QL537. C46 
Sa7 

bltfar C.1 
G360923 The Balme LibraryiB iiiia

3 0692 1078 5425 7

University of Ghana          http://ugspace.ug.edu.gh



INSECTS ASSOCIATED WITH CITRONELLA 
IN SOUTH-EASTERN GHANA, WITH PARTICULAR 

REFERENCE TO SHOOTBORERS

by

JOSEPH EASMON SARFO 
B.SC. (HONOURS) CAPE COAST (1984) 

DIP> EDUC. CAPE COAST (1984)

A thesis submitted to the Department of Zoology, Faculty of 
Science, in partial fulfilment of the requirements for the degree of 
MASTER OF PHILOSOPHY of the UNIVERSITY OF GHANA

September, 1993.

University of Ghana          http://ugspace.ug.edu.gh



To

Maame Yaa Boatemaa Sarfo 
And

Nana Akua Sarfowaa Sarfo

University of Ghana          http://ugspace.ug.edu.gh



ZOOLOGY DEPARTMENT 
UNIVERSITY OF GHANA 
LEGON

This is to certify that this thesis has 
not been submitted for a degree to any 

other University. It is entirely my 
own work and all help has been duly 

acknowledged.

Supervisor
(Dr. S.Q. Quartey)

University of Ghana          http://ugspace.ug.edu.gh



A CKNO WLED GEMENTS

I would like to express my sincere gratitude to my principal supervisor, Dr. S.Q. 

Quartey for his patience, useful criticisms and suggestions. My sincere thanks are due to 

Miss Cobblah of Zoology Department, a supervisor, who helped in the identification of 

some insects and also made available to me some useful printed materials. I am very 

grateful to Mr. Deeming (British Museum) for identifying the Chloropidae and 

Lonchaeidae and for providing me with key for the identification o f shootflies. I would 

also like to thank Dr. C.Y. Brempong-Yeboah of Crop Science Department, Faculty of 

Agriculture, who as a supervisor, enthusiastically and readily offered assistance whenever 

consulted. My sincere gratitude also goes to Dr. J.B. Ackonor (Cocoa Research Institute 

of Ghana -  CRIG) for patiently reading through the re-written script and making very 

useful suggestions. I thank Dr. N.N.N. Nuamah (ISSER, Legon) and Mr. Ken Brew 

(CRIG) for helping with some statistics. I am also grateful to Dr. (Mrs.) Botchwey and 

Mr. P. Kwapong (University of Cape Coast) for providing me with very useful printed 

material.

I appreciate greatly, the concern expressed by the Professors and lecturers of 

Zoology Department, especially about the delay of the work. I am particularly grateful to 

Dr. Wilson and Prof. Coker for their usefid suggestions. To the technicians of Zoology 

Department, especially Mr. Titiati, Mr. Ofosu Baako, Mr. Asiedu and all others, I say a 

big thank you for your patience and immense technical help.

I am greatly indebted to Kwesi of Osofoiaman, Dr. Tagoe, Madam Selina Taylor, 

Mr. Ababio and all the other farmers who allowed me to use their farms for the study.

Mr. Vincent Acheampong (University of Ghana, Academic Affairs), Mr. F. 

Aneani (CRIG), Mr. Kwame Vowotor (Food Research Institute) and Mr. Alfred Quartey 

(University of Ghana, Data Processing Centre) deserve special mention. They, despite 

the inconvenience to them, stayed for long hours and sometimes late to help me with 

most of the analysis of my data and the processing of my graphs. I also thank Willie of

University of Ghana          http://ugspace.ug.edu.gh



Survey Department, Accra, for his neat cartographic work. I appreciate very much the 

photographs on the grass-hoppers taken by Mr. & Mrs Effah-Bekoe of Amsterdam, 

Holland. I am very grateful to Mr. Williams and my colleagues, Mr. Carl Fiati and Mr. 

Owusu-Boateng for their encouragement especially at difficult times. My special thanks 

go to another colleague, Mr. Albert Addo Kwafo, Faculty of Agriculture, for coming to 

my aid by offering me accommodation at a very critical time and also for the other 

numerous acts of kindness which helped me a great deal in getting this work completed.

I am very grateful to my friends, Mr. Peter Opoku, Survey Department, Accra, 

and Mr. Paul Forson of K.Y. Merchants, Accra, for the numerous valuable assistance 

given in diverse ways to make this work a possibility.

My warmest thanks go to Miss Mercy Techie-Menson, Miss Bertha Afful, Miss 

Theodora Mifetu, Miss Akosua Henewaa Dapaah, Miss Suzzy Micah and Miss Joyce 

Anane for their encouragement, company and friendship.

Finally, I thank Miss Grace Acheampong (University of Ghana, Academic 

Affairs), Mr. Kwame Amoako and Mr. Michael Doe Kulogo (CRIG) for accepting to 

neatly type this work and patiently doing all corrections when they had other pressing 

commitments.

J .E S

V

University of Ghana          http://ugspace.ug.edu.gh



ABSTRACT

Insects associated with citronella were surveyed in the south-eastern part of 

Ghana for 11 months. Using two different sweep nets, intensive and extensive surveys 

showed that the foliage of citronella was associated with a diverse insect fauna 

comprising 10 orders and 102 families as follows: Hymenoptera (29), Diptera (32), 

Homoptera (8), Hemiptera (12), Coleoptera (10), Dictyoptera (2), Orthoptera (3), 

Lepidoptera (2), Dermaptera (1), and Neuroptera (3).

The shoot of citronella was found to be damaged by dipterous shootflies (Fam: 

Chloropidae and Fam: Lonchaeidae) and lepidopterous stemborers (Fam: Pyralidae and 

Fam: Noctuidae). There was no significant difference (P > 0.05) between the mean 

percentage infestation by the shootflies (mean =1.18%) and that o f the stemborers (mean 

= 0.59%). The chloropid shootflies were identified as Scoliophthalmns micantipennis 

Duda, Scoliophthalmus trapezoides Becker, Anatrichus pygmaeus Loew, Aprometopis 

flavofacies Becker, Elachiptera occipitalis Becker, and unidentified species. The 

lonchaeid shootfly was Silbapectita McAlpine. The pyralid and noctuid stemborers were 

Eldana saccharina Walker and Sesamia sp, respectively. Whereas S. micantipennis, A. 

flavofacies were found to cause primary damage by attacking fresh stems, S. trapezoides 

caused both primary and secondary damage by attacking fresh as well as already infested 

stems. The status of the other chloropids and the lonchaeid was not clear, since they were 

always found in association with other insects in already infested tillers. The 

lepidopterous stemborers were also found to cause primary damage. The larvae of the 

stemborers attacked the lower part of the tiller and tunnelled a length of 75.40 ± 12.79 

mm while the shootflies tunnelled 32.60 ± 3.90 mm. The larvae live inside the tillers and 

so can possibly be controlled by systemic inseticides.

Preliminary field investigation showed that the worker caste of unidentified 

termites attacked the roots and stems of citronella, and this invariably led to complete 

death of the plant. This was more prevalent in farms older than one year; for example,

University of Ghana          http://ugspace.ug.edu.gh



85.7% of the termite infested farms were older than one year. Farmers should therefore 

not cultivate one farmland for too long. Further investigations showed negligible 

symptoms of feeding activities of insect foliage pests. For example, less than 0.005%,

0.002% and 0.009%, showed necrosis, curling and defoliation respectively, from insect 

attack in any farm.

Investigation of the seasonal variation of shootborer infestation of citronella 

showed two possible peaks: one between February and March, and another around 

October for both the lepidopterous stemborers and chloropid shootflies. The 

synchronisation in the build up of the damage by both groups o f shootborers shows that 

they can be controlled simultaneously. Percentage 'deadheart' caused by lepidopterous 

stemborers had a weak linear relationship with rainfall: y = 0.0017x + 0.6298. That by 

the dipterous shootflies also had a weak linear relationship with climatic factors as 

follows:- rainfall: y = -0.0024x + 1.3413; evaporation: y = -0.8958x + 2.2639; 

temperature: y = -0.5826x + 17.15; and relative humidity: y = 0.1292x - 9.8968; where y 

= percentage deadheart' and x = climatic factor.

Comparision of sampling, using only visual 'deadheart' symptom and 

random dissection, showed that using only 'deadheart' symptom underestimated the 

infestation of shootborers by 31.8%. Shootborer infestation should therefore be 

determined from random dissection of tillers.

The pupa of female S. micantipennis measured 3.32 + ,08mm in length (n=30) 

and pupal period averaged 6.72 + 0.39 days (n = 22). The male pupa measured 3.16+_.07 

mm long (n = 30) and the mean pupal period was 7.24 + 0.05 days (n = 17). The adults 

of both sexes lived for about the same period: 8.59 + 0.77 days (n = 29) for the female 

and 9.41 + 0.60 days (n = 24) for the male. The adults exhibited sexual dimorphism with 

respect to their total body lengths, the male being shorter (1.90 ±  0.02mm; n = 30) than 

the female (2.98 + 0.03mm; n = 31).

University of Ghana          http://ugspace.ug.edu.gh



iii

iv

vi

xi

xiii

xiv

1

1

1

2

4

6

6

7

8

9

11

11

11

11

15

v i i i
TABLE OF CONTENTS

Declaration........................................................................................

Acknowledgements..........................................................................

Abstract.............................................................................................

List of Figures..................................................................................

List of Tables....................................................................................

List of Plates......................................................................................

GENERAL INTRODUCTION AND LITERATURE REVIEW

General Introduction........................................................................

Introduction of Citronella in G hana...............................................

Citronella farming in Ghana..........................................................

Why this study .................................................................................

Literature review..............................................................................

The Citronella plant..........................................................................

Importance of Citronella..................................................................

Insect interactions with the Poaceae in general............................

Insect pests of Citronella.................................................................

INSECTS ASSOCIATED WITH CITRONELLA.......................

Introduction.......................................................................................

Materials and methods....................................................................

Sampling Points and C lim ate.........................................................

Sampling procedures........................................................

