dark coloured materials. Flies attracted to 
them are either collected by hand nets, electrocuted, 
glued to the surface, trapped in water and detergent or 
killed by insecticides sprayed on the surface. Falling 
cages depend on the flies being attracted to a bait under 
a net or a cage and the flies being trapped when the cage 
is dropped. Artificial refuges can be built from various 
materials including boxes, huts and concrete pipes. Flies 
are caught with hand nets or other trapping mechanisms 
after they have entered these in search of resting sites.
University of Ghana          http://ugspace.ug.edu.gh
38
Tridimensional traps are designed in such a way that 
the fl'ies ate trapped mechanically when they enter the 
trap. Several types of such traps have been developed 
either independently or based on previous designs for 
different Glossina species at different places. Challier 
(1977) gives a list of the different trap types, most of 
them taking their names from their designers. The main 
ones include: the Harris trap (Harris, 1930), Chorley's 
trap (Chorley, 1933), Swynnerton's trap (Swynnerton,
1933; 1936), the Animal trap of Morris and Morris (1949), 
Langridgevs trap (Langridge, 1968). and Moloo's trap 
(Moloo, 1973). The more recently developed traps include, 
the biconical trap (Challier and Laveissiere, 1973; 
Challier et al., 1977), the ’A' and 'C 1 series traps of 
Hargrove (1977), the beta trap (Vale, 1982b), the F2 and 
F3 traps (Flint, 1985) the monoconical trap (Lancien, 
1981) the pyramidal trap ( Gouteux and Lancien, 1986) and 
the Vavoua trap (Laveissiere et a l ., (unpublished 
report).
Generally, tsetse traps, mechanical or otherwise, 
rely on one or a combination of the following stimuli for 
attraction; colour, shape and movement. The earliest 
traps, were designed to resemble common mammalian hosts 
in some aspects of size, shape and colour. Recent 
researches in trapping technology have adopted a more 
analytical approach, paying more attention to the various 
component parts of the trap.
University of Ghana          http://ugspace.ug.edu.gh
39
5.1. Colour stimulus
It was generally accepted by early workers that 
tsetse are more attracted to dark surfaces than light 
ones. Based on this knowledge, Maldonado (1910) made his 
plantation workers wear black cloth on their backs, 
smeared with glue, to trap G. p. palpal is on the Island 
of Principe. This was the earliest idea in trapping 
technology. Dark screens have been used by a number of 
workers including Jack (1941) and Swynnerton (1936) to 
trap various Glossina species. More recently, work by 
Barrass (1970) and Vale (1969; 1974a) has confirmed the 
greater attractiveness of dark surfaces compared to 
lighter ones. Lambrecht (1973) assessed the performances 
of solid panels of different colours and colour 
combinations to study the landing behaviour of G. m. 
centralis. He, however, found that white screens gave the 
best results probably due to better contrast with the 
background. Dransfield et a1, (1982) tested different
coloured water traps for G. m. submorsitans and G. 
tachinoides and found that white was most effective for 
the former species but black for the latter. Whereas the 
previous workers thought that colour preference by tsetse 
was a response to the brightness contrast with the 
background, Dransfield et al. (1982) suggested the
importance of the intensity of reflected light especially 
in the UV range.
University of Ghana          http://ugspace.ug.edu.gh
40
Id more recent years coloured screens (targets) have 
been tiied in a number of tsetse control experiments.
Blue and black screens impregnated with deltamethrin were 
effective in reducing the populations of G . pal pal is and 
G. tachinoides.in West Africa (Laveissiere and Couret, 
1981). Vale et al., (1986, 1988) also showed that 
insecticide impregnated targets were also effective for 
controlling G. pal 1idipes and G. morsitans in Zimbabwe, 
(see section 4.6 for more details).
Colour has also been exploited in the design of 
mechanical traps for tsetse. The first mechanical trap of 
Harris (1930) was based on the attracting screen 
principle; it was basically an attractive screen which 
was partially enclosed by a trapping device and almost 
all traps that follow are variations of this model. The 
colour component of various trap types have been 
investigated to varying extents by number of researchers. 
The improvement on the efficiency of the biconical trap 
by Challier et al. (1977), when the white lower cone was
replaced by a blue one is a classical demonstration of 
the importance of colour in trap design. Owaga and 
Challier (1981) found that a sky-blue biconical trap was 
less efficient for G. pallidipes than a dark blue one. 
Gouteux et al. (1981) compared the performances of 
several biconical traps of different colours on G. 
pal pal is and showed that royal blue was the most 
effective. Vale (1982a) used electric screens to study
University of Ghana          http://ugspace.ug.edu.gh
41
the trap oriented behaviour of tsetse around traps of 
different colours and colour combinations. He concluded 
that white on the outside and black on the inside were 
the most effective colour combinations for inducing entry 
response. These findings led to the empirical study by 
Vale (1982b) and later Flint (1985) in developing the 
beta trap and the F2 and 3 traps respectively. The 
relative performances of black, blue and white biconical 
traps were also tested by Flint (1985) for G. pallidipes 
and G. morsitans but no significant differences were 
found between trap catches.
Challier (1977) made an earlier remark that the 
attraction of Glossina to colours is a complex phenomenon 
that has not yet been satisfactorily explained and this 
has hitherto been confused by the variation in results 
obtained by different investigators. Dransfield et al. 
(1982) observed that the significance of colour to tsetse 
depended on which aspect of the behaviour is involved. 
They found that among several colours of water traps used 
for G. m. submorsitans and G. tachinoides in northern 
Nigeria, black and white coloured ones were most 
attractive. Inferring from the high proportion of blood- 
fed female flies, they suggested that most of the flies 
attracted were seeking resting sites.
University of Ghana          http://ugspace.ug.edu.gh
42
More intensive laboratory and field studies have 
since been directed at understanding the response of 
tsetse to colour using more objective techniques. Green 
and Jordan (1983) investigated the spectral response of 
G. m. morsitans using the behavioral response to 
variously coloured lamps in Insect-O-Cutor traps. They 
observed that flies were more attracted to ultraviolet 
(UV) lamps than white ones. In a more detailed study, 
using behavioral techniques, with various wavelengths of 
monochromatic light ranging from UV to red, Green and 
Cosens (1983) observed that tsetse showed colour 
preference in the decreasing order of UV blue, red, 
white, yellow and green. From further investigation with 
the electro-retinogram technique, they identified three 
main areas of maximum sensitivity; U V , blue/green and 
red. The relative position of black can not be assessed 
in this manner although it is known to be quite 
attractive to tsetse. Vale (1982a) observed that in 
addition to being attractive, black encourages a greater 
landing response in tsetse than other colours. From the 
above findings Green and Cozens, 1983 and Jordan and 
Green, 1984 remarked that the discrepancies in the 
findings from different studies in the past may be due to 
the failure to take (JV wavelengths into consideration as 
earlier on suggested by Dransfield et al. (1982).
To assess the relative attractiveness of various 
colours in the field, some preliminary field trials were
University of Ghana          http://ugspace.ug.edu.gh
43
carried out with different coloured traps on G. morsitans 
and G. pallidipes by Jordan and Green (1984). They, found 
no apparent simple relationship between trap score and 
trap colour. The most important determinant of the 
catches seemed to be the proportion of attractive blue to 
unattractive green and yellow. More recently, Green and 
Flint (1986) carried out field experiments using 53 
different colours of F2 traps selected on the bases of 
spectral reflectivity in the colour ranges visible to 
GIossina. They observed that trap catches depended on 
relative amounts of spectral reflectivity in four 
distinct wavebands corresponding to UV , blue-green, 
green-yellow, and orange and red. Blue-green and red 
reflectivity were positively correlated with trap 
catches, whilst UV and green-yellow-orange were 
negatively correlated. Thus the best trap material turned 
out to be royal blue cotton cloth reflecting blue green 
strongly and very little of UV or green-yellow-orange.
The above field results are not in total agreement 
with the laboratory observations especially on the 
relative position of UV reflectivity. Thus, although the 
importance of colour as an essential component part of 
tsetse traps has been established more research is still 
needed to explain the phenomenon of tsetse responses to 
col ours.
University of Ghana          http://ugspace.ug.edu.gh
44
5.2. Trap design
Almost all mechanical traps for tsetse combine 
colour with shape in their design. Generally, dark 
surfaces and shaded cavities and a variety of shapes that 
contrast well with the background are the characteristic 
features of tsetse traps. The essential components of a 
typical trap are described by Challier (1977). The 
trapping mechanism is based on the fly being attracted tc 
and entering the body of the trap, in search of either 
shade for 1arviposition and resting sites or food. Once 
in the dark body cavity of the trap the flies are next 
attracted towards a lighted summit of the trap into whicl 
they climb, passing a 'no-return' device and into a 
retaining cage.
The earliest trap designs, reviewed by Swynnerton 
(1933) and Buxton (1955), were roughly of the size and 
shape of small animals but today the most effective traps 
take no regard of animal shapes. However, the size of the 
device is apparently still an important factor. Vale 
(1974a) and Hargrove (1980b) used electric screens to 
study the response of tsetse to a model (a horizontal 
cylindrical drum). It was observed that if the linear 
dimension of the model was increased three-fold the 
number of flies visiting it was higher. Gouteux et al . 
(1981) observed that smaller biconical traps did not 
significantly reduce catches until the trap was reduced
University of Ghana          http://ugspace.ug.edu.gh
44
5.2. Trap design
Almost all mechanical traps for tsetse combine 
colour with shape in their design. Generally, dark 
surfaces and shaded cavities and a variety of shapes that 
contrast well with the background are the characteristic 
features of tsetse traps. The essential components of a 
typical trap are described by Challier (1977). The 
trapping mechanism is based on the fly being attracted t< 
and entering the body of the trap, in search of either 
shade for 1arviposition and resting sites or food. Once 
in the dark body cavity of the trap the flies are next 
attracted towards a lighted summit of the trap into whicl 
they climb, passing a ’no-return1 device and into a 
retaining cage.
The earliest trap designs, reviewed by Swynnerton 
(1933) and Buxton (1955), were roughly of the size and 
shape of small animals but today the most effective trap* 
take no regard of animal shapes. However, the size of th< 
device is apparently still an important factor. Vale 
(1974a) and Hargrove (1980b) used electric screens to 
study the response of tsetse to a model (a horizontal 
cylindrical drum). It was observed that if the linear 
dimension of the model was increased three-fold the 
number of flies visiting it was higher. Gouteux et al . 
(1981) observed that smaller biconical traps did not 
significantly reduce catches until the trap was reduced
University of Ghana          http://ugspace.ug.edu.gh
45
to a third.
From a more detailed study of the behaviour of 
tsetse around and inside traps, Vale (1982a) demonstrated 
that trap design is of critical importance in determining 
whether a fly will enter it or not. After having tested 
different sizes, colours and colour combinations of traps 
he concluded that in bringing flies to a trap any fairly 
large trap of any shape would be satisfactory. He, 
however, recommended a horizontal entrance located at the 
base with tunnel-like openings into the inside of the 
trap. The importance of an appropriate colour combination 
in designing an effective trap has recently been 
investigated by Green and Flint (1986). By varying the 
colours on the inside and outside of the F2 trap they 
observed that the external colour affected both the 
attractiveness and efficiency of the trap for G. 
pallidipes and G. m. morsitans. Yellow and green traps 
were unattractive but efficient, black and red were 
attractive but inefficient, blue attractive and efficient 
and white moderately attractive and very efficient. The 
landing response was strongest on black and weakest on 
white materi al.
The performances of different trap designs have been 
observed to vary between species and locality. The 
Harris’ trap was very successful for G. pallidipes in 
Zululand (Harris, 1930) and was later also used in Zaire, 
Tanzania and Botswana by other workers. However the
University of Ghana          http://ugspace.ug.edu.gh
Morris' trap which was originally designed for West 
African species had to be modified for East African 
species. The blue biconical trap (Challier et al., 1977) 
has proved quite successful for a good range of GIossina 
spp. and has become widely adopted for sampling both the 
palpalis and morsitans species in many parts of Africa. 
Takken (1984) evaluated the biconical trap as a sampling 
device for tsetse in Mozambique. The square-shaped F3 
trap of Flint (1985) has been shown to be more efficient 
than the biconical trap for G. pallidipes (Flint, 1985; 
Vale et al.. 1986). Thus, if trapping is to be adopted as
a sampling technique for tsetse, more work is required to 
produce appropriate designs for the various species in 
different localities.
5.3. Movement
The attraction of tsetse to moving objects has long 
been known and was exploited in developing the fly-round 
sampling technique. There have been attempts by various 
researchers to incorporate movement in tsetse trapping 
with varying degrees of success. Vale (1969) showed that 
mobile screens and black rotating drums caught more G. 
pallidipes than stationary ones. Slow moving vehicles 
have been used by many workers to sample some species 
(Ford et al., 1972). Mobile components have been added to 
some mechanical tsetse traps in an attempt to increase 
trap efficiency but with very little success. Two white
46
University of Ghana          http://ugspace.ug.edu.gh
47
and blue screens revolving around the base of a biconical 
trap, proved to be less efficient for G. pallidipes than 
the stationary trap but when the speed was reduced with 
intermittent stops the performance was near that of the 
stationary trap (Owaga and Challier, 1981). Gouteux et 
al. (1981) also incorporated revolving screens at the 
base of a cylindrical version (lower cone only) of the 
biconica1 trap and found no significant increases in 
catches of G. palpalis. Generally, tsetse thus seem to be 
attracted by certain movement patterns but not others and 
more knowledge is required on this before movement can be 
effectively incorporated in trap designs.
5.4. Odour bai ts
The conventional fly-round and the baited cages 
sampling techniques combined odour and visual attraction 
of the host to capture tsetse. Generally, smell plays a 
role from longer distances than vision but the relative 
importance of these two senses seem to vary from species 
to species (Challier, 1977).
Olfactory attractants have been used by various 
workers to increase trap catches. These include host 
residues, organic wash-offs from host animals and various 
chemicals. Chorley (1948) observed that odours from 
cattle dung and urine were responsible for the 
concentration of G. m. morsitans and G. pallidipes in 
host resting places and further showed that these
University of Ghana          http://ugspace.ug.edu.gh
48
residues could be used to increase the catches of these 
species. Langridge (1961) and Persoons (1967) applied 
petroleum extracts of pig washing to hessian screens of 
traps to increase the catches of G. fusclpes and G. 
pallidipes. Vanderplank (1944) and recently, Vale et 
al.(1986) showed that animal bedding is significantly 
attractive to tsetse. Frezil and Carnevale (1976) used 
dry ice (CO2 ) to increase trap catches of G. fuscipes 
quanzensis.
Ox odour was shown by Hargrove (1980a), Hargrove and 
Vale (1978) and Vale (1980) to increase catches of G. 
pallidipes and G. morsitans and the combination of ox 
odour and carbon dioxide was found to be even more 
attractive (Vale 1982a). The attractancy of various 
chemicals was tested by Vale (1980) for G. pallidipes and 
G. morsitans. He observed that short chain ketones 
especially acetone, as well as formaldehyde and 
propionaldehyde were attractive but long chain ketones, 
heptaldehyde and caproic acid were repellent. He also 
showed that acetone dispensed together with carbon 
dioxide was more effective than either compound on its 
own. Through electro-antennographic techniques, Hall et 
al. (1984) identified l-octen-3-ol (octenol) as the most 
potent attractive component of the volatiles from cattle. 
It was shown to significantly increase the upwind flight 
of G. m. morsitans at concentrations as low as 0.9 ngl-1 . 
In field trials using beta traps, octenol increased the
University of Ghana          http://ugspace.ug.edu.gh
49
catch of both G. m. morsitans and G. pal 1idipes by up to 
three times relative to catches in unbaited traps when 
dispensed at dose rates between 0.5 and 50 mg hr"1 . When 
dispensed together with odour from a pit containing an ox 
the catches at an electrocuting net were significant 1y 
increased by up to 2.5 times for G. morsitans and up to 
1.5 times for G. pal 1idipes at dose rates between 0.5 and 
5 mg hr 1 . The efficacy of l-octen-3-ol has since been 
demonstrated in the field by various workersf Vale and 
Hall, 1985; Vale et a1.,1986; Dransfield et al ,1986a; 
Bursell et al., 1988; Vale et al., 1988) for G.
pallidipes and G. morsitans in Zimbabwe and Kenya and 
for G. m. submorsitans in Burkina Faso (Politzar and 
Merot, 1984). The effects of various doses of ketones and 
octenol on the catch size and composition of G. 
pallidipes were assessed by Dransfield et al (1986b).
The attractancy of bovid urine to tsetse has been 
demonstrated by various workers. Working at Nguruman, 
Kenya, Owaga (1984) observed that buffalo urine increased 
the catches of G. pallidipes in biconical trap by a 
factor of ten. She further compared the efficacies of 
buffalo urine and cow urine (Owaga, 1985) and reported 
that the buffalo urine was about 7.1 times more effective 
than cow urine. The potency of buffalo urine as a tsetse 
attractant was also reported by Saini (1986) using 
el ectro-antenogram techniques. More intensive 
investigations were carried out on cow urine and buffalo
University of Ghana          http://ugspace.ug.edu.gh
50
urine in combination with other chemical attractants by 
Dransfield ft al. (1986b) at Nguruman. They showed that 
cow urine on its own increased the catches of G. 
pal 1idipes in biconical traps up to 4 times and 9-15 
times when dispensed together with acetone. They, however 
found no significant difference between the effects of 
cow urine and buffalo urine. In Zimbabwe, Vale et al. 
1986b also showed that both cow urine and buffalo urine 
increased the catches of G. pallidipes by several-fold 
but there was no significant difference between the two. 
The latter authors and Owaga (1985) observed that the 
potency of the urine increased with the age of the 
sample. This aging process which presumably results in 
the higher concentration of the attractive components of 
the urine, was thought to be the possible source of 
discrepancy between Owaga's finding and the others’ in 
comparing buffalo urine and cow urine (Vale et al . ,
1986b). The potent attractants in buffalo urine and cow 
urine have since been identified by Hassanali et aI.,
(1986) and Bursell et al., (1988) as derivatives of 
phenol. Among eight of the derivatives occurring 
naturally in cattle urine, four of them, 3-methyl phenol, 
4-methy1 phenol , 3-ethyl phenol and 3-n-propy1 phenol were 
electro-antennographically active and induced upwind 
flight in wind tunnel bioassays in G. pallidipes and G. 
m. morsitans. When tested with traps baited with acetone 
and octenol all except 4-methy1 phenol were attractive to
University of Ghana          http://ugspace.ug.edu.gh
51
both species in Zimbabwe, but a marked sex difference was 
observed in the response of both species to the latter 
compound (Bui sell et al . , 1988).
There aie no published reports on the response of 
fusca group flies to odour baits. Dransfield (pers. 
comm.) observed that the catches of G. longipennis by 
biconical trap at Nguruman appeared to increase when the 
trap was baited with acetone and cow urine. On the 
palpalis group, most of odours tested for G. p. palpal is 
in West Africa do not seem to be effective (Laveissiere 
et al . ,1986). However, Cheke and Garms (1988) recently 
observed that biconical traps baited with octenol or 
acetone caught twice as many G. p. palpalis in Liberia as 
an unbaited trap.
4.5. Trap efficiency
Trap efficiency, as considered by most workers, 
means the comparative catches of insects by different 
trap types and the relationships between the 
characteristics of the traps and numbers of insects 
caught. Various workers have observed that the 
performances of traps do vary with the location of the 
traps. Harris (1930) recommended that traps should be 
located in optimal light and shade conditions. Morris and 
Morris (.1949) placed traps in sites where host were 
suspected to frequent ("hunting grounds") in order to 
obtain maximum trap yields. Glasgow and Duffy (1961)
University of Ghana          http://ugspace.ug.edu.gh
52
observed thai siting traps in relation to the sun and 
bio type:.; are also important. They recorded the highest 
catches of G . pallidipes on one side of a thicket before 
noon and on the opposite side after noon. Challier and 
Laveissiere (1973) recorded maximum catches by siting 
traps in more open and even sites. Langridge (1977) noted 
that traps should not be set in shade but in direct 
sunlight to give maximum contrast of light and shade on 
the bod/ of the trap. Muirhead-Thompson (1968) however, 
thought that the exact location of the best trap site 
depends on trial and error.
Hargrove (1977) developed a method for estimating 
trap efficiency by comparing the performance of the traps 
with that of an electric net of Vale (1974b) under the 
same conditions in a randomized block design. By this 
method Hargrove (1977) assessed the efficiencies of the 
Morris and Moloo's traps as 10% and 15-25% respectively. 
He observed that the day-to-day variation in trap 
performance (the numbers of tsetse caught by traps) was 
very great thus yielding very large coefficients of 
v a r i a t ion.
Vale and Hargrove (1979) considered that the above 
method in fact estimated the relative efficiency of the 
trap instead of the absolute efficiency which they then 
defined as the percentage of the flies approaching the 
trap that are actually caught by it. To estimate this, 
they developed the technique of placing an incomplete
University of Ghana          http://ugspace.ug.edu.gh
53
ring of electric nets around the trap to be tested and 
were able to estimate the number of flies which escaped 
the trap as well as the number caught by it. Odour 
attractants were used to help concentrate flies around 
the trap, so that the absolute efficiency of the trap 
then depended solely on its ability to capture the flies 
around it. They observed that estimates obtained by this 
technique, from several replicates, had smaller 
coefficients of variation compared to the previous 
method. Using the above technique they found that the A7C 
trap of Hargrove (1977) was a very efficient trap for G. 
pallidipes and G. morsitans, as its efficiency was 
comparable to that of an electric net also placed in the 
incomplete ring. Using the incomplete ring technique, 
Hargrove (1980b) was able to rapidly assess the 
efficiencies of various new trap models as modifications 
were being made. The efficiencies of three biconical 
traps were also estimated as 40-70% for G. morsitans and 
50-80% for G. pallidipes as against 70-90% for the small 
electric net whose efficiency was also estimated by 
placing it in the incomplete ring of electric nets. 
Details of the calculations involved are given by Vale 
and Hargrove (1979) and Hargrove (1980b).
5.6. The role of traps in tsetse control
Although traps have been used extensively for 
sampling tsetse flies, they have until recently been
University of Ghana          http://ugspace.ug.edu.gh
54
least considered in tsetse control campaigns. There are a 
few records from the early years of tsetse control where 
traps were incorporated in control operations. The most 
primitive trapping technology by Maldonado (1910) 
(plantation workers wearing sticky black cloth) was the 
first to be incorporated in a control campaign on the 
Island of Principe against G. p. palpalis. Harris (1936) 
used large numbers of his trap in an attempt to eradicate 
G. pal 1idipes in Zululand but without success. The Morris 
and Morris trap was used in Nigeria against G. p. 
gambiensis to reduce the population below transmission 
level and was later also incorporated in a control 
programme when G. p. pal pal is re-invaded Principe. With 
the advent of insecticides, investigations on the 
potential of traps for control were abandoned.
However, the development of the more effective biconical 
trap by Challier and Laveissiere (1973), the increasing 
cost of insecticides and the lack of foreign exchange for 
insecticidal control led to the renewed interest in the 
possibility of using traps for control. As such, a number 
of workers have recently carried out control experiments 
with the biconical trap. Laveissiere and Couret (1981) 
deployed 600 biconical traps impregnated with 
deltamethrin along a 62 km river bordering Burkina Faso 
and Ivory Coast against G. p. gambiensis and G. 
tachinoides. The catch of G. tachinoides was reduced by 
85% after 3 days and by 98% after twelve days and
University of Ghana          http://ugspace.ug.edu.gh
55
remained so for four months. The experiment was repeated 
by Laveissiere and Couret (1981) using blue screens 
(targets) also impregnated with deltamethrin. The effect 
was slightly less but the cost was also reduced. They 
observed that the effectiveness would be less in the 
rainy season but the potential had been demonstrated and 
the economics worked out. Lancien (1981) used a cheaper 
and more simplified trap, the monoconical trap, in a 
control campaign against G. f. quanzensis in the republic 
of Congo. In the forest region of Ivory Coast, biconical 
traps (Laveissiere et al., 1980) and screens (Gouteux et 
al . , 1981) were also shown to reduce populations of G. p. 
palpal is.
Following the promising results of these preliminary 
control experiments, considerable research work has been 
directed at improving the efficiency and reducing the 
costs of traps and methods of target installation and 
insecticide application (Laveissiere and Hervouet, 1988). 
Further control experiments have been carried out 
alongside to assess some of the research findings.
Gouteux et al (1987) have demonstrated that with 
community participation, a non-insecticide impregnated 
pyramidal trap could effect 97-98% reduction in tsetse 
populations in three villages in the Ivory Coast.
With the progressive discovery of tsetse odour 
attractants the potential of traps and targets for 
control is becoming greater. The prospects for using
University of Ghana          http://ugspace.ug.edu.gh
56
stationary baits for sampling and control of tsetse was 
discussed by Vale (1981). In Zimbabwe, Vale et a l . (1986)
showed that isolated populations of G. rn. morsitans and 
G. pallidipes could be eliminated rapidly with a sparse 
placement of odour baited targets impregnated with 
insecticide (Vale et al. 1986). More recently, Vale et 
al. (1988) used black cloth targets flanked by netting 
baited with l-octen-3-ol and acetone or butanone coated 
with deltamethrin against a G. pallidipes population in a 
600 km2 on the Rifa triangle in the Zambesi Valley 
Zimbabwe. A 99.99% reduction in the population of G. 
pallidipes was achieved in about six months. They 
discussed the effectiveness, practicability and 
ecological impact of targets in a large area that had 
abundant tsetse and was subject to strong invasion. 
According to Laveissiere et al (1986), no effective odour 
baits have been found to significantly increase the 
catches of G. palpalis in West Africa.
Jordan (1986) gives a review of the role of traps in 
past and present control campaigns and also provides an 
insight into the potential of trapping technology for 
future control efforts. The potential of odour baited 
traps and targets in tsetse control has been discussed by 
Vale (1987). Generally, there is every indication from 
ongoing research that any long lasting control strategy 
for tsetse will in future incorporate some degree of 
trapping technology.
University of Ghana          http://ugspace.ug.edu.gh
CHAPTER THREE
DESCRIPTION OF THE STUDY AREA
3.1. INTRODUCTION
All the field studies were carried out at Nguruman in the 
Kajiado district of the Rift Valley Province in south- westerr 
Kenya. The area is situated in a generally semi-arid zone and 
was formerly classified as a Maasai Reserve area. A wide 
variety of game animals are found there with herds of zebra 
and wildebeest moving in during the rains from the Serengeti 
ecosystem. Today the area is still inhabited mainly by the !
Maasai who keep herds of cattle, goats, sheep and donkeys. 
However, people belonging to other Kenyan tribes have recentl 
migrated into the area and taken to small scale, irrigated 
farming, producing mainly maize and a variety of fruits and 
vegetab1es.
The area has considerable potential for development. 
Firstly, the rangeland could be used for improved animal 
production for the benefit of the local Maasai community and 
for the nation as a whole. Secondly, the irrigation scheme 
could be extended to increase agricultural crop and vegetabl 
production. Finally, the abundant game could be a tourist 
attraction if tourist facilities were to be developed in the 
area .
One of the reasons why the area has not been fully 
exploited for these resources has been the presence of large
University of Ghana          http://ugspace.ug.edu.gh
58
numbers of tsetse flies, specifically Glossina pallidipes and 
Glossina 1ongipennis. Apart from the fact that tsetse flies 
can be a nuisance to man when numbers are very high, the 
species found in the area are vectors of animal sleeping 
sickness (nagana). Considerable numbers of Maasai cattle have 
been reported to regularly die from the disease especially 
during drought periods when the resistance level of the 
animals is very low and animals expose themselves to higher 
tsetse challenge by moving into the denser woodland in search 
of food. The control of tsetse flies and the disease would 
contribute enormously to the development of the area.
Since the presence of tsetse flies was reported in the 
area by Lewis (1934), no tsetse control has been carried out 
there. A project was undertaken by ICIPE in the area to carry 
out research into the understanding of the dynamics of the 
tsetse/trypanosomiasis system in order to develop suitable 
control strategies. At the start of the project, most of the 
work was concentrated on Glossina pallidipes which was thought 
to be a more important vector than G. longipennis. It was, 
however, realized that for a complete understanding of the 
system it was vital to have adequate knowledge of both 
species. Therefore the work to be reported was undertaken to 
provide information on G. longipennis that could be useful in 
the study of the tsetse/trypanosomiasis situation in the area.
University of Ghana          http://ugspace.ug.edu.gh
10 km
i----------1
Figure 3.1: Location of the main study area (boxed) in
relation to major geographical landmarks and other 
tsetse infested areas of the Nguruman tsetse fly  
belt.
University of Ghana          http://ugspace.ug.edu.gh
60
3.2. LOCATION OF THE NGURUMAN AREA
Figure 3.1 shows the location of the study area (boxed 
and labelled 1) in relation to other tsetse infested areas 
(numbered on the map) within the Nguruman tsetse belt. The 
area which forms part of the alluvial plains of the Rift 
Valley of East Africa is located between latitude 1° 48?N and 
longitude 35° 56'E and lies at 600-700m above sea level. 
According to Sayad and Sayad (1980), the dense wooded area 
covers 330km2 measuring 35km from north to south and 10km from 
east to west.
The major landmarks are the Ewaso Ngiro river to the 
east, the Nguruman escarpment rising to about 3000ft in the 
west and Mount Shompole about 20km to the south. To the 
north, the study area extends mainly as a strip of narrow 
vegetation along the Oloibototo to join a bigger tsetse 
infested area in the Olkeramatian plains. Within the latter 
area and about 10 km from the centre of the study area are 
located the irrigated farms.
3.3. CLIMATE AND WEATHER
Climatic data for the duration of the study are shown in 
Fig. 3.2. The Nguruman area generally experiences a mean 
annual rainfall of about 500-700mrn. Mean monthly maximum 
temperatures range from 32-41°C and minimum temperatures from 
18-22°C . Over the study period minimum relative humidities 
(taken at 1500h local time) ranged from 24% in February 1986 
to 48% in May 1987. However, previous records show that
University of Ghana          http://ugspace.ug.edu.gh
61
Figure 3.2: Profiles of monthly rainfall (histogram), mean
monthly minimum and maximum temperature and mean 
monthly minimum and maximum relative humidity 
recorded at Nguruman over the study period.
Rel
ativ
e 
hum
idit
y
University of Ghana          http://ugspace.ug.edu.gh
62
minimum relative humidities can go as low as 21% in the hot 
dry months of January and February.
Rainfall generally occurs in two seasons in a year; short 
rains occurring more or less in November and December and the 
long rains from late March till May. These are separated by 
hot dry months from January to early March and cold dry months 
from June to September. During the rains, grasses and shrubs 
grow rapidly and food is generally abundant for both domestic 
and wild animals. In the dry seasons, most grasses die leaving 
the ground bare, deciduous trees and bushes loose their leaves 
and there is a general scarcity of food for animals. Some 
species of game migrate to other places whilst domestic 
animals are grazed more in the thicker woodland and eventually 
moved up the escarpment if the dry conditions persist any 
longer than usual.
The main drainage systems of the area include the Ewaso 
Ngiro river, which flows permanently and the Oloibototo, Sampu 
and Lentore rivers which are seasonal. During very heavy 
rains, the study area may get flooded as the seasonal rivers 
overflow their banks.
3.4 VEGETATION
The vegetation in the Nguruman area has been described by 
Lewis (1934), Sayad and Sayad (1980) and Dransfield et a1. 
(1986a). Figure 3.3 (taken from Dransfield et al., 1986a) 
shows the main vegetation types in the study area. In moving 
from the Ewaso Ngiro river to the base of the escarpment one
University of Ghana          http://ugspace.ug.edu.gh
63
Figure 3.3: Map showing the different vegetation types in 
the main study area.
University of Ghana          http://ugspace.ug.edu.gh
64
traverses five main vegetation types as follows:
a. The Open Plains
This starts immediately after the wooded fringes of the 
river. The section is mainly grassland with isolated trees 
mainly of Acacia seyel fistula and A. siberiana. In the dry 
season the ground here is parched and bare except for some dry 
tussocks. During the rains however the area is covered with 
tall grasses and shrubs and most animals graze in this area. 
When the general area is hit by drought people tend to move 
their animals into the thicker woodland.
b. The Acacia Woodland
This is characterized by a closer distribution of various 
Acacia spp, including A . albida, A. seyel fistula and A.
torti1 is. The rest of the ground is taken up by grasses and
thicker shrubs than found in the plains.
c. The Riverine Thicket
This forms a strip of lowland woodland around the 
Oloibototo river which runs from north to south through the 
central part of the study area. The riverine vegetation is 
characterized by a close canopy of tall trees, shrubs liannas 
and bushes. The section is generally characterized by densely 
shaded thickets which form the main breeding sites for G. 
pallidipes. The dominant trees include Euphorbia candelabrum, 
Acacia pennata and Cassine aethiopica and shrubs include
Sautia myrtina and Aloe spp.
University of Ghana          http://ugspace.ug.edu.gh
65
d. The Lower and upper woodland
This is a wooded bush-land consisting mainly of close 
stands of thick shrubs and scattered Acacia and other trees. 
The dominant shrubs in this section include Boscia coriacea 
and Aloe spp. The ground becomes covered with tall grass 
during the rains. This area provides most animals with the 
bushes for browsing during the dry season.
The vegetation becomes thinner as the area rises to 
rocky ground towards the base of the escarpment. The tall 
trees give way to shorter scattered xerophytic acacia trees 
such as Acacia mellifera and thorny bushes. The grasses here 
are shorter and the ground is generally more open than the 
previous section.
e. The Valley Woodland
This is a dense wooded valley lying at the base of the 
escarpment consisting of tall trees mainly of Ficus spp and 
various types of shrubs and grasses. The vegetation becomes 
thicker along the course of the Sampu river coming down from 
the escarpment.
Both species of tsetse are found in all these vegetation 
types in varying densities but become more evenly distributed 
during the rainy season.
3.5. MACRO-VERTEBRATES
A wide range of both domestic and wild animals form the 
macro-vertebrates population in the area. Domestic animals 
kept by the Maasai include cattle, goats, sheep and donkeys. A
University of Ghana          http://ugspace.ug.edu.gh
6 6
few dogs and cats are kept as pets. The Maasai households 
which form the Olkeramatian group ranch in the area together 
own about 7000 cattle, 18000 sheep and 19000 goats.
A variety of wild animals are encountered as one moves 
from the plains into the woodland although the distribution is 
not sharply distinct neither is it constant throughout the 
year. Generally, the open plains are characterized by herds of 
zebra (Equus zebra L.), Grants gazelle (Gazella granti 
Brooke), wildebeest (Connochaetes taurinus) and less commonly 
hartebeest (Alcelaphus buselaphus Thomas). The zebra and 
wildebeest migrate into the area during the long rains. Also 
found in the plains are smaller numbers of eland (Taurotragus 
oryx (Pallas)), waterbuck (Kobus el 1ipsiprymmus (Ogilby)) and 
warthog (Phacochoerus aethiopicus L.). The big cats, lion 
(Panthera leo (L.)), and cheetah (Acinonyx jubatus (Schreber)) 
may also be encountered in the plains as well as the 
scavengers, black-backed jackal (Canis mesomelas Schreber), 
side-striped jackal (Canis adustus Sundevall) and the spotted 
hyena (Crocuta crocuta Erxleben). Packs of African hunting dog 
(Lycaon pictus (Temmick)) also move into the area at some 
times of year. The cape hare (Lepus capensis L.) is a very 
common sight on the plains and aardvark (Orycteronyx afer 
(Pallas)) are also found there. A number of large bird species 
found mainly in this area include ostrich (Struthio camelus 
L.), Kori bustard (Ardeostis kori),and the secretary-bird 
(Sagittarius serpentarius). Guinea fowls (Numida spp.) and 
francolins (Francolinus spp.) may be encountered both in the
University of Ghana          http://ugspace.ug.edu.gh
67
plains and in the acacia woodland.
Impala (Aepyceros melampus Lichtenstein) and giraffe 
(Giraffa camelopardalis (L.)) are encountered more frequently 
at the edge of the woodland with giraffe being more frequent 
in the acacia woodland. Large troops of baboon (Papio anubis 
(Fischer)) roam between the plains and the edge of the 
woodland whilst vervet (Cercopithecus sp.) and colobus 
(Colobus spp.) monkeys are found in the thick woodland. In the 
thickest parts of the woodland are found buffalo (Syncerus 
caffer Sparrman) which tend to graze near the river bed in the 
morning and late evening and withdraw into the heavy shade 
along the river bed in the hot afternoon. The ungulates here 
include bushbuck (Trangelaphus scriptus Pallas), and dikdik 
(Madoqua guentheri Thomas). Occasional sighting of leopard 
(Panthera pardus (L.)) and porcupine (Hysterix sp.) have been 
made. Some of the animals appear to move out of the study 
area during the dry season but baboons, warthog, buffalo, 
impala and giraffe remain the whole year round.
University of Ghana          http://ugspace.ug.edu.gh
6 8
CHAPTER FOUR
DEVELOPMENT OF AN EFFECTIVE TRAP/ODOUR BAIT SYSTEM
4.1. INTRODUCTION
Trapping has several advantages over other methods of 
sampling tsetse. First, it is a more objective technique of 
sampling because it eliminates the human element which 
sometimes introduces unacceptable variability into sampling 
data. Secondly, different parts of the habitat can be sampled 
at the same time and over long periods without interruption. 
Owing to the above reasons and others outlined by Glasgow and 
Phelps (1970), trapping is the most reliable and convenient 
technique for sampling tsetse populations. However, due to the 
lack of an efficient and convenient trap for G. 1ongipennis, 
the method remained inapplicable until recently.
The development of the biconical trap (Challier and 
Laveissiere, 1973; Challier et. al., 1977) marked the 
beginning of the widespread use of traps for sampling tsetse 
populations. It was developed primarily for species of the 
palpalis group in West Africa but was later also shown to be 
effective for G. pallidipes. Owaga (1981) used the biconical 
trap for G. longipennis but recorded very low numbers.
More recently, some other trap designs have been produced 
and shown to be more effective than the biconical trap for G. 
pallidipes and G. morsitans. These include the Beta trap 
(Vale, 1982b) and the F2 and F3 traps (Flint, 1985). These
University of Ghana          http://ugspace.ug.edu.gh
69
traps are, however, less widely used than the biconical trap 
as sampling traps because they are bulkier and more expensive. 
However, their designs were based on the principles 
recommended by Vale (1982a) for producing an efficient trap, 
after intensive studies on the trap-orientated behaviour of 
tsetse. Thus, they can be regarded as model traps from which 
effective but cheaper and more convenient traps can be 
developed.
Besides the work on various trap designs, attention has 
also been directed to identifying odour attractants that could 
be used to improve trap efficiency. Intensive recent research 
has confirmed that various substances can be used to greatly 
increase trap catches for a number of Glossina species (see 
literature review on odours). So far, most of these 
attractants have only been found to be effective to varying 
degrees for tsetse species of the morsitans group. A few have 
been tested on the palpalis group and found to be less 
effective. There are no published reports on the response of 
the fusca species to these attractants. Dransfield et al. 
(unpublished data), however, recorded higher numbers of G. 
longipennis when the biconical trap was baited with both cow 
urine and acetone, although acetone alone was ineffective.
In the search for an effective trap/odour bait system for 
G. longipennis, the first step was to test a number of odour 
baits known to be effective for other Glossina species. The 
biconical trap, which is still the most widely used sampling 
trap for many tsetse species, was taken as the standard trap
University of Ghana          http://ugspace.ug.edu.gh
70
for these experiments. Having identified effective odour 
attractants, the biconical trap was compared with the F2 and 
F3 traps and with a series of new traps developed at Nguruman. 
It was hoped that this approach could produce a cheaper and 
more convenient trap. Further experiments on odour attractants 
were then carried out using the best of the new traps. 
Following the establishment of the most effective odour baits 
on the best trap design, the absolute efficiency of the system 
was estimated using the technique developed by Hargrove and 
Vale (1979). This involves the use of electric screens around 
the baited trap.
4.2. MATERIALS AND METHODS
4.2.1. Traps tested,
a) The biconical trap.
Plate 1 shows a biconical trap of the type used in this 
study. Basically, the trap consists of two cones joined 
together at their bases over a metal ring of about 60 cm in 
diameter. The inverted blue lower cone, which forms the 
attractive part of the trap, has three vertical oval entrances 
. The upper cone is of white netting. A black piece of cloth 
is sewn on to the inner walls of the trap in a cruciform 
manner such that areas of it are visible from outside through 
the entrances. These areas function as targets to attract 
flies into the body of the trap after they have been attracted 
to the vicinity by the blue. Constructional details of the 
original version of the trap are given by Challier and
University of Ghana          http://ugspace.ug.edu.gh
71
Plate 1: The blue biconical trap
University of Ghana          http://ugspace.ug.edu.gh
72
Laveissiere (1973) and of the improved version by Challier et 
al. (1977).
For this study, traps were constructed from locally 
manufactured light weight polyester/cotton for the lower cone 
(blue or white) and cotton for the black cruciform inner 
screen(target). The upper cone was of white nylon netting. The 
blue version was taken as the standard with which all test 
traps were compared.
b) The F2 and F3 traps
These are basically the same trap type, differing only in 
the colour of the trap body; white for the F2 (Plate 2) and 
blue for the F3 (Plate 3). Essentially, the trap is a 1m x 1m 
x lm cube of cloth pulled tightly over a metallic frame with a 
horizontal entrance (lm x 0.3m) at ground level on one side. A 
target of black cloth (lm x 0.3m) is located on the inner wall 
of the back directly opposite the entrance of the trap. The 
cone is recessed in the trap and is made of grey netting. 
Details of construction are provided by Flint (1985). The 
traps used in this study were supplied by Low and Bonar, 
Harare, Zimbabwe. Trap materials were of a heavier weight 
cotton than that used for the biconical traps. Unlike in the 
original design, the F2 traps used were set by tying strings 
from the trap to external poles (see Plate 2). The F3 traps 
were stretched over aluminium poles as originally designed.
University of Ghana          http://ugspace.ug.edu.gh
73
Plate 2: The Zimbabwe F2 trap.
University of Ghana          http://ugspace.ug.edu.gh
74
Plate 3: The Zimbabwe F3 trap.
University of Ghana          http://ugspace.ug.edu.gh
75
c) The NGU traps
All other traps that were tested were new trap designs, 
developed at Nguruman jointly with Mr. Brightwell and referred 
to as the NGU series of traps. These were the NG1A, NG1B,
NG2A, NG2B, NG2C, NG2D, NG2E, NG2F, NG3A, NG4A and NG4B traps. 
The code numbers were given according to the type of netting 
cone design on the trap. Thus, traps bearing the same number 
have the same netting cone type and belong to the same model. 
Different letters within the same model indicate variations in 
some other parts of the trap body. Constructional details 
accompanied by diagrams in Figures 4.1, 4.2, 4.3, and 4.4 are 
given below. In all the figures, cloth measurements are given 
in widt.h(w) units, where Iw = 1 yard (0.9m), the width of
locally purchased cloth in Kenya.
The NG1 Model
This is essentially a triangular version of the F3 but 
with netting on only two sides of the cone. Two versions were 
tested, the NG1A and NG1B. The numbered component parts of the 
NG1A shown in Figure 4.1 are as follows;
1- Body (blue): This was folded along the dotted line in 
the middle.
2- Target (black): This was the first part of the trap to 
be affixed to the body as shown in the diagram.
3- Shelf/B’ront (blue): This was fixed after the target.
University of Ghana          http://ugspace.ug.edu.gh
76
The upper half of the rectangular section was fixed along
the marked edges of the trap body as shown in Fig. 4.1(1). It
was backed on the inside with an equal area of black cloth. 
The lower half was then tucked inwards and affixed along the 
horizontal sides of the body to form a shelf above the 
entrance of the trap as shown in the side view of the trap 
front (Fig. 4.1(8)).
4 & 5 - Netting cone (white): A lw x lw piece of netting
was folded diagonally as shown in Fig. 4.1(4) to give Fig.
4.1(5). This was fixed to the body and front of the trap as 
per points TXYZ indicated on the three component parts (body, 
netting and front). A 1 cm diameter hole was then cut in the 
apex and reinforced with cotton stitches.
6- Base (blue): this was affixed last. The entrance of 
the trap was thus at ground level measuring lw x 0.5w.
7- shows an assembled NG1A.
To produce the NG1B the following modifications were
made;
i) the base was removed.
ii.) the target was rotated 90° and affixed as in 
NG2 model (see below),
iii) the shelf above the entrance was removed 
(see Fig 4.1(8)) .
University of Ghana          http://ugspace.ug.edu.gh
HUtH
z
Bottom
Figure 4.1: Construction of the NG1 trap model.
University of Ghana          http://ugspace.ug.edu.gh
78
The versions produced and tested were the NG2(A F). The 
following descriptions based on the diagrams in Figure 4.2 
refer to the NG2A.
1- Body (blue): folded along middle dotted line.
2- Target (black): affixed as shown in the diagram.
3- Front: blue on the outside backed by black on the 
inside and affixed as shown in the diagram.
4- Netting cone (white nylon): A lw xlw netting
was folded along the dotted lines shown in Fig. 4.2(5). A 
triangular 1/4 piece was cut off as in Fig.4.2(6).
Joining the cut edges produced a cone shown in
Fig. 4.2(7). The base XYZ of the cone was fixed to the
top of the trap by two sides to the blue body and 
one side to the front. A hole was cut at the apex as 
before. A diagram of an assembled NG2A trap is shown in 
Fig.4.2(8) .
The variations for the other NG2 models were as follows; 
NG2B: The front consisted only of black cloth.
NG2C: i) The front in the NG2B was extended 20cm
downwards so that the entrance was reduced.
ii ) The lower half of the target was of 
double thickness.
NG2D: The black front was affixed vertically 
instead of slanting inwards as indicated in the 
di agram.
University of Ghana          http://ugspace.ug.edu.gh
Figure 4.2: Construction of the NG2 trap model.
University of Ghana          http://ugspace.ug.edu.gh
80
Plate 4: The ICIPE NG2B trap.
University of Ghana          http://ugspace.ug.edu.gh
JNG2E: White cloth was used for the body instead of 
blue. Other parts were as for NG2B.
NG2F: Two Iw x 0.5w 'wings* of blue cloth were
added to the front sides of the NG2B. Fig. 2(9) 
shows the difference between the NG2B and NG2F 
when viewed from above.
Plate 4 shows the NG2B trap, for which Brightwell et al
(1987) provide a more detailed description of its 
construction. This includes a description of the ’control1 
version of the trap. i.e. as would be appropriate for a local 
control campaign which involves reducing the cost of the trap 
by replacing the metallic supports with wooden ones. The net
cage which was used for experiments was also replaced by a
large polythene bag.
The NG3 model,
Only one version, the NG3A, was tested. The following 
descriptions are based on Figure 4.3.
1- E^ody (blue): The folds in the body of this trap are 
determined more by the attachment to the netting cone (see 
below) .
2- Target (black): affixed as shown in the figure.
3- Front/shelf (black): affixed as shown in the figure.
4- Netting cone (white): A lw x lw square was cut 
diagonally Fig.4.3(4). One half was folded along the dotted 
line as shown in Fig.4.3(5) and the marked edges Fig.4.3(6)
81
University of Ghana          http://ugspace.ug.edu.gh
Figure 4.3: Construction of the NG3 trap model.
University of Ghana          http://ugspace.ug.edu.gh
83
were joined up. When folded again as was done Trf Fig.4.3(5), a
"hat” was produced with the edges of the base inclined upwards
as shown in Fig. 4.3(7). This was affixed to the top of the 
trap; 'af and ’b ’ to the two corners, X and Y to the centres
of front and back. A hole was cut in the apex as before. This
trap required an additional internal support between X and Y 
to hold the trap open. The centre pole, carrying the 
supporting cone underneath the netting (when the trap is set), 
passed through the position indicated on the body of the trap. 
A diagram of the assembled trap is shown in Fig.4.3(8)
The NG4 model
The versions tested were the NG4A and NG4B. The
descriptions given below are based on Figure 4.4 and refer to
the NG4A.
1- Body (blue): no distinct folds.
2- Target (black): affixed to body as indicated.
3- Shelf (black): affixed as shown in the diagram.
4- Netting cone (white): A 0.5w x 0.5w piece was affixed 
to the top of trap as per points "abed" shown in the netting
and body. The flat piece was pushed up into a convoluted cone
shape when the supporting cone was positioned centrally 
underneath it. An assembled NG4A trap is shown in fig. 4(5)
The NG4B was similarly produced by affixing an 
equilateral triangle (lw xlw xlw) of netting to an NG2B trap 
body as shown in Fig.4 (6).
University of Ghana          http://ugspace.ug.edu.gh
84
Figure 4.4: Construction of the NG4 trap model.
University of Ghana          http://ugspace.ug.edu.gh
85
When set, the netting cones of all traps were supported
by a small cone of galvanized wire. A ring opening in the apex
of the conical frame coincided with that in the apex of the 
netting cone of the trap to form the no-return exit hole for 
flies passing into the collecting cage on top.
The same type of collecting cage was used for all
experiments unless otherwise stated. Cages consisted of 
rectangular bags of nylon netting pulled over rectangular 
galvanized wire frames measuring 20 x 10 x 7cm. When set, 
cages were secured over the exit hole at the apex of the trap 
by a couple of rubber bands put around the sleeve end of the 
cage.
4.2.2. Odours Baits
Cow urine was collected from local East African Zebu 
cattle at Nguruman. Buffalo (Syncerus cafer Sparrman) urine 
was obtained from animals in captivity at the Veterinary 
Laboratories, Kabete, Kenya. Prior to use, the cow urine was 
usually stored at ambient temperatures for about three weeks 
in stoppered bottles. The sample of buffalo urine was kept at 
4°C for about one week before use. Acetone, methyl ethyl 
ketone (MEK), l-octen-3-ol (henceforth termed octenol) and 3“ 
buten-2-ol (butenol) were obtained from the commercial market.
Odour baits were dispensed from glass jars at different 
dose rates effected by varying the aperture of the container. 
The following containers were used for the various dose rates;
University of Ghana          http://ugspace.ug.edu.gh
8 6
Cow urine and buffalo urine: These were ^ trsrpensed from 
glass jars (600 ml) with an aperture 4.5 cm in diameter which 
according to Dransfield et al., (1986b) gave a dose rate of c.
lOOOmg/h at afternoon temperatures of 30-35°C- This was 
considered medium dose rate for cow urine in this study. High 
and low doses of the urine were obtained from glass jar 
dispensers with aperture diameters of 8.5 cm and 0.6 cm 
respectively.
Acetone, MEK and 3-buten-2-ol: High dose (c. 2500mg/h) 
and medium dose (c. 500mg/h) were obtained from GIBCO bottles 
with apertures of 2.2 cm and 0.6 cm respectively whilst low 
dose (c. 150mg/h) was dispensed from a 15-ml bottle with an 
0.2 cm aperture. Low dose MEK and 3-buten-2-ol were dispensed 
from the same dispenser type as used for low dose acetone.
l-Octen-3-ol: This was dispensed from a 10-ml bottle with 
the top replaced by a rubber septum kept in contact with the 
chemical by a folded pipe cleaner. This was estimated to give 
a dose rate of about 0.5mg/h (Dransfield et. al.,1986b).
Except where the effect of different odour positions were 
being tested, odour dispensers were positioned on the ground, 
30 cm from the entrance of the trap (from the pole of a 
biconical trap). Odours were sheltered from the sun and rain 
by metal tin cans, suitably cut out and clipped over the 
dispenser openings, leaving enough space underneath for the 
escape of odour.
University of Ghana          http://ugspace.ug.edu.gh
87
4.2.3. Experimental design
Each experiment was conducted using a Latin square 
design, where the different treatments under comparison were 
incorporated in replicated Latin squares of treatments x sites 
x days. This involved the daily rotation of treatments (traps, 
odours etc.) between sites such that by the end of the 
experiment each treatment operated in every site. By this 
design, data collected can be analyzed to separate the effects 
due to treatments from those due to sites and days.
Traps were normally set about 200 m apart and 30 m away 
from the vehicle track. For logistical reasons, most 
experiments were run jointly with the ICIPE research team 
collecting data on G. pallidipes. Therefore, most experiments 
were ru*n from 1530 h and trap catches collected at 1.5 h 
intervals until 1830 h when collecting cages were left on 
overnight for a last collection at 0730 h the following 
morning. These times include the active periods of both 
species.
Precautionary measures that were routinely taken to 
minimize experimental errors included;
i) Setting up the experiment well before starting time 
and double-checking to ensure that each treatment was at the 
right site.
ii) Shifting round treatments several hours before 
running time (especially when comparing odour baits) to be 
sure that there was no residual effects from previous 
treatments. For the same reason odour containers were usually
University of Ghana          http://ugspace.ug.edu.gh
8 8
kept tightly closed until the start of each experimental run 
and the rain/sun covers were continuously shifted round with 
their respective odour dispensers throughout the experiment.
iii) Collecting cages were usually checked very carefully 
for holes, before setting them on traps and when catches were 
being collected. Traps were also regularly inspected for 
holes, especially in the netting cone.
iv) All odour containers were thoroughly washed after 
every experiment to ensure that they were clean for subsequent 
use.
Unless specified otherwise, in the experiments in which 
different trap types were being compared, traps were usually 
baited with medium doses of cow urine (CU(m)) and acetone 
(A(m)), the experiments were run from 1530 h and trap catches 
collected at the time intervals previously specified.
4.2.4. Details of experiments 
Exp t. 1
This was to test the efficacy of cow urine and buffalo 
urine when dispensed on their own and in combination with 
acetone, using the standard blue biconical trap. (Acetone on 
its own was already known to have little effect on the catch; 
Dransfield, pers. comm.) Medium dose rates of the urines and
University of Ghana          http://ugspace.ug.edu.gh
89
acetone were used in all treatments. The treatments were as
foilows;
a) unbaited biconical (-)
b) baited with cow urine only (Cu)
c) baited with cow urine and acetone (Cu+Ac)
d) baited with buffalo urine only (Bu)
e) baited with buffalo urine and acetone (Cu+Ac).
Cf. Table 4.1 A and B for the results,
Expt. 2
From the above experiment no significant difference was 
observed between cow urine and buffalo urine (cf. Table 4.1A) 
and considering the ready availability of the former, it was 
selected for further testing. This experiment was to test the 
effect of octenol on catches, when dispensed together with cow 
urine and acetone. The effect of reduced dose rate of cow
urine was also tested. Again, blue biconical traps were used
with the following bait treatments;
a) medium doses of cow urine and acetone
b) medium doses of cow urine and acetone + octenol
(0.5mg/h)
c) low dose cow urine and medium dose acetone
The method of dispensing to obtain these dose rates are given 
in section 4.2.2.
University of Ghana          http://ugspace.ug.edu.gh
90
Expt. 3
The experiment on the effect of dose rate of cow urine 
was repeated with three dose levels; high, medium and low. 
Another known chemical attractant, methyl ethyl ketone (MEK) 
was also tested in place of acetone. The blue biconical traps 
were thus baited as follows;
a) medium doses of cow urine and acetone
b) high dose cow urine and medium dose acetone
c) low dose cow urine and medium dose acetone
d) medium dose of cow urine and low dose MEK
The experiment was run from 1730 h each day, one catch 
collected at 1830 h and a last collection at 0730 h the 
following morning. The results of this experiment are given in 
Table 4.1A and B .
Ex p t .  4
The effect, on trap performance, of adding a light source 
to a trap was investigated. The light source was provided from 
a special chemical light-stick ’Cyalume', obtained from 
Cyanamid Chemical Light Department, New Jersey, U. S. A. This 
is a cylindrical plastic stick (152 mm long, 90 mm in 
diameter) containing two different chemicals, one in an inner 
glass ampule. When the flexible plastic is bent slightly, the 
inner tube breaks releasing an energy-producing solution that 
mixes with a greenish fluorescense in the outer tube. A strong 
chemical reaction, based on peroxyalate chemi1uminescence, 
produces a greenish yellow light. Spectrometer analysis has
University of Ghana          http://ugspace.ug.edu.gh
90
Expt. 3
The experiment on the effect of dose rate of cow urine 
was repeated with three dose levels; high, medium and low. 
Another known chemical attractant, methyl ethyl ketone (MEK) 
was also tested in place of acetone. The blue biconical traps 
were thus baited as follows;
a) medium doses of cow urine and acetone
b) high dose cow urine and medium dose acetone
c) low dose cow urine and medium dose acetone
d) medium dose of cow urine and low dose MEK
The experiment was run from 1730 h each day, one catch 
collected at 1830 h and a last collection at 0730 h the 
following morning. The results of this experiment are given ir 
Table 4.1A and B .
Expt. 4
The effect on trap performance, of adding a light source 
to a trap was investigated. The light source was provided fron 
a special chemical light-stick 'Cyalume', obtained from 
Cyanamid Chemical Light Department, New Jersey, U. S. A. This 
is a cylindrical plastic stick (152 mm long, 90 mm in 
diameter) containing two different chemicals, one in an inner 
glass ampule. When the flexible plastic is bent slightly, the 
inner tube breaks releasing an energy-producing solution that 
mixes with a greenish fluorescense in the outer tube. A stronc 
chemical reaction, based on peroxyalate chemi1uminescence, 
produces a greenish yellow light. Spectrometer analysis has
University of Ghana          http://ugspace.ug.edu.gh
91
shown that this light gives a single emission peak at 510 nm. 
More details of this can be obtained from Service and Highton 
(1980).
The light stick was tied to one of the inner corners of 
the collecting cage frame just before setting the cage on the 
trap. Other treatments included for comparison were the white 
biconical trap, baited with two dose levels of cow urine 
(medium and low) each with medium dose acetone. All baited 
blue biconicals were with medium doses of cow urine and 
acetdne. Thus, the treatments effected in this experiment were 
as foilows;
a) unbaited blue biconical
b) baited blue biconical
c) baited blue biconical + light
d) baited white biconical
e) white biconical (low dose cow urine)
The experiment was run from 1730 h each day, and the traj 
catches were collected in the same time periods as in the 
previous experiment. Results are given in Table 4.1A and B.
Expt. 5
Having observed from the preceding experiments that cow 
urine and acetone produced a substantial increase in catch 
(cf. Table 4.IB) but neither additional odours nor light 
significantly increased catches, attention was then directed 
at testing the performances of different trap types. In this
experiment the blue and white biconical traps and the Zimbabw»
University of Ghana          http://ugspace.ug.edu.gh
F2 and F3 traps were tested. All traps were baited with medium 
doses of cow urine and acetone.
a) baited blue biconical
b) baited white biconical
c) baited F2
d) baited F3
The experiment was run from 1730 h each day and catches 
collected at the same time periods as specified above.
The results are included in Table 4.2A and B.
Exp t. . 6
This was, essentially, a repeat comparison of the 
performances of white and blue biconical traps owing to the 
low catches that were recorded in the previous experiment. 
Unbaited and baited white biconical traps were compared with a 
baited blue biconical trap in a 3 x 3 Latin square. All traps 
were baited with medium doses of cow urine and acetone. See 
Table 4.2E for results.
Expt. 7