University of Ghana          http://ugspace.ug.edu.gh



2.2.2.1 Insects on citronella shoot (leaves and stems)............................................................

2.2.2.2 Insects (pests) damaging citronella  16

2.2.2.3 Comparison of the Efficiencies of the Improved and Standard Sweep nets  19

2.2.2.4 Comparison of Sampling methods  19

2.3 Results  23

2.3.1 Insects on citronella shoot (leaves and stems)................v   23

2.3.2 Insects damaging citronella  24

2.3.3 Comparison of the Efficiencies of the Improved and Standard Sweep nets............ 26

2.3.4 Comparison of Sampling methods.............................................................................  27

2.4 Discussion.................................................................................................................... 35

3.0 INSECT PREFERENCES FOR DIFFERENT SIZES OF HOST 
PLANT AND NATURE OF SHOOTBORER DAMAGE TO THE
SHOOT OF CITRONELLA....................................................................................... 42

3.1 Introduction...............................................................................................................  42

3.2 Materials and Methods.............................................................................................  43

3.2.1 Field investigation of preference for size of Citronella tillers by shootborers......... 43

3.2.2 Boring activity of the shootborers................................................................................  43

3.2.3 Field Identification of damage from differential growth (height) of newly 
ratooned infested and uninfested tillers........................................................................ 44

3.3 Results............................................................................................................................ 46

3.3.1 Field investigation of preference for size of Citronella tillers by shootborers.......  46

3.3.2 Boring activity of the shootborers.................................................................................  46

3.3.3 Field identification of damage from differential growth (height) of newly 
ratooned infested and uninfested tillers........................................................................ 47

University of Ghana          http://ugspace.ug.edu.gh



51

54

54

54

54

56

56

56

58

67

70

70

70

74

80

82

82

83

85

91

Discussion.............................................................................................................

SEASONAL VARIATION IN INFESTATION OF CITRONELLA BY 
LEPIDOPTEROUS STEMBORERS AND CHLOROPID SHOOTFLIES ..

Introduction...........................................................................................................

Materials and methods...........................................................................................

Determination of Sample size and measurement of Incidence of 'deadheart’ 
caused by shootborers...........................................................................................

Rainfall, Temperature, R.H. and Evaporation....................................................

Results.....................................................................................................................

Determination of Sample size and measurement of incidence of'deadheart’ 
caused by shooborers...........................................................................................

Rainfall, Temperature, R.H. and Evaporation...................................................

Discussion..............................................................................................................

ASPECTS OF THE BIOLOGY OF SHOOTFLY, SCOLIOPHTHALMUS 
MICANTIPENNIS (DIPTERA: CHLOROPIDAE)...........................................

Introduction........................................................................ ;..................................

Materials and methods..........................................................................................

Results.................................................................................................................. .

Discussion............................................................................................................. .

SUMMARY AND CONCLUSIONS..................................................................

Summary.................................................................................................................

Conclusions............................................................................................................

REFERENCES.............................................

x

APPENDICES

University of Ghana          http://ugspace.ug.edu.gh



LIST OF FIGURES Page

Figure 1: Map of Southtem Ghana showing locations where sampling was done ............. 13

Figure 2: The standard and improved sweep nets.................................................................. 22

Figure 3: Infestation rates as determined by only 'deadheart' symptom and by

Random dissection.................................................................................................... 28

Figure 4: Dorso-Iateral view of female Scoliophthalmus trapezoides..............................  29

Figure 5: Dorso-lateral view of female Scoliophthalmus micantipennis..........................  30

Figure 6: Lateral view of female Anatrichus pygmaeus........................................................  31

Figure 7: Dorso-lateral view of female Aprometopis flavofacies......................................  32

Figure 8: Lateral view of female Elachiptera occipitalis.................................................... 33

Figure 9: Longitudinal section of tiller attacked by lepidopterous stemborer........................ 48

Figure 10: Longitudinal section of tiller attacked by shootfly./............................................  49

Figure 11: Daily growth of healthy and infested tillers of ratooned citronella at

Osofoiaman in one week in 1991.........................................................................  50

Figure 12: Percentage 'deadheart' caused by shootflies and lepidopterous stemborers,

Osofoiaman Farm B, 1991................................................................................  59

Figure 13: Percentage 'deadheart' caused by shootflies and lepidoperous stemborers,

Osofoiaman Farm A, Jan - June, 1991................................................................ 60

Figure 14: Mean monthly rainfall, evaporation, relaltive humidity, temperature and 

percentage 'deadheart' by shootflies and lepidopterous stemborers,

Osofoiaman, Farm B, 1991...................................................................................  61

Figure 15: Relationship between percentage 'deadheart' by stemborers and rainfall....... 62

xi

University of Ghana          http://ugspace.ug.edu.gh



Figure 16: Relationship between percentage 'deadheart' by shootflies and rainfall............  63

Figure 17: Relationship between percentage 'deadheart' by shootflies and evaporation....... 64

Figure 18: Relationship between percentage 'deadheart1 by shootflies and temperature....... 65

Figure 19: Relationship between percentage 'deadheart1 by shootflies and relative humidity 66 

Figure 20: Dorsal view of male puparium of Scoliophthalmus micantipennis with the tip

broken to provide passage for the emerging adult...........................................  77

Figure 21: Dorsal view of female puparium of Scoliophthalmus micantipennis with the

tip broken to provide passage for the emerging adult....................................... 78

xii

University of Ghana          http://ugspace.ug.edu.gh



LIST OF TABLES Pag£

Table 1: Foliar damage by insects to citronella leaves in different farms..........................  24

Table 2: Adult Diptera reared, and Lepidoptera larvae collected, from tillers showing
'deadheart' symptoms from Osofoiaman and other farms in 1991...................... 25

Table 3: Incidence of termites in sampled farms, in Greater Accra Region, 1991............  26

Table 4: Analysis of twenty (20) samples of insects each by improved and standard sweep
nets, at 10 sweeps per sample................................................................................... 27

Table 5: Summary and analysis of tillers preferred and damaged by the shootborers....... 47

Table 6: Survival and Development of larvae of shootflies collected from infested tillers 75

Table 7: Length of empty puparia of female and male Scoliophthalmus micantipennis.... 76

Table 8: Pupal period of male and female Scoliophthalmus micantipennis in the laboratory
at 27.4°C................................................................................................................  76

Table 9: Longevity of adult male and female Scoliophthalmus micantipennis in the
Laboratory at 27.4°C........................................................................................... 79

Table 10: Total body length of adult male and female Scoliophthalmus micantipennis...... 79

xiii

University of Ghana          http://ugspace.ug.edu.gh



LIST OF PLATES Page

Plate 1: Citronella farm..................................................................*..................................... .......... 3

Plate 2: Rearing of insects from cut citronella shoots........................................................  21

Plate 3: Zonocerus variegatus attacking citronella on the farm.......................................  34

Plate 4: Experiment to study the boring activity of shootfly larvae..................................  45

Plate 5: Experiment to determine oviposition site of Scoliophthalmus micantipennis

in the laboratory.........................................................................................................  73

xiv

University of Ghana          http://ugspace.ug.edu.gh



CHAPTER ONE

1.0 GENERAL INTRODUCTION AND LITERATURE REVIEW

1.1 General Introduction

1.1.1 Introduction of Citronella in Ghana

Citronella is an essential oil producing Poaceae (Gfaminae) with origins in Asia. It is 

however, now cultivated in many tropical and subtropical areas such as Taiwan, Malaysia, Sri 

Lanka, Indonesia, Seychelles Islands, Zaire, B razil, Uganda (Skerman and Riveros, 1990). 

The two main species, Cymbopogon nardus and Cymbopogon winterianus, have been 

cultivated in Ghana since 1978 (Laryea, 1988).

Citronella was introduced in Ghana from Sri Lanka by Lever Brothers Ghana Limited 

(LBG) in 1978 through Dr. David Tagoe, the then Technical Services Manager of LBG, to 

produce citronella oil locally for the company’s soap manufacturing industry. A few tillers 

each of the following five types of Cymbopogon were introduced;

i. Cymbopogon nardus

ii. Cymbopogon winterianus

iii. Citronella variety Number one

iv. Lemongrass ex Sri - Lanka 

v. East Indian Lemongrass

Initial agronomic studies at the University of Ghana showed that it was feasible to 

cultivate all the five types in Ghana. LBG, however, preferred the citronella to the lemon grass 

because it had a higher yield of oil and therefore it was of better commercial value. In 

1981, LBG’s first attempt at cultivating citronella in a pilot scheme at Banso in the Western

1

University of Ghana          http://ugspace.ug.edu.gh



Region failed because of poor soil condition due to high acidity, high incidence of stemborers 

and rust infection at the site (Laryea, 1988).

LBG attempted another pilot project at Bunso in the Eastern Region in 1982. This was 

successful and the farm served as the source of planting material for many outgrowers. As at 

1989 there was a total of 46 farms of citronella spread over Brong Ahafo, Ashanti, Western, 

Eastern, Central, Greater Accra and Volta regions (Laryea, personal communication). The 

farmers formed an association known as Citronella Oil Producers Association (COPA), and 

they are the core of citronella producers in Ghana. .According to Laryea (Personal 

communication), their number as at 1992 had risen to just over 50 with farms covering over 

200 ha from which about 10,000 kg of oil was being produced annually. The number of 

farmers could have been higher but for the ceiling put on COPA membership by LBG. There 

are numerous small scale citronella farmers outside COPA.

1.1.2 Citronella Farming in Ghana

Citronella is currently cultivated countrywide except in northern Ghana. According 

to LBG, sandy loam with fairly distributed annual rainfall of over 900mm is the most preferred 

condition for the cultivation of citronella. The most suitable time for planting is at the onset 

of the rains. The planting material is at least a 6-month old tiller pruned of the leaves and 

roots. The tillers are planted about 8 to 10cm deep in the soil and spaced about 60cm, giving 

about 27,000 tillers per hectare. Time and frequency of weeding is determined by the farmer. 

Good husbandry requires that flowers from citronella should be removed since flower 

production depletes the oil content of the plants. Harvesting of citronella leaves is done when 

the leaves become mature at which stage the tips become dry and curl backwards. In times

2

University of Ghana          http://ugspace.ug.edu.gh



3

Plate 1: Citronella Farm

University of Ghana          http://ugspace.ug.edu.gh



of adequate rainfall, this is about 6 months after planting and subsequent harvests are done at 

3-4 month intervals or shorter.

Harvested citronella is dried in the field for about 4 days to reduce the water content 

and also to reduce the load on the still. Harvesting is therefore not recommended during the 

very wet periods since the leaves rot during drying. Dried citronella is cleaned of moulds and 

other foreign bodies before being distilled by steam.

1.1.3 Why this Study

One of the reasons for the failure of the first citronella pilot programme at Banso in 

1981 was stemborer attack. Also in 1988, a team from the University of Ghana reported 

lepidopterous stemborer attack and possible leaf hopper damage at Ayikuma in the Greater 

Accra Region. According to the report, damage by the insects was so serious at Ayikuma and 

some areas in the Greater Accra region that severe crop losses and revenue were recorded. 

There is also the fear that since citronella is cultivated as a monocrop, the insect pest problem 

will become more serious as cultivation increases.

Apart from the initial investigations, no detailed study has been done on the pest and 

disease problems of citronella. This study is, therefore, aimed at providing a more detailed 

information about the insects associated with citronella, and to study the ecology and biology 

of the major pest(s), with a view to providing baseline information that will be useful in 

designing an ecologically acceptable and economically feasible pest management programme 

for the citronella cultivation in Ghana.

The objectives of the study are, therefore,

i. to collect and identify all stages of insects associated with citronella in south

4

University of Ghana          http://ugspace.ug.edu.gh



- eastern Ghana,

to identify the major and potential insect pests and their natural enemies, when 

possible,

to determine insect pest preferences for different sizes of host and to 

characterize the nature of pest damage,

to study the seasonal trends of infestation of the major insect pests of citronella 

to determine the effects of relative humidity (R.H.), evaporation, temperature 

and rainfall on the infestation of the major insect pests of citronella and 

to study the biology of the major insect pest(s) of citronella.

5

University of Ghana          http://ugspace.ug.edu.gh



1.2 Literature Review

1.2.1 The Citronella Plant

Citronella is a perennial grass and it is classified as follows :

Family : Poaceae (= Graminae)

Subfamily : Panicoideae 

Tribe : Andropogoneae

Genus : Cymbopogon

The genus Cymbopogon has about 60 species (Purseglove , 1983) which give varied products 

including essential oils, fodder, mulch and thatching material (Bor, 1960; Arber, 1965; 

Purseglove, 1983; Skerman and Riveros, 1990; Bajracharya, 1990; Singh et al, 1991). The 

species include Cymbopogon nardus (Rendle - Ceylon/Sri Lankan type) and C. winterianus 

Jowitt - Java type which produces citronella oil ( Wijesekera, 1973), C. citratus (DC) Stapf, 

which produces lemon grass oil (Arber, 1965) and C. martinii (Roxb) Wats which gives 

palmrosa oil (Bor , 1960 ).

Citronella grows in tufts and at maturity, the height varies from about 0.5 to lm. The 

plant consists of tillers which vary from two to over a hundred, depending on age and variety. 

A tiller consists of a short rhizome, made up of a series of nodes separated by short intemodes 

and relatively long sheaths crowned by long narrow leaves. The basal nodes are very close 

together and the lowest nodes normally give rise to the roots.

The morphological differences between C. nardus and C. winterianus lie in the shape 

of their leaves which, according to Senaratne (1956); are narrower in C. nardus. 

Cymbopogon nardus is hardier than C. winterianus and while the former can thrive even on

6

University of Ghana          http://ugspace.ug.edu.gh



marginal soils and produce for 10 to 15 years, the latter thrives only on well drained sandy 

loam with abundant rainfall and can produce for only three to four years (Laryea, 1988). The 

chemical difference between the two species is found in the citronellal component of the oil 

(Wijesekera, 1973). Thus, C. winterianus contains more citronellal (32.7%) than C. nardus 

(5.2%).

1.2.2 Importance of Citronella

Citronella plants are cultivated for their extensively used essential oils, which are 

mainly terpenes such as geraniol, citronellal and citronellol. The most important oils produced 

are citronella oil from C. nardus and C. winterianus (Bor , 1960); lemon grass oil from C. 

citratus (Aber, 1965); and palmrosa oil from C. martinii (Bor, 1960).

According to Wijesekera (1973), citronella oil is used in perfumes; as deodorants in 

mosquito repellants, insecticides, soaps, disinfectants, paints and polishes. It is also used in 

the preparation of aromatic chemicals derived from citronellol and geraniol. Citronella oil also 

has great potential in pest control, for it repels stored product pests, such as Tribolium 

castaneum (Hbst)(Coleoptera: Tenebrionidae) and Bruchus chinensis (L.) (Coleoptera: 

Bruchidae), as well as Periplaneta americana (L.) (Dictyoptera: Blattidae) (Saraswathi and 

Rao, 1987). The oil may also inhibit fungal growth and spore gennination (Raghavaiah and 

Jayarmaiah, 1987). In Ghana, Lever Brothers Ghana Limited (LBG) uses citronella oil to 

perfume the following brand soap products: KeyR soap, Sunlight11 soap, and GuardianR soap. 

Lemon grass oil is used in the commercial synthesis of vitamin A and also in scenting soaps, 

while palmrosa oil is acclaimed to be the best natural source of geraniol used as a soap 

perfume and also for the flavouring of tobacco (Purseglove, 1983).

7

University of Ghana          http://ugspace.ug.edu.gh



Other products of citronella are also useful. For example, water extract from C. 

flexuosus is a nematicide that suppresses root knot development in tomato and aubergine 

caused by Meloidogyne incognita (Tiyagi et al, 1990), leaves of C. nardus are used as mulch 

and as a thatching material (Skerman and Riveros, 1990) whereas the spent grass o f C. 

flexuosus is used as fodder (Bajracharya, 1990).

1.2.3. Insect interactions with the Paaceae in general

Citronella belongs to a family of pants (Poaceae) which play host to numerous insects, 

consisting of plant feeders, predators, parasitoids and casual visitors (Tams and Bowden, 

1953; Ingram, 1958; Nye, 1960a; Harris, 1962; Watts, 1963; Byers, 1967; Deeming, 1971). 

The age, morphological features, texture, turgidity and manner of growth of the plant are 

among the factors that influence insect attack of the group (Wiliams, 1954; Mulkem, 1967; 

Easwaramoorthy and Jayaraj, 1988). Shootborer activity on poaceous plants also tend to 

fluctuate under the influence of several extraneous factors including the following, wet and 

dry seasons (Sarel-whitfield, 1929; Rose, 1973; Wilson et al, 1973), cropping seasons (Harris, 

1962; Endrody-Younga, 1968; Girling, 1978) and weather and climate (Smithers 1960; Harris, 

1962; Adeyemi, 1969; Barry, 1972; Balraj et al, 1977; Clearwater and Otieno, 1977; Ogwaro, 

1979), as well as farming practices (Harris,1962; Girling,1978) and altitude (Ingram, 1958; 

Nye, 1960b; Reddy, 1982).

Insect attack on the family produces several damage symptoms such as 'white head’ 

and 'deadheart’ (Nye, 1960a; Deeming, 1971; Agen-Sampong, 1975), gouted shoots 

(Goodlife, 1942) and wilted shoots (Simmonds, 1952). It may directly result in decrease in 

yield (Poos and Johnson, 1936; Wolfenbarger, 1963; Peterson and Granovsky, 1950; Byers,

8

University of Ghana          http://ugspace.ug.edu.gh



1967) or in the transmission of virus and other pathogens for diseases, the symptoms of which 

include yellows, chiorotic streaks, necrosis of the phloem and tumor (Smith and Brierley, 

1956; Day and Venables, 1961; Kennedy et al, 1962; Watson and Plumb, 1972).

1.2.4 Insect pests of citronella

There is scanty information on insects associated with citronella because of the lack 

of systematic literature on both the crop (Gour et al, 1991) and insect pests (Sontakke et al, 

1991). Nevertheless, insects appear to be a problem to the root, stem and leaves of citronella 

(Sontakke et al, 1991, Gour et al, 1991). For instance, the worker termite Microtermes obesi 

(Isoptera; Termitidae) and the white grub, Holotrichia consanguinea (Blanchard) (Coleoptera: 

Melolonthidae), attack the root and eventually the stem leading to the death o f the plant 

(Sontakke et al, 1991). The same authors reported further that between 20 and 30% of the 

stems of young citronella are severely infested by the larvae of Chilo infuscatellus (Snellen) 

(Lepidoptera: Pyralidae), a stemborer. Gour et al (1991) identified the larva of 

Stemmatophora fuscibalis (Snellen) (Lepidoptera: Pyralidae) as a stemborer of citronella. In 

Ghana, the stem is attacked by the larvae of Eldana saccharina Walker (Lepidoptera: 

Pyralidae), Sesamia sp. (Lepidoptera: Noctuidae) and Atherigona sp. (Diptera: Muscidae) 

(Botchwey, personal communication).

The leaves are attacked by both the adult and the nymph of Anaphothrips sudanensis 

Trybom, Chloethrips holorphnus (Karney) (Thysanoptera: Thripidae), Macrosiphum 

miscanthi (Talcahashi) (Homoptera: Aphididae), the larvae of Mythimna separata (Walker) 

(Lepidoptera: Noctuidae) and the nymphs of Colemania phenarioides Bol (Orthoptera: 

Acrididae) (Sontakke et al, 1991). Infested leaves are characterised by white patches and sap

9

University of Ghana          http://ugspace.ug.edu.gh



sucking by the insects results in some leaves curling inwards longitudinally from the margin 

while others dry at the tip ( Sontakke et al, 1991).

Three species of leaf hoppers, Zygina manensis, Neophotettix nigropictus and 

Typhlocyba maculifrons (Homoptera: Cicadellidae) have also been found to be associated with 

the leaves of citronella, but the pest status and the nature of damage caused by these insects 

were not ascertained (Sontakke et al, 1991).

10

University of Ghana          http://ugspace.ug.edu.gh



CHAPTER TWO

2.0 INSECTS ASSOCIATED WITH CITRONELLA

2.1 Introduction

Grasses are generally visited by many insects, some of which are plant feeders, 

predators, parasitoids and casual visitors. For instance, Watts (1963) collected insects made 

up of 9 orders, 55 families, 109 genera and 120 species of the above categories on black grama 

grass. Other poaceous plants like maize, sorghum, millet, wheat, oats, barley, sugar cane and 

rice also interact with insects in various ways (Tams and Bowden 1953; Ingram 1958; Nye, 

1960a; Harris, 1962 and Deeming, 1971).