This was designed, mainly, to assess the performance of 
the Zimbabwe F3 trap in its complete form and in a slightly 
modified version. In one treatment the trap was set, complete 
with the blue detachable plastic floor and in another the 
floor was omitted. The first trap of the NGU series, the NG1A 
was included as a third treatment. Thus, the three traps all
92
University of Ghana          http://ugspace.ug.edu.gh
93
baited with medium doses of cow urine and acetone were;
a) F3 (with base)
b) F3 (without base)
c) NG1A
he results are included in Tables 4.2 A and B.
Expt. 8
Owing to the poor performance of the NG1A some 
modifications were made on it to produce the NG1B (see Fig. 
4.1). This was tested together with two other traps of a new 
model (NG2A and NG2B). Unbaited and baited biconical traps 
were included for comparison. Thus, in a 5 x 5 Latin square 
design, the following treatments were effected;
a) unbaited biconical
b) baited biconical
c) NOIB
d) NG2A
e) NG2B
See Tables 4.3 A and B for results.
Expt. 9
Based on the results of the previous experiment the NG2B 
was selected for further testing. Two more new NGU trap 
models, NG2C and NG3A, were then tested together with the 
NG2B. These were again compared with the baited and unbaited
University of Ghana          http://ugspace.ug.edu.gh
94
biconical traps in a 5x 5 Latin square with the following
treatments;
a) unbaited biconical
b) baited biconical
c) NG2B
d) NG2C
e) NG3A
Exp t .  10
Based on its superior performance in the previous 
experiment, the NG2B was again selected from the previous 
experiment and tested together with another couple of new 
traps (NG2D and NG4A). Baited and unbaited biconical traps 
were as usual included as controls giving rise to five 
t reatmentr ;
a) unbaited biconical
b) baited biconical
c ) br> i t ed NG2B
d) baited NG2D
e) baited NG4A
Exp t .  11
The NG2B which again showed up best in the previous 
experiment, was tested together with another set of two NGU
University of Ghana          http://ugspace.ug.edu.gh
95
trap models (NG2E and NG4B) and compared with biconical traps 
as above. Treatments were as follows;
a) Unbaited biconical
b) baited biconical
c) baited NG2B
d) baited NG2E
e) baited NG4B
Expt. 12
The results of the above experiments indicated that the 
NG2B was more effective than the biconical trap especially for 
female G. longipennis. Therefore the former trap was used in 
this experiment for a repeat test of the separate and combined 
effects of cow urine and acetone. A baited blue biconical trap 
was included for comparison. Medium doses of cow urine and 
acetone were used in the following treatments;
a) baited biconical
b) unbaited NG2B
c) NG2B baited with acetone alone
d) NG2B baited with cow urine alone
e) NG2B baited with acetone and cow urine.
The experiment was run from 1730 h each day, with one
collection at 1830 h and another at 0730 h the following 
morning.
University of Ghana          http://ugspace.ug.edu.gh
96
Expt. 13
This was to investigate the effect of different odour 
positions on catches by the NG2B trap. Using the usual medium 
doses of cow urine and acetone, four bait positions were 
selected and tested;
a) baits placed 30 cm in front of the trap
b) baits placed 90 cm in front
c) baits placed 30 cm behind
d) baits placed 30 cm in front plus another jar of cow 
urine (aJso medium dose) positioned 1 0  m in front.
A similarly baited blue biconical trap (baits in the usual
position) was included for comparison.
Expt. 14
The effect of urine temperature on trap catches was 
investigated by baiting NG2B traps with urines at different 
temperatures. The urine temperatures were maintained at 20, 3( 
and 40°C ± 1°C, by immersing the urine jars in water at these 
temperatures in 15-litre polystyrene boxes. These three 
temperature regimes were compared with urine at ambient 
temperatures in a 4 x 4 Latin square design. In all the 
treatments the urine (medium dose) operated in combination 
with acetone (medium dose) at ambient temperatures. The 
experiment was run from 1730 h each day and catches collected 
as in the previous experiments.
University of Ghana          http://ugspace.ug.edu.gh
97
Expt.15
It was often observed that some trapped flies remained 
perched on the inner walls of the netting cones of the traps. 
This led to some modification on the cone of the NG2B trap 
aimed at getting flies to move up faster into the collecting 
cage. First, a different type of cone material was introduced. 
This was polyester/cotton netting with a coarser mesh compared 
to the original cone material which was of fine nylon mesh. 
Next, a 3-inch wide strip of black cloth was sewn along the 
full length of each of the three corners of the cones (both 
old and new types). In a third treatment, the black shelf 
(above entrance) was replaced by a white one. A fifth 
treatment was to test the effect of a lower dose rate of 
acetone on catch by the NG2B trap. The five treatments were as 
foilows;
a) NG2B (original)
b) NG2B (old cone) + black strips
c) NG2B (new cone) + black strips
d) NG2B white front
e) NG2B (original), (low dose acetone)
All traps were baited with medium doses of cow urine and 
acetone except in treatment 'e' where the medium dose acetone 
was replaced by low dose.
University of Ghana          http://ugspace.ug.edu.gh
98
E x p t .  16
A direct comparison of the F3 and NG2B traps was made.
The 1bait-behind1 position with the NG2B trap was also 
repeated in this experiment as one treatment. The fourth 
treatment was the control version of the NG2B trap. In this 
version most of the metallic supports of the trap were 
replaced by wooden ones and a large polythene bag drawn out 
into a tetrahedron was used for the collecting cage.
Construct i.onal details can be obtained from Brightwell et . al . 
(1987). All traps were baited with medium dose cow urine and 
low dose acetone; (from the previous experiment it wast
observed that low dose acetone did not reduce catches and 
could thus be conveniently used in place of medium dose). The 
experimental treatments were as follows.
a) NG2B (original)
b) F3
c) NG2B (bait behind)
d) NG2B (control version).
E x p t .  17
The new cone material (polyester/cotton, open mesh) was 
tested again without the black strips along the corners. 
Another modification was made on the NG2B to increase the area 
of attractive blue; a 1 x 0.5 metre piece of blue cloth 'wing1 
was sewn on lengthwise to each of the front sides of the trap. 
The resultant version was termed the NG2F or "winged NGUM . A 
fourth treatment, was to test the efficacy of 3 -buten-2 -ol
University of Ghana          http://ugspace.ug.edu.gh
99
(0 .5mg/h), dispensed in combination with medium dose cow urine 
and low dose acetone . The other traps were baited only with 
similar doses of cow urine and acetone. The treatments 
effected were:
a) NG2B (old cone); cow urine (med. dose) + acetone (low
dose)
b) NG2B (new cone); cow urine (med. dose) + acetone (low
dose)
c) NG2F; cow urine (med. dose) + acetone (low dose)
d) NG2B; cow urine and acetone as above + 3-buten-2-ol 
(0 .5mg/h).
4.2.5. Analysis of data
Separate analyses were carried out on numbers of males 
and females so as to be able to detect any differences between 
the two sexes in their responses to the various treatments.
For each experiment, data from the two replicates of the Latin 
square design, were subjected to an analysis of variance. 
Before analysis the data were subjected to a log 
transformation (N + 1 ,to take care of zeros) to normalize the 
data. By this method, treatment effects can be separated from 
those due to days and sites. Where the treatment effect was 
significant, the differences between treatment means were 
further compared by the Duncan's multiple range test.
The relative efficacies of the different odour baits 
tested were estimated by computing an index of catch increase 
over an unbaited biconical trap; i.e. the ratio of the
University of Ghana          http://ugspace.ug.edu.gh
100
detransformed mean catch with the odour bait to that of an 
unbaited biconical trap. To assess the performances of the 
different trap types (including new trap designs), an index of 
increase over a similarly baited biconical trap was computed. 
The effects of further modifications on the NG2B were also 
assessed by comparing mean catches by the modified versions to 
those of a similarly baited original NG2B. The ovarian aging 
method (Saunders, 1960; Challier, 1965) was used to age female 
G. longipennis from all experiments. In brief, this technique 
categorizes female flies into eight age groups (Oa&b-7) 
according to the stage of the follicular development. 
Categories Oa & Ob flies are very young flies which have never 
larviposited and are termed nulliparous whilst flies in groups 
1-7 have larviposited once or more times and referred to as 
parous (See Chapter 7 for more details). Differences between 
the age compositions from different trap types and odour baits 
were tested using the chi-square analysis.
4.2.6. Test of trap efficiency
The efficiency of the best trap model (the NG2B) was 
estimated using the technique developed by Vale and Hargrove 
(1979) and later improved by Hargrove (1980b). This involved 
making an incomplete ring of electric screens round the baited 
trap. The number of flies caught by the trap and those caught 
on the outside and inside surfaces of the screen were then 
used to estimate the efficiency of the trap.
University of Ghana          http://ugspace.ug.edu.gh
101
The electric screens used for this study were of the type 
designed by Vale (1974b). The same screens, measuring 90cm x 
90cm were used for the activity pattern studies described in 
detail in chapter 5. Two NG2B traps, similarly baited with cow 
urine (c. lOOOmg/h) and acetone (c. 500 mg/h), were set about
200 m. apart in open glades in the woodland. One of the traps 
was surrounded by three screens each placed at about 3 m from 
the base of the trap, with one screen facing the entrance of 
the trap whilst the other two faced the two lateral sides of 
the trap> thus forming an incomplete ring. A shallow tray 
filled with water was placed at the base of each side of the 
screens to retrieve electrocuted flies* The other trap was 
similarly set up but without screens. At about 1700 h the 
screens were turned on, the odour baits opened and a 
collecting cage put on each of the trap. Collections were made 
at hourly intervals from the screens and the trap until 2 0 0 0  h 
when evening activity ceased. Catches were also made during 
the morning activity period. The test was repeated four times 
alternating the screens between traps on alternate days. Thus, 
for each test, the unscreened trap acted as a control. Flies 
caught from the inside and outside surfaces of the screens and 
in the retaining cage of the trap were kept separate. The data 
were, however, pooled over the four days before carrying out 
the analysis because of the small numbers caught on individual 
days but the data on males and females were kept separate.
University of Ghana          http://ugspace.ug.edu.gh
102
4.3. RESULTS
4.3.1. Efficacy of odours
Results of the initial tests of odour attractants are 
given in Table 4.1A, showing detransformed mean catches of 
male and female G. 1ongipennis with the indices of increase of 
odour baits in relation to an unbaited trap in Table 4.IB. 
Neither cow urine nor buffalo urine on their own increased 
catches over an unbaited trap. However, they both showed 
significant increases when dispensed together with acetone.
The trends were the same for both sexes. Taking the average 
for both sexes, there was a 4.9X increase with buffalo urine 
and acetone and 3.6 X increase for cow urine and acetone. There 
was, however, no significant difference between the cow and 
buffalo urine.
There were differences between the different dose rates 
of cow urine with the highest dose rate giving the highest 
index of increase, similar in fact to medium dose buffalo 
urine. There was a significant reduction in the catches of 
both sexes with low dose cow urine.
Catches were also reduced but not significantly when MEK 
instead of acetone was dispensed with medium dose cow urine. 
When l-oct.en-3-ol was tested the treatment effect was not 
significant, but the raw data showed that the octenol 
treatment yielded the highest mean catches on both sexes 
(expt. 2). It does appear that octenol has the potential to 
increase catches but the catches in this experiment were
University of Ghana          http://ugspace.ug.edu.gh
103
Table 4.1A: Total trap catches, detransformed means, and ANOVA F- 
ratios for G longipennis using various odour baits.
G. l o n g i p e n n i s
Expt. Odour
M
Tot
a l e s  
Detr. 
mean
F-ratio
&df
F
Tot
e m a l
Detr.
mean
e s
F-ratio
&df
1 . 60 3.9b 28 2 .0b
Cu(m) 56 2.9b 37 2.9b
Cu(m),A(m) 2 1 2 14.0a 11.90*** 88 7.2a 12.90***
Bu(m) 51 3.3b (4,24) 27 1.7b (4,24)
Bu(m),A(m) 240 17.8b 137 1 0 .6a
2. Cu(m),A(m) 23 2. 8 3 0.4
As above+octen 28 4.4 2.15ns 5 0.7 0.27ns
Cu(1),A(m) 12 1.3 (2,4) 3 0.4 (2,4)
3. Cu(m),A(m) 1 2 2 1 2 .6ab 52 4. 7ab
Cu(h),A(m) 218 16.6a 4.10* 62 6 .2 a 4.90*
Cu(1),A(m) 56 5.2b (3,12) 28 3.0b (3,12)
Cu(m),MEK(1) 123 9. 3ab 41 2.9b
Means followed by the same letter are not significantly different
from each other (P<0.05) ; F-ratios: *=P<0..05; **=P<0.01; ***=P<0.001
University of Ghana          http://ugspace.ug.edu.gh
104
4.IB.:Indices of increase for G. longipennis for various odour 
baits relative to an unbaited trap.
Males Females
Unbaited 1.0 1.0
Cu(m) 0.7 1.5
Bu(m) 0.8 0.9
Cu(1),A(m) 1.6 2.9
Cu(m),A(m) 3.6 3.6
Cu(h),A(m) 4.7 4.7
Bu(m),A(m) 4.6 5.3
Cu(m),MEK(l) 2.5 2.2
Cu(m),A(m)+0cten 5.8 6.5
Expt.8-11: Cu(m),A(m)
i) 15.0 6.3
ii) 4.1 4.5
iii) 3.0 6.3
iv) 6.3   4.7 ______
Weighted mean 7.5 5.6
University of Ghana          http://ugspace.ug.edu.gh
105
general]y very low. The experiment needs repeating, using the 
more efficient NG2B trap, for a confirmation of these results.
Considering various other experiments in which medium 
doses of cow urine and acetone were compared with the unbaited 
biconical trap (expts. 8 -1 1 ), the index of increase of the cow 
urine/acetone ranged from 3.0-15.0 for males and 4.5-6.3 for 
females. The weighted mean indices taken from these 
experiments were 7.5X for males and 5.6 X for females. The 
higher value for males results from the weighting factor for 
experiment 8 when the largest numbers of G. longipennis were 
caught; examination of all the data reveals little evidence 
for any difference between males and females in the response 
to cow urine and acetone.
4.3.2. Trap performance
Having identified medium doses of cow urine and acetone 
as effective attractants for G. longipennis, various 
established trap types were tested using these odour baits.
The results for these experiments are given in Table 4.2A with 
the indices of increase in Table 4.2B. The addition of the 
chemical light did not improve catches by the blue biconical 
trap. In fact the trap without light yielded higher mean 
catches on both sexes than the lighted trap.
There was much variability in the performance of the 
white biconical trap. In experiment 4, the white biconical 
trap caught significantly more males than the blue one. The 
treatment effect on the female catch was not significant but
University of Ghana          http://ugspace.ug.edu.gh
106
Table 4.2A.: Total trap catches, detransformed means, and ANOVA F 
-ratios for G. longipennis using various trap designs.
Expt Odour Tot
M a l e
Detr.
mean
4. Blue bicon 36 2 .1b
Blue bicon 
(+light)
30 1 .8b
White bicon 55 4.0a
White bicon 
Cu(1),A(m)
31 1.9b
Blue bicon 
(unbaited)
12 0.9c
5. Blue bicon 18 1 . 6
White bicon 25 1.9
F3 18 1.7
F2 18 1 . 2
6 . Blue bicon 89 11.3a
White bicon 39 4.5b
White bicon 
(unbaited)
12 1.5c
7. F3 (+Base) 48 3.5al
F3 (-Base) 101 6 .8a
NG1A 20 1 .2b
g 1 p e n n i
F e m a l e s  
F-ratio Tot Detr. F-ratio
&df mean Sdf
8 0.6
8 0.4
3.20* 22 1.2 2.45ns
(4,24) 6 0.4 (4,24)
3 0.1
5 0.4
0.25ns 17 0.7 2.55ns
(3,12) 17 1.4 (3,12)
26 1 . 6
23 3.8
18.23* 11 1.3 2.13ns
(2.4) 9 1.3 (2,4)
47 3.6ab
6.81*** 95 8.5a 12.21***
(2.4) 18 1.3b (2,4)
University of Ghana          http://ugspace.ug.edu.gh
107
Table 4.2B. Indices of increase for G. longipennis for various trap 
types relative to a biconical trap.
Males Females
Blue bicon 1.0 1.0
Blue bicon 0.9 0.7
(+1ight)
White bicon i) 1.9 2.0
ii) 1 . 2 1 . 8
iii) 0.4 _________________ 0.3
Weighted mean 1.3 1.4
F3 (+Base) 1.1 3.5
F3 (-Base) 2.1 8.3
F2 0.8 4.0
University of Ghana          http://ugspace.ug.edu.gh
108
the white biconical trap still recorded the highest catches.
In experiment 5, the white biconical trap again recorded 
slightly higher catches than the blue but the treatment 
effects were generally not significant. The white and blue 
biconical were again compared in experiment 6 , where the 
treatment effect was just significant at the 5% level only on 
the male catches but not on the females. Here, the blue 
biconical came out significantly better on the males and also 
recorded the highest female mean catch. A re-analysis after 
combining both sexes showed no significant treatment effects 
for both sexes. The weighted mean indices of increase of the 
white biconical trap over the blue (taken for experiments 4,
5, and 6 ) were 1.3 for males and 1.4 for females. Thus, the 
two colours were basically similar in performance for 
G. Iongipennis.
In the experiment that the F2 and F3 traps were tested 
(expt.5) together with the blue and white biconicals traps, 
the treatment effect was not significant. Catches were 
generally very low in the area where the experiment was 
carried out. Both the F2 and F3 traps caught similar numbers 
of males as the blue biconical trap but 3.5-4X more females. 
The F3 caught slightly more males than the F2 whilst it was 
the reverse for females. The F3 trap performed better for both 
male and female G .1ongipennis when set without the blue 
plastic base than with the base although there were no 
significant differences between their mean catches. On
University of Ghana          http://ugspace.ug.edu.gh
10 9
average, the trap was twice as effective without the base 
(expt. 7 ) .
The results on the tests of new trap designs are given in 
Table 4.3A. with the indices of increase in Table 4.3B. The
first new trap to be tested was the NG1A in experiment 7.
Compared to the biconical trap, this trap was less effective 
for males but slightly more effective for females. There was a 
considerable improvement in performance when the base of the 
NG.1A trap was removed giving rise to the NG1B (expt. 8 ). 
Whereas the NG1A was about as effective as a biconical for 
females, the NG1B was nearly 5x as effective. However in the 
same experiment the two versions of the NG2 model were even 
better than the NG1B so efforts were concentrated on 
these.(expts 8 -1 1 ).
There was no significant difference in catch between the
NG2A and the NG2B so since the NG2B used rather less material
than the NG2A, it was adopted for further experiments. 
Subsequent modifications (NG2C - E) did not increase catch 
significantly. Averaging over the six experiments when the 
NG2B was compared with the biconical, the NG2B trap was 
consistently more effective (4.OX) for females. On males, 
however, the NG2B yielded a significantly higher mean catch 
than the biconical on only one occasion (expt. 8 ), giving an 
average of 1.2X the catch in a biconical trap. The average 
percentage female catch by the NG2B trap was 46.5 ± 8 .8 % as 
compared to 24.8 ± 8.9% for the biconical trap. The NG3A and
University of Ghana          http://ugspace.ug.edu.gh
110
Table 4.3A: Total trap catches, detransformed means,and ANOVA F-ratios 
for G longipennis using the new NGU traps.
G. l o n g i p e n n i s
M a l e s F e m a 1 e s
Trap
type
Tot Detr. 
mean
F-ratio
&df
Tot Detr.
mean
F-ratio
Sdf
Bicon
Unbaited
4 0.3c 6 0.4c
Bicon 80 4.5b 18.90*** 39 2.5b 29.85***
NG1B 98 5. 9ab (4,24) 188 ’1 2 .1 a (4,24)
NG2A 108 7. 4ab 206 13.6a
NG2B 171 13.1a 288 19.2a
Bicon
(Unbaited)
35 2 .2b 16 1 . 1
Bicon 147 9.0a 5.52** 94 4.9a 13.51***
NG2B 98 8.7a (4,24) 137 10.7a (4,24)
NG2C 1 1 0 8.9a 1 2 1 9.6a
NG3A 43 3.2b 37 2 .0 b
Bicon
(Unbaited)
45 3.6b 5 0.3b
Bicon 147 1 0 .8a 8.27*** 33 1.9a 14.28***
NG2B 67 6 .2a (4,24) 43 3. 4a (4,24)
NG2D 69 6 .0 a 48 4.2a
NG4A 28 2 .1b 6 0.4b
University of Ghana          http://ugspace.ug.edu.gh
Ill
Table 4.3A (contd)
G. l o n g i p e n n i s
M a l e s  F e m a l e s
Expt. Trap Tot Detr. F-ratio Tot Detr. F-ratio
type mean &df mean &df
11. Bicon 11 0.9c 4 0.3c
(Unbaited)
Bicon 76 5.7ab 19.33*** 21 1.4b 12.17***
NG2B 102 7.7a (4,24) 53 4.1a (4,24)
NG2E 60 3.2b 63 4.7a
NG4B 50 4.5ab 21 1.7b
Means followed by the same letter are not significantly different from 
each other (P<0.05); F-ratios: *=P<0.05; **=P<0.01; ***=P<0.001
University of Ghana          http://ugspace.ug.edu.gh
112
Table 4.3BIndices of increase for G. longipennis for traps of the 
NGU series.relative to the blue biconical trap.
Mai es Females
Baited Bicon 1 . 0 1 . 0
NGlA(expt.7) 0.4 1.4
NG1B 1.3 4.8
NG2A 1 . 6 5.4
NG2B i) 2.9 7.7
ii) 1 . 0 2 . 2
iii) 0 . 6 1 . 8
iv) 1.4 2.9
12 v) 1.5 5.1
13 vi) 0.5 1 . 6
Weighted mean 1 . 2 4.0
NG2C 1 . 0 2 . 0
NG2D 0 . 6 2 . 2
NG2E 0 . 6 3.4
NG3A 0.4 0.4
NG4A 0. 2 0 . 2
NG4B 0 . 8 1 . 2
University of Ghana          http://ugspace.ug.edu.gh
113
NG4A models weie significantly less effective than the 
biconical trap whilst the NG4B gave similar catches.
4 .3.3.Efficacy of odours with the NG2B trap
Table 4. 4A shows the results of further tests, using the 
NG2B trap, on the influence of certain factors on odour bait 
efficacy, with the indices of increase in Table 4.4B. The 
apparent synergistic effect of cow urine and acetone for G. 
longipennis was confirmed in experiment. 12. Neither cow urine 
nor acetone on their own increased catch significantly, but 
together ’increases were 5.2X for males and 4.2X for females.
In the test on odour positions the treatment effect was 
not significant meaning, there were no significant differences 
between the mean catches from the different odour positions 
tested {expt. 13). However, positioning the bait 30 cm behind 
the trap did appear to increase catches markedly. When the 
test was repeated in experiment 16, this position again 
yielded higher (but again not significant) mean catches of 
both sexes.
The effect of temperature on the efficacy of cow urine 
was also not significant. However, catches at 20oC were 
consistently lower than those at 30oC whilst catches at 
ambient temperature (3 2oC) and 40oC were consistently higher. 
Thi s indicates a r educti on in catch with decreasing 
temperature on both sexes and suggesting that if the urine 
temperature goes below 30°C, catches could be reduced.
University of Ghana          http://ugspace.ug.edu.gh
114
Table 4.4A.: Total trap catches, detransformed means, ANOVA F-ratios 
for G. longipennis using different odour baits.
G. 1 o n g i p e n n i s
Expt. Odour/ 
Treament
Tot
M a l e
Detr.
mean
s
F-ratio
&df
Tot
F e m a 1 e s 
Detr. F-ratio 
mean &df
12. Unbaited 54 4.5b 56 5.4b
Cu(m) 80 6.7b 10.08*** 1 1 2 8.4b 6 .8 8 ***
A(m) 77 6 .8b (4,24) 96 7.7b (4,24)
Cu(m), A(m) 263 23.4a 289 22.9a
Bicon •
(Cu(m),fl(m))
244 15.3a 66 4.5b
13. 30cm in front 172 12.7 139 1 0 .3ab
90cm in front 208 16.4 2 .2 2ns 226 16.9a 4.17*
30cm behind 179 13.4 (4,24) 272 17.7a (4,24)
30cm in front 
(+1 0m in front)
222 15.1 200 15.3a
Bicon 483 23.6 97 6.5b
14. Ambient 72 6.5 49 4.7
Urine 20°C 33 3.0 2. 0 0ns 27 1 . 8 1 .8 6ns
Urine 30°C 47 5.6 (3,12) 33 2.6 (3,12)
Urine 40°C 62 6.4 46 4.3
Means followed by the same letter are not significantly different 
from each other (P<0.05); F-ratios: *=P<0.05, **=P<0.01, ***=P<0.001
University of Ghana          http://ugspace.ug.edu.gh
115
Table 4.4B Indices of increase for G. longipennis for various 
experimental treatments relative to an unbaited 
NG2B trap.
Mai es Females
Unbaited 1 . 0 1 . 0
Cu(m) 1.5 1 . 6
A(m) 1.5 1.4
NG2B Cu(m),A(m) 5.2 4.2
90cm in front 6 . 8 6.7
30cm behind 5.7 7 .1
30cm + 10m in front 6 . 2 6.3
Urine 20°C 2 . 6 1.7
Urine 30°C 4.7 2 . 5
Urine 40°C 5.2 3.8
University of Ghana          http://ugspace.ug.edu.gh
116
Tablf" 4 .5A summarizes the results on the effects on trap 
performance of lowering the dose rate of acetone, of using 3 
but en- 2 -o 1 (an octenol analogue) in conjunction with the cow 
urine/acetone and of further modifications on the NG2B.
Indices of increase are given in Table 4.5B. There was little 
evidence of any difference between the two dose levels of 
acetone, so subsequent experiments were carried out using the 
low dose rate of acetone. Putting the baits behind the trap 
again increased the catch but not significantly. The use of 
butenol in addition to the cow urine and acetone resulted in a 
significant reduction in female catch.. On the males the 
treatment effect was not significant but the mean catch with 
cow urine/acetone alone was still higher than that with 
butenol .
Further modifications on the cone of the NG2B trap 
(expts. 15 & 17) did not improve the performance of the trap. 
Replacing the black front with a white one gave a slight 
increase in catch on the females but not on males. The 
'control version’ of the NG2B yielded a substantial increase 
in the male catch as well as a slight increase in the female 
catch. The NG2F model also recorded a higher female catch 
(1.4X more) than the original NG2B trap but there was no 
significant difference between their mean catches. On the 
males, the treatment effect was not significant but again the 
NG2F recorded the highest catch.
Compared to the F3 trap, the NG2B trap caught 
significantly less males. The treatment effect on females was
University of Ghana          http://ugspace.ug.edu.gh
117
Table 4.5A: Total trap catches, detransformed means and ANOVA F-ratios 
for G. longipennis using various odour baits and
modifications
Expt. Odour/ Tot
Treament
15. NG2B 34
(Cu(m),A(1))
NG2B 29
(Cu(m),A(m))
NG2B+strips 39
(Cu(m),A(m))
New cone NG2B 28
+strips(Cu(m), A(m)
NG2B(white front) 34
(Cu(m),A(m))
of the NG2B.
G. 1 o n g i p
M a l e s  
Detr. F-ratio 
mean &df
2.9
2 . 6
2.7 0.20ns
2.3 (4,24)
2.5
n n i s
F e m a 1 e s 
Tot Detr. F-ratio
mean &df
23 2.0
36 2.6
26 2.3 1.54ns
26 2.2 (4,24)
37 3.4
16. NG2B(Cu(m),A(1) 24 1.8b 23 2.3
NG2B(Cu(m),A(l) 37 3.3ab 29 2.7
(Baits behind)
NG2B(Cu(m),A(l) 71 6.5a 4.5* 37 3.2 0.87ns
(control version)
F3 60 5.2a (3,12) 67 5.2 (3,12)
(Cu(m),A(1))
17. NG2B(Cu(m),A(1) 38 4.3 37 3.7a
As above +butenol 26 2.7 2.22ns 31 2.0b 7.42*
Newcone NG2B 25 2.7 (3,12) 37 3.1a (3,12)
(Cu(m),A(l)
NG2F(Cu(m),A(1) 45 5.1 5.6 5.3a
Means followed by the same letter are not significantly different 
from each other (P<0.05); F-ratios: *=P<0.05, **=P<0.01, ***=P<0.001.
University of Ghana          http://ugspace.ug.edu.gh
118
Table 4.5B.: Indices of increase for G. longipennis for 
various odour baits and trap modifications 
relative to a NG2B trap baited with low dose 
acetone and cow urine.
males females
Qdp  UESJ.
CUAc(l) 1.0 1.0
CUAc(m) 0.9 1.3
CUAc behind trap 1.8 1.1
butenol 0.6 0.5
.T i  g a t . im sj__
NG2B standard 1.0 1.0
+ black strips 1.1 0.9
+ poly/cotton cone# 0.7 0.9
+ white front 1.0 1.3
control version 3.6 1.4
NG2F 1.2 1.4
F3 2.9 2.3
N.B. Index for poly/cotton cone alone was obtained by taking 
mean of values from expts. 15 and 17.
University of Ghana          http://ugspace.ug.edu.gh
Jnot significant although the F3 trap again caught about twice 
as many as the NG2B trap. The data suggest however that a 
'control' version of the NG2F would do as well as if not 
better than the F3 trap for G. longipennis.
Figure 4.5 gives the age distribution of females from 
biconical traps baited with different chemicals. The chi- 
square test showed no significant association between age 
composition and bait treatment (X2 = 4 . 5n 8 , df = 3, P>0.05). 
Generally, there is a relatively high preponderance of Oas and 
Obs which is characteristic of the biconical trap as shown 
below.
The age distributions of females from three different 
NGU trap models and a similarly baited (cow urine and acetone) 
biconical trap are given in Figure 4.6. A chi-square analysis 
on the age compositions grouped into nulliparious (categories 
Oa and Ob) and parous (categories 1 -7) flies (Challier, 1965) 
showed a significant association between trap type and age 
composition (X? = 31***, df=3, P<0.001). Generally, the NGU 
traps caught a higher proportion of older flies than the 
biconica1 trap.
119
University of Ghana          http://ugspace.ug.edu.gh
Figure 4.5: Age distribution of female G. longipennis caught 
in the blue biconical trap baited with different 
chemicals (CU=cow urine, BU=buffalo urine and 
Ac=acetone).
1 2 0
1 2 3 4*  5*  6+  7+
Age  Category
University of Ghana          http://ugspace.ug.edu.gh
Figure 4
121
3c
(D
y
&
 ^ _
6 -  0 6  
I °A!
B ic o n ic d i  t r a p  
|| N  = 3 7  '
iifcii
N G U B  t r a p  
A -  y /y  N  = 1 2 3
-  •  i f
M U
2 4 -
1 6 -
' '  C B |  
a  O A
,* 6  N G U 2 A  t r a p  
p  N = 1 7 9
illilM_.................. ....................—
0 /  N G U 2 B  t r a p  
2 A '  N = 2 3 4
' 6 -
8 "  o e j j l l i l
o  oMmmmm__________
Age Category
6 : Age distribution of female G. longipennis caught 
in the different trap types all baited with 
similar doses of cow urine and acetone.
University of Ghana          http://ugspace.ug.edu.gh
J122
4.3.4. Efficiency of the NG2B trap.
Hargrove (1980b) provided the following formula for 
estimating the efficiency of the test trap, using the numbers 
of flies caught on the insides of the screens and those caught 
in the collecting cage of the trap:
where
y - the number of flies caught in the cage of the test trap,
x = the total number caught by the insides of the screens and
p = the pioportion of the perimeter of the ring covered by the
electric screens.
The validity of these analyses are based on the general 
assumption that flies approach and depart from the trap only 
once and in random directions.
Given that each screen measured 90 cm x 90 cm and screens 
were placed at a radius of about 3 metres (300 cm), the 
proportion of perimeter covered by the screens (p) was 
estimated as 0.14.
The same figure was used in a second approach to estimate 
the efficiency by using the total numbers caught in the 
outside of the screens as follows:
where
z = the number of flies caught on the outside of the screens. 
In both formulae the denominator gives the estimate of the 
total number of flies getting pass the screens and actually 
arriving at the trap. The efficiency is therefore the
Efficiency = y x 100 / y + (x/p) (1)
Efficiency - y x 100 / ((z/p) z) (2)
University of Ghana          http://ugspace.ug.edu.gh
123
Table 4.6.: Numbers of G. longipennis caught on the screens 
and in the cage of the NG2B trap and the estimated 
trap efficiency (E) for males and females.
Males
Day In Out
1 11 8
2 4 10
3 11 28
4 6 9
Tot. 42 55
(%fem) 56.8 6 6 . 2
E(%) 8. 8 6.5
SCREENS
Females
Cage In Out
4 13 9
8 6 9
7 4 6
3 9 4
22 32 28
46.7 43.2 33
5.00 7.0
NO SCREENS 
Males Females
Cage
4 11 8
4 2 1
2 3 4
2 11 9
12 27 22
35.3 55.1 44.9
6.9 1 1 . 0
University of Ghana          http://ugspace.ug.edu.gh
124
percentage of these flies that are caught by the trap.
Assuming that the same number of flies visiting the trap 
with screens are also visiting the one without screen, the 
trap efficiency can be estimated a third way as follows:
Efficiency = y x 100 / (z/p) (3)
where
y = the number of flies caught in the cage of the trap without 
the screens and z and p have the same notations as equation 
(2 ) above.
The numbers of G. longipennis caught in the system of 
electric screens and traps over the four days and the trap 
efficiency estimated by the above analyses are given in Table
4.6. The estimates by the three approaches all lay within the
same range. A chi-squared analysis showed that there were no
significant differences between the estimates by the different 
approaches nor between the estimates for males and females.
Taking the averages of the efficiency estimates by the three
approaches, the trap efficiency for the males was 7.3 percent 
and that for the females was 7.9 percent.
4. DISCUSSION
The finding that there was no significant difference
between buffalo urine and cow urine as odour baits agrees with
the results obtained for G. pallidipes by Dransfield et al., 
(1986b) at Nguruman and by Vale et al., (1986b) in Zimbabwe.
Owaga (1985), however, reported that buffalo urine was several
times more effective than cow urine. Recent biochemical
University of Ghana          http://ugspace.ug.edu.gh
analysis (Hassanali et al . , 1986; Bursell et. al, 1988) has 
shown that the two urines contain identical attractive 
phenolic compounds. Therefore, differences between urine 
samples are probably due to differences in the concentrations 
of these compounds. Cattle drink more water than buffalos, so 
it is likely that their urine is more dilute. It was observed 
by Owaga (1985) and Vale et al., (1986b) that the efficacy of
urine increases when allowed to age over a period of time. In 
this study the cow urine had been 'aged* three weeks at 
ambient temperatures whilst the buffalo urine was kept at 
about 4°C prior to use. Thus in this experiment it is likely 
that the phenolic concentrations were similar in the two types 
of urine.
Vale (1986b) remarked that undesirable variables are 
introduced when host residues are tested in their raw states 
as odour baits, due to variations in the batches of residues. 
He recommended that, ideally, known doses of chemically 
identified attractants should be used. However, phenolic 
components in bovid urine are expensive and at the start of 
this work had not been identified. Moreover, Dransfield et 
al . , (1986b) have shown that despite such variability, catches 
with urine baited traps are still closely correlated with 
those from unbaited traps. Hence, cow urine was selected for 
further experimentation on trap types and for regular 
sampling.
The apparently synergistic effects of cow urine and 
acetone on catch size of G. longipennis were quite different
125
University of Ghana          http://ugspace.ug.edu.gh
J126
from their effects on G. pallidipes in the same habitat. When 
baited with either cow urine or acetone alone, there were no 
significant increases in catches of G. longipennis by the NG2B 
trap; the indices of increases were 1.5x on males and 1.4X on 
females for acetone whilst those for cow urine were 1.5X and
1. 6 X respectively. When dispensed together, there were 
significant increases of 5.2X and 4.2X on males and females 
respectively. From data collected on G. pallidipes at the same 
time, Dransfield (unpublished data) observed that both acetone 
and cow urine had significant independent effects on catches 
viz. 3.IX and 4.IX with acetone and 2.8 X and 3.4X with cow 
urine for males and females respectively. When dispensed 
together the increases were 6 .5X for males and 9.OX for 
females. Thus for G. pallidipes, increases when the two were 
used together were equal to or less than would be expected, in 
contrast to G. 1ongipennis where they are more than would be 
expected. Increases with acetone and cow urine are still, 
however, lower for G. longipennis than they are for G. 
pal 1idipes.
Like the fusca group, the pal pal is group has also proved 
to be less responsive to odour baits than the morsitans group. 
Carbon dioxide was formerly the only attractant known to be 
effective for G. p. palpalis (W.H.O, 1986). More recently, 
however, Cheke and Garms (1988) have showed in Liberia that 
acetone, dispensed at lOOmg/h, doubled the catch of G. p. 
palpalis by the biconical trap. In the same experiment, they 
observed that a mixture of phenolic compounds found in cow
University of Ghana          http://ugspace.ug.edu.gh
Jurine ('TF 86/05’) was not effective, either on its own or in 
combination wilh acetone and octenol. Relative to the 
concentration of 4-methy1pheno1, *TF 86/05* contains 4-
methylphenol (100), phenol (1.4), 3 -methyl phenol (9.9), 3- 
ethy1pheno1 (1.1), 4-ethyl phenol (2.1), 3 -propyl phenol (2.5),
4-propyl phenol (0.8) and 2-methoxypheno1(0.4). The mixture has 
been shown to increase the catches of G. pallidipes in traps 
baited with acetone and l-octen-3-ol in Zimbabwe (Vale et al . , 
1988 )
Thus, the response of various species to cow urine and 
acetone appears to be quite varied. These differences could 
mean that the odours either elicit different behavioural 
responses in the different species or they are simply less 
effective for some species than for others. The differences 
between sympatric species in their response to odour baits has 
also been observed by Vale et al., (1986) in Zimbabwe. They
found that cow urine was only effective for G.pallidipes and 
not for G. m. morsitans.
In the present study, considering the flight periods of 
G. longipennis and judging from the trend of catches in the 
experiment with urines at different temperatures (expt.14), 
temperature could be having an effect on the efficacy of the 
cow urine for this species since there was a general trend of 
increasing catches with increasing temperature. Since G. 
longipennis flies at dawn and dusk, the relatively lower 
temperatures at these periods could render the urine less 
effective.
127
University of Ghana          http://ugspace.ug.edu.gh
Octenol (dispensed alone or in combination with other 
odours) has been shown to significantly increase the catches 
of G. pallidipes in Kenya (Dransfield et al., 1986b) and in 
Zimbabwe (Vale and Hall, 1985; Vale et al., 1986b). In this 
study, the octenol did cause an increase in the catches of G.
1ongipennis although the increases were not significant. More 
trials should be carried out using this attractant.
On the variability of the index of increase of baited 
traps over unbaited ones, Dransfield et al. (1986b) observed
that the index of increase of a baited (cow urine/acetone) 
biconical trap over an unbaited one increased with increasing 
temperature during the day for G. pallidipes. They suggested a 
temperature-mediated entry response, such that below 30°C 
fewer flies actually enter the biconical trap although very 
many may be attracted to it by the odour bait. More recently, 
Rogers et a1. (in press) showed from fat~haematin analysis of 
G. pallidipes from the study area that baited traps tend to 
catch a higher proportion of less hungry flies than unbaited 
traps. Since the proportion of less hungry flies is likely to 
be higher at higher temperatures (probably because hungry 
flies would not risk their low fat reserves at high 
temperatures), the index of increase of the baited trap will 
tend to increase with increasing temperature.
This is the first time that odour attractants have been 
shown to be effective for a species of the fusca group, but 
compared to G. pal 1idipes the odours so far tested appear to 
be less effective for G. longipennis. Given the crepuscular
128
University of Ghana          http://ugspace.ug.edu.gh
129
activity of the latter, one would expect olfaction to play a 
major role in host seeking. If the right host has to be 
identified within the very short activity period then the 
olfactory cues must be very specific. Therefore it is likely 
that more odours have yet to be identified. The results from 
this study indicate that the effectiveness of odours may rest 
on synergism and more trials will be required to identify 
effective combinations of attractants.
The F3 trap has been shown to be more effective than the 
biconical for G. pallidipes (Flint 1985, Brightwell et a l .,
1987). This is the first time the performance of the F3 trap 
for G. longipennis has been reported. It is clearly more 
effective than the biconical trap for this species, even more 
so when the blue floor is removed. This is probably because 
flies that entered a trap with the floor were settling on it 
and later on escaping instead of going up. It was quite common 
to observe G .1ongipennis settling on rocks or other objects on 
the surface of the ground. A similar explanation can be 
offered for the considerable improvement on the NG1A trap when 
its floor was removed as a modification to produce the NG1B.
Although the treatment effect was not significant when 
the F2 and F3 traps were compared (probably due to low 
numbers), the F2 (white) seemed to do slightly better on 
females than the F3 trap. During the main evening activity 
females tend to fly later than males(cf. Chapter5). as such 
the light intensities during the peak female activity would be 
lower than that during the male's. Thus the better perfomance
University of Ghana          http://ugspace.ug.edu.gh
of the F2(whi t e ) over the F3 (blue) for females could be due 
to the fact that the former was more visible at the time of 
peak female flight. The experiment however needs repeating for 
confi rmation.
The F3 trap and the control version of the NG2B trap 
were comparable in performance. The polythene bag cage is 
probably the main reason why the latter did so well since 
there should be better light transmission above the exit hole 
leading into the cage leading to a better entry response. The 
F3 trap uses twice as much blue material as the NG2B trap and 
is more difficult to construct and less convenient to use. The 
NG2B trap is therefore a more cost-effective trap than the F3 
trap. Brightwell et al. (1987) discussed the economics of the
NG2B trap in terms of catch per unit area of trap material.
Although some models of the NGU series of traps were 
found to be similar in performance to the NG2B, the latter 
model was selected for various reasons. The NG1B, NG2A and 
NG2.C, were eliminated because they required more material for 
construction. The NG2E trap (white cloth in place of the blue) 
was not selected because white cloth easily looses its 
brightness, becoming yellowish and soiled with the exposure to 
the sun and handling. This problem was experienced by Flint 
(1985) who had to re-paint white traps monthly with polyvinyl 
alcohol emulsion paint in order to revive trap efficiency.
Such an expensive maintenance would only be worthwhile if the 
white trap was found to be considerably more effective than 
the blue one.
130
University of Ghana          http://ugspace.ug.edu.gh
The index of increase of the baited NG2B trap over a 
similarly baited biconical trap was quite variable between 
experiments, especially on females (range 1.6 X -7.7X). 
Brightwell et al . , (1987) compared the performance of the
baited NG2B trap to that of a similarly baited biconical trap 
for G. pallidipes. They showed that below 30°C the NG2B trap 
was up to eleven times more effective than the biconical trap 
but only twice as effective above this temperature. In the 
case of G. 1ongipennis, no records were kept of the 
temperatures when the experiments were being run, but 
differences in mean temperature between different experiments 
could also be influencing the relative efficiencies of the two
trap types. On the other hand since light intensity has been
shown to be a critical factor influencing the level of
activity of this species (see chapter 3), the entry response
to the different trap designs may be influenced by the light 
intensity, which can vary considerably between seasons. Both 
these speculations need confirmation through further 
experimentation with accurate records of these physical 
factors.
Compared to the biconical traps, the NG2B and F3 traps 
were several times more effective on females. The average 
percentages of female catches in the biconical and NG2B traps 
were 24.8% and 46.5% respectively. Flint (1985) also recorded 
more female than male G. pallidipes and G. morsitans in F3 
traps in Zimbabwe. Considering that the similar odour baits 
attracted similar sections of the population to the different
131
University of Ghana          http://ugspace.ug.edu.gh
itrap typer., the higher preponderance of females in the NG2B 
and F3 traps should be due to a better entry response by the 
females to these traps.
Until traps became widely used for sampling, female 
tsetse were always poorly represented by other sampling 
methods. Rogers (1984) pointed out that the improvement of 
traps should aim at increasing the proportion of sections of 
the populations that are underrepresented. In this respect, 
the NG2B trap is an improvement over the biconical trap.
Female samples are particularly useful for various studies, 
including ovarian aging and for the establishment of 
laboratory colonies.
It was observed that difference in trap design was more 
important than different odour baits in influencing the age 
composition of catches. Dransfield et al., (1986b) also found
that the age composition of G. pallidipes caught in odour 
baited and unbaited biconical traps were not different. The 
difference in the age composition of the samples from the two 
trap types was most likely due to a difference in entry 
response but there seem to be no obvious reason why older 
flies enter the NG2B trap more readily than the biconical 
trap. The possibility that gravid females could be seeking 
1 arviposit.ion sites in the more enclosed ground shade provided 
by the NG2B trap was investigated by comparing the percentages 
of third instar bearing females in both trap types but no 
apparent difference was observed. It could be that parous 
females generally fly lower than nulliparous ones so that the
132
University of Ghana          http://ugspace.ug.edu.gh
133
anomaly was not observed in the overall male catch, although 
on two days the inside of the screens caught more males than 
the outside. It is, therefore, suggestive that more females 
than males were being killed that could have been caught by 
the trap. This implies that the females were moving round the 
trap in wider circles than the males. It is quite logical that 
male flies which may be seeking mating partners will tend to
remain closer to the host (trap in this case) for easier
location of: landing females whilst females that are in the 
refractory phase will tend to stay further away from the trap. 
The idea is supported by the higher female percentage caught 
in the trap without screens although the difference just 
failed to be statistically significant by the chi-squared 
test. This difference also resulted in the higher efficiency 
estimated by using the catch from the trap without screens.
The general impression is that the screens were too close 
to the trap and the test needs repeating probably using more 
screens placed at a greater distance. However, the difference 
this could make on the estimate is not likely to be very much
and the fact remains that there is considerable room for
improvement on either the trap design or odour baits or both.
From the results of this study it was concluded that the 
NG2B trap baited with medium dose (c. lOOOmg/h) cow urine and 
medium (c. 500mg/h) or low dose (c. 150mg/h) acetone was the 
cheapest most effective trap for sampling G. longipennis at 
Nguruman. More work on the use of octenol and the 'winged* NGU
University of Ghana          http://ugspace.ug.edu.gh
lower entrance of the NG2B trap would be more accessible to 
them than those of the biconical trap.
The efficiency of the NG2B trap recorded for G. 
longipennis was very low for both males and females (7.3% and 
7.9% respectively). In the same experiment, data were 
collected on G. pallidipes and the trap efficiency estimate 
was about 56% for both sexes (Dransfield pers. comm.). 
According to Hargrove (1980b) the efficiency estimated by this 
technique is considered the minimum efficiency of the trap 
because some flies that get intercepted by the outside of the 
screen could have been caught if they got to the trap and 
others that could have been caught by the trap were probably 
caught by the inside of the screens as they circumnavigated 
the trap. Vale and Hargrove (1979) actually observed flies 
being electrocuted by the screens as they were moving within 
the ring. The technique thus tends to underestimate trap 
ef f i ci ency.
Looking back at Table 4.6, a higher catch of female flies 
was recorded on the inside of the screens than the outside, 
indicating that flies were being killed that could have 
returned to be caught by the trap. This is under the 
assumption that all flies approached and left the trap at a 
height not greater than that of the screens. Under the 
assumptions on which the technique is based, more flies should 
be caught on the outside than the inside because the inside 
should be catching the same proportion as the outside, but 
from a smaller population of flies within the ring. This
13 If
University of Ghana          http://ugspace.ug.edu.gh
13$?
could piovide further increases in the catches of G. 
1ongipennis .
University of Ghana          http://ugspace.ug.edu.gh
136
CHAPTER FIVE 
THE ACTIVITY PATTERN OF G. LONGIPENNIS
5.1. INTRODUCTION.
The daily activity of tsetse flies, as measured by 
the flight time, is limited to relatively short periods within 
each day (Randolph and Rogers, 1978; Bursell and Taylor,
1980). In the early studies on tsetse behaviour and ecology, 
it was observed that the known vector species, belonging 
mainly to the morsitans and palpalis groups, had a diurnal 
activity pattern. Detailed laboratory studies have since been 
carried out on several of these species. Some workers also 
observed that some less well studied species of the fusca 
group were crepuscular, that is mainly active at dawn and 
dusk.
Remarks on the crepuscular activity of fusca species 
include those of Lamborn (1912) and Swynnerton (1921) on G. 
brevipalpis, Neave (1912) and Lewis (1942) on G. longipennis, 
and Chapman (1950) and Page (1959) on G. fusca. The few 
reports giving some information on the activity patterns of 
the fusca species are those of Power (1964) and Owaga (1981) 
on the activity pattern of G. longipennis, Harley (1965) on 
the activity pattern of G. brevipalpis and Kangwagye (1974) on 
G. fuscipleuris.
The lack of adequate sampling techniques has, however, 
limited such studies. Most of the field studies carried out on
University of Ghana          http://ugspace.ug.edu.gh
137
the palpalis and morsitans groups have relied on the use of 
traps which have not been used successfully for the fusca 
species. Work on these species has therefore relied on the use 
of fly rounds (Power, 1964), baited oxen techniques (Harley, 
1965; Kangwagye, 1974) and slow moving vehicles (Owaga, 1981). 
However, all these methods depend on the efficiency of human 
catchers, which is variable and impossible to standardize. 
Furthermore, for crepuscular species, the efficiency of fly- 
rounds drops to zero when it is too dark for the human eye. 
Power (1964) noted that his catchers were unable to catch 
flies; at very low light intensities, although G . 1 ongipennis 
could still be heard buzzing around. It was therefore not 
possible for him to determine how long the species remained 
active after dark.
In the present study an electric net (Vale ,1974b), known to 
kill about 95% of the flies colliding with it, was used to 
provide a more objective measure of the activity pattern of G. 
longipennis at Nguruman.
Adequate knowledge of the activity patterns of the tsetse 
species involved in any tsetse trypanosomiasis/system is 
vital. Firstly, host-vector contact is one of the factors 
determining the level of trypanosomiasis in any given area. 
This depends on the fly distribution in relation to the host 
distribution in both space and time. Avoidance of challenge 
from G. 1ongipennis may be achieved by keeping cattle out of 
the fly habitat during periods of activity. Secondly,
University of Ghana          http://ugspace.ug.edu.gh
138
knowledge of the activity pattern of a species can be 
exploited in the development of better sampling techniques.
5.2. MATERIALS AND METHODS 
5.2.1 The electric screen
The electrocuting device used for the study was designed 
by Vale (1974b) and is termed the electric net or electric 
screen. It consisted of a sheet of fine black nylon netting, 
measuring about 90 cm x 90 cm, suspended vertically from an 
aluminium frame measuring 110 cm x 120 cm. Blackened fine 
copper wires 0 . 2 2  mm in diameter were strung vertically 0 . 8  cm 
apart between the top and bottom of the frame on either side 
of the net.
When electrified, a voltage differential of about 20 to 
30 kV was created between adjacent wires so that the discharge 
was enough to kill or stun the fly when it touched the wires. 
Power was obtained from a 12 volt battery delivered through a 
specially designed high voltage pulse generator. Details of 
the construction and functioning of the device are provided by 
Vale (1974b).
The pulse generator and component parts of the electric 
net were obtained from Zimbabwe and the net was assembled in 
ICIPE, Nairobi. In the field the system was set up as shown in 
Figure 5.1, in a glade in woodland. The screen was tied in a 
vertical position between two metal poles stuck firmly into 
the ground. To increase the concentration of flies around the 
electric screen, a 2 m x 1 m piece of black cloth was
University of Ghana          http://ugspace.ug.edu.gh
139
stretched adjacent to the electric screen in the position 
shown in the figure, as advised by Vale (pers. comm.)* Stunned 
or killed flies were recovered in shallow water-filled metal 
trays placed on the ground on each side of the electric net. A 
few drops of detergent were added to reduce the surface 
tension of the water, thus preventing the recovery and escape 
of stunned flies.
The trays were placed close to the base of electric net 
to ensure the collection of all dead flies that might slide 
down the net. To eliminate the possible attractiveness of the 
shiny metal to flies, the inner surfaces of the trays were 
painted orange-brown whilst the aluminium frame was painted 
with black patches. The equipment was left in place 
throughout the period of investigation. When not in use, the 
screen was carefully covered with a heavy plastic tarpaulin 
to prevent it getting wet. The whole set-up was protected from 
animal damage by a ring of thorny branches of Acacia.
5.2.2. Meteorological data
A Grant's automatic meteorological recorder was set up 
about 1 0  metres from the experimental set-up to record air 
temperature, relative humidity and wind speed when experiments 
were being run. The probes for these physical factors were 
mounted on a pole about 1.5 metres above ground level. Light 
intensity was measured with a Metrawatt (Metrux K) luxmeter 
which was placed next to the screen on a horizontal wooden 
plank erected about 1 metre above the ground. Daily notes were
University of Ghana          http://ugspace.ug.edu.gh
140
Figure 5.1: Set-up of the electric screen and cloth target 
used in the study of activity pattern of G. 
longipennis.
University of Ghana          http://ugspace.ug.edu.gh
141
kept of the general weather conditions such as cloud cover, 
the phase of the moon and the times it shone during the 
experiments. Sunset and sunrise times were also recorded 
daily.
5.2.3. Activity Period
In a preliminary study, catches were made every hour for 
two days and two nights. It was observed that most G.
1ongipennis were caught in the hour after sunset (1815 h) and 
the 30 minutes before sunrise (0630 h) whilst activity 
throughout the rest of the night and day was negligible.
The main investigations were carried out in the dry 
months of September and October 1987, when it was hoped that 
there would be no rains to disrupt the experiments. The 
intention was to collect flies at short time intervals in 
order to provide a more precise picture of the activity 
period. Therefore, all experiments were run in the evenings 
from 1730 h and from 0530 h in the mornings. Flies were 
collected from the trays at 15-minute intervals until all 
activity ceased. Light intensity readings were also taken at 
the beginning of every collection. Two successive records of a 
zero catch was adopted as the standard to mark the end of fly 
activity.
A minimum of two people collected flies from each tray 
and this took 1 - 2  minutes during which time the screen was 
switched off. Fly collectors were required to pick out every 
insect from the tray that was about the size of tsetse. This
University of Ghana          http://ugspace.ug.edu.gh
142
was both to hasten the collecting process and to minimize the 
period of interruption and also to safeguard against the loss 
of tsetse through misidentification at very low light 
intensities. To further reduce errors in fly counts for the 
appropriate time periods and to facilitate the next 
collection, the trays were cleared of all insects at every 
collection. The pooled catch for each 15-minute period (from 
all collectors) was kept separate in an appropriately labelled 
glass tube.
5.2.3. Experiments
Although the main objective of the investigation was to 
define the activity pattern of G. 1ongipennis, a variety of 
other factors were tested with the electric screen to assess 
their effects on catches. These treatments involved materials 
related to tsetse trapping technology (screen colour and odour 
baits) which were selected with the view to obtaining 
information that could be useful in the further improvement of 
trap efficiency for this species.
Experiment 1:
The attractiveness of blue and black cotton material (the 
type used in the construction of biconical and NG2B traps; see 
chapter 4.) was compared. Treatments were alternated from day 
to day with a 2 m x 1 m piece of the material stretched 
between two poles adjacent to the screen in the position 
indicated earlier in Fig 5.1. Cow urine and acetone were 
dispensed as odour baits; the dispensers were placed next to
University of Ghana          http://ugspace.ug.edu.gh
143
the pole between the screen and the cloth. The experiment was 
run for 10 days (the 9th-19th September 1987) giving 5 morning 
and 5 evening replicates for each treatment.
Experiment 2:
This was designed to compare the catches with and without 
the cloth target placed next to the screen. Black cloth was 
used for the experiment. This was run only in the evenings 
from the 20th-25th September. Each treatment was thus 
replicated three times on alternate days.
Experiment 3:
The efficacy of warm cow urine (30 ± 1 oC) was compared 
with that of urine at ambient morning temperatures (19-22°C). 
The warm urine was maintained at the desired temperature by 
immersing a can of urine in water warmed to the required 
temperature and kept in a polystyrene box as was done in Expt. 
14 in Chapter 4. Acetone was dispensed together with both 
treatments. The experiment was run in the mornings on the same 
days as experiment 2 , giving three replicates for each 
treatment.
Experiment 4:
A comparison was made between the efficacies of cow urine 
and a recently developed chemical mixture of paracresol, 1 - 
octen-3~ol and 3n-propy1-phenol (8:4:1 ratio) (Bursell et al . ,
1988). This was obtained from The Tsetse R e s e a r ^ - ^ . .
University of Ghana          http://ugspace.ug.edu.gh
144
Laboratories, Bristol, U.K packaged in 150 micron thick porous 
polythene sachets of total surface area 50cm2 . According to 
Hall (pers. comm.) the release rates from these sachets at 
27°C with wind speed of 8 kph are 9.1, 3.0 and 0.5 mg/day for 
paracresol, l-octen-3-ol and 3 n-propylphenol respectively. In 
the field, these rates will be slightly higher in the 
afternoon when the temperatures were a little higher than 
27°C. Morning temperatures, on the other hand, ranged between 
19-22°C which will result in lower release rates. Acetone was 
dispensed with both treatments. The experiment was run for 6 
days (6th-10th October) in the evenings and mornings, giving 3 
replicates of each treatment for each time period.
A two way analysis of variance (treatments x time) was 
carried out on the data from each experiment using a log (N+l) 
transformation on the numbers caught. When the same 
experiment was run both in the morning and in the evening the 
data for these periods were analyzed separately.
5.3 RESULTS
5.3.1. Fly numbers from various treatments
Table 5.1 gives a summary of the results obtained from 
the various experiments including the fly numbers on each day 
(males and females pooled), the total catch for each 
treatment, the F-ratio from the analysis of variance (ANOVA) 
and the detransformed mean catch/per 15-minute period for each 
treatment.
Comparing the effects of cloth colours, separate analyses 
for males and females showed that the treatment effect was
University of Ghana          http://ugspace.ug.edu.gh
145
Table 5.1: Numbers of G. longipennis caught at the electric 
screens using various treatments (males and
females pooled).
Daily total Overa11 Detr.mean F-ratio
Evening
1 2 3 4 5
total per 15-min. 
period
(df)
Blue cloth 41 54 38 51 143 327 7.7 1 .2 2 "®
Black cloth 64 70 55 31 31 256 6.3 (1,40)
Cloth 31 51 54 - _ 135 5.2 15.35***
No cloth 9 25 5 - - 39 1.7 (1 ,1 2 )
Phenols
+Acetone
69 60 46 - - 157 1 1 . 1 5.35*
Cow urine 
+Acetone
53 1 1 37 1 0 1 5.7 (1/16)
Morning
Blue cloth 4 16 24 3 15 62 2 . 5 0 .0 0 1 n3
Black cloth 17 19 5 4 1 2 57 2 . 4 (1,24)
Warm urine 5 2 1 3 _ _ 29 1 . 8 0.32“®
Ambient 13 5 15 - - 33 2.4 (1,13)
Phenols 
+Acetone
19 4 7 - 30 2 . 0 0.93"s
Cow urine 
+Acetone
15 1 4
"
2 0 1 . 2 (1 ,1 2 )
(* = P<0.05 ; ** * =P<0.001; ns=not significant)
University of Ghana          http://ugspace.ug.edu.gh
146
just significant at the 5% level for males (F(i,4 0 ) = 5.06)
but not for females (F(i,4 0 ) = 0.74). The blue cloth appeared
seemed to catch significantly more males than black, mean 
catches being 5.5 for blue cloth and 3.5 for black. However, 
when the pooled data for both sexes were analyzed, there was 
no significant treatment effect.
The treatment effect in experiment 2, comparing the use 
of cloth and no cloth, was highly significant (P<0.001). The 
use of cloth increased catches by about 3 times. The analysis 
was performed on the pooled data for both sexes because the 
numbers caught without cloth were so few.
There was no significant treatment effect when warm urine 
and urine at ambient temperature were compared. Therefore, the 
efficiency of cow urine did not seem to be improved by warming 
to 30°C. There was a significant difference for the evening 
catches between the phenol/octenol mixture and cow urine.
The latter odour doubled the catch over the former. The effect 
was not significant for the morning catches but the 
phenol/octenol mixture still nearly doubled the catch.
5.3.2. The daily activity pattern
Having established that there was little effect of cloth 
colour on catches, the pooled data from experiment 1 over the 
1 0 -day period were used to describe the activity pattern of 
G . 1ongpennis. Tables 5.2 and 5.3 show the daily catches at 
different times for the evening and morning activity periods 
respectively. Fig. 5.2 shows the relative levels of activity
University of Ghana          http://ugspace.ug.edu.gh
147
between the morning and evening periods and Fig. 5.3 focuses 
on the main activity pattern in the evening.
Taking sunset at 1815 h (local time = GMT + 3h) as the 
reference point for evening activity, significant levels of 
activity were observed by 1745-1800 h, and one or two males 
were occasionally recorded earlier than this time. Male 
catches increased gradually to a peak by 1815-1845 h. 
Thereafter, a gradual decline followed till 1900 h followed by 
a sharp drop to zero catch by 1915-1930h. Females on the other 
hand were regularly recorded only after sunset, with the
exception of one very cloudy day (later on followed by rain)
when two females were caught just before sunset. A rapid rise 
in catches from 1815 h to 1845 h followed by an equally rapid 
drop thereafter, often produced a sharp female peak at 1830- 
1845 h. On one day (11th September) a broad peak extended from 
1845 h -1900 h and on another day (15th September) the peak 
catch was recorded at the earlier time of 1815-1830 h and 
catches remained high until 1845-1900 h (see discussion for 
possible reasons). On most days, male and female activity 
stopped at the same time but one female was recorded on each
of two occasions at 1930 h. Although there was very bright
moonlight on some days this did not seem to influence the time 
when activity stopped.
University of Ghana          http://ugspace.ug.edu.gh
148
Table 5.2: Catch per 15 minutes,tota1 catch and the
percentage of female G. longipennis caught at the 
electric screen during evening activity period.
Collection Time (h)
Day
1
Sex
M
1730
0
1745
0
1800
0
1815
3
1830
8
1845
9
1900
8
1915
1
1930
0
TOT.
29
%Female
F 0 0 0 0 1 7 3 0 1 1 2 29.3
2 M 0 0 0 1 1 0 4 13 2 0 30
F 0 0 0 0 3 1 2 18 0 1 34 53.1
3 M 0 2 1 5 7 14 3 1 0 33
F 0 0 0 0 3 17 1 0 0 2 1 38 . 9
4 M 0 1 1 1 1 2 9 1 0 1 0 35
F 0 0 0 0 3 2 0 1 1 1 0 35 50 .0
5 M 0 0 1 3 2 1 0 7 1 0 24
F 0 0 0 0 1 1 0 3 0 0 14 36.8
6 M 0 0 0 0 1 1 9 5. 1 0 26
F 0 0 0 0 1 2 1 1 6 0 0 29 52 .7
7 M 0 1 1 3 4 13 7 0 0 29
F 0 0 1 0 1 17 2 1 0 2 2 42.1
8 M 0 0 0 5 5 3 4 0 0 17
F 0 0 0 0 0 1 1 3 0 0 14 45.3
9 M 1 0 0 14 30 2 1 7 0 Rain 73
F 0 0 0 2 8 55 3 2 Rain 70 48 . 9
10 M 0 0 0 1 5 7 1 1 0 15
F 0 0 0 1 4 9 2 0 0 16 51. 6
Tot. M 1 4 4 36 94 99 65 8 0 311
F 0 0 1 3 36 169 52 4 2 267 46.2
University of Ghana          http://ugspace.ug.edu.gh
149
Table 5.3: Catch per 15 minutes, total catch and the percentage
of female G. 1ongipennis caught at the electric screen 
during morning activity period.
Collection Time (h)
0545 0600 0615 0630 0645 0700 0715 TOT. %Fem
Day Sex
1 M 0 1 1 0 0 0 0 2
F 0 2 0 1 0 0 0 3 60 .0
2 M 0 4 1 1 0 0 0 6
F 0 5 6 0 0 0 0 1 1 64.7
3 M 1 9 4 0 0 0 0 14
F 0 2 0 0 0 0 0 2 12.5
4 M 1 2 2 1 0 0 0 6
F 0 1 0 3 0 0 0 0 13 68.4
5 M 1 1 2 3 0 0 0 0 16
F 0 2 5 1 0 0 0 8 33.3
6 M 0 2 1 0 0 0 0 3
F 0 2 0 0 0 0 0 2 40 .0
7 M 0 1 0 0 0 0 0 1
F 0 0 1 1 0 0 0 2 6 6 . 6
8 M 0 2 0 1 0 0 0 3
F 0 0 1 0 0 0 0 1 25.0
9 M 0 6 2 1 0 0 0 9
F 0 4 2 0 0 0 0 6 40.0
10 M 0 2 4 0 0 0 0 6
F 1 3 1 1 0 0 0 6 50 .0
Tot. M 2 41 18 4 0 0 0 6 6
F 1 30 19 4 0 0 0 54 45.0
University of Ghana          http://ugspace.ug.edu.gh
Pe
rc
en
ta
ge
 