There is lack of information on the insects associated with citronella. Only few 

workers like Sontakke et al (1991) and Gour et al (1991) have provided some information on 

insects attacking the crop and from their work, the leaves, stems and roots of citronella are 

attacked by various insects in India (Section 1.2). However, information on insect abundance 

and activity on the crop can be obtained from surveying the insect complex through various 

sampling methods including light traps, Malaise traps and sweep nets (Southwood, 1968). 

Pre-proposal visits to citronella farms in the Greater Accra and Eastern Regions showed the 

presence of several insects on the crop with those of economic importance occurring in 

immature stages in the shoot. This study was therefore undertaken to find the occurrence of 

general insects, insect feeders (pests), potential feeders and the natural enemies of these 

feeders, where possible.

2.2 Materials and methods

2.2.1 Sampling points and climate

11

University of Ghana          http://ugspace.ug.edu.gh



The studies were carried out mostly on farms in the Greater Accra and Eastern 

Regions of Ghana (Fig. 1). In the Eastern Region, data were collected from farms at Omenako 

and Kpankpan-Ayikuma near Suhum. In the Greater Accra region, data were collected from 

Ashalaja, Ababio and Selina Taylor farms at Ayikuma in addition to data collected from a 

farm at Botoku, near Pokuase. Detailed studies were carried out at Osofoiaman in the Greater 

Accra region, where two farms were located. All the farms in the Eastern region were located 

in the forest zone. The rest were in the coastal savanna. *

Osofoiaman is a small village, about 15 km north-west of Accra (Fig. 1). It lies on 

latitude 5°44' N and longitude 0°.16' W. on the southern-most portion of the Akwapim 

ranges. The main occupation of the people is farming. Pineapples and citronella are the main 

cash crops which are cultivated, together with food crops like cassava, maize and vegetables. 

Farm A at Osofoiaman was located at about 200 meters south of the village. It was part of 

Kate citronella farms and it was located about 40 meters from an older and larger farm. Farm 

A measured about 88 x 18.5 m and it contained about 1800 citronella hills cultivated without 

any ploughing. It was bordered by wild grasses, maize and an abandoned farm of plantain and 

banana. The soil was mainly sandy loam and clay. The latter contributed to the flooding of 

the farm during rains since it was poorly drained. Farm B belonged to a small scale farmer. 

It was located about 300 meters N.W. of the village. The size was 58 x 42m with about 1570 

citronella hills, also cultivated without ploughing. It was bordered by pineapple and citronella 

farms, grasses, cassava and maize. The soil was sandy loam and well drained.

The climatic factors are summarized in Appendix I. Even though Osofoiaman lies 

within one of the driest zones in Ghana (The Accra plains), its climate is ameliorated by relief

12

University of Ghana          http://ugspace.ug.edu.gh



13

Fig 1: Map of Southern Ghana showing locations where sampling was done.

University of Ghana          http://ugspace.ug.edu.gh



14

rains from the Akwapim ranges. The climatic data presented in this section were obtained 

from the Ghana Meteorological Services, Legon, from their station at Pokuase, 4km west of 

Osofoiaman, for a five year period (1987 to 1991). Temperatures varied little from year to 

year. Average maximum temperatures for the period ranged between 29.3°C (in August) and 

34.7°C (in February). The average minimum temperatures ranged between 20.7°C (in 

January) and 23.6°C (in February).

The values for the relative humidities were calculated from the dry and wet bulb 

readings at 0900 hours and 1500 hours. Higher r.h. values recorded at 0900 hours ranged 

between 88 and 94%. Relative humidities at 1500 hours were lower; ranging between 70 and 

88%. The r.h. regimes exhibited an inverse relationship with those of the temperature. Thus 

the hottest period of the year had the lowest humidity with the reverse also being true for the 

coolest periods of the year.

There was a steady increase of the average evaporation rates from January (1.9 mm.) 

to April (2.6 mm.) and subsequently decreased over the months to December (1.6 mm.).

The lowest rainfall was 8.8 mm in January and the highest, 202.4 mm in May. The 

rainfall was bimodal with the first rainy season beginning in April and ending in July. The 

second one began in September even though August was not completely dry and it ended in 

November. .

The other sites showed similar climate and four distinct seasons were recognized as 

follows:

1. hot season from March to April and also from October to November,

University of Ghana          http://ugspace.ug.edu.gh



2. cool wet season from May to July and September,

3. cool dry season from August with little rainfall, and

4. hot dry season from December to February, which was the driest period with 

low rainfall and low humidity and very high temperatures and high evaporation. This 

is the harmattan period of cool nights and hot days and dry winds.

2.2.2 Sampling procedures 

2.2.2.1. Insects on Citronella Shoot (Leaves and Stems)

Insects whose interaction with the shoots of citronella were not obvious were sampled 

intensively and extensively using the sweep net, which is more appropriate for sampling 

insects on shoot of grasses because it samples about two thirds of the plant height (Nielson, 

1957). It is simple to use and provides rapid estimates of insect distribution, activity and 

abundance (Balogh, 1958). The intensive sampling involved the continual observation of the 

populations of insects that visited citronella at two areas and was carried out at Osofoiaman 

Farm A for six months (January to June, 1991.) and at Osofoiaman Farm B for 11 months 

(February to December, 1991) except in May when heavy rainfall made sweeping impossible.

Investigations on Farm A had to be discontinued in June when the field was 

completely destroyed by floods. The farms were visited weekly and sweeping done before

10.00 a.m. except when it was raining or when the visit coincided with the harvesting time of 

the farmer. Sweeping was done with a standard insect net of diameter 47 cm and depth 46 cm. 

Ten to and fro swings constituted one sample, which was transferred into a killing jar charged 

with ethyl acetate. Twenty samples, made up of five along each diagonal and each of two 

opposite sides of the farm were taken. The samples were ‘taken to the laboratory where the

15

University of Ghana          http://ugspace.ug.edu.gh



insects were identified, using keys and museum specimens, and their numbers recorded. Some 

of the insects were pinned, dried and kept in boxes while others were preserved in 70% 

alcohol.

The extensive sampling, was aimed at studies on the distribution of insects 

associated with citronella in the study area and was conducted in six farms at Osofoiaman 

(Farm C), Botoku, Omenako and Kpankpan Ayikuma near Suhum and Ayikuma near Dodowa 

(Selina Taylor and Ababio's farms). Each of these six farms was sampled with an improved 

net (Fig. 2b) at about three-week intervals, between the last week of October and the last week 

of December 1992. At each farm, 50 x 40 m was demarcated and sweeping done as in the 

intensive sampling. Insects collected were killed and identified as above. The extensive 

sampling had to be done between 11.00 a.m. and 3.00 p.m. because of problems from using 

public transport.

2.2.2.2 Insects (pests) damaging citronella

Insects damaging the leaves: Feeding on the foliage of citronella by insects may be 

manifested as defoliation of the plant, necrosis or curling of the leaf (Sontakke et al, 1991). 

Thus, this study was initiated in experimental Farm B to find the prevalence o f these 

symptoms and the insects responsible. Moving in a zig-zag fashion from one long side o f the 

farm to the other, every eighth (8th) plant was sampled and all the leaves inspected for 

defoliation, necrosis and curling. Such leaves were counted and recorded. Sampled plants 

were marked with pegs and skipped if encountered again. This was done monthly for six 

months, (February, March, June, September and November in 1991). It was repeated 

fortnightly, between the last week of October and last week of December, 1992 in Osofoiaman

16

University of Ghana          http://ugspace.ug.edu.gh



(Farm C), Botoku, Omenako, Kpankpan-Ayikuma, Selina Taylor and Ababio farms. The study 

was done on the same demarcated plots for sweeping as above. (See 2.2.2.1)

Insects damaging the stem: From preliminary studies at Osofoiaman, Ayikuma and 

Ashalaja, between November, 1990 and January, 1991, insect infestations were observed in 

the stems of citronella in all the farms visited. The following survey was thus initiated to 

find out the species of insects involved. The work was done at Osofoiaman Farms A and B, 

from January to June, 1991 and thereafter continued on Farm B and other farms at 

Osofoiaman, Botoku, Ayikuma and Ashalaja till December.

'Deadheart’ symptom, which is the gradual yellowing and browning of the innermost 

whorl of the tiller, as a result of damage by insects in the stem, was used to identify tillers 

infested by the shootborers. At Osofoiaman farms'A and B, randomly selected citronella hills 

were inspected weekly and all tillers counted. Those showing 'deadheart1 symptoms were 

removed, using a scapel and classified as either early deadheart1 tiller, advanced 

deadheart1 tiller or secondarily infested1 tiller in the case of Diptera. The inner whorl of a 

healthy tiller was pale green but infested tiller classified as early deadheart1; (i.e. freshly 

attacked) had the inner whorl just beginning to wilt and changing from pale green to greenish 

yellow. An 'advanced deadheart1, had the inner whorl wilted and the colour ranging from 

yellow to brown and may have multiple infestation by Diptera. The secondarily infested1 tiller 

was a condition where the tiller had already been attacked by a Lepidoptera before the Diptera. 

Those tillers that had been attacked by Lepidoptera only, were classified as 'Lepidoptera1. All 

the tillers showing ' deadheart' were taken to the laboratory and dissected, to determine the

17

University of Ghana          http://ugspace.ug.edu.gh



causes of the 'deadheart'. Tillers with dipterous larvae were carefully closed again and the 

portions with the larvae cut into 5-8 cm pieces, which were then stood in 10 x 2 cm tubes 

having moist cotton wool at the bottom, corked but opened frequently to allow in air (Plate2). 

The moist cotton wool prevented the cut tillers from drying quickly. The set up was inspected 

daily until adult flies emerged. If pupae of the Diptera were obtained from the tillers, they 

were placed individually in 10 x 2 cm tubes without moist cotton wool until adult flies 

emerged. Representative adults were labelled and sent to the British Museum for 

identification. Subsequently, all adults were identified using the keys provided by the British 

Museum. A key (Appendix 4), was eventually prepared for the shootflies identified in this 

study. The larvae of the Lepidoptera were either preserved or reared to adulthood and later 

identified. To preserve the larvae, they were killed in hot water of about 60°C, passed through 

40% and 60% alcohol and finally stored in 70% alcohol. The hot water ensured that the larvae 

maintained their colour patterns.