of 
tot
al 
ca
tch
 
Pe
rc
en
ta
ge
 
of 
tot
al 
ca
tc
h
150
Time (hours)
Figure 5.2: The relative levels of activity of G.
longipennis in the morning and in the evening. 
(  males; .females)
Time (hours)
Figure 5.3: Activity patterns of male( ) and female
v-y—J g. longipennis in the evening
University of Ghana          http://ugspace.ug.edu.gh
151
In g e n e r a l ,  l e s s  a c t i v i t y  was ob se r v ed  i n  the  morn ings  
than i n  the even ing s  and the morning a c t i v i t y  p e r i o d  was much 
s h o r t e r  than the even ing  one.  Over  the  ten  days  an a v e r a g e  o f  
58 f l i e s / d a y  were caught  d u r i n g  the even ing  a c t i v i t y  as  
a g a i n s t  12 f l i e s / d a y  i n  the morn ing .  P ronounced morning  
a c t i v i t y  was ob se r ved  on l y  a f t e r  0545 h a l t h ou gh  the  
o cca s i o na l  male was caught  b e f o r e  then.  Most f l i e s  were  
r e g u l a r l y  caught  between 0545 h and 0615 h ( s u n r i s e  a t  0630h)  
with the peak c a t che s  u s u a l l y  o c c u r r i n g  between 0545 h and 
0600 h. By 0630 h on ly  s i n g l e  f l i e s  were caught  on a few days  
but on ly  r a r e l y  were any f l i e s  caught  a f t e r  t h i s  t ime .  G iven  
the sho r t  p e r i o d  o f  morning a c t i v i t y ,  t h e r e  was no s i g n i f i c a n t  
d i f f e r e n c e  between the male and f ema le  a c t i v i t y  p a t t e r n  i n  the  
morn ing .
5 .3 .  A c t i v i t y  and p h y s i c a l  f a c t o r s .
For each day,  the number o f  f l i e s  caught  i n  each 15-  
minute p e r i o d  was e x p r e s s e d  as the p r o p o r t i o n  o f  the  t o t a l  
catch f o r  the day and used as an i ndex  o f  a c t i v i t y  f o r  tha t  
time p e r i o d .  S ep a r a t e  p l o t s  were  then made o f  the  a c t i v i t y  
ag a i n s t  t empe ra tu r e ,  r e l a t i v e  humid i t y  and l i g h t  i n t e n s i t y  
r ecorded  at  the c o r r e s p ond i n g  p e r i o d s .  The a r c s i n e  { P  + 0 . 5  
t r an s f o rmat i on  was used on the p r o p o r t i o n s  to  n o rma l i z e  the  
data .  In o rde r  to a v o i d  the many z e ro  c a t ch es  r e c o r d e d  i n  the  
very e a r l y  and very  l a t e  p e r i o d s ,  the  a n a l y s i s  was l im i t e d  to  
catches  made from 1800 h ( f o r  ma l e s )  and f rom 1815 h ( f o r  
f emal es )  to 1900h.
University of Ghana          http://ugspace.ug.edu.gh
152
Among the physical factors that were considered, only 
light intensity showed any significant relationship with 
activity. Figures 5.4A and 5.4B show the plots of activity 
against light intensity for males and females respectively.
For both sexes the relationship between activity (Y) and light 
intensity (X) was best described by the fitted parabola with 
the following functions:
Y = 1.03 + 0.23X - 0.08X2 (r2 = 33%) for the males
and
Y = 0.99 + 0.23X - 0.09X2 ( r 2 = 67%) for the females.
In both cases activity is seen to increase with decreasing 
light intensity up to a peak at about 100 lux. Activity then 
decreased with further decrease in light intensity. As can be 
observed from the figure the relationship was stronger for 
females than it was for males. Thus 67% of the variability in 
female catch and 33% of the male catch in the different time 
periods is explained by the above relationships.
For the morning activity the distinction in the onset of 
male and female activity was less clear and the sexes were 
pooled in order to avoid the many zero catches recorded for 
some periods. No significant relationship was however observed 
when the above analysis was carried out.
University of Ghana          http://ugspace.ug.edu.gh
Tr
an
sfo
rm
ed
 
ca
tch
 
Tr
an
sf
or
me
d 
ca
tc
h
JL O
Log(light intensity/lux)
B
Log(light intensity/lux)
Figure 5.4: The relationship between activity and light 
intensity.(For males (A ), Y = 1.03 + 0.23X -  
0.08X2 ; r 2 = 33% and for females (B ), Y = 0.99 + 
0.23X -  0.09X2; r 2 = 67%)
University of Ghana          http://ugspace.ug.edu.gh
154
5.4. DISCUSSION
The comparisons among various treatment factors bring out 
several important points that could be useful in the 
development of traps for G. 1ongipennis and for tsetse control 
in general. Assuming that the screen killed 95% of the flies 
colliding with it, as estimated by Vale (1974b), the 
differences in catches observed between treatments were due to 
differences in their effectiveness in attracting flies to the 
screen.
Blue cloth appears to be more attractive than black cloth 
to male G .1ongipennis. This could be due to the brighter 
colour of the blue and better contrast with the background. 
This brightness factor would be more pronounced during the 
male activity period when there is still sufficient 
illumination. It was, probably, less important during female 
activity due to the very low light intensity. Although the 
overall effect was not significant, it is worth noting the 
colour combinations when designing traps for G. 1ongipennis. 
The proportion of blue to black could have an effect on the 
sex ratio.
It was observed that the presence of both odour and 
visual attraction are required for optimum catches. Visual 
attraction in host finding should be less important for 
crepuscular species than for diurnal species but the 
significant reduction in catch when no cloth was used 
emphasizes the importance of visual attraction for G.
1ongipennis.
University of Ghana          http://ugspace.ug.edu.gh
155
The increase in catches when traps were baited using the 
mixture of phenols and octenol as compared to cow urine could 
be due to the phenols being superior to the cow urine. The 
phenols in the mixture are supposed to be the most attractive 
components occurring naturally in cow urine. The poorer 
performance of the crude cow urine may be due to the presence 
of other compounds in the urine which are known to have 
repellent effects on tsetse (Bursell et al . , 1988).
The presence of octenol in the mixture is however a more 
likely reason for the significant difference in the catch. 
Brightwell et al.(1989) have recently shown that if octenol 
is dispensed together with crude cow urine and acetone a 2-4 
times increase in catch is realized for G. longipennis. This 
supports the idea that the difference in catch obtained in the 
present investigation was mainly due to the presence of the 
octenol in the mixture of phenolic compounds.
Although differences in levels of catches were observed 
between some treatment factors, the onset and end of activity 
were not altered by these treatments. Basically, the various 
attractants were effective only when the fly became active. 
According to Brady (1972, 1973) and Brady and Crump (1978), 
tsetse generally exhibit spontaneous activity which is 
controlled by an endogenous rhythm, that may be influenced by 
some physical factors. Many field workers on diurnal Glossina 
species have observed that the daily activity patterns were 
dependent on the temperature profile (Challier, 1973; Turner, 
1980 and Turner, 1987).
University of Ghana          http://ugspace.ug.edu.gh
156
The l i t t l e  e a r l i e r  work on c r e p u s c u l a r  f u s c a  s p e c i e s ,  
however ,  i n d i c a t e d  that  l i g h t  i n t e n s i t y  was the  ma jo r  f a c t o r  
i n f l u e n c i n g  a c t i v i t y  p a t t e r n s .  Power ( 1964 )  o b s e r v e d  th a t  male  
G . 1ongipennis  ( n e a r  Lake J i p e ,  Kenya)  e x h i b i t e d  an a c t i v i t y  
pa t t e r n  s im i l a r  to  tha t  o b se r v e d  i n  t h i s  s tu dy .  S im i l a r  
pa t t e rn s  of  a c t i v i t y  were a l s o  o b se r ve d  by H a r l e y  ( 1965 )  f o r  
G. brevipalpis  i n  Uganda.  Both au tho r s  no ted  the  c l o s e  
r e l a t i o n s h i p  between the onset  o f  morning and even in g  a c t i v i t y  
and s u n r i s e  and s un se t ,  r e s p e c t i v e l y .  They c o n s i d e r e d  l i g h t  
i n t e n s i t y  to be the major  f a c t o r  d e t e rm in i n g  the p a t t e r n  o f  
a c t i v i t y .  A l though  Power ( 1964 )  e s t a b l i s h e d  a p a r t i a l  
r e g r e s s i o n  connec t ing  f l y  c a t ch  w i t h  l i g h t  i n t e n s i t y  and  
r e l a t i v e  humid i t y ,  he conc luded  tha t  l i g h t  i n t e n s i t y  app ea r ed  
to t r i g g e r  the onset  o f  a c t i v i t y .
The r e s u l t s  f rom the p r e s en t  s tudy  a l s o  i n d i c a t e  tha t  
l i g h t  i n t e n s i t y  was the major  f a c t o r  i n f l u e n c i n g  the a c t i v i t y  
of G . 1ongipennis  at  Nguruman. C o n s i d e r i n g  the s t e a d y  d e c l i n e  
in  l i g h t  i n t e n s i t y  around dusk,  the  sudden b u r s t  o f  a c t i v i t y ,  
e s p e c i a l l y  of  the  f em a l e s ,  s t r o n g l y  s u g g e s t s  a " t r i g g e r "  or  
th r e s ho ld .  The f a c t  tha t  the onset  o f  a c t i v i t y  and the  t im in g  
of peak c a tches  u s u a l l y  o ccu r r ed  at  the  same t ime might  
i n d i c a t e  an e n t i r e l y  endogenous rhythm. However ,  i t  was 
observed  tha t  on a few ve r y  c l oudy  days ,  when the  l i g h t  
i n t e n s i t i e s  d e c l i n e d  somewhat f a s t e r  than u s u a l ,  the  peak  
catches  occu r r ed  c o r r e s p o n d i n g l y  e a r l i e r  i n  the  ev en i n g .  On 
c l e a r  sunny days on the o the r  hand,  the  peak was r e c o r d e d  
co r r e s pond i n g l y  l a t e r  as l i g h t  i n t e n s i t i e s  f e l l  more s l o w l y .
University of Ghana          http://ugspace.ug.edu.gh
157
This suggested that if the threshold level of light intensity 
occurs at any time during the day, flies may be triggered into 
activity.
Taking  the t h r e s h o l d  l e v e l  as  the  l i g h t  i n t e n s i t y  at  
which the sudden r i s e  i n  c a t ch  s t a r t e d ,  mal es  and f ema l e s  show 
d i f f e r e n t  t h r e s h o l d  l e v e l s .  From f i g u r e  4 .3  the  t h r e s h o l d  
l e v e l  f o r  males  was 1,600 l ux  w h i l s t  tha t  f o r  the  f ema l e s  was  
250 l ux .
Th i s  can be co nv e r t ed  i n t o  p h y s i c a l  u n i t s  o f  i l l u m i n a t i o n  
energy (W a t t s / s qua r e  m e t e r ) ,  which a c c o r d i n g  to  Young and 
Gibson ( 1987)  i s  a more o b j e c t i v e  measure  o f  l i g h t  i n t e n s i t y  
from the i n s e c t s ’ po in t  o f  v i ew .  U s in g  t h e i r  c o n v e r s i o n  f a c t o r  
of 1 Wm2 = 355 l ux ,  the  t h r e s h o l d  l e v e l s  become 4,507 mWm2 f o r  
males and 704 mWm2 f o r  f ema l e s .  The peak c a t ch  f o r  bo th  s ex e s  
occu r r ed  at  about  lOO lu x ,wh i c h  works  out to  281 mWm2 . Quo t ing  
l i g h t  i n t e n s i t y  i n  Z e i s s  u n i t s ,  Power ( 1964 )  r e c o r d e d  peak  
catches  at  4 Z e i s s  u n i t s .  Othe r  than the  f a c t  th a t  he 
mentioned tha t  t h i s  o cc u r r ed  a f t e r  s un s e t ,  i t  i s  not  p o s s i b l e  
to compare t h i s  v a l u e  w i th  th a t  o b t a i n e d  i n  the  p r e s e n t  s tudy  
because the photometer  was not  c a l i b r a t e d  on the bench and the  
Ze i s s  un i t  i s  not  a r e c o g n i z e d  s t a n d a r d  u n i t  f o r  measu r ing  
l i g h t  i n t e n s i t y .
From l a b o r a t o r y  s t u d i e s  on G. m o r s i t a n s ,  Brady  ( 1987 )  
i n f e r r e d  that  the  sh o r t  b u r s t  o f  a c t i v i t y  o b s e r v e d  i n  many 
d iu rn a l  t s e t s e  s p e c i e s  when they change r e s t i n g  s i t e s  a t  dawn 
and dusk was c o n t r o l l e d  by l i g h t  i n t e n s i t y .  I t  i s  q u i t e  
p o s s i b l e  tha t  the a c t i v i t y  i n  c r e p u s c u l a r  s p e c i e s  i s  an
University of Ghana          http://ugspace.ug.edu.gh
158
extended form of  the above phenomena i n  d i u r n a l  s p e c i e s .  T h i s  
cou ld  be an a d a p t a t i o n  i n  c r e p u s c u l a r  s p e c i e s ,  making i t  
p o s s i b l e  f o r  them to use  the same t ime f o r  f e e d i n g  and f o r  
chang ing  r e s t i n g  s i t e s  as w e l l .  Brady ( 1987 )  o b s e r v e d  tha t  
the mean t h r e s h o l d  l i g h t  i n t e n s i t y  f o r  t ake  o f f  was 350 mW/m2 . 
This  i s  q u i t e  c l o s e  to  the v a l u e  o b t a i n e d  a t  peak a c t i v i t y  i n  
t h i s  s tu dy ,  soon a f t e r  which G. longipennis  r e t r e a t s .  The 
a c t i v i t y  i n v o l v e d  i n  B r a d y ' s  l a b o r a t o r y  s tudy  took ve ry  s h o r t  
pe r i o d s  which were thought  to  r e f l e c t  what happens  i n  the  
f i e l d  du r i ng  the change of  r e s t i n g  s i t e s  i n  d i u r n a l  s p e c i e s .  
Compared to the d i u r n a l  s p e c i e s ,  G. longipennis  a p pe a r s  to  
take o f f  at  a much h i g h e r  l i g h t  i n t e n s i t y  i n  o r d e r  to  f e e d  but  
withdraws  to  p r o b a b l y  change r e s t i n g  s i t e s  a t  abou t  the  same 
time as i n  d i u r n a l  s p e c i e s .
Temperature  and r e l a t i v e  humidi ty  d i d  not  appea r  to  be  
f a c t o r s  that  were l i k e l y  to de te rmine  the p a t t e r n  o f  a c t i v i t y .  
In the f i r s t  p l a c e ,  the change i n  the l e v e l s  o f  t he se  f a c t o r s  
from one c o l l e c t i o n  p e r i o d  to the o the r  were so sma l l  t ha t  
they were not  l i k e l y  to  c ause  the  ob se rv e d  changes  i n  c a t ch e s  
wi th  t ime.  Mo reover ,  t empe ra tu r e s  and r e l a t i v e  h um i d i t i e s  
dur ing  morning peak a c t i v i t y  d i f f e r e d  g r e a t l y  f rom tho se  
dur ing  evening  peak a c t i v i t y .  L i g h t  i n t e n s i t i e s ,  on the  o th e r  
hand, were w i t h i n  the same r ange  d u r i n g  bo th  morn ing  and 
evening peak a c t i v i t i e s .  I t  i s  t h e r e f o r e  more l o g i c a l  to  
r e l a t e  a c t i v i t y  to l i g h t  i n t e n s i t y  r a t h e r  than w i t h  
t emperature  and r e l a t i v e  humid i ty .  Power ( 1 9 6 4 ) ,  however ,  
sugge s t ed  tha t  the g e n e r a l l y  low dawn t empe ra t u r e s  5i i ^ 4^ h a v e
University of Ghana          http://ugspace.ug.edu.gh
159
an i n h i b i t o r y  e f f e c t  i n  the morn ing .  In the months th a t  t h i s  
study was conducted ,  morning t empe ra t u r e s  r anged  between  19°C  
and 22 °C , q u i t e  c l o s e  to  c r i t i c a l  t empe ra t u r e  o f  18°C a t  whi ch  
t s e t s e  a r e  g e n e r a l l y  thought  to be i n a c t i v e  (G l a s g ow ,  1963 ) .
I t  would ,  however ,  be d e s i r a b l e  to conduct  e xpe r iment s  a t  
o ther  t imes of  the y e a r ,  when morning t empe ra t u r e s  a r e  l owe r  
or h i g h e r ,  to  f i n d  out i f  the  l e v e l  o f  dawn a c t i v i t y  would  be  
any d i f f e r e n t  f rom what has been r e p o r t e d  he r e .
University of Ghana          http://ugspace.ug.edu.gh
160
CHAPTER SIX
POPULATION DYNAMICS OF GLOSSINA LONGIPENNIS:
I .  APPARENT DENSITIES AND MORTALITY RATES
6.1.  INTRODUCTION
Adequate  knowledge  o f  a g i v e n  t s e t s e  p o p u l a t i o n  i s  
e s s e n t i a l  f o r  u nde r s t and in g  i t s  i nvo l vement  i n  d i s e a s e  
t r a nsm i s s i on  and f o r  the deve lopment  o f  a p p r o p r i a t e  c o n t r o l  
s t r a t e g i e s .  The main i ndex  o f  the  pe r f o rmance  o f  a g i v e n  
animal p o p u l a t i o n  i s  the  p o p u l a t i o n  s i z e  which i s  d e t e rm ined  
by f ou r  pr imary  p a r amete r s :  the  b i r t h  r a t e ,  d ea th  r a t e ,  
immigrat i on  r a t e  and em i g r a t i o n  r a t e .
In the s imp l e s t  terms ,  b i r t h s  and imm ig r a t i on  l e a d  to  
i n c r e a s e s  i n  p o p u l a t i o n  s i z e  w h i l s t  d ea ths  and em i g r a t i o n  
r e s u l t  i n  d e c r ea se  i n  p o p u l a t i o n  s i z e .  In n a t u r e ,  a 
combinat ion of  t he se  f a c t o r s  a c t  on the  p o p u l a t i o n  a t  the  same 
time.  S tud i e s  of  the changes  i n  the se  pa r ame t e r s ,  and the  
unde r l y i n g  causes  o f  t he se  changes ,  form the b a s i s  o f  
popu l a t i o n  dynamics.
Changes i n  t s e t s e  p o p u l a t i o n  s i z e  a r e  i n f l u e n c e d  by bo th  
a b i o t i c  and b i o t i c  f a c t o r s .  Of the  a b i o t i c  c l im a t i c  f a c t o r s ,  
t emperature  and r e l a t i v e  humid i t y  have l ong  been shown t o  
i n f l u e n c e  t s e t s e  p o p u l a t i o n  s i z e s  th rough  t h e i r  e f f e c t  on 
b i r t h  and dea th  r a t e s .  The impo rt ance  o f  c l im a t e  f o r  the  
s u r v i v a l  of  t s e t s e  i n  the f i e l d  was f i r s t  shown by Nash ( 1937 )
University of Ghana          http://ugspace.ug.edu.gh
161
and Nash and Page (1953 )  who e s t a b l i s h e d  s i g n i f i c a n t  
c o r r e l a t i o n s  between the p o p u l a t i o n  s i z e  and m e t e o r o l o g i c a l  
data f o r  G. tachinoides  and G. m. submorsitans  p o p u l a t i o n s  i n  
nor thern  N i g e r i a .  More r e c e n t l y ,  Gouteux and Buck l and  ( 1984 )  
a l s o  showed s i g n i f i c a n t  r e l a t i o n s h i p s  between  t empe ra tu r e  and 
r e l a t i v e  humidi ty  and the appa r en t  p o p u l a t i o n  d e n s i t i e s  o f  G. 
p. p al p a l i s , G. pallicera  and G. n i g r o f u s c a  i n  Cote  d ’ I v o i r e .  
Other wo rker s  have shown s i g n i f i c a n t  r e l a t i o n s h i p s  between  
mo r t a l i t y  r a t e s  and c l im a t i c  f a c t o r s ,  e . g .  Roge rs  ( 1 9 79 ) ,  
Gouteux and L a v e i s s i e r e  ( 1982 )  and Rogers  et  a l . ( 1 984b ) .
These climatic factors act in a density independent way and 
therefore cannot regulate populations.
Long term r e c o r d s ,  however ,  ma in l y  f rom f l y  round d a t a ,  
show that  t s e t s e  numbers f l u c t u a t e  around c h a r a c t e r i s t i c  
popu l a t i o n  l e v e l s  ( G l as gow and We l ch ,  1962; F a i r b a i n  and 
Culwick,  1950; Onyiah,  1978 ) ,  i n d i c a t i n g  tha t  the  p o p u l a t i o n s  
are s u b j e c t  to t i g h t  r e g u l a t i o n .  Some wo rke r s  i n  the  p a s t  were  
of the o p i n i on  that  t h i s  was de t e rmined  by c l im a t i c  f a c t o r s .
I t  i s  now w i d e l y  a c cep te d  tha t  on l y  d e n s i t y  d e p e n d e n t - f a c t o r s  
can r e g u l a t e  animal  p o p u l a t i o n s .  Roger s  and Rando lph  (1984,  
1985) have shown tha t  d e n s i t y - d e p e n d en t  f a c t o r s  a r e  ma in l y  
b i o t i c  ( e . g .  p r e d a t i o n ,  p a r a s i t i sm  e t c . )  and th e se  r e g u l a t e  
the p o pu l a t i o n s  through f e edb ack  mechanisms.  In t h e i r  r e v i ew  
on t s e t s e  e c o l o g y ,  Rogers  and Rando lph ( 1985 )  gave  d i r e c t  and 
i n d i r e c t  e v i dence s  of  d e n s i t y - d e p e n d en t  r e g u l a t i o n  i n  t s e t s e  
popu l a t i ons  and po i n t ed  out tha t  i n  most c a se s  the c ause s  o f
University of Ghana          http://ugspace.ug.edu.gh
162
population regulation are difficult to determine. Rogers 
(1974)  thinks that pupal and adult predation are density- 
dependent. Feeding success (Vale, 1977)  and fly movement 
(Rogers et a l . ,  1984; Turner and Brightwell, 1986)  have also 
been postulated as being regulatory.
The pe r f o rmance  o f  a g i v e n  p o p u l a t i o n  ov e r  t ime i s  
a s se s s e d  by e s t im a t i n g  the  v a r i o u s  demograph ic  pa r ame t e r s .  The  
s i z e  of  a g i v e n  p o p u l a t i o n  can be e s t ima t ed  i n  r e l a t i v e  terms  
( a s  i n d i c e s  o f  abundance i n  one p l a c e  or  a t  one t ime r e l a t i v e  
to an o th e r )  or  i n  a b s o l u t e  terms ( a s  the  a c t u a l  numbers i n  a 
g i ven  a r e a  at  a g i v e n  t im e ) .  D e t a i l e d  r e v i ew s  o f  t he  methods  
used f o r  both  type o f  p o p u l a t i o n  s i z e  e s t im a t e s  a r e  g i v e n  by  
Southwood ( 1978 )  and Sebe r  ( 1973 )  p r o v i d e s  mathemat i ca l  
d e t a i l s  f o r  the se  methods.  G e n e r a l l y ,  r e l a t i v e  methods p r o v i d e  
qu icke r  e s t ima t e s  o f  p o p u l a t i o n  s i z e  and a r e  more o f t e n  used  
but f o r  some pu rpose s  a b s o l u t e  e s t im a t e s  may be r e q u i r e d .
For G. longipennis  a t  Nguruman, t r a p s  were  used  t o  o b t a i n  
r e l a t i v e  p o p u l a t i o n  e s t im a t e s  e x p r e s s e d  as  c a t ch  pe r  t r a p  pe r  
day.  The r e c e n t l y  d e v e l o pe d  NG2B t r a p  which had been shown to  
be more e f f e c t i v e  than the b i c o n i c a l  t r a p  ( s e e  c h ap t e r  4)  was  
the main t r ap  used .  However ,  to  p r o v i d e  a s t a n d a r d  f o r  
comparison,  d a t a  was a l s o  c o l l e c t e d  from s i m i l a r l y  b a i t e d  
b i c o n i c a l  t r a p s  which were i n  use  by the  ICIPE r e s e a r c h  team 
f o r  sampl ing  G. p a l l i d i p e s . A b s o l u t e  p o p u l a t i o n  e s t im a t e s  were  
ob ta ined  through  m a r k - r e l e a s e - r e c a p t u r e  ( s e e  Chap te r  8 ) .
Va r i ou s  methods have a l s o  been de v e l o p e d  to e s t ima t e  
popu l a t i o n  m o r t a l i t y  r a t e s  u s i n g  da t a  o b t a i n e d  f rom samples
University of Ghana          http://ugspace.ug.edu.gh
163
taken from the p o p u l a t i o n s .  These  i n c l u d e  the use  o f
p opu l a t i o n  age  s t r u c t u r e  and the use o f  change i n  f l y  numbers
from one s amp l ing  p e r i o d  to  the  nex t .  D e t a i l s  o f  the  above  
t e chn iques  and methods o f  e s t im a t i n g  p o p u l a t i o n  p a r amete r s  a r e  
g i ven  i n  the s e c t i o n s  tha t  f o l l o w .
F i n a l l y ,  a t tempts  were  made to  e x p l a i n  the  c ause s  o f  the  
observed  changes  i n  some the p o p u l a t i o n  pa r amete r s  by r e l a t i n g  
them to v a r i o u s  c l im a t i c  f a c t o r s  tha t  were  measured  d u r i n g  the  
study p e r i o d .
6.2.  MATERIALS AND METHODS
6 .2 .1 .  F i e l d  S tud i e s
The f i e l d  s t u d i e s  c o n s i s t e d  o f  r e g u l a r  monthly  s amp l ing  
of the f l y  p o p u l a t i o n  u s i n g  NG2B and b i c o n i c a l  t r a p s ,  s e t  i n  
de f i n ed  p o s i t i o n s  w i t h i n  the  s tudy  a r e a .  Sampl ing  was c a r r i e d
out du r ing  the f i r s t  week o f  e ve ry  month. For  e f f e c t i v e
■9
sampl ing ,  the t r a p s  were d i s t r i b u t e d  ove r  as  much o f  the  a r e a  
as p o s s i b l e  so as to  i n c l u d e  each o f  the main v e g e t a t i o n  
t ypes .  F i g u r e  6.1 i s  a map o f  the  main s tudy  a r e a ,  showing  the  
va r i ous  v e g e t a t i o n  t ypes  and the t r a n s e c t s  i n  the main s tudy  
area  where s amp l ing  was c a r r i e d  out .
A s e t  o f  f i v e  b i c o n i c a l  t r a p s  was s e t  i n  the main s tudy  
area  a l ong  t r a n s e c t  1 ( T R l )  which  runs f rom the  Ewaso N g i r o  
r i v e r  in  the e a s t  to  the b a se  o f  the escarpment  i n  the  wes t .
As can be seen i n  F i g u r e  6 . 1 . ,  the  t r a n s e c t  cu t s  a c r o s s  the
University of Ghana          http://ugspace.ug.edu.gh
ijljj-ili- D en se  w o o d l a n d
o
o Q S c a t t e r e d  t r e e s
Scrub m i ’ 08' ' ' ' " ’" ’*^  “  ^ E s c a r pm e n t  b a s s
O r a n l a n d  * ♦ 1“ T r a n s s o t
Figure 6.1: Map of transect 1 area showing the different 
vegetation types and the transect along which 
biconical and Ng2B traps were set (between point 
A and B).
University of Ghana          http://ugspace.ug.edu.gh
165
f ou r  main v e g e t a t i o n  t ype s  i n  the a r e a .  These  a r e  the  open  
p l a i n s  ( PLNS ) ,  the a c a c i a  woodland (ACWD) , a t h i c k  l owe r  
woodland (LOWD) and a more open upper  woodland (UPWD)  
occu r r i n g  i n  t h i s  sequence  f rom e a s t  to  wes t .  One t r a p  was s e t  
i n  each v e g e t a t i o n  t ype  excep t  i n  the l ower  wood l and ,  where  
two t r a p s  were s e t .  Thi s  s e t  o f  t r a p s ,  ( TR l  b i c o n i c a l  t r a p s )  
were s e t  f o r  the l a s t  3 days  o f  each t r i p  and c a t ch es  were  
a l s o  c o l l e c t e d  at  24 -hour  i n t e r v a l s .  Data  f rom th e se  t r a p s  
cover  the p e r i o d  o f  August  1985 to December 1987.
Sampl ing  w i th  the NG2B t r a p s  began i n  June 1986. F i v e  
NG2B t r a p s  were s e t  a l t e r n a t i n g  w i t h  the  TRl  b i c o n i c a l  t r a p s  
a l ong  the t r a n s e c t ,  such tha t  one NG2B t r a p  was a l s o  l o c a t e d  
in  each o f  the  d i f f e r e n t  v e g e t a t i o n  t ypes  ( two  i n  the  l owe r  
wood l and ) .  Sampl ing  w i th  the NG2B t r a p s  on l y  took p l a c e  f o r  
the l a s t  t h r e e  days  of  the monthly  t r i p .  S i nce  th e se  t r a p s  
were used on l y  f o r  s amp l ing  G. longipenn is , c o l l e c t i n g  c age s  
were put on at  about  1815 h and c a t che s  c o l l e c t e d  a t  about  
0715 h the f o l l o w i n g  morning .  Th i s  t ime p e r i o d  l im i t  i n c l u d e d  
the a c t i v i t y  p e r i o d  o f  G. 1ongipennis  but  was o u t s i d e  the  most  
a c t i v e  p e r i o d s  of  G. pallidipes  to  a v o i d  o ve r c r owd i ng  o f  f l i e s  
in c age s .
A s e t  o f  f i v e  b i c o n i c a l  t r a p s  were  a l s o  s i t e d  about  1 km 
apar t  a l ong  the bed o f  the O l o i b o t o t o  r i v e r  which  runs  from  
north to  s outh through  the t h i c k e s t  s e c t i o n  o f  wood l and .  Th i s  
set  of  t r ap s  ( h e n c e f o r t h  termed the r i v e r  t r a p s )  was a lways  
put out on the f i r s t  day o f  each s i x - d a y  monthly  t r i p  and 
catches  c o l l e c t e d  every  24 hr s  f o r  6 days .  Sampl ing  w i t h  the
University of Ghana          http://ugspace.ug.edu.gh
166
r i v e r  t r a p s  was c a r r i e d  out f rom August  1985 t i l l  December  
1987 .
In August  1986, anothe r  t r a n s e c t  ( t r a n s e c t  4)  was  
e s t a b l i s h e d  i n  an a r e a  about  7 km to the n o r t h  o f  the  main  
study a r e a .  A s im i l a r  s ampl ing  programme was s t a r t e d  i n  t h i s  
new a rea  w i t h  f i v e  b a i t e d  b i c o n i c a l  t r a p s  and f i v e  b a i t e d  NG2B 
t r ap s  s e t  a l on g  the t r a n s e c t  and t h r e e  b i c o n i c a l  t r a p s  s e t  
along  the r i v e r .  In the dry  s ea s on ,  the main conn ec t i on  f o r  
p o s s i b l e  f l y  movement between the  two a r e a s  i s  ma i n t a i n ed  by a 
narrow s t r i p  o f  v e g e t a t i o n  a l on g  the O l o i b o t o t o  r i v e r  bed  
which runs through bo th  a r e a s .  Du r ing  the  r a i n y  s e a s on ,  the  
f l i e s  s p r ea d  out more so the chance o f  movement o f  f l i e s  
between the two a r e a s  i s  h i g h e r .  F i g u r e  6.2 shows b o th  the
main s tudy a r e a  and the t r a n s e c t  4 a r e a  and the  l o c a t i o n  o f
the r i v e r  b i c o n i c a l  t r a p s  i n  bo th  a r e a s  (R1-R5 b e l o n g  to  
t r ans ec t  1 and R6-R8 b e l on g  to t r a n s e c t  4 ) .
A l l  t r a p s  were b a i t e d  w i t h  cow u r i n e  d i s p en s e d  a t  c.  
lOOOmg/h and acetone  at  c.  2500mg/h. These were  d i s p en s e d  
from g l a s s  j a r s  w i th  a p e r t u r e  d i ame te r s  o f  4 .5  cm and 2.2 cm 
r e s p e c t i v e l y -  The cow u r i n e  was aged  f o r  about  3 weeks i n
s toppe red  b o t t l e s  at  ambient  t empe r a t u r e s ,  p r i o r  t o  u se .  When
set  in  o p e r a t i o n ,  a l l  odour  b a i t s  were s h e l t e r e d  f rom sun and
ra in  by cover s  as d e s c r i b e d  e a r l i e r  i n  Chapte r  4.
Be f o r e  the c o l l e c t i n g  cage s  were put i n  p l a c e ,  a c a r e f u l  
check f o r  ho l e s  i n  the cage  was c a r r i e d  ou t .  A p i e c e  o f  wet  
cotton wool was then put in each cage  to  i n c r e a s e  humid i t y  to
improve the s u r v i v a l  o f  f l i e s  i n  the c age s  b e f o r e  c a t ch e s  were
University of Ghana          http://ugspace.ug.edu.gh
Figure 6.2: Map showing both transect 1 and 4 areas and the 
location of the River biconical traps along the 
Oloibototo river in both areas.
►
2
University of Ghana          http://ugspace.ug.edu.gh
168
c o l l e c t e d .  When the c a t ches  were b e i n g  c o l l e c t e d  the  c age s  
were a g a in  i n s p e c t e d  f o r  h o l e s  tha t  cou ld  have l e t  f l i e s  
escape .  On c o l l e c t i o n ,  each cage  o f  f l i e s  was l a b e l l e d  and 
then immedi a te l y  put under  the cover  of  a wet b l a c k  c l o t h .  
This  minimized  f l y  movement i n  the cages  and p r e v en t e d  
a d d i t i o n a l  wear  and t e a r  on the wings  tha t  would  g i v e  r i s e  to  
e r r o r s  i n  wing f r a y  a g ing  o f  f l i e s .
A f t e r  a r r i v a l  at  the camp s i t e ,  the f l i e s  were  k i l l e d  
with  ch l o ro f o rm  and then s exed  and counted .  A l l  f ema l e  f l i e s  
from the NG2B t r a p s  were s e a l e d  i n  po l y thene  s a c h e t s  and 
immediate ly  p r e s e r v e d  i n  l i q u i d  n i t r o g e n  a c c o r d i n g  to  the  
p r e s e r v a t i o n  method recommended by M i n t e r  and Goedb l oed  
( 1971 ) .  These were l a t e r  taken to  the l a b o r a t o r y  i n  N a i r o b i  
f o r  o v a r i a n  and w i n g - f r a y  a g i n g .
6 . 2 . 2 .  L a b o r a t o r y  S t ud i e s
Female f l i e s  were  d i s s e c t e d  f o r  o v a r i a n  a ge .  When taken  
out from l i q u i d  n i t r o g e n ,  f l i e s  were immed i a t e l y  t r a n s f e r r e d  
to the f r e e z i n g  compartment o f  a r e f r i g e r a t o r  to  p r ev en t  a 
sudden r i s e  i n  t empe ra tu r e  which no rma l l y  r e s u l t s  i n  the  
d i s i n t e g r a t i o n  of  f l y  t i s s u e s .  A f t e r  about  30 m i nu t e s ,  when 
the specimens were s u f f i c i e n t l y  thawed,  they were  removed one 
at a t ime and d i s s e c t e d  under  a W i l d  d i s s e c t i o n  m i c r o sc op e .
P r i o r  to d i s s e c t i o n ,  the wing on the r i g h t  s i d e  o f  each  
f l y  was t r a n s f e r r e d  on to a drop  o f  g l y c e r i n e  on a m ic r o sc ope  
s l i d e .  Sets  o f  14 wings  were a r r a n g e d  on the s l i d e  and then  
p r e s sed  between two s l i d e s  h e l d  t o g e t h e r  w i t h  c e l l o t a p e .  The
University of Ghana          http://ugspace.ug.edu.gh
169
wings ,  thus  c o l l e c t e d ,  were l a t e r  on used  to  de te rmine  the  age  
of f l i e s  a c c o r d i n g  to the wing f r a y  method p r opose d  by Jackson  
( 1946 ) .  By t h i s  method,  f l i e s  were a s c r i b e d  to  age  c l a s s e s  1 
to 6, whereby 1 was f o r  the younges t  f l i e s  and 6 f o r  the  
o ld e s t  f l i e s  b a sed  on the d e g r e e  o f  wear  on the  t r a i l i n g  wing  
marg in .
The o v a r i e s  o f  each f l y  were  d i s s e c t e d  out to  de t e rm ine  
the o va r i a n  age  a c c o r d i n g  to the method p roposed  by Saunder s  
( 1960)  and improved by C h a l l i e r  ( 1 9 6 5 ) .  The method depends  on 
the c y c l i c a l  development  of  t s e t s e  o v a r i o l e s  which e n a b l e s  
e i gh t  o v a r i a n  age  c a t e g o r i e s  to  be r e c o g n i z e d  on d i s s e c t i o n .  
This  i s  b a sed  on the r e l a t i v e  p o s i t i o n s  o f  the  s e q u e n t i a l l y  
de ve l o p i ng  f o l l i c l e s  and the p r e s en ce  o r  a bsence  o f  f o l l i c u l a r  
r e l i c s  a t t a c he d  to the d e v e l o p i n g  e g g s .  F l i e s  i n  the f i r s t  
f our  c a t e g o r i e s  ( d e s i g n a t e d  0, 1, 2 and 3)  can be aged  
a c c u r a t e l y .  In the se  c a t e g o r i e s  the  l a r g e s t  d e v e l o p i n g  egg  i s  
found i n  one o f  f o u r  d i f f e r e n t  p o s i t i o n s  w i t h  no f o l l i c u l a r  
r e l i c  a t t a c he d  to  i t .  Ca t e go ry  0 f l i e s  a r e  n u l l i p a r o u s  ( no t  
o vu l a t ed )  and i n c l u d e  two s u b c a t e g o r i e s  OA and OB. F l i e s  i n  
subgroup OA have not had a b l o o d  meal and no egg deve lopment  
i s  seen i n  the f i r s t  f o l l i c l e  i n  the  s equence .  F l i e s  o f  t he  OB 
subgroup have had a b l o od  meal and the f i r s t  egg i s  i n  some 
s t ag e  of  deve lopment .  The l a s t  f o u r  c a t e g o r i e s  ( d e s i g n a t e d  4 + 
4n, 5 + 4n, 6 + 4n and 7 + 4n a r e  compos i t e  c a t e g o r i e s  f rom 
which i t  i s  not p o s s i b l e  to  de te rmine  whether  a f l y  i s  i n  i t s  
4th,  8th,  12th . . . ,  5th,  9th,  13th . . . ;  e t c .  o v a r i a n  c y c l e .  
This  i s  b ecause ,  u n l i k e  i n  mo squ i t o e s ,  on l y  one r e l i c  i s
University of Ghana          http://ugspace.ug.edu.gh
170
normal l y  found a t t a c he d  to  the  next  d e v e l o p i n g  f o l l i c l e  
a l though  th e r e  might  have been s e v e r a l  o v u l a t i o n s  from tha t  
o v a r i o l e .
6 . 2 . 3 .  A n a l y s i s  of  Data
6.2.3.1. Fly numbers  and s ex  ratio
The f l y  numbers f rom each monthly  s amp l i ng  were  p o o l ed  
f o r  each s e t  o f  s amp l ing  t r a p s  and the appa r en t  p o p u l a t i o n  
dens i t y  e s t ima ted  by c a l c u l a t i n g  the a r i t hm e t i c  mean c a t ch  pe r  
t rap  per  day.  Th i s  g i v e s  an unb i a s e d  e s t ima t e  o f  the  
popu l a t i o n  s i z e ;  s t a nd a r d  e r r o r s  can be o b t a i n e d  by t a k i n g  {N 
(W i l l i am s  et  a l . ,  i n  p r e s s ) .  These  were  f i r s t  p l o t t e d  to  show 
the monthly changes  i n  appa r en t  d e n s i t i e s .  To show major  
t rends  i n  r e l a t i v e  p o p u l a t i o n  d e n s i t i e s  over  t ime ,  male and 
female c a t ches  were then p l o t t e d  on a l o g  s c a l e  and the cu rve  
smoothed by t a k i n g  t h r e e  monthly  running  means.
To de te rmine  the monthly  changes  i n  f l y  d i s t r i b u t i o n  
wi th in  the d i f f e r e n t  v e g e t a t i o n  t y p e s ,  the  p e r c e n t a g e  c a t ch es  
by t r a p s  w i t h i n  the v e g e t a t i o n  were  c a l c u l a t e d .  A ve r ag e s  were  
taken o f  the  pe r c en ta g e s  o f  the  b i c o n i c a l  and the  NG2B t r a p s  
that  were s i t e d  i n  the same v e g e t a t i o n  t ype .
The monthly p e r c en ta g e  f ema l e  c a t ch es  were a l s o  
c a l c u l a t e d  f o r  the d i f f e r e n t  s e t s  o f  s amp l ing  t r a p s .  These  
data over  the whole  s amp l ing  p e r i o d  were  then s u b j e c t e d  to  an 
a n a l y s i s  o f  v a r i a n c e  f o l l o w i n g  an a r c s i n e  p t r a n s f o rm a t i o n .  
This  was to de te rmine  i f  t h e r e  were  any s i g n i f i c a n t
University of Ghana          http://ugspace.ug.edu.gh
171
d i f f e r e n c e s  i n  sex  r a t i o  between t r a p  t yp e s  and between  
months .
6.2.3.2. Estimation of mortality rates  f rom o v a r i a n  age  
d i s t r i b u t i o n
The monthly o v a r i a n  age d i s t r i b u t i o n s  were  used  to  
e st imate  m o r t a l i t y  r a t e s .  In a r ev i ew  of  the  v a r i o u s  methods  
used to  e s t ima t e  m o r t a l i t y  r a t e s  f rom t s e t s e  age  d i s t r i b u t i o n  
data ,  Roger s  et  a l . ( 1984 )  p o i n t e d  out th a t  the  age
d i s t r i b u t i o n  can only  r e f l e c t  r e a l  m o r t a l i t y  when the  
popu l a t i o n  i s  s t a b l e .  However ,  they c o n s i d e r e d  th a t  changes  i n  
age s t r u c t u r e  can p r o v i d e  i n f o rm a t i o n  on changes  i n  m o r t a l i t y  
r a t e  from one sample to the n e x t .  But D r a n s f i e l d  et  a l .
( 1986a)  ob se r ve d  tha t  t h i s  i s  not  s t r i c t l y  v a l i d  when the  
adu l t  r e c ru i tment  r a t e  i s  not c ons t an t .  In t h i s  s tudy  two 
methods were used to  e s t ima t e  m o r t a l i t y  r a t e s ;  the  f i r s t  i s  
that  method p roposed  by Roger s  et  a 1. ( 1984 )  and m od i f i e d  by
D r a n s f i e l d  et  a 1. ( 1 9 86 a ) .  The m od i f i e d  method e x c lu de s
ca tego ry  0 ( n u l 1i p a r o u s ) f l i e s  which a r e  not  sampled a t  the  
same e f f i c i e n c y  as the o th e r  age  c a t e g o r i e s .  Data  on l a t e r
University of Ghana          http://ugspace.ug.edu.gh
172
c a t e g o r i e s  on ly  were  t h e r e f o r e  used  and an i n t e r - l a r v a l  p e r i o d  
of 9 days was assumed.  Acc o rd in g  to the method the p r o p o r t i o n  
of pa rous  f l i e s  in Ca t e go ry  1 (PI) i s  g i v e n  by 
PI  = ( l - e - 9ro) 
where 'm' i s  the  m o r t a l i t y  r a t e .
S im i l a r l y ,  f o r  the o ther  age c a t e g o r i e s  
P2 = e~ 9m ( l - e ~  9m)
P3 = e 18 m ( 1 - e ' 9m)
P4+ = e- 27m ( 1 - e - 9m )
P5+ = Re- 3 6m 
P6+ = Re 45m 
P7+ = Re- 5 4m
where
R = ( 1 - e " 9m) / ( l - e -  36m)
A n o n - l i n e a r  l e a s t  s q u a r e s  f i t  method was then used  to  
es t imate  ,ml .
The second method i s  the  Robson-Chapman s u r v i v a l  method  
which was f i r s t  a p p l i e d  on f i s h  by Chapman ( 1965 )  and r e c e n t l y  
a p p l i e d  to  t s e t s e  by D r a n s f i e l d  et  a l . ( 1 9 8 6 a ) .  C a t e go r y  0
f l i e s  were a l s o  exc lu de d  f rom the a n a l y s i s .  The method i s  
based on th r ee  as sumpt ions :
1. S, the p r o p o r t i o n  o f  a g i v e n  a g e - g r o u p  s u r v i v i n g  f rom one  
age - g roup  to the n e x t ,  i s  the  same f o r  each a g e - g r o u p  and  
remains cons t an t  over  t ime.
2. Nx , the  number o f  an ima ls  o f  age  x i n  the p o p u l a t i o n  a t  the  
time of  r e p r o du c t i o n  i s  cons t an t  f rom one p e r i o d  to  the next  
and No i s  the  b i r t h - r a t e ,  ( p e r  9 days  i n  the  c a se  o f  t s e t s e ) .
University of Ghana          http://ugspace.ug.edu.gh
173
3. Sampl ing  f rom the p o p u l a t i o n  i s  random w i t h  r e s p e c t  t o  a g e .
From assumpt i on ( 2 )  the  number o f  an ima l s  N* g r ow in g  out  
of  age x w i l l  be b a l an c ed  by the number S N ( x - i )  g r ow ing  i n t o  
t h i s  a g e - g r o u p .
Hence
Nx = SNx- i  = SxNo ( 1 )
and
N = ^N x  = No ( 1+S+S2 . . . ) ( 2 )
S ince  the sum o f  the i n f i n i t e  s e r i e s  i n  b r a c k e t s  = 1 / ( 1 - S )
N= No/ ( 1 - S ) ; ( 3 )
The r e f o r e  s u b s t i t u t i n g  f o r  No i n  ( 1 ) ,
N* = ( 1 - S )  S* N ( 4 )
and the p r o p o r t i o n  i n  each age  g r oup  i s  g i v e n  by
Nx/N = ( 1 - S )  Sx ( 5 )
Based on assumpt i on ( 1 )  Robson and Chapman showed t h a t  a
minimum-var iance  unb i a s e d  e s t ima t e  o f  S i s  g i v e n  by
S = X/ ( n +X - 1 ) ( 6 )
where n = ( £nx ) i s  a random sample  i n  whi ch  nx a r e  o b s e r v e d  to
be of  age x; ( x  = 0 , 1 ,  2 . . . r )  and
X = t x n x ( 7 )
University of Ghana          http://ugspace.ug.edu.gh
174
When o l d e r  a g e - g r o u p s  a r e  b e i n g  p o o l ed  the f o rmu l a  i s  
mod i f i e d .  Suppos ing  tha t  a l l  an ima ls  y + 1 y e a r s  o l d  and above  
are  p o o l e d . L e t
ny =^ _nx ( 8 )
X = 2,xn* + ( y  + 1)  ny ( 9 )
then the maximum-1i k e l i h o o d  e s t ima t e  of  S i s  now
Spool = X/ (n -  ny + X) ( 10 )
In t h i s  s tudy  the a g e - g r o u p s  po o l ed  were  4+ to 7+.  
Assuming a 9 day i n t e r - l a r v a l  p e r i o d ,  SPo0 l the  s u r v i v a l  r a t e  
i s  equa 1 to e" 9,n so
m = - In (Spool ) / 9 .  (11)
6.2.3.3. Estimation of mortality rates  f rom Moran  cu r ve s  
Rogers  ( 1979)  p roposed  a method f o r  e s t im a t i n g  the  
o v e r a l l  d e n s i t y - i n d e p en d en t  m o r t a l i t y  r a t e  o f  the  p o p u l a t i o n  
from the change i n  appar en t  d e n s i t y  f rom one s ampl ing  o c c a s i o n  
to the o ther  u s in g  Moran c u r v e s .  To o b t a i n  the  Moran cu rv e ,  
the l og  d e n s i t i e s  i n  one month were  p l o t t e d  a g a i n s t  t ho se  i n  
the p r e v i o u s  month. A boundary  l i n e  was then f i t t e d  f o r  the  
maximum growth r a t e  o f  the p o p u l a t i o n  which f o r  t s e t s e ,  i n  the  
absence of  any m o r t a l i t i e s ,  would  be d o u b l i n g  per  month.  
Assuming an ex pon en t i a l  p o p u l a t i o n  g rowth ,  the  r e l a t i o n s h i p  
between d e n s i t i e s  i n  suc ce ed in g  months remains  l i n e a r  ( a f t e r  
the l og  t r a n s f o rm a t i o n )  u n t i l  d e n s i t y - d e p e n d en t  m o r t a l i t i e s  
set  i n ;  the curve  then bends t owa rds  an e q u i l i b r i um  l e v e l  
g iven  by the 45° l i n e  th rough  the o r i g i n  where  r a t e  of
University of Ghana          http://ugspace.ug.edu.gh
i n c r e a s e  shou ld  be z ero  and d e n s i t i e s  i n  s u c c e ed i n g  months a r e  
the same. Thus,  i n  any one month the d i f f e r e n c e  between  the  
expect ed  d e n s i t y  and the o b s e rv e d  d e n s i t y  g i v e s  the  d e n s i t y -  
i ndependent  m o r t a l i t y  r a t e  f o r  tha t  month. Th i s  may be  
e s t imated  d i r e c t l y  f rom the p l o t ,  as  the  d i s t a n c e  o f  the  
obse rved  l o g  d e n s i t y  b e l ow  the maximum growth r a t e  cu rv e .
M o r t a l i t i e s  o b t a i n ed  by t h i s  method a r e  the o v e r a l l  o r  
g ene ra t i o n  m o r t a l i t y  r a t e s  as opposed to  those  f rom o v a r i a n  
age dat a  which e s t ima t e  a d u l t  m o r t a l i t y  r a t e s .  Because  o f  the  
popu l a t i o n  s u p p r e s s i o n  e x e r c i s e  tha t  was go in g  on i n  the TRl  
area  at  the  t ime,  on l y  the da t a  from TR4 which was o u t s i d e  the  
s up p r e s s i on  a r e a  were used  f o r  the  a n a l y s i s .
6 .3.  RESULTS
6 .3 .1 .  Apparent  d e n s i t i e s
Monthly  changes  i n  appa r en t  d e n s i t i e s  ( l o g  
c a t ch / t r a p /d ay ) of  G. l o n g i p e n n i s  i n  the b i c o n i c a l  and NG2B 
t r ap s  s e t  a l ong  t r a n s e c t  1 i n  the  main s tudy  a r e a  a r e  shown i n  
F igu re s  6 . 3A and 6.3B.  To b r i n g  out the ma jo r  t r e n d s ,  the  
curves  have been s u b j e c t e d  t o  a t h r e e - p o i n t  smooth ing  and the  
two t r ap  t ypes  super imposed  f o r  males  and f ema l e s .  G e n e r a l l y ,  
both s exe s  showed s im i l a r  t r en d s  f o r  each t r a p  t y p e .  Peak  
catches  f o l l o w i n g  the l ong  r a i n y  s ea son  o f  1986 were  o b se r v e d  
in  June/July i n  both  the b i c o n i c a l  and the  NG2B t r a p s .  Peak  
catches  du r ing  the sh o r t  r a i n s  were  r e c o r de d  i n  
November/December w i t h  both  t r a p  t yp e s  but  the  b i c o n i c a l  t r a p s  
showed peak c a t ches  which were  about  3X l ower  than the peaks
University of Ghana          http://ugspace.ug.edu.gh
Month
Figure
Month
6 '3 »?i°“X 5y C w 9es 111 the apparent densities of 
male (A) and female (B) G. longipennis in the
Tnmsect ) 311(1 NG2B (* x) traps set
University of Ghana          http://ugspace.ug.edu.gh
;ca t cher i ^ cor d rl aftoj M»* 1 nnq rp ins whilst the NG2B recorded
similai nppat. r,nt densil irtis in both seanonr. The lowest catches 
were t f c  ■ f -1 rcl in Fobr up r y/Ma r ch / Ap r i 1 in both trap types. In 
1987 a I ewer peak was recorded at the end of the long rainy 
season than during the short rainy season with the biconical 
tr.ips whilst it was the reverse with the NG 2 B traps.
Fin'thei moire, these peaks were recorded about a month earlier 
with the biconical traps in both seasons.
The monthly catches in the river biconical traps are 
given in Figure 6 .4A and 6 . 4B. for males and females 
respectively. To better compare trends in these traps with 
those set along the transect, the plots of catches by the 
transect biconical traps are superimposed on those of the 
river biconical traps. Generally, the highest catches in the 
river biconical traps were observed in the hottest and driest 
periods of the year and rainy season catches were rather low. 
For example, marked peaks were observed in August-October in 
the river biconical traps when catches in the traps along the 
transect were <«n the decline. These trends were similar for 
both sexes.
Chanqes in the apparent densities recorded by the traps 
set along the transect 4 are shown in Figures 6 .5A and 6 .5B. 
Generally, peat, catches here were also observed towards the 
end of \ he dry season and that of the rainy season but the 
peaks on r’R4 preceded those of TRl by one month. For example, 
the dry season peaks in 198f» and 1087 occurred in September- 
Oclober on TR4 but in November-December on TRl . JJ^^s^iny
1.77
University of Ghana          http://ugspace.ug.edu.gh
Lo
g(
ca
tc
h/
tra
p/
da
y)
 
Lo
g(
ca
tc
h/
tra
p/
da
y)
± / o
Month
Figure 6.4: Monthly changes in the apparent densities of 
male (A ) and female (B ) G. longipennis in the
transect biconical (_____) and the river
biconical traps (*..*) of transect 1 area.
University of Ghana          http://ugspace.ug.edu.gh
Lo
g(
ca
tc
h/
tra
p/
da
y)
 
Lo
g(
ca
tc
h/
tra
p/
da
y)
179
Month
Month
Figure 6.5: Monthly changes in the apparent densities of 
male (A) and female (B) G. longipennis in the
biconical (--- ) and NG2B (*..*) traps set along
Transect 4.
University of Ghana          http://ugspace.ug.edu.gh
Jseason peak occurred in May-June on TR4 but it was observed in 
June-July on TRl. Whereas very low catches were recorded in 
February-March on TRl ,  catches were on the rise on TR4 i n  
these months, this difference being more pronounced between 
the biconical traps of the two transects.
Catches  i n  the r i v e r  b i c o n i c a l  t r a p s  o f  TR4 a r e  g i v e n  i n  
F i gu r e  6.6 (ma le s  and f ema l e s  t o g e t h e r ) .  The t r e n d s  i n  t h e s e  
appear  to  r e f l e c t  a s p e c t s  o f  the  t r end s  i n  bo th  TRl  and TR4.  
I n c r e a s e s  i n  the se  t r a p s  s t a r t e d  e a r l i e r  than they d i d  on TRl  
but at  the same t ime as on TR4. However ,  t h e se  t r a p s  showed  
broad  peaks  tha t  cove r ed  the months i n  which peak c a t che s  
occur red  on both  t r a n s e c t s .  The low ca t che s  o b se r v e d  i n  
February  and March w i th  the se  t r a p s  a r e  c h a r a c t e r i s t i c  o f  
t rends  i n  TRl  t r a p s .  Th i s  i s  b ecau se  two o f  the  r i v e r  t r a p s  
a s c r i b e d  to  TR4 a r e  a c t u a l l y  s i t e d  i n  a s t r e t c h  o f  v e g e t a t i o n  
between the c e n t r a l  p a r t  o f  TR4 and TRl .  Thus ,  the  
i n t e rmed i a r y  p o s i t i o n  o f  the  t r en d s  shown by th e se  t r a p s  may 
be r e f l e c t i n g  movement o f  f l i e s  between the two t r a n s e c t s . ( s e e  
D i s c u s s i o n ) .
When the monthly mean ca t ch  pe r  t r a p  pe r  day i n  the NG2B 
and b i c o n i c a l  t r a p s  a r e  compared,  a number o f  o b s e r v a t i o n s  can 
be made w i th  r e g a r d s  to  the s amp l ing  e f f i c i e n c i e s  o f  the  two 
t r ap  t ype s .  G e n e r a l l y ,  c a t che s  i n  the NG2B t r a p s  were  h i g h e r  
than those  i n  the b i c o n i c a l  t r a p s .  On TRl  f o r  i n s t a n c e ,  the  
ave rage  mean c a t ch e s / t r a p /d ay  + 1 s . e .  o ver  the  s amp l ing
180
University of Ghana          http://ugspace.ug.edu.gh
Lo
g(
ca
tc
h/
tra
p/
da
y)
181
Figure 6.6: Monthly changes in the apparent densities of
male (_____) and female (#..)() G. longipennis in
the river biconical traps of transect 4 area.
University of Ghana          http://ugspace.ug.edu.gh
1 8 2
p e r i o d  were 5.3 + 0 .3  males  and 2.3 ±  0 .06  f ema l e s  i n  the NG2B 
t r ap s  compared to 1.2 ±  0 .03 males  and 0 .6  ±  0 .02  f ema l e s  i n  
the b i c o n i c a l  t r a p s .
The monthly mean p e r c e n t a g e s  o f  f ema l e s  caught  i n  the  
t r a n s e c t  1 b i c o n i c a l  and NG2B t r a p s  a r e  shown in  F i g u r e  6.7  
and 6.8 r e s p e c t i v e l y .  G e n e r a l l y ,  the  p r o p o r t i o n  o f  f ema l e s  i n  
the c at ch  seemed to f o l l o w  the same t r end s  as  the ap pa r en t  
d e n s i t i e s  w i th  peaks t owa rds  the end of  the  dry  and the r a i n y  
s ea sons .  The h i g h e s t  p e r c en ta g e  o f  f ema le s  was r e c o r d e d  i n  
July  and the l owes t  at  the  b e g i nn i n g  o f  the  dry  s ea s on  i n  
August and a t  the b e g i nn i n g  of  the r a i n y  s ea s on  i n  A p r i l .  The 
t rends  were s im i l a r  i n  bo th  the b i c o n i c a l  t r a p s  and the  NG2B 
t r ap s  .
The o v e r a l l  mean f ema le  p e r c e n t a g e s  were  46.7 ±  3.4% f o r  
the NG2B t r a p s ,  27.1% ±  2 .3  f o r  the TRl b i c o n i c a l  t r a p s  and 
31.5% ±  2.5 f o r  TRl  R i v e r  t r a p s  . An a n a l y s i s  of  v a r i a n c e  
per formed on the monthly  f ema l e  p e r c e n t a g e s  ( f o l l o w i n g  an 
a r c s i n e  {P t r a n s f o rm a t i o n )  showed tha t  t h e r e  were s i g n i f i c a n t  
d i f f e r e n c e s  between f ema le  p e r c e n t a g e s  i n  the t h r e e  t r a p  s e t s .  
Comparing the means u s i n g  the Duncan ’ s M u l t i p l e  Range T e s t ,  
the mean pe r c en ta ge  o f  f ema l e s  caught  i n  the s e t  o f  NG2B t r a p s  
was s i g n i f i c a n t l y  h i g h e r  than tho se  i n  the t r a n s e c t  and r i v e r  
b i c o n i c a l  t r ap s  but  those  f rom the l a t t e r  two s e t s  o f  t r a p s  
did  not d i f f e r  s i g n i f i c a n t l y .
University of Ghana          http://ugspace.ug.edu.gh
Pe
rc
en
ta
ge
 
fem
ale
s 
Pe
rc
en
ta
ge
 
fe
m
al
es
183
Figure 6.7: Monthly changes in the percentage of females 
caught by the biconical traps.
\<\ S G  Month ^ * § 1
Figure 6.8: Monthly changes in the mean percentage of 
females caught by the NG2B traps.
University of Ghana          http://ugspace.ug.edu.gh
184
6 . 3 . 2 .  F l y  d i s t r i b u t i o n  i n  d i f f e r e n t  v e g e t a t i o n  t y p e s .
Changes i n  the r e l a t i v e  d i s t r i b u t i o n  o f  G. 1ongipennis  i n  
the d i f f e r e n t  v e g e t a t i o n  t ype s  i n  the TRl a r e a  a r e  shown i n  
F i g s .  6 . 9A and 6 . 9B f o r  males  and f ema l e s  r e s p e c t i v e l y .  These  
f i g u r e s  were produced  f rom the  a v e r a g e s  taken  o f  the  da t a  
grouped i n t o  s e t s  o f  t h r ee  months to r ou gh l y  c o r r e s p ond  w i t h  
the d i f f e r e n t  annual  s e a sons  i n  the s tudy  a r e a :  J u l y - S e p t embe r  
( c o l d  dry  s e a s o n ) ,  Oc tobe r -Decembe r  ( s h o r t  r a i n s ) ,  January  to  
March ( ho t  dry  s ea s on )  and A p r i l  to  June ( l o n g  r a i n s ) .
G e n e r a l l y ,  the  h i g h e s t  p e r c e n t a g e  c a t ch es  were  r e c o r d e d  i n  the
lower  and upper  woodlands  and the  l owes t  i n  the  open p l a i n s .
In the c o o l e r  wet s ea sons  a more even d i s t r i b u t i o n  was 
obse rved  as f l i e s  t ended  to s p r e a d  i n t o  a l l  v e g e t a t i o n  t y p e s .  
In the dry  s e a s ons ,  however ,  f l i e s  became c on c en t r a t e d  i n  the  
dense v e g e t a t i o n  ( r i v e r i n e  t h i c k e t  and l owe r  w o od l and ) .  Both  
sexes  f o l l o w e d  s im i l a r  t r en ds  o f  movement i n  r e l a t i o n  to  
s easons .  I t  i s  c l e a r  f rom the  f i g u r e s  th a t  f l i e s  c o n c en t r a t e d  
more i n  the r i v e r i n e  t h i c k e t s  i n  the hot  d ry  s ea s on  ( J a n u a r y -  
March)  o f  1986 than they d i d  i n  the same s ea s on  o f  1987.  Th i s  
i s  p r o b a b l y  b ecause  i t  was c o o l e r  d u r i n g  t h i s  s e a s on  i n  1986
than i n  1987, the a ve r ag e  maximum t empe ra t u r e  f o r  th e se  3
months i n  1986 be in g  37 .9 °C  as a g a i n s t  36 .3 °C  f o r  1987.
The r e l a t i v e  d i s t r i b u t i o n s  i n  the R i v e r  t r a p s  a r e  shown 
i n F i g .  6.10A f o r  males  and F i g .  6.10B f o r  f em a l e s .  The da t a  
were a g a i n  g r ouped  i n t o  s ea sons  as d e s c r i b e d  above .  I t  shou l d  
be po in t e d  out tha t  a l th ou gh  a l l  the R i v e r  t r a p s  were
University of Ghana          http://ugspace.ug.edu.gh
185
AUGUST 19 86-MARCH 1987
EZZZplns KX3 a c w d E iil riv C ZD low d  ^ u p w o
AUGUST 1986-MARCH 1987 
E Z 3  PLNS K X 3  acwd 1 = 3  riv CZD lowo ES53  up w d
Figure 6.9: Monthly changes in the relative distribution of 
male (A) and female (B) G. longipennis in 
different vegetation types along transect 1 .
University of Ghana          http://ugspace.ug.edu.gh
j.y b
MONTHS
R 1 * R ?  ( D e n s e )  E 3  R 3 ( l e s s  d e n s e )  I \ R * » » R 5 ( L e a s t  d e n s e )
M O N T H S
R 1 ♦ R? ( D e n s e )  E 3  R 3 { l e s s  d e n s e ) C D  R<* ♦ R5 ( l e a s t  d e n s e )
Figure 6.10: Monthly changes in the relative distribution of 
male (A) and female (B) G. longipennis in 
different vegetation types along the river bed of 
transect 1 area.
University of Ghana          http://ugspace.ug.edu.gh
187
g e n e r a l l y  l o c a t e d  i n  the t h i c k e s t  s e c t i o n  of  the  h a b i t a t  
( a l o n g  the r i v e r  b e d ) ,  t h e r e  were l o c a l  d i f f e r e n c e s  be tween  
the t r a p  s i t e s  w i t h  r e s p e c t  t o  v e g e t a t i o n  c o v e r .  Tr aps  RI  and  
R2 were l o c a t e d  in r e l a t i v e l y  dense  wood l and  to the s ou th  o f  
R3 which i t s e l f  was i n  a l e s s  dense  s i t e ,  w h i l s t  R4 and R5 
were i n  more open s i t e s  to  the no r t h  o f  R3 .
The r e l a t i v e  p e r c e n t a g e  c a t ch es  i n  t h e s e  t r a p s  show th a t  
f l i e s  a l s o  showed p r e f e r e n c e  f o r  d i f f e r e n t  v e g e t a t i o n  co ve r  i n  
d i f f e r e n t  s ea sons  i n  a manner s im i l a r  to  th a t  a l on g  the  
t r a n s e c t .  G e n e r a l l y ,  f l i e s  were abundant  i n  the more open  
woodland (R4 and R5)  i n  the coo l  d ry  s e a son  and a p p a r e n t l y  
moving i n t o  the dense r  v e g e t a t i o n  i n  the hot  d ry  and becoming  
more even l y  d i s t r i b u t e d  i n  the r a i n y  s e a s on s .  The l e s s e r  
c on c en t r a t i o n  i n  the dens e r  woodland i n  the  January -March  o f  
1987 compared to  1986 can a l s o  be ob se r v ed  he r e  f o r  the  r e a son  
s ug ge s t e d  above .  These t r end s  were  s im i l a r  f o r  both  mal es  and  
f em a l e s .
F i g u r e s  6.11A and 6.11B show the monthly  p e r c e n t a g e  
d i s t r i b u t i o n  o f  c a t ches  i n  r i v e r  t r a p s  o f  TR4 f o r  mal es  and 
females  r e s p e c t i v e l y .  The f i g u r e s  do s u g g e s t  a s e a s o n a l  
movement o f  f l i e s  between R6 ( n e a r e s t  t r a p  to TR l )  and R8 ( i n  
the c en t r e  of  TR4) .  The c a t che s  i n  R8 r o s e  s t e a d i l y  as  the  dry  
season p r o g r e s s e d  from Ju ly  to Oc tobe r  1986 and d e c r e a s e d  
du r ing  the sh o r t  r a i n s  i n  Novembe r -Decembe r . A s t e a d y  i n c r e a s e  , 
i s  a g a in  ob se r ve d  i n  R8 d u r i n g  the hot d ry  s ea s on  f rom January  | 
to  March,  f o l l o w e d  by a g r a du a l  d e c l i n e  d u r i n g  the  l ong  r a i n s .  
Thi s  p a t t e r n  of  f l y  movement between TRl  ^nd TR4 deduced f rom
University of Ghana          http://ugspace.ug.edu.gh
P
E
R
C
E
N
T
A
G
E
 