Farms at Ayikuma, Botoku and Ashalaja were also inspected monthly till December, 

1991, for 'deadheart' tillers and treated as those obtained from Osofoiaman Farms A and B. 

Insects damaging citronella root: Farms at Osofoiaman, Botoku, Ayikuma and Ashalaja 

were visited every two months between January and December, 1991, and citronella plants 

inspected for damage by root pests. Ten blocks of plots measuring 4 x 4 m were randomly 

selected and all the hills inspected. The presence of termitaria around the plants, and or the 

wilting of the leaves other than the inner whorl of the tiller, indicated the presence of root 

feeders, especially termites, in the roots. Such plants were uprooted and the underground parts 

examined. Termites were the only insects seen and some of them were collected for

18

University of Ghana          http://ugspace.ug.edu.gh



identification.

Natural enemies: Thirty- eight adults, 57 pupae and 172 larvae of the shootflies and 5 

lepidopterous stemborer larvae which died during rearing and 320 termite pests collected were 

kept in glass tubes sterilized by boiling and inspected for parasitoids and fungal growth. 

Predators were not screened but observations were made of insect activities in the field for 

predation.

2.2.2.3 Comparison of the Efficiencies of Improved and Standard Sweep Nets: It was 

observed from the intensive survey that the standard sweep net (Fig. 2a) caught rather low 

numbers of insects, especially smaller insects. The opening of the net was so wide that before 

the contents could be removed, swifter as well as and smaller insects would escape. An 

improved net was therefore designed and its efficiency compared with that of the standard 

sweep net. The improved net (Fig. 2b) had a narrower opening of diameter, D, 36.5 cm and 

a total depth, P, of 74 cm as against 52.5cm and 55.5cm, respectively, of the standard net. The 

nets were made of grey baft but the improved one had a detachable sac, T, of muslin material 

at the bottom of the net. After a collection has been made the detachable sac is closed by 

constricting the mouth with string, S. To determine their respective efficiencies, twenty 

samples were taken for each of the nets on a plot of citronella at Pokuase similar to what was 

done in the intensive sampling. The number of insects caught in one sample was counted, the 

time to make ten sweeps and get the insects into the killing bottle determined with stop watch, 

and insects caught per unit time, calculated for each net.

2.2.2.4 Comparison of Sampling Methods: Very low levels of infestation were recorded 

using 'deadheart’ symptoms alone to score infestation. In this experiment, therefore, results

19

University of Ghana          http://ugspace.ug.edu.gh



from this method were compared with those of another one to find out if  there was any 

difference. Based on the assumption that infested tillers that manifested in deadheart' 

symptom and those that did not, stood equal chance of being sampled, two methods for 

sampling shootborer infestation, viz, sampling using visual 'deadheart' symptom and sampling 

by random dissection were compared as follows; some 366 tillers were randomly removed 

and inspected for 'deadheart' symptom. Then they were all dissected carefully and inspected 

for shootborers. Percentage infestation as determined from the incidence of 'deadheart' was 

compared with that from random dissection, the null hypothesis being that there was no 

difference between the two. The comparison was replicated in 4 different farms.

20

University of Ghana          http://ugspace.ug.edu.gh



21

Plate 2: Rearing of insects from cut citronella shoots.

University of Ghana          http://ugspace.ug.edu.gh



22

L
a

Standard Sweep net

Improved Sweep net

7

The standard and Improved sweep nets

University of Ghana          http://ugspace.ug.edu.gh



2.3 Results

Intensive sampling: A total of 1093 insects of 20 families, belonging to eight Orders; 

Hemiptera, Coleoptera, Hymenoptera, Diptera, Dermaptera, Orthoptera, and Dictyoptera were 

collected in the 11 months study period at Osofoiaman Farm B and were associated with the 

shoot of citronella (see Appendix 2). Homoptera was the most abundant followed by 

Hymenoptera and then Hemiptera. The populations of the families: Cercopidae (Homoptera) 

and Formicidae (Hymenoptera) were consistently high throughout the survey. ForSculidae 

(Dermaptera), Pyrgomorphidae (Orthoptera), Diopsidae (Diptera), Lagriidae and 

Chrysomelidae (Coleoptera) also occurred in most months of the year but in low numbers. 

Extensive sampling: A total of 4567 insects, belonging to 10 Orders and 102 families were 

recorded associated with the shoot citronella in the other six farms (see also Appendix 3). The 

breakdown is as follows; 751 insect belonging to 8 Orders and 52 families were recorded at 

Osofoiaman farm C, 951 insects (9 Orders and 49 families) from Botoku and 682 insets (10 

Orders and 49 families) from Abobio farms. The rest were 716 insects of 10 Orders and 47 

families from Selina Taylor farms, 649 (9 Orders and 53 families) and 907 insects (10 Orders 

and 55 families) from Omenako and Kpankpan-Ayikuma farms respectively. The spectrum 

of diversity of insect fauna in farms in the forest zone was similar to that in the savanna. Also 

the member of families representing an order was in no way an indication of the degree to 

which such groups contributed to the total insect population. For example, Homoptera had 

eight families yet contributed 1613 insects but Hymenoptera with 12 families and Coleoptera 

with 10 contributed 160 and 83 insects, respectively.

23

2.3.1 Insects on citronella shoot

University of Ghana          http://ugspace.ug.edu.gh



Insects feeding on citronella leaves: The results (Table 1) showed that percentage necrosis 

in citronella leaves at the 7 different farms ranged between 0.0001% and 0.004%; that of 

curling was 0.001 and 0.008%. and defoliation between 0.0002% and 0.008%. This shows 

that negligible proportions of citronella leaves were attacked by insects.

Zonocerus variegatus

Zonocerus variegatus was the main species of Pyrgomorphidae (Orthoptera) collected in this 

study. It occurred in 8 months of the year in the intensive survey (Appendix 2) and in all 

farms in the extensive one, forming 73% of the Orthoptera group (Appendix 3). During the 

study, no feeding activity by the insect was observed in any of the farms visited. After the 

study period, however, it caused damage to citronella at a farm at Botoku near Pokuase. The 

damage was done by the adults and nymphs alike and large numbers of the species attacked 

the leaves at the edge of the blade and ate progressively towards the midrib which was left 

intact (plate 3).

24

2.3.2 Insects damaging citronella

Table 1: Foliar damage by insects to citronella leaves in different farms.

Farm Total number of 
leaves inspected

Percentage leaves showing
Necrosis

(%)
Curling
(%)

Defoliation
(%)

Osofoiaman farm B 18,314 0.00047 0.00025 0.0003

Botoku 17,917 0.00049 0.00012 0.008

Osofoiaman farm C 15,564 0.0002 0.00089 0.0004

Omenako 13,709 0.0033 0.0013 0.0003
Kpankpan-Ayikuma 16,315 0.0041 0.0079 0.0009
Selina Taylor’s 31,794 0.00084 0.00047 0.0003
Ababio's 27,842 0.00011 0.000411 0.0002

University of Ghana          http://ugspace.ug.edu.gh



Insects in citronella stem: The results (Table 2) showed 8 identified and 2 unidentified 

species of Diptera and also 2 identified ones of Lepidoptera. The Lepidoptera was present 

only in freshly infested tillers while the Diptera occurred in fresh as well as already infested 

ones. From Chi-square analysis of the data (Appendix 5), the numbers of different insects 

reared from the different stages of ’deadheart’ were significantly different (X2 = 192.29; p <

0.05). The predominant insect therefore was Scoliophthalmus trapezoides (Fig. 4) followed 

by Anatrichus pygmaeus (Fig.8b), S. micantipennis (Fig. 5), species D, E, Sesamia sp., 

Aprometopis flavofacies (Fig. 7), Silba pectita, Elachiptera occipitalis (Fig. 8) and Eldana. 

saccharina.

25

Table 2: Adult Diptera reared, and Lepidoptera larvae collected, from tillers showing 
'deadheart1 symptoms from Osofoiaman and other farms in 1991.

Insects Total number of insects 
reared/ collected

Number of insects from 'deadheart'

'Early' 'Late' or 'advanced' ’2°'

Diptera: Chloropidae Scoliophthalmus 64 21 35 S
trapezoides Becker (48) (17) (26) (5)

Scoliophthalmus micantipennis Duda 45 39 6 0
(39) (35) (4) (0)

Anatrichus pygmaeus Loew 46 0 14 32
(17) (0) (6) (11)

Elachiptera occipitalis Becker 3 0 0 3
(2) (0) (0) (2)

Aprometopis flavofacies Becker 5 5 0 0
(5) (5) (0) (0)

D 32 0 8 24
(18) (0) (5) (13)

Diptera: Lonchaeidae Silba pectita 4 0 3 1
McAlpine (2) (0) (1) (1)
Diptera: Phoridae 15 0 10 5

E (7) (0) (5) (2)
Lepidoptera: Noctuidae Sesamia sp 11 11 0 0

(11) (11) (0) (0)
Lepidoptera: Pyralidae Eldana saccharina 2 2 0 0
Walker (2) (2) (0) (0)

( ) Number of 'deadheart1 tillers.

University of Ghana          http://ugspace.ug.edu.gh



The results have been summarised in Table 4. Of the 12 farms visited, seven showed 

termite infestation; three severe, i.e. >10% and four mild i.p. <10%. Six of the infested farms 

were more than one year old, while one was less than one year old. The termites were not 

identified but their damage, which was caused by the workers, resulted in wilting and 

complete death. The hills were girdled with soil sheathing at, or just below, ground level. 

Examination of uprooted affected hills showed that the workers had eaten around the roots and 

into the stems, and had them filled with soil.

26

Insects damaging citronella root

Table 3: Incidence of termites in sampled farms, in Greater Accra Region, 1991.
Location Farm Age of Plants Total No. of hills inspected Percentage

infestation
(%)

> lyr. < I yr.

Ayikuma Tagoe farms II 1972 37.2

Vas farms I II 1916 0
.. „  j j l< 2304 4.0

Ashalaja New Match 
farms I

II 1920 . 14.0

New Match 
farms II

II 1692 4.3

Botoku Farm I II 1713 0
“ II II 1841 0

" III II 2024 3.1
Osofoiaman Kate farms I II 1724 0

* " II II 2074 23.0
Experimenta 
1 farm B I

II 1676 0

Experimenta 
1 farm B II

II 1978 0.7

Natural enemies: No parasitoids were observed form the dead insects, neither were there 

any pathogenic fungi. Also no predatory activity was observed in the field.