P
E
R
C
E
N
T
A
G
E
188
KX2 R6 I 1 R 7 EH3 R 8
MONTHS
633 R6 □  R7 m  R 8
Figure 6.11: Monthly changes in the relative distribution of 
male (A ) and female (B ) G. longipennis in 
the river traps of transect 4 area.
University of Ghana          http://ugspace.ug.edu.gh
189
the r e l a t i v e  p e r c en t a g e  c a t che s  i n  t r a p s  R6, R7 and R8 c ou ld  
a l s o  r e s u l t  f rom the s e a r c h  f o r  d i f f e r e n t  v e g e t a t i o n  co ve r  i n  
the two a r e a s  at  d i f f e r e n t  t imes  o f  the  y e a r .
6 . 3 . 3 .  Age s t r u c t u r e
The age d i s t r i b u t i o n  o f  f ema l e  G. longipennis  c augh t  by  
the NG2B t r a p s  a r e  shown in  F i g u r e s  6.12 and 6.13 f o r  
t r a n s e c t s  1 and 4 r e s p e c t i v e l y .  From June 1986 to  January  
1987, when the p o p u l a t i o n  was not  b e i n g  s u p p r e s s e d  ( s e e  
chapte r  9)  t h e r e  was a r e l a t i v e l y  low p e r c e n t a g e  o f  c a t e g o r y  0 
f l i e s .  Th i s  o b s e r v a t i o n  i s  t r u e  f o r  bo th  t r a n s e c t s  e x cep t  when 
the sample s i z e  was v e ry  sma l l  as i n  the  c a se  o f  Oc to b e r  1986 
( on TR1) and January 1987 ( on  TR4) .  As a l r e a d y  r e p o r t e d  
e a r l i e r ,  the  NG2B t r a p  shows a b i a s  f o r  o l d e r  f l i e s  when 
compared to  the b i c o n i c a l  t r a p  ( s e e  c h ap t e r  4 ) .
Soon a f t e r  the  s t a r t  o f  the  s u p p r e s s i o n  programme,  t h e r e  
was a c o n s i d e r a b l e  i n c r e a s e  i n  the p r o p o r t i o n  o f  young f l i e s  
e s p e c i a l l y  tho se  b e l o n g i n g  to  c a t e g o r y  O and a c o r r e s p on d i n g  
r educ t i on  i n  the p r o p o r t i o n  o f  o l d e r  f l i e s .  A l t ho ugh  the  
s u p p r e s s i o n  e x e r c i s e  was b e i n g  c a r r i e d  out i n  the  t r a n s e c t  1 
( w i t h  b a r r i e r  t r a p s  between TR1 and TR4) the  above  change i n  
age compos i t ion  was a l s o  n o t i c e a b l e  i n  the  t r a n s e c t  4 s amp l e s .  
Thi s  im p l i e s  tha t  the  p o p u l a t i o n  o f  f l i e s  i n  the l a t t e r  a r e a  
was a l s o  b e in g  a f f e c t e d  by the s u p p r e s s i o n  o p e r a t i o n ,  whi ch  i s  
expect ed  i f  t h e r e  i s  c o n s i d e r a b l e  movement o f  f l i e s  between
University of Ghana          http://ugspace.ug.edu.gh
JUN 1986
190 
n a  s f t
l b .
JU L .1986
AUG.1986
n=  52
T l > R - r n
n =  2 «
SEP 1986
n r  53
OCT. 1986
n =  &
n h
NOV. 1986 ^
n r  26
J L J J
DEC.1986
JAN. 1986
n =  3 5
d C L
r\z 21
nz 2tf
T r - n
nr 38
—
n= « 3
1 n  n  _
n= 38
T t - m
r u  92
n =  s i
T - d T k
n * . f c i
EL
—  nz 56
T t t t l
. 4 0
“ 2 0
n
r u  u - i
F E B . 1987
MARCH 1987
APRIL  19 87
MAY 19 87
JUNE  1987
JULY 19 87
AUG 19 87
SEPT.  19 87
OCT.  19 87
0  I  2 3'k' 5 * 6 *  7  * 0  1 2 3  5 ‘ 6 ‘  T
O v a r i a n  a g e
Figure 6.12: The age distribution of female G. longipennis 
caught by the NG2B traps of transect 1 area.
University of Ghana          http://ugspace.ug.edu.gh
SEPT  87
O C T .  8 7
NOV. 87 
DEC.87
JAN.'87
n-5«-
m = L
r u s o
J V v f f K L
M
n=&i
n= 9
h ...
0  1 2 3 •** 5 * 6* 1'
n =22 FEB.  19.87
^ tU T U  ._
n* m- MARCH' 87
1  ] _ □  -
n= ^  APRIL '8 7  
[ t b z ^
n.- 53 MAY ' 87
i T T T ~ L n _
n: 61
 , JUNE ’ 87JVrlTL—
■ n=24 JULY ' 87
i - r H r u ^
n=35 AUG. ' 87jTkfh
n= 33 SEPT ' 87
J T ~ T Ln.
7.4020
2 8
OCT. 87
0  1 2 3 5 * 6* T
Ovarian age
Figure 6.13: The age distribution of female G. longipennis 
caught by the NG2B traps of transect 4 area.
University of Ghana          http://ugspace.ug.edu.gh
192
the two a r e a s .  T h e r e f o r e  f l i e s  m i g r a t i n g  f rom TR4 (wh i ch  a r e  
more l i k e l y  to be o l d  f l i e s )  a r e  not  l i k e l y  to  r e t u r n  as  they  
w i l l  g e t  t r apped  i n  TRl .
However ,  f rom May 1987 up to Ju l y  1987 the  p r o p o r t i o n  o f  
o l d e r  f l i e s  i s  seen to  i n c r e a s e  a g a i n  i n  the main s tudy  a r e a  
which was p a r t l y  due to  heavy immig r a t i on  i n t o  the  s u p p r e s s i o n  
zone and p a r t l y  to the s u p p r e s s i o n  t r a p s  b e i n g  l e s s  e f f e c t i v e .  
The above  p e r i o d  i s  j u s t  a f t e r  the  r a i n s  when t h e r e  i s  
abundant  v e g e t a t i o n  cover  and f l i e s  s p r e a d  out i n t o  more open  
a r e a s  where the d e n s i t y  o f  s u p p r e s s i o n  t r a p s  i s  v e r y  low.  The 
months o f  August  to Oc tobe r  a r e  d r i e r  months when f l i e s  tend  
to w i thdraw  i n t o  the t h i c k e r  wood l ands  and one can s ee  the  
r e du c t i o n  a g a i n  i n  the p r o p o r t i o n  o f  o l d e r  f l i e s  i n  the se  
months as the co n t r o l  t r a p s  become more e f f e c t i v e  and  
immigr a t i on  i s  r educed .  Th i s  l a t t e r  e f f e c t  was not  immed i a t e l y  
n o t i c e a b l e  i n  the T r an se c t  4 where the p r o p o r t i o n  o f  o l d  f l i e s  
remain r e l a t i v e l y  h i gh  u n t i l  Oc to be r  ( s e e  c h ap t e r  9 f o r  more 
d e t a i I s ) .
The o v a r i a n  a g in g  method was compared to  the  wing  f r a y  
method by p l o t t i n g  the o v a r i a n  age a g a i n s t  wing  f r a y .  F i g u r e s  i 
6.14 and 6.15 show the r e l a t i o n s h i p s  between  o v a r i a n  age  and 
the wing f r a y  f o r  samples  f rom TRl and TR4 r e s p e c t i v e l y .  The 
a n a l y s i s  was l im i t e d  to  f l i e s  i n  o v a r i a n  age  c a t e g o r i e s  0 to  
I I I  b ecause  o f  the u n c e r t a i n t y  i n  a g i n g  f l i e s  above  age  g r oup  , 
I I I .  The f o l l o w i n g  e s t im a t e s  o f  the  c a l e n d a r  a ges  ( i n  d a y s )  oi
University of Ghana          http://ugspace.ug.edu.gh
Wi
ng
 
fra
y 
ca
te
go
ry
 
Wi
ng
 
fra
y 
ca
te
go
ry
Age(days)
Figure 6.14: The relationship between ovarian w ing -fray  age  
and ovarian age of female G. longipennis from  
transect 1 area (Y = 1.70 + 0.064X; r  = 0.0844, 
P<0.001).
0 20
Age(days)
Figure 6.15: The relationship between w ing-fray  age
and ovarian age of female G. longipennis from  
transect 4 area (Y = 1.59 + 0.073X; r  = 0.0815, 
PC0.001).
University of Ghana          http://ugspace.ug.edu.gh
194
the v a r i o u s  o v a r i a n  age c a t e g o r i e s  were  used  f o r  the  a n a l y s i s .  
1 day f o r  c a t e g o r y  Oa f l i e s ,  5 days  f o r  Ob, 12.5 f o r  c a t e g o r y  
I ,  21.5 days  f o r  I I  and 30.5 days  f o r  c a t e g o r y  I I I  f l i e s .  The  
r e g r e s s i o n s  o f  wing f r a y  c a t e g o r y  on o v a r i a n  age were  
s i g n i f i c a n t  f o r  samples  f rom bo th  t r a n s e c t s ;
Y = 1 .70+0.064X,  ( r  = 0 . 844 ,  P<0 .001 )  f o r  TRl
Y = 1 . 5 9 + 0 .073X ( r  = 0 . 0 . 8 1 5 ,  P<0 .001 )  f o r  TR4.
An a n a l y s i s  was then c a r r i e d  out to de te rmine  i f  t h e r e  
was any d i f f e r e n c e  i n  the r a t e  o f  wing f r a y  be tween  the f l i e s  
f rom two t r a n s e c t s .  F o l l o w i n g  the method used  by Ryan et  a l . 
( 1980)  and Rogers  ( 1 9 8 4 ) ,  the  s l o p e  o f  a r e g r e s s i o n  o f  wing  
f r a y  on o v a r i a n  age shou ld  g i v e  an e s t im a t e  o f  the  r a t e  of  
wing f r a y  o f  the  p o p u l a t i o n .  The s l o p e s  (± 1 s . e )  o f  the  
r e g r e s s i o n s  l i n e  were 0 .064 ±  0.0055 f o r  TRl and 0 .072 ±
0.0070 f o r  TR4. A t - t e s t  showed tha t  t h e r e  was no s i g n i f i c a n t  
d i f f e r e n c e  between the s l o p e s  and hence the r a t e  o f  wing f r a y  
between f l i e s  f rom the two t r a n s e c t s  ( t  = 0 . 2 3 ,  P > 0 . 0 5 ) .  
Con s i d e r i n g  the ev i denc e  tha t  t h e r e  i s  some movement o f f l i e s  
between the two t r a n s e c t s  the above  f i n d i n g  s u p p o r t s  the  v i ew  
that  the same p o p u l a t i o n  i s  a c t u a l l y  b e i n g  sampled  i n  bo th  
s i t e s .
The dat a  were next  examined to  f i n d  out i f  t h e r e  were  any  
s i g n i f i c a n t  r e l a t i o n s h i p s  between the r a t e  o f  wing f r a y  and 
c l im a t i c  f a c t o r s  ( t empe r a t u r e  and r e l a t i v e  h um i d i t y ) .  Th i s  was  
done by de t e rmin in g  the r a t e s  o f  wing f r a y  f o r  each month and 
then c a r r y i n g  out a c o r r e l a t i o n  a n a l y s i s  o f  t h e se  w i t h  maximum
University of Ghana          http://ugspace.ug.edu.gh
195
t empe ratu re  and minimum r e l a t i v e  humid i t y .  S i nce  s amp l i ng  was  
u s u a l l y  c a r r i e d  out i n  the f i r s t  week o f  e ve ry  month i t  was  
assumed tha t  f l i e s  e x p e r i e n c e d  env i ronmenta l  c o n d i t i o n s  o f  the  
p r e v i o u s  month and of  the month of  s amp l i ng .  T h e r e f o r e  means 
of  the se  monthly t empe ra tu r e s  and r e l a t i v e  h um i d i t i e s  o f  t h e se  
two months were taken f o r  the a n a l y s i s .  However  no s i g n i f i c a n t  
a s s o c i a t i o n  was found between the r a t e  o f  wing f r a y  and t h e s e  
c l im a t i c  f a c t o r s .
6 . 3 . 4 .  M o r t a l i t y  r a t e s  f rom o v a r i a n  age  d i s t r i b u t i o n .
F i g s .  6.16 and 6.17 show the monthly  changes  i n  a d u l t  
m o r t a l i t y  r a t e s  on TRl and TR4 r e s p e c t i v e l y ,  e s t im a t e d  from 
the o v a r i a n  age d i s t r i b u t i o n .  The e s t im a t e s  were  made f rom  
b imonthly  running  sums o f  o v a r i a n  age d i s t r i b u t i o n s  to  a v o i d  
the prob lem of  v e ry  smal l  samples  i n  some months.  Both  
f i g u r e s  i n c l u d e  m o r t a l i t y  r a t e s  e s t ima t ed  by the  two methods  
d e s c r i b e d  i n  s e c t i o n  2 .3 (Roge r s  et  a l  . , 1984 and D r a n s f i e l d  
et a l  . , 1986a ) .  Both methods gave  v e r y  s i m i l a r  v a l u e s  o f  
mo r t a l i t y  r a t e s  r an g i n g  from 0 .0 22 - 0 .0 59  pe r  day.
On TRl the m o r t a l i t y  r a t e s  r o s e  t owa rds  the  end o f  the  
co l d  dry  s ea son from Oc tober  to  a peak i n  F e b r u a r y -Ma rc h  w i t h  
a smal l  drop between December and January .  On TR4 m o r t a l i t y  
r a t e  a l s o  i n c r e a s e d  from Oc tobe r  but  r e ached  a much h i g h e r  
peak i n  December than i t  d i d  on TRl ,  d e c l i n e d  to  i t s  l owes t  
l e v e l  peak i n  Ja nu a r y - F e b r u a r y  and r o s e  a g a i n  to  a sm a l l e r  
peak i n  March.  F o l l o w i n g  the Fe b ru a ry  peak ,  t h e r e  was a 
d e c l i n e  on bo th  t r a n s e c t s  which s t opp ed  i n  A p r i l  on TRl  to  be
University of Ghana          http://ugspace.ug.edu.gh
Ad
ult
 m
ort
ali
ty 
rat
e/
da
y 
Ad
ult
 m
ort
ali
ty 
ra
te
/d
ay
Figure 6.16: Monthly changes in adult mortality rate
estimated from ovarian age structure on transect 1.
F i gur e  6 .17:  Monthly changes in  ad u l t  m o r t a l i t y  r a t e
e s t imated  from ova r i an  age s t r u c t u r e  on t r a n s e c t  4. 
(  method 1; x — xmethod 2 ) .
University of Ghana          http://ugspace.ug.edu.gh
197
f o l l o w e d  by a sha rp  r i s e  u n t i l  Augus t .  On TR4 the  d e c l i n e  
cont inued  u n t i l  May and was then f o l l o w e d  by a r i s e  a t  a l owe r  
l e v e l .  Thi s  d i f f e r e n c e  i n  l e v e l s  o f  m o r t a l i t y  r a t e s  between  
the two t r a n s e c t s  which show up a f t e r  Fe b ru a ry  p r o b a b l y  
r e f l e c t s  an a d d i t i o n a l  m o r t a l i t y  a r i s i n g  f rom the  p o p u l a t i o n  
s u p p r e s s i o n  o p e r a t i o n  ( u s i n g  odour  b a i t e d  t r a p s )  which was  
s t a r t e d  i n  Feb rua ry  1987 i n  the t r a n s e c t  1 a r e a .
Comparing the t r end s  i n  appa r en t  m o r t a l i t i e s  t o  th o se  i n  
apparent  d e n s i t i e s  on the two t r a n s e c t s ,  some g e n e r a l  
o b s e r v a t i o n s  can be made. On bo th  t r a n s e c t s ,  p e r i o d s  o f  low 
m o r t a l i t i e s  seem to be f o l l o w e d  by a r i s e  i n  appar en t  
d e n s i t i e s  i n  subsequen t  months and v i c e  v e r s a .  For  i n s t a n c e ,  
on TRl ,  the  low m o r t a l i t i e s  f rom June to  September  were  
f o l l o w e d  by r i s e s  i n  appa r en t  d e n s i t i e s  f rom Oc t ob e r  to  
December.  Then the h i g h  m o r t a l i t i e s  f rom December to  F e b r ua r y  
were f o l l o w e d  by d e c l i n e  i n  appa r en t  d e n s i t i e s  f rom January  to  
A p r i l .  Low m o r t a l i t i e s  f rom March t i l l  May were  f o l l o w e d  by  
r i s e s  i n  appa rent  d e n s i t i e s  f rom A p r i l  to  June.  However ,  
c o r r e l a t i o n s  between appar en t  m o r t a l i t y  i n  one month and the  
apparent  d e n s i t y  i n  the next  month were not  s i g n i f i c a n t .
Ana l y s e s  were c a r r i e d  out to de te rmine  the r e l a t i o n s h i p  
between a d u l t  m o r t a l i t y  r a t e  and c l im a t i c  f a c t o r s .  W i t h  the  
data  from TRl ,  a s i g n i f i c a n t  p o s i t i v e  c o r r e l a t i o n  was o b s e r v e d  
between m o r t a l i t y  r a t e  and the maximum t empe ra t u r e  ( F i g .
6 . 1 8 ) .  The r e g r e s s i o n  e q u a t i on  o f  m o r t a l i t y  r a t e  a g a i n s t  
temperature  was:
Y = - 0 . 114  + 0 . 004X ( r  = 0 . 8 3 ,  P<0 .001 ,  n = 15)
University of Ghana          http://ugspace.ug.edu.gh
Ad
ul
t 
mo
rta
lit
y 
rat
e/
da
y 
Ad
ul
t 
mo
rta
lit
y 
ra
te
/d
ay
Fig lire 6.18: The relationship between adult mortality rate 
on transect 1 and the mean monthly maximum 
temperature (Y = -0.114 + 0.004X; r =0.83, 
PC0.001, n = 15).
Min.  relative humidity
Figure 6.19: The relationship between adult mortality rate 
on transect 1 and the mean monthly minimum 
relative humidity (Y = 0.074 -  0.001X; r  = -0.54, 
P<0.05, n = 15).
University of Ghana          http://ugspace.ug.edu.gh
199
A weaker  but s t i l l  s i g n i f i c a n t  r e l a t i o n s h i p  was shown 
between m o r t a l i t y  r a t e  and r e l a t i v e  humidi ty  ( F i g .  6 . 1 9 ) .  The 
r e g r e s s i o n  e qua t i on  was:
Y = 0 . 0 7 4 - 0 . 001X ( r  = 0 . 5 4 ,  P<0.05 n = 15)
Weaker r e l a t i o n s h i p s  were  ob se r v e d  between the  m o r t a l i t y  
r a t e s  on TR4 and the c l im a t i c  f a c t o r s .  The r e l a t i o n s h i p  w i t h  
temperature  was j u s t  s i g n i f i c a n t  at  the  5% l e v e l  ( r  = 0 . 6 4 ,  n 
= 12) ( F i g .  6 . 2 0 ) .  The c o r r e l a t i o n  w i th  r e l a t i v e  humid i t y  was  
not s i g n i f i c a n t  ( F i g .  6 . 2 1 ) .
When the da t a  on TR1 was r e - a n a l y z e d  j u s t  f o r  the  p e r i o d  
b e f o r e  the p o p u l a t i o n  s u p p r e s s i o n  o p e r a t i o n ,  the s t r e n g t h  o f  
the r e l a t i o n s h i p s  was r educed  to the same l e v e l s  as  f o r  tho se  
f o r  TR4. The c o r r e l a t i o n  w i th  maximum t empe ra t u r e  was s t i l l  
s i g n i f i c a n t  ( r  = 0 . 64 ,  P<0 .05 ,  n = 9) but  tha t  w i t h  r e l a t i v e  
humidi ty  was no t .  The im p l i c a t i o n  i s  tha t  the  a d d i t i o n a l  
mo r t a l i t y  due to  the t r a p s  d u r i n g  the p o p u l a t i o n  s u p p r e s s i o n  
was s t r o n g l y  c o r r e l a t e d  w i th  c l im a t i c  f a c t o r s  e s p e c i a l l y  
tempe rature  ( s e e  D i s c u s s i o n  f o r  more d e t a i l s ) .  The da t a  f o r  
October  were exc luded  from the above a n a l y s i s  b ecau se  o f  the  
very smal l  sample s i z e  o f  f l i e s  d i s s e c t e d  f o r  tha t  month.
6 .3 . 5 .  M o r t a l i t y  r a t e s  f rom Moran Curves
F i g u r e s  6.22A and 6.22B show the p l o t s  on the  Moran  
curves  f rom which g e n e r a t i o n  m o r t a l i t y  r a t e s  were  e s t ima t ed  
and F i g .  6.23 shows the monthly  changes  i n  m o r t a l i t y  r a t e s .  
Like  the m o r t a l i t y  r a t e  e s t im a t e s  f rom o v a r i a n  age  d a t a ,  h i gh  
m o r t a l i t i e s  were g e n e r a l l y  o b s e r v e d  i n  the  hot  d ry  and c o l d
University of Ghana          http://ugspace.ug.edu.gh
Ad
ul
t 
mo
rta
lit
y 
ra
te/
da
y 
Ad
ult
 m
ort
ali
ty 
ra
te
/d
ay
Figure 6.20: The relationship between adult mortality rate 
on transect 4 and the mean monthly maximum 
temperature (Y = -0.114 + 0.004X; r  =0.83, 
P<0.001, n = 15).
Min. relative humidity
Figure 6.21: The relationship between adult mortality rate 
on transect 4 and the mean monthly minimum 
relative humidity (Y = 0.074 -  0.001X; r  = -0.54,
P<0.05, n = 15).
University of Ghana          http://ugspace.ug.edu.gh
logNtvj
F i g u r e  6.22As S c a t t e r p l o t s  o f  male ap p a ren t  d e n s i t i e s  
f rom t r a n s e c t  4 and the  f i t t e d  Moran c u r v e s .
University of Ghana          http://ugspace.ug.edu.gh
?F i gure  6.22B:  Scatterplots of female apparent d e n s i t i e s  
from t r a n s e c t  4 and the  f i t t e d  Moran c u r v e s .
University of Ghana          http://ugspace.ug.edu.gh
Ge
ne
ra
tio
n 
mo
rta
lit
y 
ra
te
/d
ay
203
Figure 6.23: Monthly changes in generation mortality rate
estimated from Moran curves (  males ;
y.-y. females).
University of Ghana          http://ugspace.ug.edu.gh
204
dry s ea sons  which d e c l i n e d  i n  the  l ong  r a i n y  s ea s on  (Ma rch ,  
A p r i l  and May ) .  However ,  t h e r e  was no s i g n i f i c a n t  c o r r e l a t i o n  
between the g e n e r a t i o n  m o r t a l i t y  r a t e  and the a d u l t  m o r t a l i t y  
r a t e  f rom the o v a r i a n  a g i n g .  Fu r the rmore ,  t h e r e  were  no 
s i g n i f i c a n t  c o r r e l a t i o n s  between the m o r t a l i t y  r a t e s  f rom the  
Moran curve s  and c l im a t i c  f a c t o r s  ( t empe r a t u r e  and r e l a t i v e  
hum id i t y ) .  Because  o f  the  p rob lem o f  imm ig r a t i on ,  th e se  
mo r t a l i t y  r a t e s  may be l e s s  r e l i a b l e  than those  e s t im a t ed  f rom  
the o v a r i a n  age s t r u c t u r e  ( s e e  D i s c u s s i o n ) .
6 . 3 . 6 .  M o r t a l i t y  r a t e s  and changes  i n  the p e r c e n t a g e  of  
n u l l i p a r o u s  f l i e s .
The monthly changes  i n  the p e r c e n t a g e  o f  n u l l i p a r o u s  
f l i e s  a r e  shown in  f i g u r e s  6.24 and 6.25 f o r  TRl  and TR4 
r e s p e c t i v e l y .  When the changes  i n  a d u l t  m o r t a l i t y  r a t e s  
( e s t im a t e s  f rom o v a r i a n  age s t r u c t u r e )  a r e  compared w i t h  t ho se  
i n  the p e r c en t a g e  o f  n u l l i p a r o u s  f l i e s ,  some g e n e r a l  t r en d s  
can be o b se r ve d .  A h i g h e r  p e r c en t a g e  of  n u l l i p a r o u s  f l i e s  
f o l l o w s  p e r i o d s  o f  v e ry  h i gh  or  v e r y  low m o r t a l i t i e s .  On TRl  
the low m o r t a l i t i e s  f rom June to Oc tobe r  a r e  immed i a t e l y  
f o l l o w e d  by an i n c r e a s e  i n  p e r c e n t a g e  o f  n u l l i p a r o u s  f l i e s  
from September to O c to be r .  The r i s e  i n  m o r t a l i t i e s  f rom  
November to a peak i n  Fe b ru a ry -Ma rch  i s  a l s o  f o l l o w e d  by a 
co r r e sp ond i ng  r i s e  i n  the p e r c e n t a g e  o f  n u l l i p a r o u s  f l i e s  w i th  
a peak i n  M a r c h - A p r i l .  S im i l a r  t r en d s  occu r  on TR4.
Norma l ly  low m o r t a l i t y  r a t e s  sho u l d  r e s u l t  i n  low 
pe rc en ta ge s  of  n u l l i p a r o u s  f l i e s  s i n c e  pa rous  f l i e s  would  be
University of Ghana          http://ugspace.ug.edu.gh
Pe
rc
en
ta
ge
 
nu
lli
pa
ro
us
 
Pe
rc
en
ta
ge
 
nu
lli
pa
ro
us
205
Figure 6.24: Monthly changes in the percentage of 
nulliparous flies in the transect 1 area.
Figure 6.25: Monthly changes in the percentage of 
nulliparous flies in the transect 4 area.
University of Ghana          http://ugspace.ug.edu.gh
206
l i v i n g  l o ng e r  and v i c e  v e r s a .  But a c c o r d i n g  to  Roge rs  et  a I .  
( 1984)  the h i gh  p e r c en ta ge  o f  young f l i e s  o c c u r r i n g  when 
m o r t a l i t i e s  a r e  low im p l i e s  an expanding  p o p u l a t i o n  or  an 
em ig r a t i on  o f  o l d e r  f l i e s  f rom the t r a p p i n g  a r e a .
R e c a l l i n g  the p a t t e r n s  o f  f l y  movement be tween  the two 
t r a n s e c t s  s u g g e s t e d  i n  s e c t i o n  6 . 3 . 2 ,  em i g r a t i o n  may p a r t i a l l y  
e x p l a i n  the changes  i n  p e r c e n t a g e  young f l i e s .  Peak  
pe rc en ta ge s  of  n u l l i p a r o u s  f l i e s  seem to  co r r e s pond  to  p e r i o d s  
of  em i g r a t i o n  f rom the r e s p e c t i v e  a r e a s .  For  i n s t a n c e  the  
movement of  f l i e s  f rom TRl to TR4 from August  to  Oc to be r  i s  
a s s o c i a t e d  w i th  a r i s e  i n  the  p e r c e n t a g e  o f  young f l i e s  on TRl  
and low p e r c en ta ge s  of  young f l i e s  on TR4. The movement i n  the  
oppo s i t e  d i r e c t i o n ,  f rom Oc to be r  to December ,  r e v e r s e s  th e se  
changes on the two t r a n s e c t s .  The net  movement o f  f l i e s  to TR4 
from December to  March r e s u l t s  i n  a r i s e  i n  n u l l i p a r o u s  f l i e s  
on TRl as e xpec t ed  but  the p e r c e n t a g e  o f  n u l l i p a r o u s  f l i e s  on 
TR4 remained r e l a t i v e l y  h i g h .  Thi s  im p l i e s  th a t  young f l i e s  
were b e in g  r e c r u i t e d  to  t h i s  p o p u l a t i o n
The g r a du a l  d e c l i n e  i n  the p e r c e n t a g e  n u l l i p a r o u s  f l i e s  
from May to June on TRl can be e x p l a i n e d  by net  immig r a t i on  
i n t o  the a r ea  du r i n g  t h i s  p e r i o d .  However ,  on TR4 where t h e r e  
was appa rent  em i g r a t i o n  d u r i n g  t h i s  p e r i o d ,  the  p e r c e n t a g e  o f  
n u l l i p a r o u s  f l i e s  was on the d e c l i n e  which can be e x p l a i n e d  by  
the g e n e r a l l y  low m o r t a l i t y  r a t e s  d u r i n g  th e se  months so tha t  
parous  f l i e s  l i v e d  l o n g e r .
University of Ghana          http://ugspace.ug.edu.gh
207
6 .4 .  DISCUSSION
A l though  t r a p p i n g  i s  now the p r e f e r r e d  t e c hn i q u e  f o r  
sampl ing  t s e t s e  p o p u l a t i o n s ,  the  method i s  s t i l l  not  
s u c c e s s f u l  f o r  some s p e c i e s .  The s uc c e s s  o f  u s i n g  a c e t one  and  
cow u r in e  to b a i t  the t r a p s  f o r  s amp l ing  G. longipennis  i n  the  
same l o c a l i t y  i n  which p a s t  a t t empts  have f a i l e d ,  c l e a r l y  
demonst r at es  the advan tage  o f  employ ing  an e f f e c t i v e  odour  
b a i t .  Un ba i t ed  t r a p s  were  p r e v i o u s l y  found to  be i n e f f e c t i v e  
(Owaga , 1980 ) .
S ince  e f f e c t i v e  s amp l ing  r e q u i r e s  more than  j u s t  c a t c h i n g  
l a r g e  numbers o f  f l i e s ,  the  two t r a p  t ypes  were  used  
s imu l t a n eou s l y  to  compare t h e i r  e f f i c i e n c i e s  f o r  s amp l in g .  
Al though s im i l a r  t r ends  i n  appa r en t  d e n s i t i e s  were  r e c o r d e d  by  
both t r a p s ,  some impo r t ant  d i f f e r e n c e s  i n  the  s amples  t aken  by  
the two t r a p  t ypes  a r e  worth  n o t i n g .  F i r s t l y ,  h i g h e r  f l y  
numbers were  g e n e r a l l y  r e c o r d e d  by the  NG2B t r a p s  than the  
b i c o n i c a l  t r a p s ,  i n  agreement  w i t h  p r e v i o u s  f i n d i n g s .  Thus ,  i n  
s i t u a t i o n s  of  v e ry  low p o p u l a t i o n  d e n s i t i e s ,  the  b i c o n i c a l  
t r a p  may not  c a t ch  any f l i e s  a t  a l l  and w i l l  t h e r e f o r e  be  
i n adequate  f o r  s amp l i ng .  S ec ond l y ,  a h i g h e r  p e r c e n t a g e  o f  
f emales  was o b t a i n ed  w i t h  the NG2B t r a p  than w i t h  the  
b i c o n i c a l  t r a p .  Going  by the ex pe c t e d  p e r c e n t a g e  o f  70%-80% 
f emales  i n  a w i l d  p o p u l a t i o n  ( J ackson ,  1949 ) ,  the  f ema l e  
pe rcen ta ge s  r e c o rd ed  i n  t h i s  s tudy  were  g e n e r a l l y  b e l ow  the  
expect ed  v a l u e .  T h e r e f o r e ,  a l t h ou gh  bo th  t r a p s  were  p r o b a b l y  
und e r - e s t im a t i n g  the f ema le  p o p u l a t i o n ,  t h i s  was more so  w i t h  
the b i c o n i c a l  t r a p .
University of Ghana          http://ugspace.ug.edu.gh
208
Thi s  u n d e r l i n e s  the o b s e r v a t i o n  made by Roge r s  ( 1984 )  
that  a l l  t r a p s  a r e  b i a s e d  i n  one way or the  o th e r  and tha t  
these  b i a s e s  shou ld  be q u a n t i f i e d  i f  p o p u l a t i o n  
c h a r a c t e r i s t i c s  a r e  to  be deduced from t r a p  c a t c h e s .  Trap  
b i a s e s  may be q u a n t i f i e d  by u s i n g  the model p r opose d  by Roge r s
(1984)  b a sed  on the f e e d i n g  c y c l e  o f  the  s p e c i e s  i n v o l v e d ,  
a l though  t h i s  i s  s t i l l  unknown f o r  G. longipennis. 
A l t e r n a t i v e l y ,  t r ap  c a t ches  cou ld  be compared to a b s o l u t e  
po pu l a t i o n  e s t ima t e s  f rom m a r k - r e l e a s e - r e c a p t u r e  methods made 
at  the same t ime as the s amp l ing  took p l a c e  to  de te rmine  the  
r e l a t i o n s h i p  between appa r en t  and a b s o l u t e  e s t im a t e s  ( s e e  
chap te r  9 ) .
In the absence  o f  such c o r r e c t i o n  f a c t o r s ,  the  q u e s t i o n  
a r i s e s  as to whether  the se  changes  r e f l e c t  a c t u a l  changes  i n  
the p o p u l a t i o n  s i z e  or i f  they  a r e  due to  f l y  a c t i v i t y  and 
hence a v a i l a b i l i t y .  In t h i s  s tudy ,  bo th  t r a p  t ypes  showed  
s im i l a r  t r end s  i n  p o p u l a t i o n  changes  ( d e s p i t e  the d i s p a r i t y  i n  
r e l a t i v e  numbers )  and,  assuming tha t  the v e ry  s h o r t  f l i g h t  
p e r i o d  o f  G. 1ongipennis  does  not  change th roughou t  the  y e a r ,  
the ob se rve d  changes  p r o b a b l y  i n d i c a t e  changes  i n  the  
po pu l a t i o n  s i z e  r a t h e r  than f l y  a v a i l a b i l i t y .
Assuming the above as sumpt i ons  h o l d ,  i t  can be conc luded  
that  the p o p u l a t i o n  o f  G. l o n g i p e n n i s  i n c r e a s e d  to  a major  
peak towards  the end of  the long  r a i n y  s ea s on ,  d e c l i n e d  i n  the  
dry s eason ,  and r o s e  to a sm a l l e r  peak du r i n g  the  s h o r t  r a i n y
University of Ghana          http://ugspace.ug.edu.gh
209
s eason .  S im i l a r  t r ends  were ob se r v e d  by D r a n s f i e l d  et
a l . ( 1 9 8 6 a )  f o r  the G. pallidipes  p o p u l a t i o n  i n  the  same s tudy
s i t e  over  the  same p e r i o d .
Looking  at  o the r  f u s ea  s p e c i e s ,  Kangwagye ( 1974 )  r e c o r d e d  
peak ca t che s  o f  G. fusciple u r i s  towards  the end o f  the  r a i n s  
in  wes t ern  Uganda.  In Cote  d ' I v o i r e ,  Gouteux and Buck land
(1984)  ob se r ve d  major  peak c a t che s  o f  G. n i g r o f u s c a  at  the  end 
of  the r a i n y  s ea son  and a sm a l l e r  peak t owa rds  the end o f  the  
dry s eason .  They a t t r i b u t e d  the smal l  d ry  s e a son  peak to the  
emergence of  pupae d e p o s i t e d  by f l i e s  d u r i n g  peak d e n s i t i e s  a t  
the end of  the r a i n s .  C h a l l i e r  (1982 )  r e v i ewed  a wide  r ange  o f  
p o pu l a t i o n  f l u c t u a t i o n  p a t t e r n s  r e c o r de d  by v a r i o u s  wo rke r s  i n  
d i f f e r e n t  l o c a l i t i e s .  He o b se rv e d  tha t  f l u c t u a t i o n  p a t t e r n s  i n  
t s e t s e  p o p u l a t i o n s  a r e  r ou gh l y  r e l a t e d  to  the  r a i n f a l l  
d i s t r i b u t i o n  throughout  the year  w i th  the b a s i c  f e a t u r e  b e i n g  
an i n c r e a s e  du r i n g  the r a i n s  and a d e c r e a s e  d u r i n g  the  dry  
s eason .  However the g e n e r a l  p a t t e r n  may be m o d i f i e d  by the  
amount and monthly d i s t r i b u t i o n  o f  the  r a i n s .  V a l e  et  a l .
( 1985)  a t t r i b u t e d  the d e c l i n e  i n  the d ry  s ea s on  to the  f a c t  
tha t  f ema l e  f l i e s  do not  l i v e  l ong  enough to  p roduce  
rep lacement  o f f s p r i n g .
The ob se rve d  s ea s ona l  changes  i n  the r e l a t i v e  
d i s t r i b u t i o n  of  G. 1ongipennis  between the d i f f e r e n t  
v e g e t a t i o n  t ypes  has been r e p o r t e d  by a number o f  wo rke r s  f o r  
o ther  t s e t s e  s p e c i e s .  In the same s tudy  a r e a  D r a n s f i e l d  et  a l . 
( 1986a )  ob se r ved  tha t  G. pallidipes  a l s o  t ended  to  s p r e a d  out  
du r i ng  the r a i n s  and c o n t r a c t e d  i n t o  the  more dense  v e g e t a t i o n
University of Ghana          http://ugspace.ug.edu.gh
210
du r i ng  the hot and dry  s e a s on s .  A s im i l a r  o b s e r v a t i o n  was 
r e p o r t e d  on G. m. submo r s i t an s  by D r a n s f i e l d  et  a l  . ( 1982 )  i n
no r the rn  N i g e r i a  w i t h  f l i e s  s p r e a d i n g  out i n t o  the  more open  
woodland i n  the r a i n y  s ea son  but  c o n c e n t r a t i n g  i n  the  f o r e s t  
i n  the dry  s e a son .  A cc o r d i n g  to  t h i s  a u t ho r ,  the  t s e t s e  
h a b i t a t  may be d e s c r i b e d  as a pa tchwork  o f  a r e a s  w i th  
s e a s o n a l l y  changing  p r o b a b i l i t i e s  of  s u r v i v a l  and r e p r o d u c t i v e  
s u c c e s s ;  hence f l i e s  tend to c o n c en t r a t e  i n  the most s u i t a b l e  
h a b i t a t .
H a b i t a t  s u i t a b i l i t y  may be de t e rmined  by bo th  a b i o t i c  and 
b i o t i c  f a c t o r s .  C on c en t r a t i o n  of  f l i e s  i n  the t h i c k e r  
v e g e t a t i o n  du r i n g  the  hot and d ry  s e a sons  may s e r v e  to  a v o i d  
l e t h a l  l e v e l s  of  t empe ra tu r e  and humid i t y .  At the  same t ime i t  
cou ld  be tha t  f l i e s  a r e  f o l l o w i n g  t h e i r  h o s t s  whi ch  would  tend  
to f r e qu en t  the t h i c k e r  v e g e t a t i o n  i n  s e a r c h  o f  f ood  which  
would be more a v a i l a b l e  he r e  than i n  the  more open p a r t s  o f  
the h a b i t a t  du r i n g  the dry  s eason .
Whatever  the r e a sons  may be ,  the  d i f f e r e n c e s  i n  c a t che s  
between the v a r i o u s  v e g e t a t i o n  t ypes  demons t r a t e  the  
importance  of  s amp l ing  over  the  f u l l  r ange  o f  the  d i s t r i b u t i o n  
of  a " p o p u l a t i o n "  of  t s e t s e  when c a r r y i n g  out p o p u l a t i o n  
dynamics s t u d i e s  ( G r u v e l ,1975;  Ha r g ro ve  and V a l e ,  1980;  
D r a n s f i e l d  et  a l . ,  1986a ) .  The l a t t e r  au tho r s  remarked tha t  a 
wrong p i c t u r e  may be o b t a i n e d  i f  s amp l ing  i s  not  done  
over  a l l  the v e g e t a t i o n  t yp e s .
The appa rent  l a g  i n  phase  between p o p u l a t i o n  changes  on 
the two t r a n s e c t s  cou ld  be due to  movement o f  f l i e s  be tween  1
University of Ghana          http://ugspace.ug.edu.gh
211
them. A net  movement of  f l i e s  f rom one a r e a  to the o the r  wou ld  
account  f o r  the  subsequen t  r i s e  i n  appa r en t  d e n s i t i e s  i n  the  
l a t t e r  and a c o r r e s p ond i n g  d e c l i n e  i n  the  f o rmer .  From the  
above argument ,  the t r end s  i n  appa r en t  d e n s i t i e s  seem to  show 
a net  movement o f  f l i e s  f rom TR1 to TR4 from August  to  O c t o b e r  
and i n  the r e v e r s e  d i r e c t i o n  from Oc tobe r  to December .  A g a i n  
the r e  was an appa r en t  net  movement o f  f l i e s  i n t o  TR4 from 
January to March and back to TR1 f rom March t i l l  Ju l y .
The a d v e r s e  e f f e c t  o f  c l im a t i c  f a c t o r s  on a d u l t  s u r v i v a l  
i s  shown by the s i g n i f i c a n t  p o s i t i v e  c o r r e l a t i o n  between  
mo r t a l i t y  r a t e  f rom o v a r i a n  a g i n g  and the maximum t empe ra t u r e  
e xp e r i e nc ed  by f l i e s .  A weaker  r e l a t i o n s h i p  was shown between  
mo r t a l i t y  r a t e  and minimum r e l a t i v e  humid i ty .  Th i s  s u g g e s t  
that  h i g h e r  env i ronmenta l  t empe ra tu r e s  a r e  more l e t h a l  to G. 
longipennis than env i ronmenta l  d r y n e s s .  The f a c t  th a t  the  f l y  
l im i t s  i t s  a c t i v i t y  to  dawn and dusk i s  p r o b a b l y  an a d a p t a t i o n  
to a vo id  l e t h a l  l e v e l s  o f  t he se  f a c t o r s .  Du r ing  the  r e s t  o f  
the day i t  may be tha t  the f l y  i s  a b l e  to  s eek  out m i c r o ­
h a b i t a t s  tha t  a r e  humid enough,  but  l e s s  a b l e  to  a v o i d  l e t h a l  
hi gh  t empe ra t u r e s .  I t  shou ld  be no t ed ,  however ,  t h a t  low 
hum id i t i e s  may w e l l  have s u b l e t h a l  e f f e c t s  on G. 1ongipennis 
( s e e  Chapter  7 ) .
The s t r o n g e r  r e l a t i o n s h i p  between a d u l t  m o r t a l i t y  r a t e  
and the se  c l im a t i c  f a c t o r s  o b se r ve d  on TR1 when the  da t a  
i n c l u d ed  the months o f  p o p u l a t i o n  s u p p r e s s i o n  may be an 
a r t e f a c t  c r e a t e d  by the m o r t a l i t y  due to the t r a p p i n g .  In  
d e p l o y i n g  t r ap s  f o r  the  p o p u l a t i o n  s u p p r e s s i o n  o p e r a t i o n  most
University of Ghana          http://ugspace.ug.edu.gh
212
of  the t r a p s  were l o c a t e d  i n  the  t h i c k e r  wood l and  ( s e e  c h ap t e r  
9 ) ,  where f l i e s  norma l l y  c o n c en t r a t e  when i t  i s  v e r y  hot  or  
ve ry  d ry .  Thus i t  i s  l i k e l y  th a t  a d u l t  m o r t a l i t y  r a t e  due to  
the s u p p r e s s i o n  o p e r a t i o n  i n  t h i s  a r e a  was h i g h e r  d u r i n g  t h e s e  
p e r i o d s .  Th i s  cou ld  then r e s u l t  i n  the s t r o n g  c o r r e l a t i o n  
between a d u l t  m o r t a l i t y  and t empe ra tu r e .  When the a n a l y s i s  was  
l im i t e d  to  on l y  n a t u r a l  m o r t a l i t i e s  by e x c l u d i n g  the months o f  
p o pu l a t i o n  s u p p r e s s i o n  the r e l a t i o n s h i p s  were  compa rab le  to  
those  of  TR4 which was o u t s i d e  the s u p p r e s s i o n  zone.
The l a ck  o f  c o r r e l a t i o n  between the m o r t a l i t y  r a t e  
e s t ima t e s  f rom Moran cu rve s  and tho se  f rom the  o v a r i a n  age  
s t r u c t u r e  im p l i e s  some i n a c c u r a c i e s  i n  the e s t im a t e s  by one o f  
the methods or  bo th .  Both methods a r e  s u b j e c t  to  e r r o r s  when 
f l i e s  a r e  m i g r a t i n g  i n t o  the s amp l ing  a r e a ,  but  the o v a r i a n  
age method may s t i l l  be r e l i a b l e  i f  the  immigrant  f l i e s  have  
the same age d i s t r i b u t i o n  as the r e s i d e n t  p o p u l a t i o n .  S ince  
immigrant  f l i e s  a r e  more l i k e l y  to  be o l d  f l i e s  ( e x c l u d i n g  
ca t eg o r y  O f l i e s )  and the m o r t a l i t y  r a t e  e s t im a t e s  i n  t h i s  
study exc lu ded  c a t e g o r y  O f l i e s ,  the e s t im a t e s  c o u ld  be  
r e l i a b l e .
The Moran cu rve  method,  on the  o the r  hand,  w i l l  c e r t a i n l y  
be r ender ed  i n a c c u r a t e  by immi g r a t i on .  Roger s  ( 1979 )  o b s e r v e d  
that  p o i n t s  tha t  l a y  o u t s i d e  the r e a l i s t i c  r ange  of  
r e p r o du c t i v e  r a t e  i n  the Moran p l o t  were a s s o c i a t e d  w i t h  
p e r i o d s  of  immig r a t i on  i n t o  the s amp l ing  a r e a  o r  wer e  due to  
sampl ing  e r r o r s .  In the p r e s en t  s tu dy ,  t h e r e  was ev i d en c e  o f  
immigr at i on  from the Moran p l o t s .  O u t l y i n g  p o i n t s  were
University of Ghana          http://ugspace.ug.edu.gh
213
obse rved  f o r  the  month of  September  1986 f o r  bo th  s ex e s  and  
September 1987 f o r  the males  o n l y .  From the  s u g g e s t e d  p a t t e r n  
of  f l y  movement, t h e r e  was movement o f  f l i e s  i n t o  TR4 f rom  
September  to  O c to be r .  The z e ro  m o r t a l i t y  r a t e s  r e c o r d e d  i n  
September  f o r  bo th  1986 and 1987 were  t h e r e f o r e  due to  the  
e f f e c t  of  immig r a t i on .
Rogers  (1979 )  a l s o  remarked tha t  Moran cu rv e s  g i v e  more  
r e l i a b l e  r e s u l t s  f o r  d a t a  c o l l e c t e d  over  a l o n g e r  p e r i o d  of  
t ime.  In v iew o f  t h i s ,  the  p r e s e n t  da t a  s e t  i s  p r o b a b l y  too  
smal l  f o r  the  a c c u r a t e  i d e n t i f i c a t i o n  o f  the  key d e n s i t y  
i ndependent  f a c t o r s .  C o n s i d e r i n g  the  above ,  the  m o r t a l i t y  r a t e  
e s t ima te s  f rom o v a r i a n  age s t r u c t u r e  may be more r e l i a b l e  than  
those  from the Moran c u r v e s .  Th i s  i s  s up po r t e d  by the  f a c t  
tha t  e s t ima t e s  by the o v a r i a n  age  method show s i g n i f i c a n t  
r e l a t i o n s h i p s  w i th  the c l im a t i c  f a c t o r s  w h i l s t  t ho se  f rom the  
Moran cu rve s  do n o t .
The r e l a t i o n s h i p  between  c l im a t i c  f a c t o r s  and t s e t s e  
po pu l a t i o n s  has been i n v e s t i g a t e d  by a number o f  w o rk e r s .  For  
example ,  a s i g n i f i c a n t  r e l a t i o n s h i p  between  a d u l t  m o r t a l i t y  
r a t e  and maximum t empe ra tu r e  was o b s e r v e d  f o r  G. palpal is by 
Rogers  et  a l . ( 1 9 84 ) .  However ,  the  l a t t e r  au tho r s  on l y  
e s t a b l i s h e d  s i g n i f i c a n t  n e g a t i v e  c o r r e l a t i o n s  between  appa r en t  
d e n s i t y  and the m o r t a l i t y  r a t e  o f  the  p r e v i o u s  month i n  some 
sampl ing  a r e a s  but  not  o t h e r s .  I t  was r e a l i z e d  tha t  the  a r e a s  
f o r  which th e r e  were l a ck  o f  s i g n i f i c a n t  c o r r e l a t i o n s  were  
s u b j e c t  to cons t an t  immig r a t i on  p r e s s u r e .  A s im i l a r  
e x p l a n a t i o n  cou ld  account  f o r  the  n o n - s i g n i f i c a n t  c o r r e l a t i o n :
University of Ghana          http://ugspace.ug.edu.gh
214
between m o r t a l i t y  r a t e  and appa r en t  d e n s i t i e s  o b s e r v e d  i n  t h i s  
s t u d y .
Apar t  f rom the movement o f  f l i e s  between  the  two s amp l i n g  
a r e a s  ( s u g g e s t e d  e a r l i e r ) ,  t h e r e  i s  a l s o  e v i d en c e  f rom o th e r  
s t u d i e s ,  c a r r i e d  out i n  the s tudy  a r e a  ( D r a n s f i e l d  et  a l  . , 
u n p u b l i s h e d ) ,  t ha t  f l i e s  do m i g r a t e  f rom the top o f  the  
escarpment i n t o  the s amp l ing  a r e a .  W i th  imm i g r a t i on ,  the  
e f f e c t  o f  m o r t a l i t i e s  e x p e r i e n c e d  by pa r en t  f l i e s  may not  be  
s i g n i f i c a n t l y  r e f l e c t e d  i n  the  sub sequen t  s amples  s i n c e  
immigrant  f l i e s  would  a l s o  be c o n t r i b u t i n g  to the changes  i n  
the p o p u l a t i o n  l e v e l s .
A l though  the a n a l y s e s  p r e s e n t e d  he r e  have p r o v i d e d  
i n f o rma t i o n  on the d e n s i t y  i ndependent  m o r t a l i t y  f a c t o r s ,  they  
have shed l i t t l e  l i g h t  on the  n a t u r e  o f  the  d e n s i t y  dependent  
l o s s e s  which r e g u l a t e  the p o p u l a t i o n  s i z e .  From the  Moran  
cu rv e s ,  such l o s s e s  would  appea r  to  be o p e r a t i n g  p r i m a r i l y  i n  
the months of  October  and May/June.  The h i g h  m o b i l i t y  o f  the  
f l i e s  s u g g e s t s  t h a t ,  at  l e a s t  at  a l o c a l  l e v e l ,  movement may 
be an import ant  r e g u l a t o r y  f a c t o r .  S t u d i e s  a t  Nguruman on G. 
pallidipes  have a l s o  s u g g e s t e d  tha t  pupa l  l o s s e s  and p r e d a t i o n  
of  the a d u l t s  cou ld  be impo r t an t ,  and more work needs  to  be  
c a r r i e d  out on G. longipennis  on the se  a s p e c t s .
University of Ghana          http://ugspace.ug.edu.gh
215
CHAPTER SEVEN
POPULATION DYNAMICS OF GLOSSINA LONGIPENNIS:
II. REPRODUCTION AND SIZE
7 .1 .  INTRODUCTION
In the ab sence  o f  immig r a t i on ,  t s e t s e  p o p u l a t i o n s  a r e  
dependent  on the r e p r o d u c t i v e  r a t e  to b a l a n c e  the m o r t a l i t y  
r a t e s  c o n s i d e r e d  i n  the p r e v i o u s  c h a p t e r .  The f i r s t  f a c t o r  
a f f e c t i n g  t h i s  i s  the i n s em in a t i o n  r a t e  which i s  i n f l u e n c e d  by 
the male i n s em in a t i o n  c a p a c i t y  and the f ema l e  r e c e p t i v i t y .  
Va r i ous  wo rke r s ,  r e v i ewed  by C h a l l i e r  ( 1 9 8 2 ) ,  have shown tha t  
the t iming  o f  i n s em in a t i o n  v a r i e s  w i t h  d i f f e r e n t  s p e c i e s .
The o the r  f a c t o r  a f f e c t i n g  r e p r o d u c t i v e  r a t e  i s  the  
du r a t i on  and s u r v i v a l  r a t e  o f  the  l a r v a  s t a g e  i n  u t e r o .  The 
i n t e r - l a r v a l  p e r i o d  has been shown to be dependent  on 
tempe rature  (G l a s gow ,  1970, C h a l l i e r ,  1973 ) .  C e r t a i n  
r e p r o du c t i v e  a b n o rm a l i t i e s  have been o b s e r v e d  by C h a l l i e r  
( 1973)  i n  G. p. g amb i ens i s  which r e s u l t e d  i n  d e l a y  i n  
r e p r o du c t i v e  c y c l e ,  d e g e n e r a t i o n  o f  e g g s ,  n o n - v i a b l e  l a r v a e  
and a b o r t i o n s .  He and o the r  i n v e s t i g a t o r s  ( e . g .  Ma du buny i , 
1978; Turner  and Snow, 1984)  a r e  o f  the o p i n i o n  tha t  
a b o r t i o n s  a r e  the main s ou r ce  o f  r e p r o d u c t i v e  l o s s  and Jordan  
(1962b)  sug ge s t ed  tha t  they may p l a y  a s i g n i f i c a n t  r o l e  i n  
keep ing  t s e t s e  p o p u l a t i o n s  at  low d e n s i t i e s .  Th i s  was b a s ed  on 
a v e ry  h i gh  mean a b o r t i o n  r a t e  o f  about  60% o b s e r v e d  i n  a 
p o pu l a t i o n  of  G. p a l p a l i s  i n  m idwes t e rn  N i g e r i a .  Othe r  wo rke r s
University of Ghana          http://ugspace.ug.edu.gh
216
have, however, reported rather low abortion rates (below 1 0 % 
on average) in various tsetse populations (Okiwellu, 1976,  
1977; Madubunyi, 1975, 1988;  Dransfield et al . ,  1 986 a ) .  Hence, 
Madubunyi ( 1988 )  expressed doubts as to whether abortions can 
significantly affect tsetse population levels. However, it has 
been suggested (Dransfield et al., in press) that abortions 
may only represent a small percentage of reproductive loss. 
Production of undersized larvae resulting from parental stress 
may result in subsequent losses at the pupal and emergent 
adult stages.
Thi s  l o s s  may be i n v e s t i g a t e d  by m on i t o r i n g  the  s i z e  o f  
the a d u l t s .  I t  has been shown f o r  some s p e c i e s  t h a t  f l y  s i z e  
as i n d i c a t e d  by wing v e i n  l e n g t h  i s  i n f l u e n c e d  p r im a r i l y  by  
the env i ronmenta l  c o n d i t i o n s  e x p e r i e n c e d  by the pa r en t  d u r i n g  
pregnancy  ( a l t h o u g h  t empe r a tu r e  a t  the  pupa l  s t a g e  can a l s o  
have an e f f e c t ;  B u r s e l l  and T a y l o r ,  1960 ) .  Jackson ( 1953a )  
e s t a b l i s h e d  s i g n i f i c a n t  n e g a t i v e  c o r r e l a t i o n s  between  the  
monthly mean s i z e  o f  male G. p a l l i d i p e s , G. swynne r t on i  and G. 
mors i t ans  i n  Tanzan i a  and s a t u r a t i o n  d e f i c i t  and maximum 
t empe ratur es  e x p e r i e n c e d  by the  pa r e n t  f em a l e s .  At  Nguruman,  
D r a n s f i e l d  et a l . ,  (1988 i n  p r e s s )  found s i m i l a r  c o r r e l a t i o n s .  
There i s  e v i denc e  tha t  v e ry  smal l  f l i e s  a r e  a t  a d i s a d v a n t a g e  
when they emerge.  Phe l p s  and C l a r k e  ( 1974 )  e s t im a t ed  t h a t  up 
to 75.5% of  G. m. mo r s i t an s  p o p u l a t i o n  was l o s t  at  emergence  
due to smal l  s i z e  i n  the hot  dry  s ea s on .  D r a n s f i e l d  et  a 1. 
(1988,  i n  p r e s s )  a l s o  found s i z e  dependent  m o r t a l i t y  o f  G. 
pallidipes  at  emergence.  A f i r s t  i n d i c a t i o n  o f  whe the r  such
University of Ghana          http://ugspace.ug.edu.gh
217
mortality could be occurring in G. 1ongipennis populations 
would be to establish whether size of adults of this species 
does vary seasonally and whether such changes are indeed 
correlated with environmental conditions.
Another point of interest is whether wing vein length of 
the progeny can be used as a simple index of the parental 
density-independent mortality rate. Dransfield et aI. (1988, 
in press) suggest that this may hold true for G. pallidipes, 
which could be of use when determining the additional levels 
of mortality needed from trapping to suppress a tsetse 
populati on
In the light of the above, studies were carried out to 
determine the changes in the monthly insemination and abortion 
rates and in the size of both adult and immature stages of G. 
longipennis. Investigations were carried further to determine 
the influence of climatic factors on the changes in these 
population parameters.
7.2. MATERIALS AND METHODS
All data were collected from the same samples that were 
dissected for ovarian age determination (see Chapter 6). 
Following age categorization, the uterine content of each fly 
(empty, egg, 1st, 2nd, or 3rd instar larva) was recorded. All 
flies belonging to category Oa or Ob were examined for the 
presence of sperm in the spermathecae. Measurements were then 
taken of the length of uterine content and the length of the 
largest developing follicle.
University of Ghana          http://ugspace.ug.edu.gh
218
Adu l t  f l y  s i z e  was e s t ima t ed  a c c o r d i n g  to  the  method  
proposed  by Jackson ( 1 949 ) .  Th i s  i s  by measu r ing  the  l e n g t h  o f  
the midd l e  p a r t  o f  the  f o u r t h  l o n g i t u d i n a l  v e i n  o f  the  w ing .  
These measurements were  taken o f  the  same w ing s  t h a t  were  
c o l l e c t e d  p r i o r  to  o v a r i a n  age  d i s s e c t i o n  and used  f o r  wing  
f r a y  a g ing  as  d e s c r i b e d  i n  Chapte r  6. A l l  measurements  o f  w ing  
ve in  l e n g th ,  u t e r i n e  content  and f o l l i c u l a r  l e n g t h  were  made 
us ing  a p r e - c a l i b r a t e d  eye p i e c e  g r a t i c u l e .
7 .3 .  RESULTS
7 .3 .1 .  I n s em in a t i on  r a t e
The i n s em in a t i o n  r a t e  o f  n u l l i p a r o u s  G. longipennis  was  
obse rved  to  be v e ry  h i g h .  Throughout  the s tu dy  p e r i o d ,  a l l  
f l i e s  o f  the  Ob c a t e g o r y  were  found to  be i n s em in a t ed .  The 
i n s emina t i on  r a t e  f o r  c a t e g o r y  Oa f l i e s  was a l s o  100% in  a l l  
months except  i n  June 1986, A p r i l  1987 and May,  1987 when the  
i n s emina t i on  r a t e s  were 85, 66 and 75% r e s p e c t i v e l y .
7 . 3 . 2 .  A b o r t i o n  r a t e
The pe r c en ta g e  d i s t r i b u t i o n  o f  i n t r a - u t e r i n e  con ten t s  
r ecorded  f rom r o u t i n e  d i s s e c t i o n s  o f  f ema le  f l i e s  caught  by  
the NG2B t r a p s  on TR1 and TR4 a r e  g i v e n  i n  t a b l e s  7 .1  and 7 .2  
r e s p e c t i v e l y .  The t a b l e s  a l s o  show the monthly  changes  i n  the  
pe rc en tag e  of  empty u t e r i  ( assumed to  be ma in l y  a b o r t i o n s )  f o r  
the two t r a n s e c t s  ( a l s o  shown i n  F i g . 7 . 1 ) .  The v a l u e s  r e c o r d e d  
here  a r e  most l i k e l y  h i g h e r  than a c t u a l l y  o c c u r r i n g  i n  n a t u r e  
s in ce  t r a p - i n d u c e d  a b o r t i o n s  c ou ld  not  be d i s t i n g u i s h e d  from
University of Ghana          http://ugspace.ug.edu.gh
219
na t u r a l  a b o r t i o n s .  The e s t im a t e s  f o r  Oc tober  1986 ( f o r  T R l ) 
and January 1987 (TR4)  a r e  e s p e c i a l l y  h i gh  and may be
u n r e l i a b l e  due to the v e ry  sma l l  sample s i z e s .
However ,  assuming t r ap  i nduced  a b o r t i o n s  a f f e c t e d  the  
v a l u e s  f o r  a l l  months i n  the same way,  some i n f o rm a t i o n  can  
s t i l l  be g a in ed  from the r e l a t i v e  monthly a b o r t i o n  
f r e q u e n c i e s .  G e n e r a l l y ,  the p e r c en t a g e  o f  empty u t e r i  was 
h i ghe r  i n  the c o l d  dry  and hot  d ry  s ea sons  than i n  the  o th e r  
months o f  the y ea r .  However ,  t h e r e  were s l i g h t  d i f f e r e n c e s  i n  
the t r ends  on the two t r a n s e c t s .  Whereas  on TRl  the  p e r c e n t a g e  
of a b o r t i o n s  d e c l i n e d  from Oc tobe r  to  z e ro  i n  December ,  the  
pe rc en tag e  of  a b o r t i o n s  was on the r i s e  on TR4 d u r i n g  t h i s  
p e r i o d  and reached a peak i n  December -January  ( a l t h o u g h  the  
January sample was v e ry  s m a l l ) .  The mean a b o r t i o n  r a t e s  on the  
two t r a n s e c t s  were 7.7 and 11.1% f o r  TRl  and TR4 r e s p e c t i v e l y  
which d i d  not d i f f e r  s i g n i f i c a n t l y  by the c h i - s q u a r e d  t e s t .
R e l a t i o n s h i p s  between monthly  a b o r t i o n  r a t e  and maximum 
t emperature  and minimum r e l a t i v e  humidi ty  were  examined  
through c o r r e l a t i o n  a n a l y s i s ,  a f t e r  p o o l i n g  da t a  f rom bo th  
t r a n s e c t s .  There  was a s i g n i f i c a n t  n e g a t i v e  c o r r e l a t i o n  
( F i g . 7 . 2 )  between the p e r c e n t a g e  empty u t e r i  and minimum 
r e l a t i v e  humid i ty .  The r e g r e s s i o n  e q u a t i o n  o f  p e r c e n t a g e  empty 
u t e r i  ( Y )  on minimum r e l a t i v e  humid i t y  (X )  was:
Y = 30 -  0.64X ( r  = - 0 . 5 3 ,  P<0 .05 )
There was no s i g n i f i c a n t  c o r r e l a t i o n  between  p e r c e n t a g e  empty 
u t e r i  and maximum or mean t empe ra tu r e .
University of Ghana          http://ugspace.ug.edu.gh
2 2 0
7 . 3 . 3 .  Frequency d i s t r i b u t i o n  and s i z e  of  immature s t a g e s
T a b l e s  7 .1  and 7 .2  a l s o  show the monthly  p e r c en ta g e  
d i s t r i b u t i o n  of  the  d i f f e r e n t  pregnancy  s t a g e s  o f  f l i e s  f rom  
the two t r a n s e c t s .  F l i e s  w i t h  the e g g - s t a g e  were  g e n e r a l l y  
most abundant  f o l l o w e d  by tho se  w i t h  f i r s t  i n s t a r  l a r v a .
There was no c o n s i s t e n t  d i f f e r e n c e  between the f r e q u e n c i e s  o f  
the second and t h i r d  i n s t a r  l a r v a e .  The mean p e r c e n t a g e s  f o r  
the d i f f e r e n t  s t a g e s  taken over  the whole  s amp l ing  p e r i o d  were  
s im i l a r  on both  t r a n s e c t s .
T ab l e  7 .3  g i v e s  the monthly  mean l e n g t h s  o f  the  f o u r  
deve l opmenta l  s t a g e s  r e c o r de d  f rom the  u t e r i .  Data  f rom the  
two t r a n s e c t s  were po o l ed  and a v e r a g e s  taken becau se  o f  the  
very  smal l  sample s i z e s  i n  each c a t e g o r y .  No obv i o us  s e a s ona l  
t r ends  can be ob se r ve d  i n  the v a r i a t i o n  o f  the  u t e r i n e  con ten t  
1e n g t h .
Tak ing  the a v e r ag e s  f o r  the  two t r a n s e c t s ,  the  r e l a t i v e  
pe rc en ta ge s  ( ±  1 S .E )  o b t a i n ed  f o r  the f o u r  p r egnancy  s t a g e s  
and the mean l e ng th s  (mm ±  SE)  f o r  the  deve l opmenta l  s t a g e s  
a re  g i v en  t o g e t h e r  i n  T ab l e  7 . 4 .
The mean l e n g th s  o f  the  egg f o l l i c u l e  next  i n  o v u l a t i o n  
sequence or l a r g e s t  d e v e l o p i n g  f o l l i c l e  at  the  v a r i o u s  
pregnancy s t a g e  a r e  g i v en  i n  T a b l e  7 . 5 .  These a r e  t aken  f o r  
the whole s ampl ing  p e r i o d  and f o r  the two t r a n s e c t s .
I t  can be ob se r ve d  tha t  the mean s i z e  o f  the  l a r g e s t  
de v e l o p i n g  f o l l i c u l e  at  the L I I I  s t a g e  i s  v e ry  c l o s e  i n  v a l u e  
to the mean egg-measurement  i n  the u t e r u s  i n d i c a t i n g  th a t  the  
egg i s  ready to be o v u l a t e d  soon a f t e r  d e p o s i t i o n  o f  the
University of Ghana          http://ugspace.ug.edu.gh
2 2 1
Table 7 . 1 :  P e r c e n t a g e  d i s t r i b u t i o n  o f  u t e r i n e  con t en t s  i n  monthly  
samples  o f  G longipennis  w i t h  the NG2B t r a p s  on TR1.
MONTH
1986 
June 
July 
August 
September 
October  
November 
December
1987 
January 
February  
March 
Apr i l  
May 
June 
July 
August  
Sept . 
October  
MEAN 
S.E.
EMPTY
2 . 3
4.8
15.0  
13. 6
75.0  
8 . 0  
0 . 0
0 . 0
2 0 . 0
4.3  
0.0
3.9  
1.8 
8.5
3.2
6.4
3.3  
7.0
1.9
UTERINE CONTENT 
EGG LARVA I LARVA I I  LARVA I I I
45.2
53.6  
40 .0
34.0
25.0
56.0
29.6
28 . 5 
33 . 3 
30 . 4 
37 . 5
49.0  
40 .0 
40 .0
38.7
38.7  
50 .0 
40 . 8
2.2
26.1
9.7
15.0  
27 . 2
0 . 0  
8.0 
22 . 2
14.2  
33 . 3
34.7
50.0  
21. 5 
15 .0
25.7  
0.0
16.1  
30.0
21.7  
2 . 6
8.5  
26.8  
20 .0
11.3  
0.0 
4.0
25. 9
21.4
6.6 
8 . 6
12.2
11.7
15.0 1 
2 . 8
32 . 2
9.3  
13.3
15.0
2.3
16.6
4.8  
10 .0 
13. 6
0 . 0  
24.0  
22 . 2
35.7  
6 . 6
21.7  
0 . 0  
13 . 7
8.8
22 . 8  
25.8  
19.3
3 .3  
16.2
2.3
54
50
23
51
8
24 
35
21
24
38
14
51
92
83
33
33
4 1
University of Ghana          http://ugspace.ug.edu.gh
2 2 2
Table 7 . 2 :  P e r ce n t a ge  d i s t r i b u t i o n  o f  u t e r i n e  co n t en t s  i n  monthly  
s amples  o f  G longipennis  w i t h  the NG2B t r a p s  on TR4.
MONTH
EMPTY
1986
September 7.1  
October 2.7  
November 25.0  
December 20.0  
1987 
January 
February  
March
Apri 1
May
June
Jul y
August
Sept.
October
MEAN
S.E
60 .0 
5.5  
11.1 
4.3
3 .9  
0.0
11.1
3.2
2.9  
0 . 0
11.2
1.1
UTERINE
EGG
14 . 2 
43 . 2
41.6  
40 .0
20 .0 
38 . 8
33.3
34.7
49.0
54.7
44.4  
38 .7
44.1  
47 .0 
38 . 2
1 . 5
CONTENT 
LARVA I
35.7
29.7  
8.3
20 .0
0 . 0  
22 . 2
33.3  
47 .8 
21. 5 
22.0 
16.6
0 . 0
17.6
29.4  
21.1
4.2
LARVA II LARVA III
35.7
13.5
8.3  
0.0
20 .0
16.6  
0.0
4.3
11.7
13.5  
11.1 
32.2
17.6  
5.8  
13. 6
2.7
7 .1 
10 . 8 
16.6  
20 .0
0 . 0
16.6
22.2
8 . 6
13.7  
18 . 6 
16.6
25.8  
17 . 6 
17 . 6 
15.1
1 . 6
16
36
12
16
12
25 
14 
44
59
67
26 
35 
39 
28
University of Ghana          http://ugspace.ug.edu.gh
223
Table 7.3 Monthly means of uterine content lengths
for G .1ongipennis (Jan.-Sept. 1987)
UTERINE CONTENT LENGTH (mm)
EGG LARVA I LARVA 11 LARVA I
MONTH
January 1. 91 2 . 32 4 . 48 -
February 2 . 0 1 2 . 28 4.24 5.44
March 1.85 2 . 60 4.30 5.19
Apri 1 1. 99 2.78 3.84 5 . 45
May 1. 99 2.54 4.29 5.73
June 2 . 0 0 2 .07 3.17 5.43
Jul y 2 . 0 0 2. 64 3.25 5.01
August - - - -
Sept. 1. 99 2 . 1 1 3 . 50 5.21
October 2 . 0 0 2 . 30 3 . 23 5 . 40
MEAN 2 . 0 0 2 . 35 3.31 5.20
S.E 0 . 0 1 0 . 19 0 . 23 0 . 23
N 203 106 62 6 8
University of Ghana          http://ugspace.ug.edu.gh
224
Table 7.4: Mean percentages (± 1 S.E) and lengths of
uterine contents (mm 
pregnancy stages.
± 1 S.E.) at the four
Pregancy stage Percentage Mean length
Empty Uteri 8 . 6 -
Egg stage 40.0 ± 1.5 1.97 ± 0.02
Larval 21.4 ± 2.6 2.40 ± 0.19
LarvalI 14.3 ± 2.3 3.81 ± 0.23
Larval 11 15.7 ± 2.3 5.35 ± 0.25
7.5: Mean length of the largest developing follicule a
the di f ferent pregnancy stages.
Pregnancy
confidence
Sample Mean Standard 95%
stage size 1 ength(mm) error limits
E 192 0 . 93 0.03 0.87-0.99
LI 108 1.48 0.05 1.38-1.58
L 11 59 1.85 0.04 1.77-1.93
LI 11 58 1. 96 0.02 1.92-2.00
University of Ghana          http://ugspace.ug.edu.gh
Pe
rc
en
ta
ge
 
em
pty
 
ute
ri 
Pe
rc
en
ta
ge
 
em
pty
 
ut
er
i
£ £ 0
Month
Figure 7.1: Monthly changes in the apparent abortion rate. 
(_____ transect 1; y..jf.transect 4).
30 40 50
Min. relative humidity
Figure 7.2: The relationship between monthly abortion rate
and minimum relative humidity (Y = 30 -  0 64X' r  = 
-0.53, PC0.05). '
University of Ghana          http://ugspace.ug.edu.gh
226
larva. The largest follicule of nulliparous flies in ovarian 
age category Oa measured 0.82 ± 0 .0 1 mm and those of Ob flies 
measured 1.67 ± 0.02mm.
The rate of growth of the follicle between pregnancy 
stages may be roughly estimated by taking the differences 
between two successive stages and dividing by the 
corresponding developmental period between the two pregnancy 
stages. Recalling the developmental duration of the different 
pregnancy stages (3.5 days for egg, 1.5 days for larva I, 2 
days for larva II and 2 days for Larva III), the growth rate 
of the follicule was estimated as follows: 0.16mm/day between 
egg and larva I, 0.25mm/day between larva I and larva II and 
0.06mm/day between larva II and larva III. The growth rate of 
follicle is therefore slow in the beginning, accelerating 
later on and slowing down again towards the end. The observed 
trend in the growth rate of the follicle agrees with the trend 
of follicular growth in tsetse proposed by Saunders (1960) and 
also observed by Madubunyi (1978) for G. m. morsitans in 
Zambia.
7.3.4. Size of adult flies
Fig. 7.3 shows the monthly changes in mean wing vein 
length of female flies on the two transects. The changes in 
the wing vein length followed similar trends on both transects 
with quite close corresponding values. Mean wing vein length 
range from 2.201 mm to 2.257 mm on TR1 and 2.205 mm to 2.275 
mm on TR4.
University of Ghana          http://ugspace.ug.edu.gh
Figure 7.3: Monthly changes in the wing vein length on
transects 1 and 4 in relation to changes in mean 
monthly minimum and maximum temperature and 
relative humidity.
University of Ghana          http://ugspace.ug.edu.gh
228
The effect of climatic factors (temperature and relative 
humidity) on fly size was examined. Changes in these factors 
are also shown in Figure 7.3. On examining the changes in wing 
vein length and those of temperature and relative humidity 
there are some apparent related trends. Periods of high 
relative humidity (and of low temperature) seemed to be 
followed (about two months later) by increases in wing vein 
length. Relationships were thus investigated by plotting the 
wing vein length against minimum relative humidity of the 
previous month but one. Figures 7.4A and 7.5A show the 
scatterplots for TRl and TR4 respectively. A linear fit showed 
no significant relationship. The function best describing the 
relation is a parabola in each case. The function indicates 
that below relative humidity of 40% the wing vein length 
increases with increasing relative humidity but above 40%, the 
wing vein length decreases. The breakdown of the linear 
relationship was possibly due to invasion of the area by flies 
which are smaller. These periods obviously occur during the 
long rains and short rains when there is every evidence that 
fly movement is very high (see discussion for more details). 
When the points for these months are removed from the plot, 
significant linear relationships are shown on both transects 
(Figs. 7.4B and 7.5B). The regression equations of wing vein 
length on minimum relative humidity below 40% were:
Y = 1.78 + 0.004X (r = 0.75, P<0.01) for TRl
and
Y =2.06 + 0.005X (r = 0.82, P<0.01) for TR4.
University of Ghana          http://ugspace.ug.edu.gh
Wi
ng
-ve
in
 
len
gth
 
(m
m.
) 
W
in
g-
ve
in
 
len
gth
 
(m
m
.)
229
Minimum relative humidity
Figure 7.4A: The relationship between the wing vein length  
in one month on transect 1 and the minimum 
relative humidity of the previous month but one.
Minimum relative humidity
Figure 7.4B: Linear relationship without data points for
months of suspected fly  immigration to transect 1 
(Y = 1.78 + 0.004X; r = 0.75, P<0.01).
University of Ghana          http://ugspace.ug.edu.gh
A larger data set of wing vein length measurements collected 
from the study area from June 1983 to October 1986 (courtesy 
Dransfield et al.,unpublished) was also subjected to the 
analysis described above. A significant, though still weak, 
positive linear correlation was observed between the wing vein 
length and the minimum relative humidity of the previous month 
but one. The regression of wing vein length (Y) on relative 
humidity (X) was as follows;
Y = 2.206 + 0.001X; (r= 0.35, P<0.05)
The possible association between mortality of the 
parental generation and the size of the progeny was also 
examined. Correlations were carried out between the mean wing 
vein length and mortality rates (both from ovarian aging and 
Moran curves) of the previous month but one, but no 
significant relationship was established.
4. DISCUSSION
The observed rates of insemination are much higher than 
those observed for G. pallidipes in the same study site 
(Chaudhury, pers. comm.). He observed that very few of Oa (1-3 
days old) female G. pal 1idipes were inseminated compared with 
about 8 8 % of the Ob (4-8 days old) flies. Vale et al.(1976) 
observed that no G. pallidipes females were inseminated before 
they had fed. Compared to G. pallidipes, higher insemination 
rates were recorded for teneral G. m. morsitans (Okiwellu, 
1977) and G. p. gambiensis (Challier, 1973).
230
University of Ghana          http://ugspace.ug.edu.gh
231
The higher insemination rate observed in young flies, as 
in the case of G. longipennis implies, that the species mates 
not long after emergence as opposed to G. pallidipes which is 
reported to mate only after 4 to 6 days following emergence 
(Snow, 1980). Mating in G. longipennis may be taking place 
either at'the breeding site or at the host during the first 
blood meal. It was a common sight in the study area, to 
observe mating pairs of G. longipennis coming to and getting 
caught in traps or electric screens. The apparent decrease in 
the insemination rate during the rainy season may be due to 
the fact that fly contact is reduced by dispersal of the 
population during this period.
The occurrence of peak abortion rate at different time 
periods on the two transects could imply significant 
differences between the microclimates of the two areas during 
this period. Looking back at the suggested pattern of fly 
movement between the two areas based on Figs. 6.11A and B, 
flies appear to be moving out of TR4 from October to December 
(at the time percentage abortions are increasing on this 
transect). On the other hand there was apparent movement 
towards TRl during these months when abortions were on the 
decline in this area. This probably indicates that flies 
tended to move away from areas where the microclimate was 
unfavourable at one time (as indicated by increased abortions) 
to where the microclimate was more favourable. The fly 
movement and difference in peak abortion rate could also be 
partially associated with host movement between the two
University of Ghana          http://ugspace.ug.edu.gh
232
transects. Flies moving from one area to the other during the 
above mentioned periods may be following hosts. Since 
malnutrition is regarded as a major cause of abortion the 
decrease in host population in one area could therefore result 
in an increase in the abortion rate in that area and vice 
versa.
The relationship between abortion rate and climatic 
factors indicated that atmospheric dryness appears to 
adversely affect the pregnancy more than high temperatures, 
although the two factors are often closely linked. Low 
relative humidity could be having a direct physical stress on 
the flies. On the other hand it could again be an indirect 
relationship associated with the weather and hence vegetation 
changes that bring about the movement of host form the study 
area as a whole.
Most of the work on tsetse reproduction both in the 
laboratory and in the field indicates a 9-10 day reproductive 
cycle with the duration of the egg and the three larval stages 
being 3-4, 1-1.5, 2.5 and 2 days respectively. If a 9-day 
cycle is also assumed for G. longipennis, the duration of the 
egg and larval stages can be fixed at 3.5 days (egg), 1.5 days 
(first instar), 2 days (second instar) and 2 days (third 
instar). If the probability of capturing a female in one 
pregnancy stage is proportional to the duration of each stage, 
then these probabilities should be 0.389, 0.167, 0 . 2 2 2  and
0.222. Comparing the proportions obtained in this study to 
these expected values, there is some degree of agreement. The
University of Ghana          http://ugspace.ug.edu.gh
observed value for the egg-stage (40%) is quite close to the 
expected value. However, the larva I stage is over-represented 
(21.4%) whilst larva II and III stages are under-represented 
(14.3 and 15.7% respectively) This might mean that the 
assumed duration of larval stages for G. 1ongipennis are 
incorrect. An alternative explanation is that trap induced 
abortions are more likely to occur in flies bearing later 
instar larvae than in those bearing early instar larvae. Thus, 
some of the flies which would have been classified under 
abortions on dissection would have actually been trapped as 
second or third instar stage flies. This would therefore lower 
the observed frequencies of these categories of flies. The 
high value for flies bearing first instar larvae could have 
arisen because there were occasions during dissection when it 
was quite difficult to distinguish between first and second 
instar larvae. This could account for the apparent reversal in 
the expected values for these two categories of flies.
The average percentage flies bearing third instar larvae 
recorded in this study is markedly higher than those recorded 
for G. pallidipes at Nguruman in the same type of traps 
(Dransfield pers comm.) and in Lambwe Valley, Kenya (Turner 
1987 ). On average about 5% of G. pal 1idipes caught in the NG2B 
traps are in the 3rd instar stage. Even lower percentages are 
recorded in biconical traps. In this study, only 9% of G. 
longipennis caught in the biconical traps had third instar 
larvae which is still higher than that for G. pallidipes (1%). 
The results from studies on other tsetse species indicate
233
University of Ghana          http://ugspace.ug.edu.gh
234
flies bearing third instar larvae do not actively seek blood 
meals and are thus not readily caught in traps (which are 
known to attract mainly host-seeking flies). The results of 
this study suggests that G. longipennis still feeds readily 
during the 3rd instar stage or that the female flies are 
attracted to the trap for purposes other than host-seeking. 
Consequently, trap samples of G. longipennis are less biased 
with respect to the pregnancy stage than is the case for 
several other species.
A significant correlation between wing vein length and 
relative humidity has been observed for a number of tsetse 
populations by previous workers. Jackson (1953a) working in 
Tanzania found that G. pallidipes showed the greatest 
variation in wing vein length with seasons and G. swynnertoni 
the least. He suggested as a result that the latter species
might be better adapted to the habitat than the former. In
this study, the relationship observed for G. longipennis was 
weaker than that found for G. pallidipes by Dransfield et a 1. 
(1988, in press). They, however, used the wing vein lengths of 
only ovarian age category O flies whereas in this case the 
mean age of the whole monthly sample was used (the numbers of 
category 0 flies for G. 1ongipennis were too low to be
considered separately). It was therefore not possible to
accurately define the exact period when the fly would have 
been in utero in the parent female, and hence the period to 
take for climatic data. Another explanation for the weaker 
relationship is that a higher proportion of the population may
University of Ghana          http://ugspace.ug.edu.gh
235
have been composed of immigrants from top of the escarpment 
where climatic conditions are quite different.
The latter speculation is supported by the observation 
that the apparent outliers to a significant linear 
relationship (months with over 40% relative humidity) are data 
points for the cool months during which immigration rate is 
high. The reduction in mean wing vein length could mean that 
immigrant flies are smaller implying their origin from a 
habitat with quite different environmental conditions. The 
most immediate source of such a habitat for which there is 
evidence of invasion is the top of the escarpment where 
temperatures are much lower than in the study area. Lower 
temperatures certainly results in longer developmental periods 
(Glasgow, 1963), but whether this results in smaller flies is 
not certain. This requires a separate study, that involves 
monitoring the changes in fly size with time in comparison to 
that of surrounding fly populations.
The seasonal variation in size of G. 1ongipennis and its 
correlation with environmental factors, does suggest that 
size dependent mortality may be occurring at emergence for 
this species as has been found for other species. Whether such 
mortality does occur, should be established in future by 
comparing the size of flies emerging from field collected 
pupae with the size of tenerals caught in the field at the 
same time.
University of Ghana          http://ugspace.ug.edu.gh
236
CHAPTER EIGHT
MARK RELEASE RECAPTURE STUDIES
8.1. INTRODUCTION
Mark release recapture (MRR) is widely used for 
estimating the population size of mobile animals such as 
insects. Le Cren (1965) traced the technique back to Petersen 
in 1886 who used the recapture technique to calculate the 
mortality rate in fish. Lincoln (1930) used it to estimate the 
population size of a species of North American duck.
The method is based on the following principles.
A number of animals 'r ' are captured, marked and released. 
After a period, when these have re-mixed with the rest of the 
population, a further sample 'n* is captured and the number of 
marked individuals in the latter sample (m) noted, i.e. an 
unknown fraction of the population is first marked and on the 
second occasion a sample of this fraction is taken. Provided 
the effect of sampling error is negligible *Nf the population
A
size can be estimated by N in the following formula: 
r/N = m/n
and
/i
N = rn/m 
A
Where N is an estimate of the true population size N.
University of Ghana          http://ugspace.ug.edu.gh
237
This formula is variously referred to as the 'Petersen 
estimate' or 'Lincoln index'.
If N is to be a suitable estimate of N the following 
assumptions must hold:
1. The population should be closed i.e there must be no 
addition of individuals (birth or immigration) to the 
population between the first and the second capture occasions. 
Loss from the population will not affect the m/n fraction 
because marked individuals will die or move out at the same 
rate as unmarked ones, provided condition (2 ) is fulfilled.
2. Marked individuals must behave in every respect like 
unmarked ones. In particular, they must suffer no damage and 
must not be subject to greater risk of predation or capture 
than the unmarked. Individuals must also not loose their marks 
between the first and second capture occasions.
3. Marked individuals must evenly mix with the unmarked 
population
4. The second sample is a random sample i.e. each individual 
has an equal chance of being caught; all marks must also be 
reported on recovery.
Since many animal populations are not closed, a number of 
workers have developed more complex analyses for mark-release- 
recapture data to allow for births, deaths and movement of 
marked individuals, with the underlying principles still based 
on the Petersen estimate and the rest of the assumptions. Mark 
release recapture techniques were first used to estimate
University of Ghana          http://ugspace.ug.edu.gh
238
tsetse populations by Jackson (1933). Later on Jackson (1937,
1939) extended the analysis to allow for birth rate and death
rate and the problem of immigration was considered in Jackson 
(1941, 1944, 1948). Similar analyses for a series of mark- 
release-recapture data were developed by Dowdeswell, Fisher 
and Ford (1940, 1949), Fisher and Ford (1947) and Leslie and 
Chitty (1951) for various animal populations. Bailey (1951, 
1952) examined the precision and biases of population
estimates by these methods and developed analyses for
calculating standard errors. The estimation of population 
parameters based on multiple recaptures was independently 
developed by Seber (1962), Jolly (1965), and Manly and Parr 
(1968).
The various methods developed for handling mark release 
recapture data have been reviewed by Southwood (1978)and Begon 
(1979). Seber (1973) provided the details of the mathematics 
involved in these methods. The methods for analyzing mark 
release recapture data have been grouped into five main ones 
named after the pioneers who developed them. These include the 
Jackson's positive and negative methods, the Bailey's triple 
catch method, the Fisher-Ford method,the Joly-Seber stochastic 
method and the Manly-Parr method. In addition to the general 
assumptions underlying all mark release recapture operations, 
each of these methods is based on a number of assumptions 
peculiar to it. The choice of any of them for application to 
mark release recapture data is dependent on whether the
University of Ghana          http://ugspace.ug.edu.gh
239
underlying conditions have been met. Details of these are 
provided in the reviews already mentioned above.
8 .2 .  MATERIALS AND METHODS
M a r k - r e l e a s e - r e c a p t u r e  s t u d i e s  were  c a r r i e d  out i n  two  
major  e xpe r iment s .
a .  In monthly marking e xp e r imen t s ,  whereby  f l i e s  were  
marked on the f i r s t  day ( on  T r an se c t  1)  and on the second  day  
(on T r an se c t  4)  o f  e ve ry  6 - d a y - l o n g  monthly  v i s i t  t o  the  
f i e l d .
b .  In a con t inuous  7 - day  m a r k - r e l e a s e - r e c a p t u r e  
exper iment  unde r t aken  once i n  the cou r s e  o f  the  s tudy  p e r i o d .
8 . 2 . 1 .  Monthly  m a r k - r e l e a s e - r e c a p t u r e  exp e r iment s .
These  were  s t a r t e d  i n  Fe b r ua ry  1986 i n  the t r a n s e c t  1 
a r ea  and i n  August  1986 i n  the  t r a n s e c t  4 a r e a .  On the  f i r s t  
and second even ing  o f  each monthly  v i s i t  to  the f i e l d ,  a 
number o f  odour b a i t e d  F3 t r a p s  were  s e t  up i n  the  t r a n s e c t  1 
and t r a n s e c t  4 a r e a s  r e s p e c t i v e l y  and the c a t ch e s  o f  G. 
longipennis  were  c o l l e c t e d  a t  about  0700h the  f o l l o w i n g  
morning f o r  mark ing .  Because  o f  the  g e n e r a l  s amp l i ng  s c h e d u l e ,  
f l i e s  cou ld  not be marked immed i a t e l y  on c o l l e c t i o n  but  to  
minimize  s t r e s s  and improve s u r v i v a l ,  the  f l i e s  were  a lways  
kept coo l  under  mois t  b l a c k  c l o t h  f o r  the  hour  o r  so  b e f o r e  
marking took p l a c e .
The f l i e s  were marked u s i n g  a r t i s t ’ s o i l  p a i n t  (W i n s o r  
and Newton ) ,  w i th  a d i f f e r e n t  c o l o u r  o f  p a i n t  used each month.
University of Ghana          http://ugspace.ug.edu.gh
240
By using six different colours of paint in a known sequence, a 
particular colour could only be repeated after six months. It 
was assumed that this time interval was long enough for flies 
marked with a given colour to have died.
The main colours of paint used were ultramarine blue, 
undercoat white, vermillion, red and chrome-yellow. The 
various colours used for marking were then composed from these 
by mixing the above in the appropriate proportions. The 
colours used for marking flies were blue, green, yellow, 
orange, pink and turquoise. A few drops of linseed oil was 
used when necessary to bring the paint to a consistency that 
was neither too thick nor too thin; thick paint marks were 
likely to flake off whilst very thin paint marks were more 
likely to smudge.
In both transect 1 and transect 4 areas, marking and 
release took place in sites that were roughly in the centre of 
the area. During marking, each fly was carefully removed from 
the cage, held between the fingers and marked with two 
distinct spots of the paint on the dorsal side of the 
mesothorax, care being taken to keep marks as small as 
possible. This was to ensure that marked individuals were not 
more conspicuous to predators than unmarked ones. Flies marked 
on transect 1 had the two spots positioned side by side whilst 
those marked on transect 4 had their spots positioned one 
above the other. In both areas, teneral flies were 
distinguished by a third spot on the scutellum.
University of Ghana          http://ugspace.ug.edu.gh
241
Marked flies were immediately transferred to a larger 
retaining cage of mosquito netting that was kept cool and damp 
under cover of a wet black cloth. As marking proceeded, 
batches of marked flies remained in the retaining cages for 
15-20 minutes before they were released. This was to allow 
flies to settle down following handling so as to minimize an 
escape response. For release, the cage of marked flies was 
taken into a well shaded thicket by the Oloibototo River and 
the cage opened for flies to fly out. Flies that failed to fly 
out of their own accord were considered moribund and were 
ki1 1 ed.
Recaptures of marked flies were then obtained from 
catches by all traps that were operating in the study area. 
Until February 1987, these traps were mainly those used for 
the 6 -day regular population sampling and those being used in 
experiments that were run within the six days of every monthly 
trip. From February 1987, a population suppression operation 
using odour baited NG2B traps was started in the transect 1 
area. To monitor the changes in the fly populations during 
this operation, a continuous sampling programme was started 
whereby samples were regularly collected from 2 0  of the 
control traps. This therefore provided a more extended period 
of continuous sampling, instead of just five days, after every 
monthly marking occasion.
University of Ghana          http://ugspace.ug.edu.gh
242
8.2.2. Seven-day continuous marking experiment
Because recapture rates for the monthly marking were very 
low, an experiment was carried out whereby flies were 
continuously marked and released on a daily basis for seven 
days from the 28th January to the 3rd February 1987- Date 
specific marks were obtained by using a different colour each 
day. In order to distinguish flies marked in this experiment 
from those marked on the monthly basis, the two spots of paint 
were applied diagonally on the thorax; one spot in the top 
right hand corner of the pronotum and the other in the bottom 
left hand corner of the metanotum. The colours used were blue, 
green, yellow, orange, pink, turquoise and white.
F3 traps were again used to collect flies each day at 
about 1730h but in this case the trap catches were collected 
at 1930h the same evening. The flies were then marked the same 
evening and retained in a large cage under the cover of damp 
black cloth to reduce activity and improve survival of flies. 
At about 0530h the following morning the cages of marked flies 
were taken to the same site as previously and the cage left 
open for flies to escape. For every day that flies were being 
collected for marking, the recaptures of the previous days' 
marks were recorded. The week of continuous mark-release- 
recapture was immediately followed by the regular monthly trip 
for the month of February thus providing another week of daily 
sampling in which recaptures continued to be recorded. In 
addition, recaptures were also obtained ftfom the fly samples 
taken in by the 20 NG2B traps involved in the population
University of Ghana          http://ugspace.ug.edu.gh
243
suppression operation which started on the 4th February 1987 
(one day after the last marking day in this experiment).
8.2.3. Analysis of data
Because of the low number of recaptures obtained from 
each of the monthly marking occasions, the data were pooled 
over several months before any attempts could be made at 
estimating any population parameters. The data from February 
1986 to January 1987 were pooled together because they were 
obtained before the population suppression operation. During 
the suppression period, monthly marking occasions from 
February to May 1987 and June to October 1987 were pooled for 
males, as were February to October 1987 for females. The data 
for males and females were always kept separate. Whilst it 
would have been preferable to analyze months separately, it 
was felt that the above procedure would at least give average 
estimates of population size for the different time periods.
An average recapture rate was calculated for each 
sampling period after marking. To allow for the variation in 
the samples sizes on the different sampling occasions, the 
recaptures were corrected according to the method proposed by 
Jackson (1948). When data for more than one marking occasion 
are pooled, as in this case, the method is further modified 
(Rogers and Randolph, 1986) to give the following formula for
University of Ghana          http://ugspace.ug.edu.gh
244
average corrected recapture rate (Yn).
Yn = 5 Rni . 106 /J>CoiCru
where
i ranges from 1 to k marking occasions that are being pooled 
Rni is the number of flies marked on the ith marking occasion 
that were recaptured on day n,
Coi is the number of flies released on day 0 of the ith 
occasion and
C m  is the number of flies caught on day 'n* of the ith 
occasion
For the 7-day daily marking experiment the data were first 
grouped into four marking occasions, by considering every two 
consecutive days of marking as one marking occasion. This 
yielded 3 pairs of marking occasions for the first six days 
plus the 7th day taken alone as the fourth. Again, because of 
very few recaptures, samples taken after marking had to be 
pooled over a number of days to produce the unit periods after 
marking, for each of which the average recapture rates were 
then calculated. The first two periods consisted of 2 days 
each, days 0 and 1 being the first sampling period after 
marking and days 2 and 3 forming the second. Periods 3 and 
above each consisted of data pooled at one weekly intervals. 
Finally, the average corrected recapture rate for all the four 
marking occasions was calculated for each period after marking 
using the same formula as already given above.
University of Ghana          http://ugspace.ug.edu.gh
245
Estimation of population size
Given the data obtained from the monthly marking 
experiments, the method of analysis found most suitable for 
estimating population sizes was that based on a model 
suggested by Parker (1955), details of which are given by 
Seber (1973). In brief the model is applicable when mark 
release is carried out on a single occasion followed by a 
series of 'instantaneous’samples (sampling time being 
negligible) being removed from the population. It is similar 
to Jackson's positive method described by Begon (1979) and 
allows variable mortality and recruitment in the population. 
The model is based on the equation:
E [mi/ni] = M i/Ni 
where mi is the number of marked individuals in the ith 
sample, ni is the size of the ith sample, Mi is the size of 
marked individuals just before the ith sample and Ni is the 
size of the total population just before the ith sample, i.e. 
the expected ratio of marked to unmarked individuals in the 
ith sample is the same as the ratio of all marks in the 
population just before the ith sample. Thus, if recruitment is 
taking place, Mi/Ni will decline with time. Parker suggested 
that if mi/ni is plotted against ti the curve can be extended 
to t = 0, to obtain an estimate M 0 /N0 and hence, N0 . It is 
argued that such a method will always give a graphical 
estimate of No irrespective of the changes that take place in 
the population and provided there is a reasonable smooth trend 
in the mi/ni and the above equation is valid. When possible a
University of Ghana          http://ugspace.ug.edu.gh
246
suitable transformation m i / m  may be used to linearize the 
regression so that No may be estimated by least squares 
method.
In the present experiments the equivalent of nu/ni, 
corrected average recapture rate was plotted against days 
after marking. The logarithmic transformation of the corrected 
recapture rate was used in an attempt to stabilize the 
variance and linearize the regression. A population size 
estimate was only made if the regression was significant 
indicating a smooth trend of decline in the corrected 
recapture rate.
For comparison, the Jackson’s positive method was also 
used to estimate the population from data obtained from the 7- 
day continuous marking experiment. It is essentially the same 
as Parker's method except the grouping has been carried out 
differently, and weighting is used to cope with small numbers 
of recaptures. In this study the whole 7-day marking was 
considered as one marking occasion i. e. date specific marks 
were ignored after the last day of marking and the total 
number of marked individuals released was adjusted by 
subtracting all recaptures that were made before the last day 
of marking. Although samples after marking were obtained from 
various traps that were emptied at different intervals, all 
samples were pooled and grouped as if sampling were done at 
weekly intervals.
University of Ghana          http://ugspace.ug.edu.gh
247
According to the method:
if qi is the proportion of the ith sample that is marked 
qi = mi/ni
where
mi = the number of marked individuals in the ith sample, 
ni= the size of the ith sample.
This proportion should not be affected by looses from the 
population since marked individuals are assumed to die and 
emigrate at the same rate as unmarked ones. If, however, there 
is recruitment to the population then qi will decline with 
time since only unmarked individuals will be added to the 
population.
The aim is to estimate N o , the population at the beginning of 
marking before any additions to it and this can be achieved by 
estimating q0 the marked proportion of a hypothetical random 
sample taken on day 0 , which should be the same as the marked 
proportion in the total population: 
qo = ro /No /
where r0 is the number of individuals marked and released on 
day 0 .
Therefore 
No = r0 /qo .
To estimate q0 a birth or gain rate fb' is defined.
By calculating the proportion of marked and unmarked 
individuals surviving from one day to the next, it can be 
shown that:
q± = ^ ( l - b ) 1 equation (2 )
University of Ghana          http://ugspace.ug.edu.gh
248
lnqi = i(ln(l-b)+lnq0 equation (3)
this is a regression equation of lnqi on i, both of which are 
known from the data and the regression constants ln(l-b) and 
lnq0 can be calculated.
To take care of less accurate estimates of qi s from small 
recapture values which are susceptible to chance effects the 
mi values can be used as weights to obtain the following 
equations:
ln(l-b) = ^ m i (lnqi-lnq)(i-i)/£jn* (i-i ) 2 equation (4)
lnq0 = lnq-ln(1-b)i. equation (5)
University of Ghana          http://ugspace.ug.edu.gh
249
8.3. RESULTS
8 .3.1 .Monthly marking experiments
Tables 8.1 and 8.2 summarize the results of the monthly 
mark release recapture experiments carried out between 
February 1986 and January 1987 on males and females 
respectively. The corrected daily recapture rates, shown at 
the bottom of each table, were then plotted against days after 
marking to give Fig. 8.1 for the males and Fig. 8.2 for the 
females .
The male data suggests increases in recapture rate on days 3 
and 5 implying increased activity on these days following 
marking. The plot for the females shows a general decline in 
the recapture rate with small increases on days 2 and 4.
On population estimates, the regression of male recapture 
rate against days after marking was not significant mainly 
because of the enormous increase on the fifth day. No attempt 
was therefore made to estimate population size for males. For 
the plot on the female recapture rates, a regression line was 
fitted to all the points over the five days. The female 
population size was estimated at 20,892, which represents the 
average population size from February 1986 to January 1987, 
since the recaptures rates were averages obtained for the 
marking in these months. From the slope of the regression line 
the dilution rate of the population was estimated as 3 4 .5 % per 
day. This, however, is almost certainly an overestimate of the 
dilution rate since it is affected by fluctuations in the 
recapture rate due to the feeding cycle.
University of Ghana          http://ugspace.ug.edu.gh
250
Table 8.1: Numbers of marked male G. longipennis released each month, 
percentage recaptures within five days, numbers recaptured 
and total numbers caught on various days after marking in 
each month from February 1986 to January 1987.
MONTH NO. DAYS AFTER MARKING(n=l-5)
(i) RELEASED (Recaptures(Rn) and sample size(Cn))
(Co) 1 2 3 4 5
FEB 116 1 1 1 0 2
172 177 187 136 182
MAR 41 1 0 0 0 0
59 175 107 159 96
APR 105 3 0 0 0 0
539 580 451 124 180
MAY 314 3 1 3 0 3
253 311 177 169 131
JUN 38 0 0 0 0 0
50 25 50 22 33
JUL 159 0 0 0 0 1
85 64 85 108 157
AUG 73 0 0 0 0 0
34 80 104 113 87
SEPT 63 0 0 0 0 0
54 89 78 136 131
OCT 45 0 0 0 0 0
37 39 65 52 33
NOV 67 0 0 0 0 0
19 35 52 44 34
DEC 57 1 0 0 1 0
68 77 78 106 57
JAN 7 0 0 0 0 0
13 12 109 49 49
Average corrected recapture rate (CRR) for, i=February 1986--Jan1
Days after marking 1
SRiU 9 2 4 1 6
Rm /£Coi Cni xlO6 48.24 9.20 23.84 7.43 45.43
(CRR)
Log CRR 1.68 0.96 1.38 0.87 1.66
University of Ghana          http://ugspace.ug.edu.gh
251
Table 8.2: Numbers of marked female G. longipennis released each month, 
percentage recaptures within five days, numbers recaptured 
and total numbers caught on various days after marking in 
each month from February 1986 to January 1987.
MONTH NO. DAYS AFTER MARKING(n=l-5)
(i) RELEASED (Recaptures(Rn) and sample size(Cn))
(Co) 1 2 3 4 5
FEB 112 0 0 0 0 0
39 29 60 38 74
MAR 47 0 0 0 0 0
34 103 70 58 17
APR 120 0 0 0 1 0
190 483 283 1 2 0 108
MAY 472 2 5 1 0 0
199 203 156 140 158
JUN 39 0 0 0 0 0
21 4 42 6 36
JUL . 177 0 1 0 0 1
46 51 70 152 188
AUG 168 2 0 0 0 0
18 51 45 39 69
SEPT 39 0 0 0 0 0
OCT 54 0 0 0 0 0
NOV 97 0 0 0 0 0
DEC 40 0 0 0 0 0
JAN 3 0 0 0 0 0
Average corrected recapture rates (CRR) for, i=February 1986-Aui
Days after marking
^  Rm
Rni/^COiCnixlO6
(CRR)
Log CRR
4
29.65
1.47
6
33.41
1.52
1
7.20
0.86
1
8.25
0.92
1
7.0
0.85
University of Ghana          http://ugspace.ug.edu.gh
Log
 