2.3.3. Comparison of Efficiencies of the Improved and Standard Sweep Nets: The results

University of Ghana          http://ugspace.ug.edu.gh



are summarised and analysed in Table 4 below. There were significant differences in insects 

caught in 20 samples at 10 sweeps per sample (t = 4.14; P < 0.05), in time to handle one 

sample (t = 4.54; P < 0.05), and in the rate of catch between the improved and standard nets 

(t = 6.30; P < 0.05). The former caught more insects and also it was quicker to handle its catch 

than the latter.

27

Table 4: Analysis of twenty (20) samples of insects each by Improved and Standard
Sweep nets at 10 sweeps )er sample.

Type of 
Sweep net

Range Mean 
x ± S.E.

3̂Sd f. (0 05)

Time for 1 sample 
(seconds)

Improved 58.0-75.3 65.62±.99 *4.54

Standard 77.0-164.0 100.18±7.55

Number of insects 
caught

Improved 10-45 23 .O il.9 *4.14

Standard 3 - 31 12.3±1.7

Rate
(Catch per second)

Improved 0.133-0.703 0.353±.029 *6.30

Standard 0.034-0.295 0.133±.029

* Significant at p < 0.05.

2.3.4 Comparison of Sampling Methods: The results are summarized in Appendix 6 and 

further illustrated in figure 3. The percentage infestation estimated from 'deadheart' symptoms 

alone underestimated the real infestation level which is determined from random dissection. 

The difference between the two methods was higher at Osofoiaman and Ashalaja than the 

remaining 2 farms.

University of Ghana          http://ugspace.ug.edu.gh



Pe
rc

en
ta

ge
 

in
fe

st
at

io
n

28

Fig.3

□  Only deadheart symptom 

El Random dissection

Omenako

Farm

Ababios

Infestation rates as determined by only "deadheart" symptom and by random 
dissection

University of Ghana          http://ugspace.ug.edu.gh



29

Fig 4: Dorso-lateral view of female Scoliophthalmus trapezoides

University of Ghana          http://ugspace.ug.edu.gh



30

X 35

Fig. 5: Dorso-lateral view of female Scoliophthalmus micantipennis

University of Ghana          http://ugspace.ug.edu.gh



l a *etai

University of Ghana          http://ugspace.ug.edu.gh



32

Fig. 7: Dorso-lateral view of female Aprometopis flavofacies

University of Ghana          http://ugspace.ug.edu.gh



33

X 4 0

Fig 8: Lateral view of female Elachiptera occipitalis

University of Ghana          http://ugspace.ug.edu.gh



34

Plate 3: Zonocerus variegatus attacking citronella on the farm.

University of Ghana          http://ugspace.ug.edu.gh



2.4 Discussion

The relative abundance of insects with respect to months might not be reliable in the 

intensive survey because of the inconsistency in the number of collections made per month; 

For instance, no collection was made in May, 1991 because sampling coincided either with 

rain or harvesting. Nevertheless, the results show that the vegetative growth of citronella 

supported a remarkably rich and diverse insect fauna. Thus from both intensive and extensive 

surveys, a total of 5660 insects belonging to 10 Orders and 103 families were found associated 

with the shoot of citronella and fairly distributed over the study area. The insects are mainly 

grass feeders, their parasitoids and predators.

The grass feeders include Cercopidae, Cicadellidae and Aphididae (Homoptera) whose 

feeding activities result in viral transmission (Watson and Plumb, 1972; Borror et al, 1976). 

Included also are Coccinellidae (Coleoptera) which attack leaves of maize and rice in Ghana 

(Schumutterer, 1969; cited by Scheibelreiter and Inyang, 19-74); Pyrgomorphidae (Orthoptera), 

Chloropidae and Lonchaeidae (Diptera) some species or which have been found to interact 

with citronella in this present study. Most of the grass feeders appear innocuous presently but 

they may subsequently assume important status as citronella cultivation continues. They are 

therefore a potential threat.

The predators include Chaemaeyiidae (Diptera), Sphecidae (Hymenoptera) and 

Mantidae (Opthoptera). Several members of Chaemaeyiidae and Sphecidae prey upon the 

grass feeders Aphididae, Cicadellidae and Delphacidae (Homoptera) (Watts, 1963; Thomas, 

1968, Evans, 1968) which are found in large numbers on citronella. Significantly the

35

Insects on citronella shoot

University of Ghana          http://ugspace.ug.edu.gh



species of Mantidae were green in colour or a combination of that colour to offer them good 

camouflage within the citronella leaves for their predatory activities. Among the parasitoids 

are Diapriidae, Braconidae, Eulophidae and Ichneumonidae (Hymenoptera), some species of 

which have been found by Simmonds (1952) to parasitise on other chloropids such as 

Oscinella frit. These parasitoid families are therefore important because they could be 

investigated for use in the biological control of the Chloropidae that damage citronella.

Insects damaging the leaves

Though the results showed large number of grass feeding families like Cercopidae, 

Cicadellidae, Psyllidae, Delphacidae, Aphididae (Homoptera), Pentatomidae, Lygaeidae, 

Tingidae (Heteroptera), Corimelaenidae, Curculionidae, Cerambycidae, Lagriidae, Meloidae, 

Elateridae, Pselapidae (Coleoptera) and Pyrgomorphidae, Tettigonidae (Orthoptera), none of 

their adults was observed feeding on the leaves of citronella during the survey. Results from 

further investigation also showed negligible symptoms of feeding activities by insect foliage 

pests. For example, less than 0.005%, 0.002% and 0.009%, showed necrosis, curling and 

defoliation respectively, resulting from insect attack. The present findings are similar to those 

made by Sontakke et al (1991) who, however, failed to indicate the extent of the damage. The 

absence of major foliage insect pests might be due to the repellent properties of the essential 

oils present in the leaves (Saraswathi and Rao, 1987; Bosch, 1971).

However, Zonocerus variegatus L. (Orthoptera: Pyrgomorphidae) is a potential 

foliage pest of citronella because of the isolated case of the species feeding extensively on the 

leaves at Botoku in 1992 (Plate 3) after the study period. Even though the insect ate any 

vegetation available and not only citronella it still poses a threat because of the extent o f its

36

University of Ghana          http://ugspace.ug.edu.gh



Insects damaging citronella stem

The results showed that the major insects that damaged citronella were found in the 

stem. The dipterous shootflies (Chloropidae), are particularly dominant. The presence of 

Chloropidae in citronella in this study shows their affinity for poaceous plants and wide 

distribution, especially in Africa. This is because species of the family, including 

Scoliophthalmus trapezoides, Scoliophthalmus micantipennis, Anatrichus pygmaeus and 

Aprometopis flavofacies have also been reported (Deeming, 1971; Le Pelley,1959) as attacking 

sorghum, maize, millet, guinea com and wild grasses in Nigeria, Kenya, Tanzania, Uganda, 

Cameroun, Zimbabwe, Senegal and Nigeria.

The results also suggest that the shootborers occupy at least three different feeding 

niches in citronella, primary feeding, secondary feeding and scavenging. Follow up 

experiments to determine the specific species occupying the different niches were unsuccessful 

because the insects failed to oviposit in captivity. However, species that damage fresh tillers 

are considered primary feeders if there is no evidence of the presence of other species, whereas 

those damaging previously infested tillers are secondary feeders and those found in tillers 

already attacked by Lepidoptera, apart from possibly being secondary feeders, might be 

scavengers or saprovores, or parasitoids or predators (Deeming, 1971).

In line with Demming’s (1971) classification therefore, A. flavofacies, Sesamia sp. 

and E. saccharina recorded in the present study, may be considered primary feeders of 

citronella, since they attacked only fresh tillers. The conclusions on Sesamia sp and E.

37

damage and its strong presence in citronella farms.

University of Ghana          http://ugspace.ug.edu.gh



saccharina agree with observations elswhere (Tams and Bowden, 1953; Ingram, 1958; Nye, 

1960a and Harris, 1962). S. micantipennis is also considered primarily as a primary feeder 

because as high as 86.7% of them attacked fresh tillers. The conclusions on the status o f S. 

micantipennis and A. flavofacies agree with those made by Deeming (1971) who, working in 

Samaru in Nigeria, concluded that these insects cause primary damage to sorghum and millet 

tillers, respectively.

Scoliophthalmus trapezoides was considered both a primary feeder and a secondary 

feeder since it damaged both fresh (32.8%) and already infested tillers (54.7%). The status of

S. trapezoides as a secondary feeder, agrees with the conclusion made by Deeming (1971), 

who found the insect causing secondary damage to sorghum and maize tillers. The status of 

A. pygmaeus, S. pectita, E. occipitalis and chloropid sp. D, was not clear since they were 

frequently found in both tillers already infested by other Diptera and Lepidoptera. However, 

the larvae of A. pygmaeus have been reported by Sabrosky (1962), cited by Deeming (1971), 

as being predaceous on other larvae.

The role of the Phoridae reared from citronella could also not be determined, but some 

species of the family have been recorded (Ingram, 1958) as being parasitoids of the 

Lepidoptera, Busseola fusca (Fuller) (Lepidoptera: Noctuidae). In this study, they were 

probably attracted by the presence of the Lepidoptera, E. saccharina or Sesamia sp in the 

citronella tillers.

The primary feeders are the most important insects since they initiate the attack on the 

plants (Deeming, 1971). Of the primary feeders observed in the present study, S. 

micantipennis constituted about half the number and it, is therefore considered the most

38

University of Ghana          http://ugspace.ug.edu.gh



important feeder of citronella. This observation agrees with a conclusion made by Deeming 

(1971) that S. micantipennis is one of the most important primary feeders of sorghum.

Insects damaging citronella root

Results from the study showed that only unidentified species of termites belonging to 

the family Termitidae (Isopterea) cause damage to the roots of citronella This observation is 

in line with that made by Sontakke et al (1991) in India. .The damage observed was not as 

widespread as that by the shootborers, as the termites were found in only 58.3% of the farms 

visited whilst the shootborers were present in all the farms. Despite their limited distribution, 

the termites pose a great threat to the citronella industry because their damage could result in 

the destruction of a whole hill, leading to herbage loss and reduction in oil yield.