CR
R 
L°
9 
CR
R
February 1986 to January 1987 (males)
Days  after marking
Figure 8.1: Changes in the recapture rate of male G. 
longipennis with days after marking. 
(February 1986- January 1987)
February  1986 to January  19 8 7  (females)
Days  after marking
Figure 8.2: Changes in the recapture rate of female G. 
longipennis with days after marking. 
(February 1986- January 1987), Regression : 
Y = 1.68 - 0.18X; r = 0.86, PC0.001).
University of Ghana          http://ugspace.ug.edu.gh
253
The results for monthly marking starting from February 
1987 to October 1987 are given in Tables 8.3 and 8.4 for males 
and females respectively. Marking in these months, was
immediately followed by a continuous regular sampling
programme which provided recaptures for more extended periods 
than just five days within the month of marking. The daily 
recapture rates were therefore calculated for up to 45 days 
after marking.
Data for males were pooled into two groups; one group 
from February to May and the other from June to October. Plots 
for the average daily corrected recapture rates for the two 
groups of data are shown in Figs. 8.3 and 8.4 respectively. 
Since most of the samples were pooled over about 7 days before
calculating the recapture rate there are no marked periodic
increases in recapture rate shown in these plots that reflect 
feeding cycles. The population estimates from the regressions 
were 10,471 males for February to May with a dilution rate of 
8.0% per day and 25,703 for June to October 1987 with a 
dilution rate of 12.9% per day.
Figure 8.5 shows the plot of the daily recapture rate of 
females against days after marking. Despite the log 
transformation, the decline in the recapture rate was not 
linear. Recapture rates were very high the first three days 
but then dropped rapidly. This implies that the marks were 
rapidly diluted out. A curvilinear fit was found to be 
significant and the average population size for February to 
October 1987 estimated as 15,848.
University of Ghana          http://ugspace.ug.edu.gh
254
Table 8.3: Numbers of marked male G. longipennis released each month, 
percentage recaptures within 45 days, numbers recaptured 
and total numbers caught on various days after marking in 
each month from February 1987 to October 1987.
MONTH NO. PERIODS AFTER MARRING(n=l-8 )
(i) RELEASED (Recaptures(R) and sample size(Cn))
(Co) 1 2 3 4 5 6 7 8
FEB 85 2 1 1 0 1 0 0 1
290 166 580 501 560 539 513 184
MAR 23 2 0 0 0 0 0 0 0
203 179 236 184 427 432 490 436
APR 23 1 1 1 1 0 0 0 0
152 47 490 436 434 449 590 383
MAY 45 1 0 2 0 0 0 0 0
206 156 593 383 646 493 763 541
JUN 73 0 0 0 0 0 0 0 0
372 278 558 363 530
JUL 66 3 1 0 0 0 0 0 0
420 235 641 522 645
AUG 160 4 0 0 1 0 0 0 0
497 318 522 625 352
SEPT 223 10 0 1 1 0 0 0 0
395 169 508 331 362
OCT 7 1 1 0 0 0 0 0 0
117 250 350 374 246
Average corrected recapture (CRR)rates for: 
i=February-May
Days after marking 1 3 10 17 24 31 38 45
6 2 4 1 1 0 0 1
Rni/£.CoiCnixl06 142. 6 76.0 43.2 13.5 10.4 0 0 17.0
(CRR)
i=June-October
Days after marking 1 3 10 17
'iRni 18 2 1 2
Rni /£ Cot CmxlO6 80.6 15.9 3.5 8.4
(CRR)
University of Ghana          http://ugspace.ug.edu.gh
255
Table 8.4: Numbers of marked female G. longipennis released each month, 
percentage recaptures within 45 days, numbers recaptured 
and total numbers caught on various days after marking in 
each month from February 1987 to October 1987.
MONTH NO. PERIODS AFTER MARKING(n=l-8
(i) RELEASED (Recaptures(Rn) and sample size(Cn))
(Co) 1 2 3 4 5 6 7 8
FEB 95 5 0 0 0 1 1 1 0
318 116 560 454 413 460 458 533
MAR 38 1 0 0 0 0 0 0 0
223 183 274 533 308 231 494 349
APR 14 0 0 0 0 0 0 0 0
103 41 394 349 445 442 665 455
MAY 63 1 0 1 0 0 0 0 1
222 145 665 455 732 877 1064 545
JUN 157 12 2 1 0 0 1 0 0
6 8 6 376 545 330 572 1114 770 495
JUL 69 4 0 0 0 0 0 0 0
482 208 770 495 526 527 596 456
AUG 105 1 0 1 0 0 0 0 0
216 185 456 627 268 328 383 263
SEPT 139 6 2 0 0 0 0 0 0
274 105 373 263 304 418 440 250
OCT 8 2 0 0 0 0 1 0 0
11 2 187 315 355 250 184 644 744
Average corrected recapture (CRR) rates for:
i=February-May
Days after marking 1 3 10 17 24 31 3
"^ Rni 7 0 1 0 1 1 1
Rni /£Coi Cm xlO6 129. 4 0 9.3 0 9.7 8 . 8 7
45
1
9.6
(CRR)
i=June-October
Days after marking
Rni /£ CoiCm xlO6 
(CRR)
1 3 10 17 24 31
25 2 1 0 0 1
23. 4 18,.4 4.2 0 0 3.3
i=February-October
Days after marking 1 
£Rm 32
Rni/£CoiCm xlO6 124.6 
(CRR)
3 10 17 24 31 38 45 52
4 3 0 1 3 1 1 1
29.3 8.5 0 3.3 7.2 2.5 3.6 3
59
1
3
University of Ghana          http://ugspace.ug.edu.gh
Log
 
CR
R 
Log
 
CR
R
February to May 1987 (males)
Days  after marking
Figure 8.3: Changes in the recapture rate of male G. 
longipennis with days after marking. 
(February 1987 - May 1987), Regression:
Y = 1.98 - 0.036X; r = - 0.90, P< 0.001) 
dilution rate = 8 .0 %.
Figure 8.4: Changes in the recapture rate of male G
longipennis with days after marking.(June - 
October 1987) , Regression: Y = 1.59 - 0.060X; 
r 0.82, P< 0.001 dilution ratp = 1 9 q»
University of Ghana          http://ugspace.ug.edu.gh
257
February to October 1987 (females)
Days  after marking
Figure 8.5: Changes in the recapture rate of female G.
longipennis with days after mar king. (February - 
October 1987) , Regression:
Y = 1.43 - 0.045X + 0.0005X2; r = - 0.83,
P< 0.001, dilution rate = 9.8%.
University of Ghana          http://ugspace.ug.edu.gh
258
8.3.2. Seven-day marking experiment.
Data for the seven-day marking experiment (grouped into 
the four marking occasions) are given in Tables 8.5 and 8 . 6  
for males and females respectively. The average daily 
recapture rates were calculated up to 25 days for males and up 
to 29 days for females. The plots of recapture rates against 
days after marking with the fitted regression lines are given 
in Fig.8 . 6  for males and Fig. 8.7 for females.
Both sexes showed a decline in the daily recapture rate 
with time with damped fluctuations at more or less regular 
intervals. For males, the first increase in recapture rate 
occurred on day 3 with subsequent ones at 4-5 day intervals.
On the females a peak was also recorded on day 3 after marking 
but more regular increases occurred at 9-10 day intervals.
Population estimates were carried out in the same way as 
before by extrapolating the regression line to day 0 and 
reading off the corrected recapture rate. The populations at 
the beginning of marking (late January 1987) were estimated at
15,848 males with a dilution rate of 8.2% per day and 14,125 
females with a dilution rate of 6.7% per day. This would give 
a female percentage in the population of 47%.
Recaptures made over 4 weeks were used in estimating 
population sizes by the Jackson’s positive method. The data 
are given below in Tables 8.7 and 8 . 8  (for males and females 
respectively) showing the number of recaptures per week, the 
proportion of marked flies in the weekly samples, and
University of Ghana          http://ugspace.ug.edu.gh
259
Table 8.5: Numbers of marked male G. longipennis released in 
successive marking occasions, numbers recaptured 
and total numbers caught on subsequent days after marking 
from 28th January to 3rd February, 1987.
PER NO. NO. PERIOD
IOD RELEASED CAUGHT
(i) (Co) (Cn)
1 2 3 4 5
1 231 281 3 7 0 2 1
2 165 184 2 2 2 5 1
3 105 107 2 4 0 0 0
4 31 383 0 0 1 0  0
5 209
6 290
7 166
8 186
9 158
10 236
11 62
12 132
13 141
14 134
15 2 2 0
16 133
17 173
AFTER MARKING(n=l-17) 
(Recaptures)
6 7 8 9 10 11 12 13 14 15
2 0 0 1 1 0 0 0 0 0
2 1 0 0 0 1 0 0 1 1
0 0 0 0 0 1 0 0 0 0
0 0 0 0 0 0 0 0 0 0
Average corrected recapture rates(CRR)
Days after marking
1 2 5 7 9 11 13 15 17 19 21 23 25 27 29
1  Rn 7 13 3 7 2 4 1 0 1  1 2 0 0 0  0
CRR 59.1 123.4 26.1 46.2 17.8 35.3 11.3 0 12.6 12.9 39.3 0 0 12.6 11.1
University of Ghana          http://ugspace.ug.edu.gh
260
Table 8 .6 : Numbers of marked female G. longipennis released in 
successive marking occasions, numbers recaptured 
and total numbers caught on subsequent days after marking 
from 28th January to 3rd February, 1987.
PER NO. NO. PERIOD AFTER MARKING(n=l*-17)
IOD RELEASED CAUGHT (Recaptures)
(i) (Co) (Cn)
1 2 3 4 5 6 7 8 9 10 11 1 2 13 14 15 16 17
1 198 326 3 4 2 0 4 4 1 2 0 0 0 0 1 0 1 0 2
2 216 219 1 7 2 1 2 1 2 1 1 1 1 0 0 0 0 0 0
3 136 146 4 5 0 1 2 1 1 0 0 0 0 0 0 0 0 0 0
4 24 197 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
5 212
6 318
7 116
8 308
9 141
10 240
11 61
12 134
13 137
14 104
15 184
16 91
17 130
Average corrected recapture rates(CRR)
Days after marking
1 2 5 7 9 11 13 15 17 19 21 23 25 27 2
iRn 8 16 4 2 8 7 4 3 1 1 1 0 1 0 1
CRR 58.6 152.1 38.7 16.2 62.8 56.5 37.7 26.3 12.0 13.2 17.7 0 14.4 0 14.
University of Ghana          http://ugspace.ug.edu.gh
Log
 
CR
R 
L°
9 
CR
R
261
Days  after marking
Figure 8 .6 : Changes in the recapture rate of males with days 
after marking. (28th January - 3rd February 1987). 
Regression: Y = 1.80 - 0.37X; r=0.83, P<0.001, 
dilution rate = 8.3%.
Days  after marking
Figure 8.7: Changes in the recapture rate of females with
^  markmg- (28th January - 3rd February 
1987). Regression: Y = 1.85 - 0.30X: r=0 7 4  
P<0.001, dilution rate = 6 7%
University of Ghana          http://ugspace.ug.edu.gh
2 6 2
estimates of the population size and gain rate (b). Population 
size estimates were very close to those made by Parker’s 
method.
Table 8.7: Weekly recaptures of male G. longipennis marked 
from the 23rd January-3rd February 1987
Week(i) 0 1 2  3 4
ri 510
No. caught(ni) 754 580 501 560
No. recaptured(mi) 16 3 1 2
qi(mi/ni) 0.0212 0.0052 0.0020 0.0036
lnqi -3.853 -5.264 -6.217 -5.635
A
0.554 N = 13,421
Table 8 .8 : Weekly recaptures of female G. longipennis
marked from the 28th January to 3rd February. 
Week(i) 0 1 2  3 4
ri 647
ni 671 568 454 413
mi 18 7 3 3
qi(mi/ni) 0.0268 0.0123 0.0660 0.0073
lnqi -3.518 -4.396 -5.019 -4.925
b = 0.416 AN = 15,081
University of Ghana          http://ugspace.ug.edu.gh
263
Table 8 .8 : Weekly recaptures of female G. longipennis
marked from the 28th January to 3rd February.
Week(i) 0 1 2 3 4
n  647
ni 671 568 454 413
mi 18 7 3 3
qi (mi / m  ) 0.0268 0.0123 0.0660 0.0073
lnqi -3.518 -4.396 -5.019 -4.925
0 . 416 ft = 15,081
University of Ghana          http://ugspace.ug.edu.gh
264
8.3.3. Relationship between relative and absolute population 
estimates
Table 8.9 shows the apparent and absolute population 
size estimates for the same periods. Relative densities are 
the averages over the same months for which the average 
absolute population size was also estimated.
There does appear to be quite a good correlation between 
relative and absolute population estimates although more data 
would be required to establish this statistically.
Table 8.9: Relative and absolute population estimates for G.
1 ongipennis.
Month(s)
Relative 
estimate 
males females
Absolute 
estimate 
males females
20,892Feb. 1986-Jan. 1987 
Jan. 1987
4.0
3.7 
2 . 5 
5 . 2
4.0
4.4 
2.1 
2.2 
5 . 9 
4.2
15,848 14,125
10,471
25,703
17,340 15,848
Feb. 1987-May 1987
Jun. 1987-Oct. 1987 
Feb. 1987-Oct. 1987
University of Ghana          http://ugspace.ug.edu.gh
265
8.3.4. The assessment of fly movement between Transect 1 and 
Transect 4 areas.
The monthly marking experiments on transect 4 yielded 
widely scattered recaptures which were considered inadequate 
for population size estimates. The data were however useful 
for assessing fly movement between the two transects. Only 
data from the monthly marking occasions from February 1987 
onwards were used for assessing fly movement between the two 
areas, because during these months there was more continuity 
in the sampling between marking occasions than before.
To quantify the movement of flies between the two areas 
the methods proposed by Randolph and Rogers (1984) were 
adopted. According to these authors the exchange of flies 
between two areas can be quantified in two different ways.
a) by calculating the probability of capturing a marked 
fly either at the site of marking (Rnsitel/Cnsit©l ) or after 
transfer to the other site (Rnsitel/Cnsite2 ) or
b) by calculating the percentage of the total number of 
recaptured flies marked at one site that have moved to the 
other site before recapture (RnSite2 x 100/Rnaitei + Rns i t• 2
Estimates of fly movement by the two methods are given in 
Table 8.10. By the first method, the probability of capturing 
marked flies on transect 4 was generally higher than it was on 
transect 1 and this included the capture of flies that crossed 
over from one transect to the other. By the second method, 
estimated percentage cross-overs were: 31.6% of the marked and 
recaptured transect 4 males and 42.9% of the females moved
University of Ghana          http://ugspace.ug.edu.gh
266
Table 8 .t0' The assessment of movement of G. longipennis between 
Transect 1 (TR1) and Transect 4 (TR4) from 
February 1987-September 1987.
Site
Marked
Site
Caught
Co Cn Rn Rn/Cn Rn«it«2 xl00/ 
RnB i t•l+Rns it# :
Hales
TRl TRl 698 16182 42 0.003
TR1 TR4 698 1849 2 0 . 0 0 1 4.54
TR4 TR4 630 1849 13 0.004
COiiX
TR4 TRl 630 16182 6 0.0004 31.58
Females
TRl TRl 675 17090 46 0.004
TRl TR4 675 1465 3 0 . 0 0 2 6 . 1 2
TR4 TR4 719 1465 24 0 . 0 1 1
TR4 TRl 719 17090 18 0 . 0 0 1 42.86
Co=total number marked and released 
Cn=total number of flies captured 
Rn=total number of flies recaptured 
***p<0 .001
University of Ghana          http://ugspace.ug.edu.gh
267
into transect 1 over the period whilst only 4.5% and 5.8% 
females moved in the reverse direction. A chi-square analysis 
showed that these differences were highly significant 
(X2=8.25, PcO.Ol for the males and 17.20, P<0.001 for the 
females) .
University of Ghana          http://ugspace.ug.edu.gh
268
8.4. DISCUSSION
The expression recapture rate is often used loosely to 
refer to the percentage of marked flies that have been 
recaptured. The value of this is dependent mainly on the 
efficiency of the sampling system and the behaviour of the 
flies after release. The corrected recapture rate proposed by 
Jackson (1948), on the other hand, is dependent on the size of 
the total fly population as well as on the other factors 
mentioned above. In this study, the corrected recapture rates 
appeared to be generally low which is most likely due to a 
high dispersal rate. Inferring from the rate of fly movement 
in the transect 4 direction alone, it appears that G. 
longipennis actually moved freely within a much larger habitat 
than was covered by the trap sampling arrangement, so that 
marked flies appeared to move out of the sampling area 1-3 
days after they were released. They may then return into the 
sampling area after some time. This seems to explain the 
tendency to a curvilinear decline in recapture rate with 
time, i.e. high recapture rates in the first few days after 
marking dropping to very low rates soon after.
Phelps and Vale (1978) also obtained low percentage 
recaptures of marked G. pallidipes in Zimbabwe which they 
attributed to several factors including high mortality rate of 
marked flies caused by handling and/or starvation. In the 
present study, flies were trapped (most of them unfed) on the 
evening of the marking occasion at about 1845h, detained over 
night and released the following morning at about 0530h (if
University of Ghana          http://ugspace.ug.edu.gh
269
marked the previous night) or at 0900hr after marking. Such 
interruption with the feeding activity of the flies could 
possibly lead to increased mortality compared to unmarked 
individuals. Another factor could be their failure to find 
suitable daytime resting sites when released outside their 
normal activity periods. The latter factor was circumvented in 
the 7-day marking experiments by releasing flies at dawn.
Periodic increases in activity, as indicated by increased 
corrected recapture rate, are associated with the feeding 
interval (Glasgow, 1961; Rogers, 1977; Randolph and Rogers, 
1978 and Rogers and Randolph, 1986). However, the latter 
authors have pointed out, as earlier observed by Glasgow 
(1961), that the activity of female tsetse appears to be more 
influenced by the 9-10-day pregnancy cycle in which case one 
would require more than 5 sampling days after marking to 
detect such periodicity. However, the curve on the male 
recapture rate suggested a 2-3 day feeding cycle and this was 
supported by the data from the 7-day marking experiment which 
were collected over a longer period after marking. The 7-day 
experiment also provided much better data on the females and 
peak recapture rates did indeed occur at 9-10 days intervals 
after an initial peak on day three.
The above values are comparable to those reported by a 
number of workers for other Glossina spp. Jackson (1933) 
recorded increased activity of marked G. m. morsitans and G. 
swynnertoni at 4-5 days intervals. Glasgow (1961) showed that 
the feeding interval of male G. swynnertoni averaged 3.5-4.5
University of Ghana          http://ugspace.ug.edu.gh
270
days. Rogers (1977) found a 4-day periodicity in the recapture 
rates of male G. fuscipes which he thought reflected the 
feeding cycle. In a more detailed analysis using 
autocorrelations and spectral analysis on long term recapture 
data on G. p. palpal is, Rogers and Randolph (1986) found mean 
periodicities of 4 days and 7 days in two different villages 
in Cote d'Ivoire for males and a 9-10 day periodicity for 
females. Turner (1987) also reported a 4-day feeding cycle in 
male G. pallidipes in the Lambwe Valley, Kenya and detected a 
3-day feeding cycle for females which he thought occurred at 
specific stages of the developing larva. The results in this 
study strongly support the idea that the female activity cycle 
is mainly influenced by a 9-10-day pregnancy cycle and that of 
the males purely by the hunger cycle.
There are no previously published reports on the feeding 
intervals of G. 1ongipennis. The only mark release recapture 
experiment ever carried out on the species was done by Power 
(1964) on males only near Lake Jipe, Kenya.
He did not analyze the data for the detection of feeding 
cycles. The data have been re-analyzed here for comparative 
purposes. Table 8.11 shows the daily recaptures for the 
various marking occasions, the numbers of flies released and 
caught in subsequent samples. The average corrected daily 
recapture rates over all the marking occasions (now 
calculated) are shown at the bottom of the table.
University of Ghana          http://ugspace.ug.edu.gh
271
Table 8.11: Recaptures of male G. longipennis marked and 
released near Lake Jipe, Kenya (1964).
Days after marking 
(recaptures)
No. N o .
Caught Released 0 1 2 3 4 5 6 7
6 6 6 6 0
89 8 6 0 3
54 49 0 0 5
83 77 1 2 2 1
108 92 4 7 1 3 1
42 39 0 1 1 0 0 1
0 0 0 0 0 0 0 0 0
48 47 0 0 0 1 0 0 0 0
152 140 2 1 0 1 1 2 2 1
Average corrected dai 1 y recapture rates(CRR)
Days after
marking 0 1 2 3 4 5 6 7 8
Avg.CRR 1 .7 4.1 4 .6 2 . 2 0.7 1 . 8 1.7 0 . 6 1
University of Ghana          http://ugspace.ug.edu.gh
272
Starting from day 0, peak recapture rates were observed 
on days 2, 5 and 8 . These suggest a 2-3- day feeding 
interval, as was also observed in the present study.
On the estimates of population sizes from the monthly 
marking experiments, the values obtained are subject to large 
errors, first because of the very low recapture rates and 
secondly because the data have been pooled over several months 
with very varied conditions. Those from the 7-day marking 
experiment appear to be more reliable as they show a smoother 
decline in the recapture rate with time. Even the pooled 
estimates may be of value since there is some degree of 
agreement between:
a) the estimates made by the two different methods (Parker's 
and Jackson’s positive)
b) the estimates made for the same period using the monthly 
marking data and those made using the 7-day marking data and
c) the trend of population changes, as shown by the absolute 
estimates at different periods and by the apparent densities 
(catch/trap/day) for the same periods.
The results on fly movement indicate that the probability 
of capturing a marked fly was higher on TR4 than it was on 
TR1, which is surprising because there were fewer traps on TR4 
and hence it should be a less efficient trapping system. 
Randolph and Rogers (1984) pointed out that this probability 
in one area or the other is much influenced by the samples 
released and/or subsequently caught in the two areas. In this 
case the numbers released were about the same on bojth-—^
University of Ghana          http://ugspace.ug.edu.gh
273
transects for both sexes but the samples (Cn) taken on 
transect 1 were about 1 0 times larger than those taken on 
transect 4.
One explanation is that the large number of traps for 
suppression on TRl were causing a higher mortality rate of 
marked flies than on TR4, hence the lower recapture rates over 
a period of time. The results also show that there is a 
greater movement of flies from TR4 to TRl than in the reverse 
direction. If this movement into TRl indicates a general 
higher immigration rate (which brings in unmarked flies), then 
marks on this transect will get more diluted than those on 
TR4. This could also explain the lower probability of 
capturing marked flies on TRl.
But what should cause a higher movement of flies into 
TRl? If the data were taken from a single marking occasion, it 
could be that TRl offered more suitable micro-habitat at the 
time. Since these data cover a period of over a year, an 
explanation must be sought elsewhere. The fly population on 
TRl was being reduced by the trapping (although this is not 
apparent from the MRR estimates because the periods chosen 
included times of re-invasion - see Chapter 9) The apparent 
densities shown by the NG2B and biconical traps on the two 
transects (Figs. 6 .1-6.7) indicate trends of general decline 
on TRl but of increase on TR4. The movement of flies into TRl 
could, therefore,be in response to the vacant niche being 
created on this transect. Such movement of tsetse has
University of Ghana          http://ugspace.ug.edu.gh
274
also been suggested by Turner and Brightwell (1986),for G. 
pallidipes between an adjacent coniferous forest and a thicket 
in the Lambwe Valley, Kenya, after the latter habitat 
underwent insecticidal spraying.
Transects 1 and 4 are about 7km apart, taken from the 
points of release in the two areas. Movement of flies from one 
transect to the other could take as short as one day. For 
instance, a female fly marked on transect 4 at about 0900h on 
the 7th of June 1987 was captured in the next morning’s trap 
collections made on transect 1 at about 0700h about 8 km away 
from transect 4. No vehicle moved between these areas during 
this period. Considering that this fly was released during 
the inactive period, it probably made the 8 -kilometre journey 
during the evening activity period which lasts for about one 
hour. Although this might have been an exceptional case, many 
flies were observed to move up to 2 km in one evening and most 
flies took 2-3 days to make the crossover.
For future marking experiments on G. 1ongipennis the recapture 
rates could be improved by marking and releasing flies in the 
same evening that they were caught (as was done on the seven 
day marking experiment) so as to minimize interruption with 
the feeding activity. Secondly, for a habitat such as Nguruman 
where the physical boundaries of the population are difficult 
to define, one needs to mark
University of Ghana          http://ugspace.ug.edu.gh
275
very many flies and extend the trapping arrangement to cover 
as wide an area as possible.
University of Ghana          http://ugspace.ug.edu.gh
276
CHAPTER NINE
SUPPRESSION OF A G. LONGIPENNIS  POPULATION 
WITH BAITED NG2B TRAPS
9.1. INTRODUCTION
In the past decade, research has been intensified to 
develop tsetse control methods that are less dependent on the 
use of insecticides. Considerable progress has been made in 
using traps as a means of controlling tsetse populations 
rather than being just sampling tools. Insecticide impregnated 
targets have also been developed. The identification of odour 
baits that can be used to greatly increase catches has 
increased the potential of traps and targets for controlling 
tsetse populations.
An account has already been given in Chapter 2 (section 
2.5.7) of the trial deployment of traps in some parts of 
Africa to reduce tsetse populations to very low levels. This 
started with the use of insecticide-impregnated biconical 
traps and cloth targets against populations of G. palpalis and 
G. tachinoides in Cote d'Ivoire (Laveissiere, 1980;
Laveissiere and Couret, 1981). More research has since been 
carried out in the same country to improve the effectiveness 
of the targets and the methods of insecticide application 
(Laveissiere et al., 1988). Vale et al . (1988) also deployed
insecticide-impregnated targets to reduce populations of G. 
pallidipes in Zimbabwe.
University of Ghana          http://ugspace.ug.edu.gh
277
Research work at Nguruman was aimed at developing 
improved methods of trypanosomiasis control through a better 
understanding of tsetse population dynamics and development of 
more appropriate control technologies. By February 1987, it 
was considered that an appropriate stage was reached for a 
trial deployment of odour baited NG2B traps to suppress the 
local G. pallidipes population both to improve understanding 
of the system and to test out the new control technology. 
Unlike the trials carried out in the above mentioned cases, no 
insecticide was to be used in this operation. The effect of 
this exercise on the population of G. 1ongipennis is reported 
in this chapter.
9.2. MATERIALS AND METHODS
One major factor that was considered in this operation 
was the cost involved. Efforts were therefore made to reduce 
the cost of the trap and the odour bait to as low a level as 
possible without reducing the efficiency of the trap.
Brightwell et a1. (1987) describe the initial experiments
leading to the development of the NG2B prototype trap. 
Brightwell et a1. (in press) give an account of experiments 
that were conducted to transform the NG2B prototype into a 
operational control version and to identify the optimum dose 
rates of the odour baits. The optimum dose rate of acetone was 
found to be 150 mg/h dispensed from a medicine bottle with a 
0.2 cm diameter hole punched in the cap. Cow urine was 
dispensed from a 1 kg used cooking fat tin with the opening
University of Ghana          http://ugspace.ug.edu.gh
278
covered with polythene and a 2 x 4 cm slot cut below the rim; 
this was found to give a dose rate of about 1 0 0 0  mg/h.
The operation was limited to the main study area which 
includes transect 1 and covers about 100 km2 (see Fig. 3.3). 
Although not completely isolated from other tsetse infested 
areas, it was hoped that immigration could be reduced by 
setting up barrier traps along the invasion routes.
Based on information on the natural mortality rates and 
estimates of the population size of G. pal 1idipes it was 
estimated that an average trap density of one trap/km2 would 
increase the mortality rate enough to reduce the population to 
very low levels. It should be pointed out here that most of 
the operation was based on studies on the G. pal 1idipes 
population as there were only limited data on G. longipennis. 
Prior to trap deployment, a network of tracks were cut within 
the area with the help of the local community. To further 
involve the community, a training session on making the trap 
was given in each homestead (manyatta). Within two weeks the 
required number of 1 0 0  traps had been made by the community 
and were ready for deployment.
The month of February was chosen for the deployment of 
the traps because this was normally within the hot dry season 
when natural mortalities were highest for both G. pallidipes 
and G. longipennis. It was therefore hoped that a greater 
initial impact on the population could be achieved if the trap 
mortality was added at this time. Furthermore, it is known 
that the flies concentrate in the thicker part of the woodland
University of Ghana          http://ugspace.ug.edu.gh
279
during this season thus giving the opportunity to place most 
traps in the thicker woodland to maximize the effect.
The traps were deployed over several days in the first 
week of February. Each trap was placed in a moderately shaded 
area to avoid direct exposure to sunlight which can cause 
rapid destruction of the nylon netting of the trap. The sites 
were cleared of surrounding vegetation to improve visibility 
of the trap to flies. To the north of the area, a higher 
density of traps was placed along the narrow strip of 
vegetation leading to transect 4 which was considered to be 
one of the main sources of immigration. To the south of the 
area lay more open woodland except for a narrow neck of 
thicker vegetation along the banks of the Oloibototo river 
within which several traps were set at 1 km intervals.
The odour baits were placed 30 cm behind each trap and 
firmly tied with wire to thick wooden pegs hammered into the 
ground. This was to prevent the baits from being knocked over 
by animals.
Out of the 100 traps, 20 of them were selected to be 
sampled regularly to monitor changes in the population. The 
selection was made in such a manner that each section of the 
area covered by the operation was represented and the number 
of traps sampled from each section was proportional to the 
density of the traps in the area. Catches from these traps 
were collected at least every 3 days and once every week the 
polythene bag cages were replaced by net cages to collect 
samples for ovarian age dissection. Grease was put around trap
University of Ghana          http://ugspace.ug.edu.gh
280
supports to prevent ants from damaging samples from these 
traps.
Initially, trap servicing was carried out every two weeks 
on all hundred traps. This involved visiting each trap in turn 
to ensure that it was functioning. The trap was carefully 
inspected for holes, odour baits checked and topped up if 
necessary and all damage and their possible causes were 
recorded. For logistical reasons the frequency of this 
exercise was soon reduced to monthly intervals.
The impact of the suppression exercise was also monitored 
at monthly intervals with the biconical traps previously 
described (Chapter 4, sect 4.2). Since these extended outside 
of the suppression zone, an estimate could be made of 
percentage reduction taking into account seasonal changes.
9.3. RESULTS
The changes in the population levels of male and female 
G. 1ongipennis from February to December 1987 as indicated by 
the monitoring NG2B traps are shown in Figures 9. 1A and 9.IB 
respectively. Averages of all the monitoring traps, except 
trap 18, were taken over every 8 days. Trap 18, which was the 
outermost trap in the southern barrier, was catching virtually 
no tsetse flies and had to be moved to a different site; hence 
its omission from the analysis.
Population levels of both males and females rapidly 
declined over the first 2 months with some fluctuations from 
the general trend. Within this period trap catches declined
University of Ghana          http://ugspace.ug.edu.gh
Lo
g(
ca
tc
h/
tra
p/
da
y)
 