The incidence of termites in citronella farms is probably related to the history of 

cultivation. In particular, long term cultivated farms have higher incidence than short term 

ones. For this reason 85.7% of infested farms had more than one year of cultivation as against 

only 16.7% which had less than one year. This observation agrees with the one by Wood et 

al (1977) who found long term cultivated soils to have higher population densities of termites 

than soils in the first year of cultivation. The reason found for this is that as the farms are 

cultivated, the availability of roots of natural vegetation decreases in the soil with the duration 

of cultivation (Black and Wood, 1989). The termites are therefore forced to forage more 

intensively on the surface (Wood et al 1980)

The exact relationship between age of citronella farm and severity of termite damage 

could not be established, because the exact ages of the farms could not be determined. It is 

however, suggested that the area be further investigated, since it could help determine how

39

University of Ghana          http://ugspace.ug.edu.gh



long citronella farm could be kept in the savanna without attracting too many termites.

Natural enemies

The insects collected from the shoot included predator and parasitoid families but no 

parasitoids were found in the field or from laboratory screening. This is in sharp contrast to 

the results by Simmonds (1952), in which numerous hymenopteran parasitoids from 

Chalcidae, Pteromalidae, Diapriidae, Prototrupidae and Cynipidae were found associated with 

Oscinella frit (Diptera, Chloropidae). Predators were not screened for, however, the absence 

of parasitoids and pathogenic fungi is probably because the shootborer populations were very 

low, as depicted by the low infestations (Fig. 13) and that not many insects could be screened.

Comparison of Sweep Nets: Results of the comparison of the sweep nets showed 

that the standard sweep net is less efficient for assessing insect population as observed by 

Strickland (1961) also. Thus the standard net caught about a third (0.13 + 0.02) of the insects 

caught by the improved one (0.35 + 0.02). Also it took longer time to collect a sample using 

the standard net (100 ± 7.6s) than using the improved one (65.6 ± 0.1s). Furthermore, the 

improved net caught the smaller insects more than the standard one. For example as many as 

25 small adult chloropid shoot flies whose larvae damage citronella tillers were caught with 

the improved net as against none by the standard net at Osofoiaman Farm B. The improved 

net should therefore be developed further for use in sampling insects on the Poaceae.

Comparison of Sampling Methods: T-test analysis, after square root transformation 

of the data showed that the percentage infestation determined from only 'deadheart' symptom 

(mean, 0.77%) was significantly lower than that determined from random dissection (mean, 

1.13%) (t = 3.18; P < 0.05). The null hypothesis is therefore rejected. Indeed there is a

40

University of Ghana          http://ugspace.ug.edu.gh



function of y = 1.078x + 0.3 between the two rates where y = percentage infestation by 

dissection and x that by only 'deadheart'. Thus, shootborer infestation determined from only 

visual 'deadheart' symptom underestimate the actual infestation by (1.078x + 0.3) -  x, about 

31.8% on the average, probably because freshly attacked tillers without 'deadheart’ symptoms 

were missed. The results are by no means conclusive, considering the small number of 

replicates, i.e. 4, and also the fact that it was not possible to repeat the work in the replicates. 

Therefore, there is the need to investigate further to establish a very reliable relationship 

between the two methods for predictive purposes.

In conclusion, the major insects damaging citronella were found to be dipterous 

shootflies and lepidopterous stemborers. The minor ones were termites which attacked the 

root and the stem. No natural enemies were found. The dipterous shootflies belonged mainly 

to Chloropidae. Zonocerus variegatus is a potential pest of the leaves even though the leaves 

are not seriously damaged by insects. The major feeders are more important, hence a further 

look is taken of their damage in the next chapter.

41

University of Ghana          http://ugspace.ug.edu.gh



CHAPTER THREE

3.0 INSECT PREFERENCES FOR DIFFERENT
SIZES OF HOST PLANT AND 

NATURE OF SHOOTBORER DAMAGE
3.1 Introduction

In the previous chapter, the most significant insects causing damage to citronella were 

found to be lepidopterous stemborers and dipterous shootflies which bored into the stem. 

Earlier investigations found that the major pests of citronella were those that bored into the 

stem and led to an externally visible symptom known as the 'deadheart' symptom. The damage 

may be severe as in the case of Chilo infuscatellus Snellen (Lepidoptera: Pyralidae) (Sontakke 

et al, 1991) or mild as in the case of Stemmatophora fuscibasilis Snellen (Lepidoptera: 

Pyralidae) (Gour et al, 1991). Severe borer infestation may result in serious losses in crop and 

revenue, as occurred with Eldana saccharina Walker (Lepidoptera: Pyralidae), Sesamia sp. 

(Lepidoptera: Noctuidae) and Atherigona sp. (Diptera: Muscidae) (Botchwey, personal 

communication).

Infestation by a particular insect may be influenced by several factors including size 

and age of the host plant. Thus, bigger and taller hosts are preferred by lepidopterous stem 

borers (Patch, 1929; Chiang et al, 1960; Walker, 1963) while shootflies prefer smaller as well 

as shorter tillers (Deeming, 1971; Simmonds, 1952). In pre-proposal studies conducted at 

Mayera, Ashalaja and Ayikuma, smaller and shorter tillers appeared to be preferred by the 

chloropid shootflies. In this work, therefore, field studies were carried out to investigate insect 

preferences for different sizes of host plants. This could hfelp in the selection of materials for 

planting. A laboratory investigation of the boring of the insects was also carried out where

42

University of Ghana          http://ugspace.ug.edu.gh



3.2 Materials and methods

3.2.1 Field investigation of preference for size of citronella tillers by shootborers.

This preliminary investigation was conducted in the field to find out if the insects 

exhibited any preference in their choice of citronella tillers for attack.

A total of 70 tillers in two groups of 35 tillers each, already infested by each group of 

shootborers were randomly removed from farms at Omenako, Ashalaja and Ayikuma from 

October to December 1992. Measurements were taken of the tiller heights, diameters and 

lengths of tunnels in the laboratory. The lengths and heights were measured with straight- 

edged ruler while the diameters were measured with a micrometer screw guage. Measurement 

of the diameters were taken from two opposite sides of the base of the sheath in the region of 

the apex immediately above the uppermost exposed node. The results were analysed 

statistically, following the procedure by Walker (1963) and Endrody-Younga (1968).

3.2.2. Boring activity of the shootborers

In the previous chapter, adult shootborers were reared from infested tillers collected 

from the field. This experiment was therefore designed to find out how the insects bored into 

the tiller, their activity inside it and the effects of their activity on the tiller. Different larval 

instars of the shootborers were collected from field infested tillers and put on leaves and 

sheaths of potted citronella in the laboratory (Plate 4). They were inspected daily and boring 

activities such as point of entry into stem, direction of boring and effect on leaves were 

studied. After a larva had bored into a tiller and the inner whorl began to wilt, the tiller was 

dissected and the active larva transferred onto another tiller. In this way 13 lepidopterous

43

possible to study the nature of damage done to the tiller.

University of Ghana          http://ugspace.ug.edu.gh



larvae were made to bore into 35 tillers. Shootfly larvae obtained either died or pupated 

without attacking the plant.

3.2.3. Field identification of damage from differential growth (height) of newly ratooned 

infested and uninfested tillers

In the field, damage by shotborers has been most easily identified by only one index

i.e. 'deadheart’. This limits the ability of the farmer to detect early infestation in situations 

where 'deardheart’ symptom may not be showing particularly in freshly ratooned crop. This 

study was therefore undertaken to detect early infested crop without 'Deadheart’ symptom. 

A farm measuring 50 m by 40 m and lying adjacent to Farm B at Osofoiaman, was harvested 

five months after planting. Every fifth hill was selected along two opposite sides of the farm 

and also along the two diagonals. Five tillers were randomly selected from each hill and 

labelled. Thus a total of 60 hills and 300 tillers were sampled. The heights of the labelled 

tillers were measured with straight-edged ruler every morning for seven consecutive days until 

the infested tillers began to show 'deadheart1 symptom. The average growth of the infested 

tillers was compared with that of the healthy tillers.

44

University of Ghana          http://ugspace.ug.edu.gh



Plate 4: Experiment to study the boring activity of shootfly larvae.

University of Ghana          http://ugspace.ug.edu.gh



3.3 Results

3.3.1 Field investigation of preference for size of citronella tillers by shootborers

The results have been summarised and analysed in Table 5. The mean tiller heights 

of infested tillers by the shootborers were very close; so were the mean diameters. The range 

of heights of tillers infested were also similar. They ranged from short to tall tillers. Thus the 

shootflies attacked tillers ranging from 255 mm to 413 mm while the lepdopterous stemborers 

attacked those from 278 mm to 413 mm. Testing at 5% level of significance there was no 

difference between the means. Therefore the larvae of lepidopterous stemborers attacked 

citronella tillers that were similar in height and diameter to those attacked by the shooflies.

3.3.2 Boring activity of the shootborers

Lepidopterous stemborers: In the laboratory, boring was done by larvae of Sesamia 

sp and E saccharina. Both entered or left the tiller through entry/exit holes at a mean height 

of 21.0 + 14.7mm (Table 5) above the apex of the rhizome at the base of the tiller. On 

entering the tiller, a larva tunnelled vertically upwards and7or downward producing a lot of 

frass and creating a cavity at the apex of the rhizome (Fig. 9). Tunnels measured from field 

sampled tillers had mean length of 75.4 ± 23.6mm (Table 5). Tunnelling destroyed the apical 

region of the rhizome; it also caused the inner whorl to wilt to produce the 'deadheart’ 

symptom.

Shootflies: No boring was done in the laboratory. Therefore, the point of initial attack could 

not be determined. However, from field observations, the larvae tunnelled 32.7 ± 23.6mm; 

(Table 5) or ate the tissue of the inner whorl which became rotten to produce the 'deadheart’ 

symptom. The rotten material normally extended along the inner whorl to the apex of the

46

University of Ghana          http://ugspace.ug.edu.gh



rhizome which became destroyed. The attack also resulted in constrictions within the inner 

whorl (Fig. 10). If the constriction was pushed out of the tiller on top of a healthy growing 

inner whorl, the plant overcame the attack of the larva.

47

Table 5: Summary and analysis of tillers preferred and damaged by the shootborers.

Mean Tiller height 
(mm) 
x ± s.d.

Mean Tunnel 
length (mm) x ± 
s.d.