Lo
g(
ca
tc
h/
tra
p/
da
y)
281
Months
Figure 9.1. Changes in apparent densities of male G.
longipennis in monitoring NG2B traps: A: Excluding 
barner traps and B : all traps except trap 18.
University of Ghana          http://ugspace.ug.edu.gh
282
from about 5.8 males and 4.6 females/trap/day to about 2.1 
males and 1.5 females/trap/day giving a 64% reduction on the 
males and 67.4% on the females. With the onset of the rains, 
the numbers of both sexes increased and by the end of May the 
population levels were back to almost where they had started.
In fact the female population reached a higher level (6.3 
f1ies/trap/day) whilst that of the male was just below its 
starting level (4.8 f1ies/trap/day). A rapid decline followed 
over the next month to about 3.4 flies/trap for both sexes, 
increasing again to about 4.8 flies/trap on the average by the 
end of July. From then on there was a steady decline with some 
oscillation in the female population and by the end of October 
the traps were recording about 1.6 flies/trap for each sex. 
Rapid fluctuations were observed in November and December but 
the fly numbers were down to averages of 1.9 males and 1.3 
females/trap /day by the end of December 1987. Thus over the 
11 months the male population had fallen by 67.2% whilst that 
of the females was down by 73.9%.
Average catches from the barrier traps alone are shown in 
Figures 9.2A and 9.2B for males and females respectively. 
Catches in these traps generally follow the same trends as 
described above for all traps. The levels of population 
reduction shown in these traps are about the same as those 
shown in the general area (77% for males and 64% for females 
after the first two months). Similar increases were also 
observed in May-June and November-December except that the 
increases in the latter months were much greater in the
University of Ghana          http://ugspace.ug.edu.gh
283
I W
figure y .Z: Changes in apparent densities of male (A) and 
female (B) g. longipennis in 7 of the monitoring 
NG2B traps in the barrier.
University of Ghana          http://ugspace.ug.edu.gh
284
barrier traps than those shown in the general area. By the end 
of December the reduction levels in the barrier traps were 
51.8% for males and 60% for females.
Figure 9.3 shows the population reduction levels 
estimated from the biconical trap catches within the 
suppression area relative to those outside the area. Levels 
of population change were similar to those shown by the NG2B 
monitoring traps. Thus, there was an average 64% reduction of 
both sexes in the first two months with populations bouncing 
back to almost initial levels in June-July. By December- 
January reduction levels were about 60% for males and 90% for 
females.
The monthly age distributions and the estimated mortality 
rates from them are shown in Figure 9.4. Attention should be 
drawn to the fact that the mortality rate estimated from the 
ovarian age of flies caught in any month actually reflects the 
effects of mortality factors operating in the previous month. 
Moreover, if there is a steady decline in population size, the 
mortality rate must be corrected (Van Sickle and Phelps,
1988). The corrected mortalities are shown in the figure for 
those months in which a steady decline in population was 
observed. High mortality rates were observed in the first two 
months which then decreased over the next three and rose again 
in the following months. The overall mortality rate in 
February-March was the same (4.2%) as it was in July to 
September. The monthly age structure clearly reflects the 
corresponding levels in mortality rates i.e. there is a higher
University of Ghana          http://ugspace.ug.edu.gh
% 
R
ED
U
C
TI
O
N
285
D /J  F /M  A /M  J / J  A / S  O /N  D /J  F /M
1987
Figure 9.3: Bimonthly percentage reduction in the population
levels of male and female G. longipennis estimated 
from apparent densities from the river biconical 
traps operating within and outside the suppression 
zone.
University of Ghana          http://ugspace.ug.edu.gh
2 8 6
proportion of young flies when mortality rates are high 
(February to March and July to September).
4. DISCUSSION
It is obvious that there is a considerable seasonal 
variation in the degree of suppression of the G. 1ongipennis 
population which is linked to climatic conditions. The impact 
of trapping on the population was very great in the dry 
seasons (February-March and July-September) whilst in the 
rainy season it had little or no effect on the population, and 
numbers increased to their former level.
Three factors probably contribute to the rapid decline in 
the dry seasons. First there is an increase in the natural 
mortality rate with increasing temperature and decreasing 
relative humidity as has already been established in Chapter 6 
This however can only account for a very small part of the 
decline in 1987 since there was a considerable percentage 
reduction in the suppression zone relative to populations 
outside the zone which experienced similar climatic 
conditions.
Secondly there was a decrease in the efficiency of the 
suppression traps during the rainy seasons. It was also shown 
in Chapter 6 that flies spread out to open areas during the 
rainy season and concentrate in the thicker woodland during 
the dry season. Since there was a higher concentration of the 
suppression traps in the thicker woodland the impact of the
University of Ghana          http://ugspace.ug.edu.gh
Figure 9
287
n= 15ft
Mortality rate per day 
Estimate Corrected
FEB 19 87 - o . o 3 o - 0 . 0 4 0
ZhU
n* no
MARCH 1987 - 0 . 0 3 4  - 0 . 0 4 4
n = 13^ APRIL 19 87 -0.030
UrO
n= 73
i d
MAY 1987 - 0 . 0 2 9
n= 66
JUNE 1 9 87 -0.029
bl=L
n : 87
TTrn-.
JULY 19 87 - 0 . 0 3 4  - 0 . 0 4 1
n= 50
AUGUST 1 987 -0.032 - 0 . 0 3 9
3ZL
0 40
20
n= 76
3m SEPT. 1987 - 0  .038 - 0 . 0 4 5
0  1 2 3  u* 5 V  7
Ovar ian  age
4: The age distribution and mortality rate 
estimates of G. longipennis caught in the 
monitoring traps during population suppression 
operation.
University of Ghana          http://ugspace.ug.edu.gh
288
traps was greater during the time the flies are within this 
vegetation type.
The third and possibly most important factor influencing 
the whole operation is the seasonal immigration of flies into 
the area which was observed to be closely linked with climatic 
conditions. Although barrier traps were set up to reduce 
immigration, the barriers become less effective in the rainy 
season because the increased dispersal of the flies took them 
into very open vegetation types where there were few or no 
barrier traps. The increases in population levels in the rainy 
seasons are probably therefore mainly due to an increase in 
the immigration rate owing to a breakdown of the barrier. The 
relatively higher peaks observed in the barrier traps during 
the rains is strong evidence for immigration during these 
periods. Further evidence of immigration is the sudden rise in 
the percentage of older flies which could not have been 
derived from just the young flies of the previous months.
Therefore, the traps did have an effect on the population 
of G. 1ongipennis but to a lesser extent than they had on the 
population of G. pallidipes. This is because firstly, the trap 
and odour baits are less effective for G. 1ongipennis and 
secondly G. 1ongipennis appear to occupy more open areas than 
G. pallidipes. This means that the trap densities in these 
areas are insufficient for G. longipennis. For the same reason 
it is more difficult to set up effective barriers for this 
species because barrier traps limited to only the thicker 
woodland would only be effective for very short periods.
University of Ghana          http://ugspace.ug.edu.gh
289
The problem of immigration which applies to the 
population of G. pallidipes as well has also been experienced 
by other workers who have carried out trial deployment of 
traps to control tsetse populations (viz. Laveissiere et al . , 
1988; Vale et al., 1988). It thus appears that success in this 
technology in a very small area relies on how well isolated it 
is from other infested areas. In the absence of sufficient 
isolation the alternative is to cover as much of the infested 
area as possible and thereby lessen the immigration pressure.
University of Ghana          http://ugspace.ug.edu.gh
290
CHAPTER TEN
GENERAL DISCUSSION 
There is increasing awareness that sustained control of 
tsetse and trypanosomiasis is dependent on acquiring a sound 
knowledge of the vector/trypanosome complex. Progress in the 
past decade on the development of better control strategies 
for the vector has been the result of intensive research work 
on their behaviour and ecology and there is every indication 
that with a proper understanding of vector populations their 
control is possible.
There is, however, the problem that in many localities 
where different tsetse species occur together, attention is 
still only being directed at the so-called main vector 
species. In most cases, the position of these main vectors 
were established by previous workers who may have laid down 
background information on methods of studying them. Therefore, 
they probably are more amenable to study than species which 
have been worked on less or not at all. As a result of this 
bias there remains lack of knowledge on 'minor vectors' such 
that in most cases, there is no evidence for their non­
involvement in disease transmission. In view of this, if the 
move for vector control is to achieve its goal it is vital 
that proper knowledge be obtained on all tsetse species in any 
given system.
University of Ghana          http://ugspace.ug.edu.gh
291
GIossina 1ongipennis, one of the supposed minor vectors 
was found to be infected with trypanosomes at Nguruman.
Further studies on the species were limited by the lack of an 
effective sampling device. In such a situation it would be 
unsafe to carry out a control programme without taking this 
species into consideration. The need for more information 
about G. 1ongipennis was therefore vital for the development 
of a control strategy that would be effective for the system 
as a whole. The development of a sampling trap which was 
achieved in this project and the subsequent study on some 
aspects of the population dynamics of the species therefore 
provide a working base for developing such a control strategy.
The difference between the apparent densities recorded by 
the biconical traps and those recorded by the NG2B trap 
emphasizes the difference between the sampling efficiency of 
the two traps. This underlines the point that populations 
could remain undetected, especially at very low levels, if an 
ineffective sampling device is used. By the time field work on 
this project was completed, it was evident that the efficiency 
of the NG2B trap could be increased through further 
modification on the design and the use of more effective odour 
baits. The NG2F ('winged NG2B1), a modification from the 
original design, proved to be an improvement. On odour baits, 
the addition of octenol to the cow urine/acetone bait system 
showed a potential to increase catches but the treatment 
effect was not significant because of the generally low
University of Ghana          http://ugspace.ug.edu.gh
292
numbers of flies that were caught. Brightwell et al. (in 
press) have since conducted further experiments to show that 
the use of octenol does indeed increase the catches of G. 
longipennis by 2-4 times.
For population studies the present trap/odour bait system 
was considered adequate for obtaining basic information on the 
population dynamics of G. longipennis. The observed trends in 
population change over the study period were similar to those 
observed by Dransfield et a1. (pers. comm.) for G. pallidipes 
in the same habitat. Basically, both species attained peak 
population levels towards the end of the short and long rains
and declined to low levels in the dry seasons. This indicates
that both populations are probably responding to the same 
environmental factors. It was shown that high ambient 
temperatures and environmental dryness have adverse effects on 
the survival rate of both species. However, it is clear from
the apparent densities of the two populations, supported by
absolute population estimates, that the population levels of 
G. 1ongipennis were always much lower than those of G. 
pallidipes.
The immediate question is what maintains the difference 
in population equilibrium levels in favour of G, pallidipes. 
Two factors could be maintaining the lower population of G. 
longipennis. Firstly, it could be that G. longipennis is less 
adapted to the environment than G. pallidipes. Secondly, or as 
a consequence of the first, G. longipennis is probably being 
out-competed by G. pallidipes for some essential environmental
University of Ghana          http://ugspace.ug.edu.gh
Jresource(s). If G. longipennis was forcibly isolated by 
geographical barriers from a forest dwelling habitat as 
proposed by Machado (1959) and Evens (1953), there is the 
likelihood that it is less adapted to savannah conditions than 
most savannah species. If the two species are in competition, 
the most likely place where this occurs would be at the host. 
Vale (1974) showed that defence mechanisms when a host is 
being bitten by flies, such as skin rippling and general 
agitation,, can reduce the feeding success of tsetse. Being a 
larger species than G. pallidipes, G. longipennis would 
require a larger blood meal for larval development, in which 
case it might have more difficulty getting adequate feeds. The 
problem would be worse if they are feeding on the same host at 
the same time and using the same feeding sites on the host.
Some behavioral characteristics of G. longipennis appear 
to be adaptive efforts towards increasing the chances of 
survival under the conditions speculated above. Firstly, the 
crepuscular behaviour could be a move to both avoid adverse 
climatic conditions (high temperatures and low relative 
humidities) and to isolate itself temporally from G. 
pal 1idipes. Secondly, the high dispersal rate of G.
1ongipennis suggests an active hunting life to increase its 
chances of encountering a host. Thirdly, the high insemination 
rate of young flies (Oa and Ob) is probably an adaptation to 
increase the chances of larval production.
293
University of Ghana          http://ugspace.ug.edu.gh
294
Front the high abortion rates, it appears that feeding 
success is an important limiting factor on the population size 
of G. longipennis. Abortion rates were observed to be 
particularly high in the hot and dry seasons probably because 
fly movement is more restricted during these periods, and 
hence there are fewer chances of getting adequate feeds. 
Furthermore, although the crepuscular behaviour may be a move 
to avoid competition and the rigors of climatic conditions, 
the rather short period does not offer the best chances for 
feeding success. Abortions may therefore be playing an 
important role in maintaining the low population levels of G.
1ongipennis as was also suggested for G. palpalis in Nigeria 
by Jordan (1962b).
From the point of view of tsetse control operations, it 
would be important to establish whether competition is an 
important factor in determining the population levels of 
sympatric tsetse species. It is essential to know if the 
population levels of G. longipennis in the same habitat could 
rise to any higher levels in the absence of G. pallidipes. 
Could selective control of G. pallidipes lead to a replacement 
by G. longipennis?. Evidence for the possibility of such a 
phenomenon occurring has already been reported on some West 
African tsetse species by Laveissiere et al. (1988) in
experimental control operations (see literature review 4.6 for 
details). Therefore, more information on this would be useful 
to tsetse control technology in general
University of Ghana          http://ugspace.ug.edu.gh
295
The results of the 11 months of population suppression 
showed that the system was more effective for G. pallidipes 
than it was for G. 1ongipennis. Evidence from marking 
experiments with both species from within and outside the 
suppression zone showed that the trap barriers (at least the 
northern one) were more effective for G. pallidipes than they 
were for G. longipennis. Therefore, it is difficult to judge 
whether the lesser reduction in the population of G.
1ongipennis was due to invasion or due to trap inefficiency. 
Whatever the case, the solutions to both problems are vital 
for the effective suppression of G. longipennis populations.
This would require a further improvement on the present 
system and the coverage of more area. However, at this stage 
the question of cost-effectiveness has to be considered. It 
was for this reason that the research into trap design and 
odour attractants concentrated as much as possible on the use 
of cheap and/or locally available materials and odour baits, 
such as locally produced cotton and cow urine. Chemicals like 
acetone, octenol, etc. although imported, were normally 
dispensed at very low dose rates so that very small quantities 
were required.
The trap/odour bait system which was developed turned out 
to be quite cheap requiring only 2 metres of blue cloth, 1 
metre of black and 0.75 metres of netting. An inexhaustible 
supply of cow urine was available at no cost whilst acetone 
was required at only 150mg/day. The recent addition of an 
additional metre of blue cloth and octenol to the population
University of Ghana          http://ugspace.ug.edu.gh
suppression traps was more for the population of G. 
longipennis and in terms of trap material and odour baits the 
system now costs about US$9 per trap per year. It is 
impossible to tell how much more material and/or odour bait 
would be required to develop a more efficient system. The 
coverage of more area would require more traps as well as 
increase the cost of trap servicing.
At this point the question arises as to whether the 
baited trap is the best tool for effective control of G. 
longipennis. Are there other already developed methods of 
tsetse control that are worth considering? The results from 
the electric screen test for trap efficiency indicated that a 
lot more flies come to the trap than actually get caught. 
Furthermore, it was a common observation that quite a number 
of flies landed and stayed on the outside of the trap for a 
considerable length of time before taking off again although 
many of such flies may eventually get caught.
In view of the above, insecticide impregnated traps 
and/or targets are worthy of consideration. However, the 
present trap/odour bait system is a standard against which any 
alternative approaches must be weighed in terms of cost. 
Ideally, a decision for any change or addition should be based 
on comparing the increase in efficiency over the present 
system and the extra cost involved. However, assuming that G. 
longipennis is as susceptible to insecticides as other tsetse 
species then impregnated targets may be more economical than 
traps because of the reduction in material c o s t ^ t e  target
University of Ghana          http://ugspace.ug.edu.gh
principle was implied when the effect of cloth and no cloth 
was tested on the catch of an odour baited electric screen 
(Chapter 5). It was shown that although the black or blue 
cloth does not have the shape of the trap it caused a 
significant (3x) increase in catch over the odour bait alone. 
Targets of the Zimbabwe design have been tested against G.
1ongipennis at Galana Ranch, Kenya, but as with traps they 
appear to be having a lesser effect than on G. pallidipes 
(Dransfield, pers. comm.).
Apart from the cost involved certain, other points should 
be considered before using insecticide impregnated targets in 
place of traps. Firstly, despite the controlled use of 
insecticides are used on targets, traps are still 
environmental safer. This point is especially important in a 
place like Nguruman where wild life is still abundant and 
Masai children roam the bushes with their cattle. Secondly, 
traps are self-monitoring and the effect of control can be 
closely followed without having to set up a separate sampling 
method as would be necessary with targets. Furthermore, the 
need to involve local communities is important for long term 
control campaigns as intended for the Nguruman project. In 
view of this, trap catches, which serve as convincing evidence 
of the possibility to control tsetse, play a significant role 
in getting co-operation from the community. However, targets 
are simple to construct and set up and require less servicing 
than traps do. It would therefore be easier to deploy targets 
over a larger area than traps.
297
University of Ghana          http://ugspace.ug.edu.gh
298
In view of the pros and cons for traps and targets, the 
following approach is suggested for the control of G. 
longipennis. Since the traps presently being used are quite 
effective in the dry season, when fly movement is restricted 
to the thick woodland and taking into account the foreign 
exchange requirement of insecticides, a combination of traps 
and targets could be the most cost-effective approach to the 
control of the species. The currently baited modified NG2B 
trap can be concentrated in the thicker woodland and invasion 
routes whilst similarly baited targets should be placed in the 
more open woodland which forms the largest part of the 
habitat. Considering the high mobility of the species, targets 
could be effective at very low densities so that it may not 
require very many of them to cover a large area. The traps can 
be used to monitor population changes and detect routes of 
invasion. It may be necessary to increase the density of 
targets during the rains. Given the already low population 
levels which is apparently maintained by unfavourable 
environmental conditions, the suggested approach could keep 
the population below any significant levels.
University of Ghana          http://ugspace.ug.edu.gh
299
SUMMARY
1. Investigations were carried out at Nguruman to develop an 
effective trap/ odour bait system that could be used for 
sampling populations of Glossina longipennis Corti. The trap 
was then used to carry out studies on the population dynamics 
of this species. Lastly, traps were deployed over about 100 
km2 in an attempt to suppress the population.
2. Acetone dispensed together with cow urine or buffalo urine
increased biconical trap catches by about 4-6X over unbaited 
traps. The two chemicals appeared to act in a synergistic way 
since neither chemical on its own produced a significant 
increase in catch. Use of odour attractants did not affect the
age distribution of flies caught in the trap.
3. The Zimbabwe F3 trap was found to be more effective than 
the biconical trap, especially for females. Several new traps 
were developed at Nguruman (NGU traps) and tested for G. 
longipennis. The best of these,the NG2B, was comparable with 
the F3 and, in view of material cost and ease of construction, 
was identified as the best sampling trap for this species. The 
NG2B caught a higher proportion of older flies than the 
biconical trap.
University of Ghana          http://ugspace.ug.edu.gh
300
4. The crepuscular activity pattern of G. longipennis was 
critically investigated using an odour baited target adjacent 
to an electric screen. Morning peak catches were recorded from 
0600-0615 hours, shortly before sunrise at 0630 hours. Evening 
catches were greater with males active before females. For 
males the peak was from 1815 - 1845 hours whilst for females 
it was from 1830 - 1845 hours. No flies were caught after 0700 
hours in the morning and after 1900 hours in the evening.
Light intensity was the main factor affecting flight activity.
5. The population dynamics of G. longipennis were studied 
through regular sampling with the biconical and NG2B traps. 
Trends in apparent densities showed two annual peaks, a major 
peak in June-July following the end of the long rains and a 
minor one during the short rains in November-December. Flies 
were observed to spread out into all vegetation types during 
the rainy seasons but were concentrated within the thick 
woodland during the dry seasons.
6. Daily mortality rates estimated from ovarian age structure 
were highest during the hot, dry seasons. They showed a 
positive correlation with maximum temperature and a negative 
correlation with minimum relative humidity.
University of Ghana          http://ugspace.ug.edu.gh
301
7. The insemination rates of young female G. longipennis were 
very high throughout the sampling period being 100% in non- 
teneral nulliparous flies and over 80% in teneral flies.
8. An average abortion rate of 6% was recorded over the study 
period with a range of 0% in the rainy season to 60% in the 
dry season. Abortion rates showed a significant negative 
correlation with minimum relative humidity.
9. A significant relationship was established between adult 
fly size, as indicated by a measure of the wing vein length, 
and minimum relative humidity two months previously.
10. The absolute population size of G. longipennis in the 
study area was estimated using mark-rel ease-recapture at an 
average of 17,300 males and 16900 females. Dilution rates were 
high indicating constant movement of flies into and out of the 
sampling aiea. Absolute population estimates at different 
times of year followed similar trends to apparent densities 
from trap catches.
12. Changes in the recapture rate of marked male flies 
indicated that they fed at 2-3 days intervals whilst the 
recapture iate of females showed 9-10 days intervals 
reflecting the pregnancy cycle.
University of Ghana          http://ugspace.ug.edu.gh
302
13. A trial tsetse population suppression operation with odour 
baited NG2B traps carried out in the study area revealed that 
the system is more effective for G. pallidipes than it is for 
G. longipennis. Even so, the G. longipennis population was 
reduced by 60 -90% relative to populations outside the 
suppression zone during the dry season.
14. Barrier traps are less effective for this species due to 
its high mobility and wide distribution in more open 
vegetation types. Effective control of this species will 
probably require use of traps over a much larger area or the 
integration of insecticide impregnated targets with the traps.
University of Ghana          http://ugspace.ug.edu.gh
303
REFERENCES
Adabie, D. A. (1987). Pupal ecology and role of predators 
and parasitoids ir» natural population regulation of 
Glossina pallidipes Austen (Diptera: Glossinidae) at 
Nguruman, Kenya. PhD. thesis, University of Ghana, 
Legon.
Allsopp, R. (1984). Control of tsetse flies (Diptera: 
Glossinidae) using insecticides: a review and 
prospects. Bull. ent. Res. 74, 1-23.
Bailey, N. T. J. (1951). On estimating the size of mobile 
populations from capture-recapture data. Biometrika
38, 293-306.
Bailey, T. N. J. (1952). Improvement in the interpretation 
of recapture data. J. Anim. Ecol. 21, 120-127.
Barrass, R. (1960). The settling of tsetse flies Glossina 
morsitans Westwood (Diptera, Muscidae) on cloth 
screens. Entomologia exp. Appl. 3, 59-67
Barrass, R. (1970). The flight activity and settling
behaviour of Glossina morsitans Westw. (Diptera: 
Muscidae) in laboratory experiments. Bull. ent. Res. 
59, 627-635.
University of Ghana          http://ugspace.ug.edu.gh
304
Begon, M. (1979). Investigating animal abundance: capture 
-recapture for biologists. London, Edward Arnold.
Brady, J. (1970). Characteristics of spontaneous activity in 
tsetse flies. Nature, Lond. 228, 286-287.
Brady, J. (1972). Spontaneous, circadian components of
tsetse fly activity. J. Insect Physiol. 18, 471-484.
J . (1973). The pattern of spontaneous activity in the
tset se fly GI ossina morsitans Westw. (Diptera:
GI os sinidae at low temperatures. Bull . ent. Res,, 63,
441 -444 .
Brady, J.(1987). The sunset activity of tsetse flies: a
light threshold study on Glossina morsitans. Physiol. 
Entomol. 12, 363-372.
Brady, J. and Crump, A. J. (1978). The control of circadian 
activity in tsetse flies: environment or physiological
clock. Physiol. Entomol. 3, 177-190.
Brightwell, R., Dransfield, R. D., Kyorku, C. A.,
Golder, T. K., Tarimo, S. A. and Mungai, D. (1987). A 
new trap for Glossina pallidipes. Trop. Pest 
Manage. 33, 151-159.
University of Ghana          http://ugspace.ug.edu.gh
305
Brightwell, R., Dransfield, R. D., Kyorku, C. (1989).
Development of a low-cost trapping technology for the 
tsetse flies Glossina pallidipes and G. longipennis. 
Med. vet. Entomol. (in press)
Bursell, F.. (1984). Effects of host odour on the behaviour
of tsetse. Insect Sci. Applic. 5, 345-349.
Bursell, E. and Taylor, P. (1980). An energy budget for
Glossina. (Diptera: G 1ossinidae). Bull. ent. Res. 70, 
187-196.
Bursell, E., Gough, A. J. E., Cork, A., Beevor, P. S.,
Hall, D. F. and Vale, G. A. (1988). Identification of 
components of cattle urine attractive to tsetse flies, 
Glosssina spp. (Diptera: GIossinidae). Bull. ent. Res. 
78, 31-49
Buxton, P. A. (1955). The natural history of tsetse flies.
An account of the biology of the genus Glossina 
(Diptera). Mem. Lond. Sch. Hyg. trop. Med. No. 10. 
London, H.K. Lewis pp. 816
Challier, A. (1973). Ecologie de Glossina palpalis gambiense 
Vanderplank 1949 (Diptera: Glossinidae) en savanne 
d'Afrique Ocidentalle. Mem. O.S.T.O.R.M. No. 64 
Paris, pp. 27 4.
University of Ghana          http://ugspace.ug.edu.gh
306
Challiei, A. (]965). Amelioration de la methode de
determination l’age physio1ogique des glossines.
Etudes faites sur Glossina pa 1 pal is Vanderplank.
Bull. Soc. Path. exot. 58, 250-259.
Challier, A. (1977). Trapping Technology. In Tsetse: The 
Future for Biological Methods in Integrated Control 
(Laird M., Ed.) IDRC, Ottawa, pp. 109-123.
Challier, A. (1982). The ecology of tsetse (Glossina spp.)
(Diptera, Glossinidae): A review (1970-1981). Insect 
Sci. Applic. 3, 97-143.
Challier- A. and Laveissiere, C. (1973). Un nouveau piege 
pour la capture des glossines (GIossina: Diptera, 
Muscidae): description et essais sur le terrain. Cah. 
ORSTOM, s o l . Ent. med. Parasitol. 11, 251-262.
Challier, A., Eyraud M., Lafaye A. and Laveissiere C.
(1977). Amelioration au rendement du piege biconique 
pour glossines (Diptera, Glossinidae) par l'emploi 
d 1 un cone inferieur bleu. Cah. O.R.S.T.O.M., ser.
Ent. med. Parasitol. 15, 283-286.
University of Ghana          http://ugspace.ug.edu.gh
307
Chapman, D. G. (1965). The estimation of mortality and 
recruitment from a single tagging experiment. 
Biometrics 21, 529-542.
Chapman, R. F. (1950). A note on Glossina medicorum in 
Ghana. Bull. ent. Res. 51, 435
Cheke, R. A. and Garms, R. (1988). Trials of compounds to
enhance trap catches of Glossina pal pal is palpal is in 
Liberia. Med. vet. Entomol . 2, 199-200.
Chorley. T. W. (1933). Traps for tsetse flies of the
f,Crinol ine" and "Venti 1 ator" forms. Bull. ent. Res.
24, 315-317.
Choiley, T. W. (1948). Glossina pallidipes Austen attracted 
by the scent of cattle-dung and urine (Diptera). Proc. 
R. Entomol. Soc. (Lond.) (A). 23, 9-11.
Crump, A. J. and Brady, J. (1979). Circadian activity
patterns in three species of tsetse fly: Glossina 
palpalis, austeni and morsitans. Physiol. Entomol. 4,
311-318.
University of Ghana          http://ugspace.ug.edu.gh
308
Cuisance, D., Politzar, H., Fevrier, J., Bourdoiseau, G. and 
Sellin, E. (1980). Association d ' un traitment 
insecticide avec la methode du male sterile contre 
Glossina palpalis gambiensis : interet de la mise en
oeuvre de plusieurs methodes Revue Elev. Med. vet.
Pays trop. 33, 127-133.
Dame, D . A. and Jordan, A. M. (1981). Control of tsetse
flies Glossina spp. Adv. vet. Sci. comp. Botswana 8, 
11-14.
Dame, D. A., Williamson, D. L., Cobb, P. E., Gates D. B., 
Warner, P. V., Mtuya, A. G. and Baumgartner, H.
(1980). Intergration of sterile insects and pesticides 
for the control of Glossina morsitans morsitans. In 
Isotopes and radiation research on animal diseases 
and their vectors. Proceedings of a Symposium,
Vienna, 7-11 May 1979.
Dowdeswell, W. H., Fisher, R. A. and Ford, E. B. (1940). The 
quantitative study of populations in the Lepidoptera.
I. Polyommatus icarus Rott. Ann. Eugen. Lond. 10, 123 
-136 .
Dowdeswell, W. H., Fisher, R. A. and Ford, E. B. (1949). The 
quantitative study of populations in the Lepidoptera.
II. ManioJa jurtina L. Heredity 3, 67-84
University of Ghana          http://ugspace.ug.edu.gh
309
Dransfield, R. D. (1984). The range of attraction of the 
biconical trap for Glossina pallidipes and G. 
brevipa 1 pi s . Insect. Sci . Aplic. 5, 363-368.
Dransfield, R. D ., Brightwell, R., Onah, J. and Okolo C. J. 
(1982). Population dynamics of Glossina morsitans 
subinorsi tans Newstead and G. tachinoides Westwood 
(Diptera: Glossinidae) in sub-Sudan savanna in 
northern Nigeria. I. Sampling methodology for adults 
and seasonal changes in numbers caught in different 
vegetation types. Bull. ent. Res. 60, 225-235.
Dransfield, R. D., Chaudhury, M. F., Tarimo, S. R., Golder, 
T. K. and Brightwell, R. (1986a). Population dynamics 
of Glossina pallidipes under drought conditions at 
Nguiuman, Kenya. In OAU/STRC, 1986, Publication No.
4612 .
Dransfield, R. D., Brightwell, R., Chaudhury, M. F., Golder, 
T. K. and Tarimo, S. A. R. (1986b). The use of odour 
attractants for sampling Glossina pal 1idipes Austen 
(Diptera: Glossinidae) at Nguruman, Kenya. Bull. ent. 
Res. 76, 607-619.
University of Ghana          http://ugspace.ug.edu.gh
310
Dransfield, R. D., Brightwel1, R., Kiilu, J., Chaudhury, M.
F ., Adabie, D. A., Tarimo, S. A. and Golder, T. K. 
(1988). Size and mortality in Glossina pallidipes 
Austen at Nguruman, south-western Kenya. Med. vet.
Ent. (in press)
Duggan, A. J . (1970). A historical perspective. In The
African Trypanosomiasis (Mulligan Ed.). George Allen & 
Unwin, Lond., pp. 81-108
Evens, F. (1953). Dispersion geographique de glossines au 
Congo Beige. Mem. Instr. r. Sci. nat. Belg. 2nd Ser. 
Fasc. 48, pp. 70
Fairbain, H. and Culwick, A. T. (1950). Some climatic 
factors influencing the populations of Glossina 
swynnertoni. Ann. Trop. med. Parasitol. 44, 27-33.
Fisher, R. A. and Ford, E. B. (1947). The spread of a gene 
in natural conditions in a colony of the moth Panaxia 
dominula L. Heredity 1, 143-174.
Flint, S. (1985). A comparison of various traps for Glossina 
spp. (GIossinidae) and other Diptera. Bull. ent. Res.
75, 529-534.
University of Ghana          http://ugspace.ug.edu.gh
311
Ford, J. (1970). The geographical distribution of Glossina.
In The African Trypanosomiases.(Ed. Mulligan, H. W. ), 
George Allen& Unwin, London, pp. 27 4.
Ford, J. (1971). Role of Trypanosomiasis in African Ecology. 
A Study of the Tsetse Fly Problem. Claredone Press, 
Oxford, pp. 568
Ford, J. and Katondo, K. M. (1977a). Maps of tsetse fly
(Glossina) distribution in Africa, 1973, according to 
subgeneric groups on scale of 1:5,000,000 Bull. Anim. 
Hlth. Production 25, 187-193.
Ford, J. and Katondo, K. M. (1977b). Revision of the
Glossina distribution map of Africa. In International 
Scientific Council for Trypanosomiasis Research.14th 
Meeting, Dakar, 1975. OAU/STRC publication no. 109.
Ford, J. and Katondo, K. M. (1979). Tsetse distribution maps 
in Africa. In International Scientific Council for 
Trypanosomiasis Research. 15th Meeting, Banjul 1977. 
OAU/STRC publication No. 110
Ford, J., Glasgow, J. P., Johs, D. L. and Welch, J. R. 
(1959). Transect fly rounds in field studies of 
Glossina. Bull. ent. Res. 50, 275-280.
University of Ghana          http://ugspace.ug.edu.gh
312
Frezil, J. L. and Carnevale, P. (1976). Utilisation de la 
carbonglace pour la capture de GIossina fuscipes 
quanzensis Pires, 1948, avec le piege Challier 
-Laveissiere. Consequences epidemiologiques. Cah.
ORSTOM, ser Ent. med. Parasit. 14, 225-233.
Glasgow, J. P. (1961). The feeding habits of Glossina 
swynnertoni Austen. J. Anim. Ecol. 30, 11-lb.
Glasgow, J. P. (1963). The Distribution and Abundance of 
Tsetse. Pergamon Press, Oxford, pp. 252.
Glasgow, J. P. and Duffy, D. J. (1961). Traps in field
studies of Glossina pallidipes Austen. Bull. ent. Res. 
52, 795-814.
Glasgow, J.P. and Phelps, R.J. (1970). Methods for the
collecting and sampling of Glossina. In The African 
Trypanosomiases (Ed. by Mulligan H.W.), George Allen & 
Unwin, London, pp. 395
Glasgow, J. P. and Welch, J. R. (1962). Long-term
fluctuations in the numbers of the tsetse fly Glossina 
swynnertorii Austen. Bull, ent Res. 53, 129-137.
University of Ghana          http://ugspace.ug.edu.gh
313
Gouteux, J.-P. (1983). Ecologie des glossines en secteur
pre-forestier de Cote d'Ivoire. 7. Analyse de la 
distribution spatiale des glossines en activite dans 
une plantation de cafeiers. Cah. ORSTOM, ser Ent. 
med. Parasitol. 21, 231-239.
Gouteux, J.-P. and Buckland, S. T. (1984). Ecologie des
glossines en secteur pre-forestier de Cote d'Ivoire. 
Dynamique des populations. Cah. ORSTOM, ser Ent. 
med. Parasitol. 22, 19-34.
Gouteux, J.-P. and Lancien, J. (1986). Le piege pyramidal a 
tsetse (Diptera: Glossinidae) pour la capture et la 
lutte. Essais comparatifs et description de nouveux 
systemes de capture. Trop. med. Parasit. 37, 61-66.
Gouteux, J.-P. and Laveissiere, C. (1982). Ecologie de
glossines en secteur preforestier de Cote d'Ivoire: 4. 
Dynamique de 1'ecodistribution en terroir villageois. 
Cah. O.R.S.T'O.M, Ser ent. Med. Parasitol. 20, 199-229
Gouteux, J.-P., Couret, D. and Bicaba, A. (1981).
Observations sur 1es glossines d 1un foyer forestier de 
trypanosomiase humaine en Cote d'Ivoire.: 2. Effectifs
des populations et effets du piegeage. Cah.
ORSTOM, ser Ent. med. Parasitol. 29, 209-222
University of Ghana          http://ugspace.ug.edu.gh
314
Gouteux, J.-P., Bansimba, P ., Bissandidi, N. and Noireau, F.
(1987). Le prise en charge de la lutte contre les 
tsetse par les communautes rurales: premiers essais
dans cinq villages congolais. Ann. Soc. Beige Bed. 
trop. 67, 37-49.
Green, C. H. and Cozens, D. (1983). Spectral responses of 
the tsetse fly Glossina morsiptans morsitans. J.
Insect. Physiol. 29, 795-800.
Green, C. H. and Flint, S (1986). An analysis of colour 
effects in the performance of the F2 trap against 
Glossina pallidipes Aust. and G. morsitans morsitans 
Westwood. (Diptera: GIossinidae). Bull. ent. Res. 76, 
409-418.
Green, C. H. and Jordan, A. M. (1983). The responses of 
GIossina morsitans morsitans to a commercial light 
trap. Entomologia exp. appl. 33, 336-342.
Gruvel, J. (1975). Structures des population de Glossina
tachinoides W. a la reserve de Kalamaloue (Vi). Revue 
Elev.. Med. vet. Pays trop. 28, 195-215.
Hadaway, A. B. (1977). Resting behaviour of tsetse flies and 
its relevance to control with residual insecticides. 
Miscellaneous Report No. 36, COPR, London.
University of Ghana          http://ugspace.ug.edu.gh
315
Hall D .R ., Beevor P.S., Cork A., Nesbitt B.F. and Vale G.A.
(1984). 1-octen-3-o1 A potent olfactory stimulant and
attractant for tsetse isolated from cattle odours. 
Insect. Sci . Applic. 5, 335-339.
Hargrove, J. W. (1977). Some advances in the trapping of
tsetse (Glossina spp.) and other flies. Ecol. Ent. 2, 
123-137 .
Hargrove, J. W. (1980a). The effect of model size and ox
odour on the alighting response of Glossina morsitans 
Westwood and G. pallidipes Austen (Diptera: 
Glossinidae). Bull. ent. Res. 70, 229-234.
Hargrove, J. W. (1980b). Improved estimates of the
efficiency of traps for GIossina morsitans morsitans 
Westwood and G. pallidipes Austen (Diptera: 
Glossinidae), with a note on the effect of 
concentration of accompanying host odour on 
efficiency. Bull. ent. Res. 70, 579 
-■587 .
Hargrove, J.W. and Vale, G.A. (1978). The effect of host 
odour concentration on catches of tsetse flies 
(Glossinidae) and other Diptera in the field. Bull, 
ent. Res. 68, 607-612.
University of Ghana          http://ugspace.ug.edu.gh
316
Hargrove, J . W. and Vale G. A. (1980). Catches of Glossina 
morsitans inorsitans Westwood and G. pallidipes Austen 
(Diptera : Glossinidae) in odour-baited traps in 
deciduous woodlands in the Zambesi Valley of Zimbabwe. 
Bull. ent. Res. 70, 571-578.
Harley, J. M. B. (1965). Activity cycles of Glossina
pal 1idipes Austen, G. palpal is fuscipes Newst. and G. 
brevipalpis Newst. Bull. ent. Res. 56, 141-160.
Harris, R. H. T. P. (1930) Report on the bionomics of the 
tsetse fly (Glossina pallidipes Aust.) and a 
preliminary report of a new method of control, 
presented to the provincial administration of Natal. 
Pietermaritzburg (quoted from Challier, 1977).
Hassanali, A., McDowell, P. G., Owaga, M. L. A., and 
Saini, R. K. (1986). Identification of tsetse 
attractants from excretory products of a wild 
host animal, Syncerus caffer. Insect. Sci.
Applie. 7, 5-8 .
Jack, R. W. (1941). Notes on the behaviour of Glossina 
pallidipes and G. brevipalpis and some common 
comparisons with G. morsitans. Bull. Ent. Res. 31, 
407-430.
University of Ghana          http://ugspace.ug.edu.gh
317
Jackson, C. H. N. (1933). On the true density of tsetse 
flies. J. Anim. Ecol. 2, 204-209.
Jackson, C. H. N. (1937). Som new methods in the study of
Glossina morsitans. Prroc. Zool . Soc. Lond. 4, 811-96.
Jackson, C. H. N. (1939). The analysis of an animal 
population. J. Animal Ecol. 8, 238-246.
Jackson, C. H. N. (1941). The analysis of a tsetse-fly 
population. Ann. Eugen. Lond. 10, 332-369.
Jackson, C. H. N. (1944). II. The analysis of a tsetse-fly 
population. Ann. Eugen. Lond. 12, 176-205.
Jackson, C. H. N. (1948). III. The analysis of a tsetse-fly 
population. Ann. Eugen. Lond. 14, 91-108
Jackson, C. H. N. (1953a). Seasonal variations in the mean 
size of tsetse flies. Bull. ent. Res. 43, 703-706.
Jackson, C. H. N. (1953b). A mixed population of Glossina 
morsitans and G. swynnertoni. J. Anim. Ecol. 22, 78 
-86 .
University of Ghana          http://ugspace.ug.edu.gh
318
Jackson, H. C. N. (1955). The Pattern of Glossina communities 
Bull. ent. Res. 46, 517-530.
Jolly, G. M. (1965). Explicit estimates from capture 
recapture data with both death and immigration 
-stochastic model. Biometrika 52, 225-247.
Jordan, A. M. (1962a). The distribution of the fusca group 
of tsetse flies in Nigeria and West Cameroun. Bull, 
ent. Res. 54, 307-316.
Jordan, A. M. (1962b). The pregnancy rate of Glossina
palpalis (R-D) in southern Nigeria. Bull. ent. Res.
53, 387-393.
Jordan, A. M. (1965a). The status of Glossina fusca Walker 
(Diptera, Muscidae) in West Africa. Ann. trop. Med. 
Parasit. 59, 319-325.
Jordan, A. M. (1965b). Long-term fluctuations in the numbers 
of a population of Glossina palpalis (R. D.): sixteen 
years' observations, pp. 85-90 in International 
Committee for Trypanosomiasis Research, Tenth Meeting, 
Kampala, 24-28 October, 1964.
University of Ghana          http://ugspace.ug.edu.gh
Jordan, A. M. (1974). Recent developments in the ecology and 
methods of control of tsetse flies (Glossina spp.) 
(Diptera: GIossinidae)-a review. Bull. ent. Res. 63, 
361-399.
Jordan, A.M. (1986). Control of African Trypanosomiasis. 
Longman Inc., New York, 357p.
Jordan, A.M. and Green C. H. (1984). Visual responses of
tsetse to stationary targets. Insect Sci. Applic. 5, 
331-334.
Kaminsky, R. (1987). Tsetse ecology in Liberia rain-forest 
of Gambian sleeping sickness. Med. Vet. Entomol. 1, 
257-264.
Kangwagye, T. N. (1971). Observations on Glossina fuscipes 
Newstead, G. fuscipleuris Austen and G. pallidipes 
Austen in western Uganda. In Internationa1 
Committee for Trypanosomiasis Research, Thirteenth 
Meeting, Lagos, 1971.
Kangwagye, T. N. (1973). Diurnal and nocturnal biting
activity of flies (Diptera) in western Uganda. Bull, 
ent. Res. 63, 17-29.
319
University of Ghana          http://ugspace.ug.edu.gh
320
Kangwagye, T. N. (1974). The seasonal incidence of biting
flies (Diptera) in Ruwenzori National Park and Kigezi 
Game Reserve, Uganda. Bull. ent. Res. 63, 535-549.
Katondo, K. M. (1984). Revision of second edition of tsetse 
distribution maps. An interim report. Insect Sci.
Applic. 5, 3 81-388.
Lamborn, W. A. (1912). A preliminary report on the problem 
of controlling Glossina in Nyasaland. Bull. ent. Res.
6, 59.
Lamborn, W. A. (1921). 1922 Report of the Medical
Entomologist. Ann. Med.Rep. of the Nyasaland Prot. for 
the year ended in 1921.
Lambrecht, F. L. (1964). Aspects of evolution and ecology of 
tsetse flies and trypanosomiasis in prehistoric 
African environment. J. Afric. History. 5, 19-29.
Lambrecht, F. L. (1973). Colour attraction of Glossina
morsitans morsitans in N'Gamiland, Botswana. J. trop. 
Med. Hyg. 76, 94-96.
Langridge, W. P. (1968). Tsetse fly traps and trapping
methods. In International Scienti fic Committee for 
Trypanosomiasis Research, 12th Meeting Bangui, 
OAU/ISCTRC, Publication No. 'J.
University of Ghana          http://ugspace.ug.edu.gh
321
Langridge, W. P. (1.961). Scent attractants for tsetse flies, 
pp. 253-241 In International Scientific Committee for 
Trypanosomiasis Research, Jos, 19-23 July, 1960, pp. 235
Langridge, W. P. (1977). Design and operation of Langridge's 
tsetse trap. In International Scientific Committee for 
Trypanosomiasis Research, 14th Meeting Dakar, 1975. 
OAU/STRC, publication No. 109 pp. 57.
Lancien, J. (1981). Description de piege monoconique utilize 
pour elimination de glossines en Republique Populaire 
du Congo. Cah. ORSTOM, ser Ent. med. Parasit. 19,
235-238.
Laveissiere C., Gouteux J.-P. and Couret D. (1980). Essais 
de methodes de lutte contre les glossines en zone 
preforestiere de cote d ’Ivoire: 5. Note de syntese.
Cah. ORSTOM ser. Ent. med. Parasit. 18, 323
-328 .
Laveissiere, C. and Couret, D. (1981). Luttes contre les 
glossines riveraines a laide de pieges biconiques 
impregnes d 1 insectticide, en zone de savanne humide.
5. Note de synthese. Cah. ORSTOM, ser Ent. med.
Parasitol. 19, 49-54.
University of Ghana          http://ugspace.ug.edu.gh
Laveissiere, C. and Couret, D. (1983). Consequences de 
essais de la lutte repete sur les proportions de 
glossines riveraines. Cah. ORSTOM, ser Ent. med. 
Parasitol. 14, 63-67-
Laveissiere, C. and Hervouet, J-P. (1988). Essais de
methodes de lutte contre les glossines en secteur pre 
-forestier de Cote d'Ivoire. Etudes et Theses ORSTOM 
(in press)
Laveissiere, C. and Couret, D. and Eouzan, J. P. (1986). La 
campaigne pilote de lutte contre la trypanosomaise 
humaine dans le foyer de Vavoua (Cote d'Ivoire): 3. 
Resultats des evaluations entomologiques. Cah.
ORSTOM, ser Ent. med. Parasitol. 24, 7-20.
Le Cren, E. D. (1965). A note on the history of mark-
recapture population estimates. J. Anim. Ecol. 34,
453-459.
Leslie, P. H. and Chitty, Denis (1951). The estimation of 
population parameters obtained by means of the 
capture-recapture method. Biometrika 38, 268-292.
Lewis E. A. (1934). The behaviour of the larvae of tsetse 
flies before pupation. Bull. ent. Res. 25, 195-199.
322
University of Ghana          http://ugspace.ug.edu.gh
323
Lewis, E. A. (1939). Observations on Glossina fuscipleuris
and other tsetse in Onyani Valley, Kenya Colony. Bull. 
ent. Res. 30, 303-308.
Lewis, E. A. (1942). Notes on Glossina longipennis and its 
breeding places. Bull. ent. Res. 32, 303-307.
Lincoln, F. C. (1930). Calculating water-fowl abundance on 
the basis of banding returns. Circ. U.S. Dep. Agric. 
n o .118, p p . 46.
Machado, A. de B. (1959). Nouvelles contributions a 1 * etude 
systematique et biogeographie des glossines (Diptera).
Publ. cult. Co. Diamantes de Angola, 46, 15-90.
(quoted from Ford 1971).
Madubunyi, L. C. (1975). A technique for detecting abortions 
in wild populations of Glossina spp. In Sterility 
Principle for insect control or Eradication. 
Proceedings of Symposium on the Sterility Principle 
for Insect Control or Eradication Jointly Organized by 
the IAEA and FAO, Innsbruck, 22-26 July 1974. IAEA, 
Vienna, STI, Publication NO. 377.
University of Ghana          http://ugspace.ug.edu.gh
324
Madubunyi, L. C. (1978). Relative frequency of reproductive
abnormalities in a natural population of Glossina
morsitans morsitans Westwood (Diptera: Glossinidae) in 
Zambia. Bull. ent. Res. 68, 437-442
Madubunyi, L,. C. (1988). The seasonal breeding rate of
Glossina tachinoides Westwood (Diptera: Glossinidae) 
in peridomestic agrosystems of Nsukka area, Nigeria. 
Bull. ent. Res. 78, 143-150.
Maldonado, (1910). (English abstract of Portuguese texts of 
1906 and 1909). Sleeping sickness Burueau Bull. 2,
26. (quoted from Challier, 1977).
Manly, B. F. J. and Parr, M. J. (1968). A new method of 
estimating population size, survivorship and birth 
rate from capture-recapture data. Trans. Soc. Brit. 
Ent. 18, 81-89.
Minter, D. M. and Goedbloed, E. (1971). The preservation in 
liquid nitrogen of tsetse flies and phlebotomine 
sandflies naturally infected with trypanosomatid 
flagellates. Trans. R. Soc. Trp. Med. H yg. 65, 175 
-181.
Moggridge, J. Y. (1949). Climate and the activity of the
Kenya coastal Glossina. Bull. ent. Res. 40, 307-321.
University of Ghana          http://ugspace.ug.edu.gh
325
Mo1oo, S. K. (1973). A new trap for Glossina pallidipes
Aust. and G. fuscipes Westw. (Diptera, GIossinidae). 
Bull. ent. Res. 63, 231-236.
Moloo, S. K ., Kutuza, S. B. and Borham, P. F. L. (1980).
Studies on Glossina pallidipes, G. fuscipes and G. 
brevipalpis in terms of epidemiology and epizootiology 
of trypanosomiases in south-eastern Uganda. Ann. Trop. 
Med. Parasit. 74, 219-237.
Morris, K. R. S. and Morris M. G. (1949). The use of traps 
against tsetse in West Africa. Bull. ent. Res. 39, 
491-526.
Muirhead-Thomson, R. C. (1968). Ecology of Insect Vector 
Populations. Academic Press, London and New York.
Mulligan, H. W. Ed. (1970). The African Trypanosomiasis. 
George Allen and Unwin Lond. 950pp.
Nash, T. A. M. (1930). A contribution to our knowledge of 
the bionomics of Glossina morsitans. Bull. ent. Res. 
21, 201-205
University of Ghana          http://ugspace.ug.edu.gh
326
Nash, T. A. M. (1931). The relationship between Glossina 
morsitans and evaporation rate. Bull. ent. Res. 22,
383-389.
Nash, T. A . M. (1937). Climate, the vital factor in the 
ecology of Glossina. Bull. ent. Res. 28, 75-121.
Nash, T. A. M. (1952). Some observations on resting tsetse 
populations and evidence that Glossina medicorum is a 
carrier of trypanosomes. Bull. ent. Res. 43, 33-42
Nash, T. A. M. (1960). A review of African trypanosomiasis 
problem. Trop. Dis. Bull. 57, 973.
Nash, T. A. M. and Davey (1950). The resting habits of G. 
medicorum, G. fusca and G. longipalpis. Bull. ent. 
Res. 41, 153-157.
Nash, T. A. M. and Page (1953). THe ecology of Glossina 
palpalis in northern Nigeria. Trans. R. Soc. Lond. 
104, 71-169.
Neave, S. A. (1912). Notes on blood sucking insects of
eastern tropical Africa. Bull. ent. Res. 3, 275-323.
University of Ghana          http://ugspace.ug.edu.gh
327
OAU/STRC (1971). International symbols and colours for
recording on maps the distribution of tsetse species 
and subspecies. Annex to publication No. 105, Report 
of 13th Meeting ISCTRC, Lagos, 1971, pp. 99.
Okiwellu, S. N. (1976). Seasonal variation in age
-composition and survival of a natural population of 
female Glossina morsitans morsitans Westwood at the 
Chakwenga Game Reserve, Republic of Zambia. Zambia J. 
Sci. Technol. 1, 48-58.
Okiwellu, S. N. (1977). Insemination, pregnancy and 
suspected abortion rates of GIossina morsi tans 
morsitans (diptera: Glossinidae) in the Republic of 
Zambia, Africa. J. med. Ent. 14, 152-155.
Onyiah, J. A. (1978). Fluctuations in numbers and eventual 
collapse of a Glossina palpalis (R.-D.) population in 
Anara forest Reserve of Nigeria. Acta trop. 35, 253 
-2 61 .
Owaga, M. A. L. (1980). Relative efficiency of some
mechanical traps used in the study of the tsetse 
Glossina pallidipes Austen. Insect Sci. Aplic. 1, 97 
-101.
University of Ghana          http://ugspace.ug.edu.gh
328
Owaga, M. L. A. (1981). Ecological studies and laboratory
rearing of the tsetse species Glossina 1ongipennis in 
Kenya. Insect Sci. Applic. 2, 197-200.
Owaga, M. L. A. (1984). Preliminary observations on the
efficiency of olfactory attractants derived from wild 
hosts of tsetse. Insect Sci. Appl. 5, 87-90.
Owaga, M. A. L., (1985). Observations on the efficacy of
buffalo urine as a potent attractant for Glossina 
pallidipes Austen. Insect Sci. Applic. 6, 561-566.
Owaga, M. L. A. and Challier, A. (1981). Sampling techniques 
for Glossina pallidipes. ICIPE Eighth Annual Report
1980 pp. 55-57.
Page, W. A. (1959). Some observations on the fusea group of 
tsetse flies in the south of Nigeria. Bull. ent. Res. 
50, 633-646.
Parker, R. A. (1955). A method for removing the effect of
recruitment on Petersen-type population estimates. J. 
Fish. Res. B d . Canada. 12, 447-450.
University of Ghana          http://ugspace.ug.edu.gh
329
Persoons, C. J. (1967). Trapping G. pallidipes and G.
palpalis fuscipes in scented traps. In International 
Scientific Council for Trypanosomiasis R e s e a r c h r 11th 
Meeting, Nairobi, 31st October - 4th November, 1966, 
PP. 127.
Phelps R. j. and Clarke, G. P. Y. (1974). Seasonal
elimination of some size classes in males of Glossina 
morsitans morsitans Westwood (Diptera: GIossinidae). 
Bull. ent. Res. 64, 313-324.
Phelps R. J. and Vale G. A. (1978). Studies on populations 
of GIossina morsitans morsitans and G. pallidipes 
(Diptera: Glossinidae) Rhodesia. J. appl. Ent. 15, 
743-760.
Pilson, R. D. and Pilson, D. M. (1967). Behaviour studies of 
Glossina morsitans Westw. in the field. Bull, ent.
Res. 57, 227-231.
Politzar, H. and Cuisance, D. (1982). SIT in the control and 
eradication of Glossina palpalis gambiensis. In Pro. 
Symp. Sterile insect technique and radiation on insect 
control. Neuterberg 29 June-3 July 1981, IAEA, 101 
-109.
University of Ghana          http://ugspace.ug.edu.gh
330
Politzar, H. and Merot, P. (1984). Attraction of the tsetse 
fly GIossina morsitans submorsitans to acetone, 1 
-octen-3-ol and a combination of these compounds in 
West Africa. Revue d'Elevage et de Med. vet. des Pays 
Trop. 37, 468-473
Potts, W. H. (1930). A contribution to the study of the
numbers of tsetse fly (glossina morsitans Westw.) by 
quantitative methods. S. Afric. J. Sci. 27, 491.
Potts, W. H.(1937). The distribution of tsetse flies in 
Tanganyika Territory. Bull. ent. Res. 28, 129-137.
Potts W. H. (1953-1954). Distribution of tsetse species in 
Africa, compiled from information collated by W. H. 
Potts. London. Directorate of colonial surveys.
Potts, W. H. (1970). Traps and attractants. In The African 
Trypanosomiasis, Mulligan (Ed.) Gorge Allen & Unwin, 
London, p p . 458.
Power R.J.B. (1964). The activity pattern of Glossina
longipennis Corti. Proc. R. Ent. Soc. London. A 39, 
5-14.
University of Ghana          http://ugspace.ug.edu.gh
331
Randolph, S. E. and Rogers, D. J. (1978). Feeding cycles 
and flight activity of field population of tsetse 
(Diptera: GIossinidae). Bull. ent. Res. 68, 655-671.
Randolph, s . E. and Rogers, D. J. (1984). Movement patterns 
of the tsetse fly Glossina palpalis palpalis 
(Robineau-Desvoidy) (Diptera: Glossinidae) around 
villages in pre-forest zone of Ivory Coast. Bull. ent. 
Res. 74, 689-705.
Rogers, D. J. (1974). Natural regulation and movement of 
tsetse populations. In Les moyens de luttes contres 
les trypanosomes et leurs vecteurs (Control Programs 
against Trypanosomes and Their Vectors). Actes du 
Collogue, Paris, 12-15 Mars 1974. Office International 
de epizooties. Revue Elev. Med. vet. Pays trop.
Suppl . , 35-38 .
Rogers, D. J. (1977). Study of a natural population of
GIossina fuscipes fuscipes Newstead and a model of 
movement. J. Anim. Ecol. 46, 309-330.
Rogers, D. J. (1979). Tsetse population dynamics and
distribution: A new analytical approach. J. Anim.
Ecol. 48, 825-849.
Rogers, D. J. (1984). The estimation of sampling biases for 
male tsetse. Insect Sci. Applic. 6, 369-373.
University of Ghana          http://ugspace.ug.edu.gh
332
Rogers, d . J. and Randolph, S. E.. (1978a). A comparison of
electric trap and hand net catches of Glossina 
palpalis palpalis ( Robineau-Desvoidy) and G. 
tachinoides Westwood (Diptera: Glossinidae) in the 
Sudan vegetation zone of northern Nigeria. Bull. ent. 
Res. 68, 283-297.
Rogers, D. J . and Randolph, S. E.. (1984). A review of
density dependent processes in tsetse populations. 
Insect Sci. hplic. 5, 379-402.
Rogers, D. J. and Randolph, S. E. (1985). Population ecology 
of tsetse. Ann. Rev. Entomol. 30, 197-216.
Rogers, D. J. and Randolph, S. E.(1986). Detection of the
activity cycles of tsetse from capture-recapture data. 
Ecol. Entomol. 11, 95-109.
Rogers, D. J. and Smith, D. T. (1977). A new electric trap 
for tsetse flies. Bull. ent. Res. 67, 153-159.
Rogers, D. J. and Randolph, S. E. and Kuzoe, F. A. S.
(1984). Local variation in the population dynamics of 
Glossina palpalis palpalis (Robineau-Desvoidy) 
(Diptera: Glossinidae). I. Natural population 
regulation. Bull. ent. Res. 74, 403-423.
University of Ghana          http://ugspace.ug.edu.gh
333
Ryan, L., Molyneux, D. H. and Kuzoe, F. A. (1980).
Differences in rate of wing fray between Glossina 
species. Tropenmed. Parasit. 32, 145-148.
Saini, R. K. (1986). Antennal responses of Glossina 
morsitans morsitans to buffalo urine a potent 
olfactory attractant of tsetse. Insect Sci. Aplic. 1, 
771-775.
Saunders D. S. (1960). The ovulation cycle of Glossina
morsitans Westwood (Diptera: Muscidae) and a possible 
method of age determination for female tsetse flies by 
the examination of their ovaries. Trans. R. ent. Soc. 
Lond. 112, 221-238
Saunders, D. S. (1964). The effect of site and sampling
method on the composition of catches of tsetse flies 
Glossina and Tabanidae. Bull. ent. Res. 55, 483-497.
Saunders, D. S. and Phelps, R. J. (1970). Reproduction of 
Glossina: breeding sites. In The African 
Trypanosomiasis (Ed. Mulligan, H. W.). Gorge Allen & 
Unwin, London, pp.327.
University of Ghana          http://ugspace.ug.edu.gh
334
Sayad, v. c. and Sayad, C. (1980). Ecological survey of the 
Nguruman forest, Kenya. 1980 Report on Natural 
history. National Museum,Kenya.
Seber, G. A. F. (1962). The multiple-sample single recapture 
census. Biometrika 49, 330-349.
Seber, G. A. F. (1973). The estimation of animal abundance 
and related parameters. Griffin, London.
Service, M. W. and Highton, R. B. (1980). A chemical light 
trap for mosquitoes and other biting insects. J. Med. 
Entomol. 17, 183-185.
Smith, I. M. and Renninson, B. D. (1961). Studies of the
sampling of G. pallidipes Aust. I. The number caught 
daily on cattle in Morris traps and on a fly-round.
II. The daily pattern of flies caught on cattle, in 
Morris traps and on a fly-round. Bull. ent. Res. 52, 
165-189.
Snow, W. F. (1980). Tsetse ecology on the Kenyan Coast. In 
ICIPE 7th Annual Report 107-108.
Southwood, T. R. E. (1978). Ecological methods.-2nd edn, 
Lond. Chapman and Hall.
University of Ghana          http://ugspace.ug.edu.gh
335
Swynnerton, C. F.M. (1921). Examination of the tsetse
problem in North Mossurrisse, Portuguese East Africa. 
Bull. ent. Res. 11, 315-324.
Swynnerton, C. F. M. (1933). Some traps for tsetse flies. 
Bull. ent. Res. 24, 69-102.
Swynnerton, C. F. M. (1936). The tsetse flies of East
Africa: a first study of their ecology, with a view to 
their control. Trans. R. Ent. Soc. Lond. 84, 1-579.
Takken, W. (1984). Studies on the biconical trap as a 
sampling device for tsetse (Diptera :
Glossinidae) in Mozambique. Insect Sci. Applic.
5, 357-361.
Turner, D. A. (1980). Tsetse ecological studies in Niger and 
Mozambique. II. Resting behaviour. Insect Sci. Applic. 
1, 15-21.
Turner, D. A. (1987). The population ecology of Glossina
pallidipes Austen (Diptera: Glossinidae in the Lambwe 
Valley, Kenya: I. Feeding behaviour and activity
patterns. Bull. ent. Res. 11. 317-333.
University of Ghana          http://ugspace.ug.edu.gh
336
Turner, D. A. and Snow, W. F. (1984). Reproductive
abnormality and loss in natural populations of 
Glossina pallidipes Austen (Diptera: Glossinidae) in 
Kenya. Bull. ent. Res. 74, 299-309.
Turner, D. A. and Brightwell, R. (1986). An evaluation of 
the sequential spraying operation against GIossina 
pal 1jdipes Austen. (Diptera: Glossinidae) in the 
Lambwe Valley, Kenya: Aspects of post-spraying 
recovery and evidence of natural population 
regulation. Bull, ent Res. 76, 331-349.
Vale, G. A. (1969). Mobile attractants for tsetse flies 
Arnoldia 33, 1-6
Vale, G. A. (1974a). The responses of tsetse flies (Diptera, 
Glossinidae) to mobile and stationary baits. Bull, 
ent. Res. 64, 545-588.
Vale, G. A. (1974b). New filed methods for studying the
responses of tsetse flies (Diptera, Glossinidae) to 
hosts. Bull. ent. Res. 64, 199-208.
Vale, G. A. (1980). Field studies of the responses of tsetse 
flies (Glossinidae) and other Diptera to carbon 
dioxide acetone and chemicals. Bull. ent. Res. 70, 
563-570.
University of Ghana          http://ugspace.ug.edu.gh
337
Vale, G. A. (1981). Prospects for using stationary baits to 
control and study populations of tsetse flies in 
Zimbabwe. Zimbabwe Sci. Newsl. 15, 181-186 (quoted 
from Challier 1982).
Vale, G. A. (1982a). The trap-orientated behaviour of tsetse
flies (Glossinidae) and other Diptera. Bull. ent. Res.
72, 71-93.
Vale, G. A. (1982b). The improvement of traps for tsetse
flies (Diptera: Glossinidae). Bull. ent. Res. 72, 95 
-106.
Vale, G. A. (1987). Prospects for tsetse control. Int. J.
Parasit. 17, 665-670.
Vale, G. A. and Hargrove J.W. (1979). A method of studying 
the efficiency of traps for tsetse flies (Diptera: 
Glossinidae) and other insects. Bull. Ent. Res. 69, 
183-193.
Vale, G. A. and Flint, S. and Hall, D. R. (1986). The field
responses of tsetse flies, Glossina spp. (Diptera:
Glossinidae), to odours of host residues. Bull. ent. 
Res. 76, 685-693.
University of Ghana          http://ugspace.ug.edu.gh
338
Vale, G. A. and Hall, D. R. (1985). The use of l-octen-3 
-ol acetone and carbon dioxide to improve baits 
for tsetse flies, Glossina spp.
(Diptera:GIossinidae). Bull, ent Res. 75, 219 
-231.
Vale, G. A., Bursell, E. Hargrove, J. W. (1985). Catching 
out the tsetse fly. Parasitology Today 1, 106-110
Vale, G . A ., Hargrove, J. W . , Cocbi11,■G . F. and Phelps, R . 
J. (1986). Field trials of odour baits to control 
populations of Glossina morsitans morsitans Westw. and
G. pallidipes Aust. (Diptera: GIossinidae). Bull. ent. 
Res. 76, 179-193.
Vale, G. A., Lovemore, D. F., Flint, S. and Cockbill, G. F.
(1988). Odour-baited targets to control tsetse flies 
Glossina spp. (Diptera: Glossinidae) in Zimbabwe.
Bull. ent. Res. 78, 31-49.
Van Sickle, J. and Phelps, R. J. (1988.) . Age distributions 
and reproductive status of declining and stationary 
populations of Glossina pallidipes Austen (Diptera: 
Glossinidae) in Zimbabwe. Bull. ent. Res. 78, 51-61.
University of Ghana          http://ugspace.ug.edu.gh
339
Vanderplank, F. L. (1944). Studies of the behaviour of the 
tsetse-fly (Glossina pallidipes) in the field . the 
attractiveness to various baits. J. Anim. Ecol. 13,
39-48.
Vanderplank, F. L. (1948). Studies on the behaviour of the 
tsetse-fly, Glossina pallidipes, in the field: 
influence of climatic factors on activity. J. Anim. 
Ecol. 17, 245-256.
Weitz, B. G. F. (1970). Hosts of Glossina. In The African 
Trypanosomiases (Mulligan, H. W. Ed.) George Allen & 
Unwin, London pp. 317.
Weitz, B. G. F., and Glasgow, J. (1956). The natural hosts 
of Glossina in East Africa. Trans. R. Soc. Trop. Med. 
Hyg. 50, 593-601.
Weitz, B.G. F., Langridge, W. P., Bax, P. N. and Lee-Jones, 
F. (1958). The natural hosts of Glossina longipennis 
Corti and some other tsetse flies in Kenya. In 
International Scientific Committee for Trypanosomiasis 
Research, Brussels, 7th Meeting, Publication No.
41, pp. 303.
University of Ghana          http://ugspace.ug.edu.gh
W. H. O., (1986). Epidemiology and Control of
Trypanosomiasis. Report of a WHO expert 
Technical Report Series No. 739. Geneva 
Organi zati on.
340
African 
committee. 
World Health
Young, S. David, C. T. and Gibson, G. (1987). Light 
measurement for entomology in the field and 
laboratory. Physiol. Entomol. 12, 373-379.
University of Ghana          http://ugspace.ug.edu.gh