Mean height o f  
entry/exit hole (mm) 

’ x ± s .d .

Mean Diameter o f  
infested Tillers (mm) 
x ± s.d.

Lepidoptera 355.3±37.7 75.4±23.6 21.0±14.7 10.3±2.5

Diptera 324.5±36.9 32.7±23.6 - 10.2±2.7

'34df (0.05) N.S. *6.55 N.S.

* Significant at p < 0.05

3.3.3 Field identification of damage from differential growth (height) of newly ratooned 

infested and uninfested tillers.

Three hills, i.e. numbers 13, 17 and 30, had both infested and healthy tillers present 

and both types of tillers increased in height with time (Fig. 11; Appendix 7). The curves 

coincided closely showing that the height of the infested and uninfested tillers had similar 

growth. Nevertheless, a slight deflection of the curves oflhe infested tiller numbers 13 and 

17 after the fifth day showed that they had a slight delay in growth.

University of Ghana          http://ugspace.ug.edu.gh



48

Fig. 9: Longitudinal section of tiller attacked by lepidopterous stemborer.

University of Ghana          http://ugspace.ug.edu.gh



49

Fig 10
Longitudinal section „ f  (iller a „ acked by s h o o (%

University of Ghana          http://ugspace.ug.edu.gh



H
ei

gh
t 

of
 

til
le

rs
 

of
 

hi
ll 

30 
(m

m
) 

H
ei

gh
t 

of
 

til
le

rs
 

of
 

hi
ll 

17 
(m

m
) 

H
ei

gh
t 

of
 

til
le

rs
 

of
 

hi
ll 

13 
(m

m
)

56

Day

Fig. 11: Daily growth of healthy and infested tillers of ratooned citronella at 

Osofoiaman in one week in 1991.

University of Ghana          http://ugspace.ug.edu.gh



The results showed that the incidence of shoot flies and lepidoterous stemborers do not 

depend on tiller diameter and/or height. This is in contrast with the suggestion by Patch 

(1929) and Chiang et al (I960), that lepidopterous stemborers prefer taller hosts, as well as the 

observations by Simmonds (1952) and Deeming (1971) that shoot flies prefer smaller, shorter 

tillers for infestation.

Infestation by a mixture of stemborer populations of Busseola and Sesamia and also 

a purely Busseola population depend on the size of the hosts but that by a purely Sesamia 

population shows no preference for size (Walker, 1963; Endrody - Younga, 1968). It is 

therefore in consonance with the above findings that as observed in this study, a stemborer 

population almost entirely of Sesamia (Table 2) shows no preference for size. The results also 

confirm that the stemborer population in citTonella in South-Eastern Ghana consists mainly 

of Sesamia as indicated in Table 2.

However, the results are preliminary and further investigations ought to be carried out, 

especially at different ages of the plant for definite conclusions to be drawn.

Damage of the shootborers .

Correlation analysis showed that for the two shootborers the length of the tunnel did 

not depend on the height of the tiller. However, t-test analysis on field data (Table 5) showed 

that the length of the tunnel by lepidopterous stemborers (75.4±23.6mm) was significantly 

longer than that by the shootflies (32.7±23.6mm) (t = 6.55; P<0.05). Indeed, tunnels by the 

shootflies were less than half the length of those made by the lepidopterous

51

3.4 Discussion

Field investigation of preference for size of citronella tillers by shootborers.

University of Ghana          http://ugspace.ug.edu.gh



stemborers. This may probably be due to the bigger size and more powerful mouth parts of the 

lepidopterous larva. Sesamia sp and E. saccharina entered straight into the stem through the 

sheaths, and this is in conformity with the observation made by Ingram (1958) and Harris 

(1962), who reported that the young Lepidoptera larvae, especially of Sesamia sp, bore straight 

into the stems of cereal crops.

The point of entry of the shootflies could not be determined either in the field or 

laboratory because of high mortality in the larvae but according to Simmonds (1952), in 

wheat, larvae of other Chloropidae e.g. Oscinella frit bore into any part o f the stem or 

between the tissue layers of leaves and descend to the lower part of the shoot.

The entry/exit points of the lepidoptera larvae were lower down the tiller close to the 

apical region of the rhizome. This observation agrees with that by Girling (1972), who 

reported that the larvae of E. saccharina bore into the lower part of the sugar-cane plant. The 

apex of the citronella rhizome is very soft and this might have served as a good source of food 

for the larva. It was also observed that the larvae frequently rested in the cavities they made 

in the rhizome. At that low position in the rhizome, they could easily be missed during 

harvesting, making it possible for them to re-infest the ratoon crop, which therefore suggests 

this might be an adaptive behaviour.

Damage by the larvae of the shootborers caused 'deadheart' symptom which results 

from the mechanical damage to the scattered vascular bundles during the boring activity (Nye 

1960a.). In addition to this, observations in the present study suggest that the deadheart' 

could also have been caused by the eating of the meristematic tissue at the apex of the rhizome 

by the larvae and also from the rotting of that point. The production of 'deadheart1 symptom

52

University of Ghana          http://ugspace.ug.edu.gh



by the Chloropidae has also been abserved by Deeming (1971) on other cultivated cereals and 

wild grasses. Nonetheless, there may be other symptoms from Chloropidae damage. For 

example, Goodlife (1942) found them to produce swollen nodes and intemodes referred to as 

gouted shoots while Simmonds (1952) reported the wilting of shoots of cereals from their 

damage.

Whereas the ftass and entry/exit holes are very characteristic of the lepidopterous 

stemborer attack, attack by larvae of the shootflies could be differentiated by the rotting and/or 

constrictions in the inner whorl. Agents causing the rotting were not determined but the 

destruction of the innermost part of the tiller agrees with Deeming’s (1971) observation on

S. trapezoides and S. micantipennis in sorghum and millet tillers. It also agrees with the 

report by Simmonds (1952) that O. frit destroyed the central part of the wheat stem.

Field identification of damage from differential growth (height) of newly 

ratooned infested and uninfested tillers.

Dissection of infested tillers showed the destruction of the apical meristem of the stem. It 

follows then that growth in infested tillers as expressed in their heights will slow down even 

if it does not cease completely. Therefore, assuming that harvesting affected both infested and 

healthy tillers equally; that tagging did not affect the physiological state of the plant; and that 

all healthy tillers in a citronella hill grow almost at the same rate, the differences in height 

would indicate infested and healthy tillers. The result in this study only showed slight 

differences in growth of infested and uninfested tillers after harvesting, i.e. after six days (Fig. 

11). The differences were too small to be recognised visually without measuring. Therefore 

they can not be used to detect early infestation which is not showing “deadheart”.

53

University of Ghana          http://ugspace.ug.edu.gh



CHAPTER FOUR 

4.0 SEASONAL VARIATION IN INFESTATION OF CITRONELLA  
BY LEPIDOPTEROUS STEMBORERS AND 

CHLOROPID SHOOTFLIES 

4.1 Introduction

Infestation of crops by insect pests tend to fluctuate in response to several factors both 

biotic and abiotic (see section 1.2). However, weather and climate are commonly accepted 

as dominant influences on the behaviour, abundance and distribution of insects. Thus, Reddy 

(1982) found the distribution of shootborer species to be influenced by rainfall, while the 

fluctuations in the incidence of pests with rainfall pattern has been reported by Adeyemi 

(1969), Harris (1962), Smithers (1960) and Easwaramoorthy and Jayaraj (1988). Temperature 

also has an effect on the dynamics of insect numbers (Balraj et al, 1977; Reddy, 1982); so 

also is relative humidity (Girling, 1978; Easwaramoorthy and Jayaraj, 1988). Climatic factors 

may also induce diapause in shootborers (Barry, 1972; Clearwater and Otieno, 1977; Ogwaro, 

1979).

There is lack of information on the effect of climatic factors on shootborer Infestation 

of citronella but such information is required for use in control planning. Therefore these 

studies on the progress of infestations in relation to rainfall, evaporation, temperature and 

relative humidity were carried out with this objective in mind.

4.2 Materials and methods

4.2.1 Determination of sample size and measurement of incidence of deadheart’ caused 
by shootborers

54

University of Ghana          http://ugspace.ug.edu.gh



Sample size: Two citronella farms at Osofoiaman were used in this study. Farm A measured 

88m x 18.5m and Farm B, 58m x 42m. All the citronella hills in the farms were pegged and 

numbered. The tiller was the sampling unit and preliminary studies were carried out to 

determine the sampling size, following the procedure by Hansen et al, (1962).

ie. K2 NV2

n = -----------

ND2 + K2 V2

where n = sampling size,

N = total number of tillers on farm,

V = ratio of failure to success in finding infestation in the tiller,

K = the probability that the sample result will have no error greater than + D, 

and D = allowable error.

Incidence of “deadheart”: The incidence of 'deadheart' by the shootborers was measured 

for both farms A and B. The work started in January 1991 and February 1991 in farms A and 

B, respectively. Farm A was monitored until June when it was abandoned because the farm 

became flooded and most of the citronella hills got destroyed.

From the procedure above, sampling size of between 1100 and 1200 tillers (see 4.3.1) 

were randomly selected from a box of numbers for a farm. Each farm was visited weekly. For 

any citronella hill sampled, the leaves were carefully inspected for eggs, larvae and adults as 

well as external damage symptoms. The tillers were counted and those showing 'deadheart'

55

University of Ghana          http://ugspace.ug.edu.gh



symptoms were recorded, removed with a scapel and dissected in the laboratory to determine 

the cause of the 'deadheart' symptom. Percentage infestations by the respective shootborers 

were then calculated.

4.2.2. Rainfall, Temperature, R.H. and Evaporation

Figures for rainfall, evaporation and temperature including wet and dry bulb readings 

were obtained from the meteorological substation at Pokuase. Mean relative humidity values 

were calculated from wet and dry bulb readings at 0900 Hours and 1500 Hours GMT. A 

stepwise regression analysis was done between infestation and climatic factors after square 

root transformation of percentage infestation.

4.3 Results

4.3.1 Determination of sample size and measurement of incidence of'deadheart' caused 
by shootborers.

Sample size: The total number of tillers and the infestation levels in both farms were

similar from the studies, so results from one farm were used to determine the sample size for

the two, as follows:

Total number of hills on farm =1530

Mean hill size = 34.6 tillers

Total number (N) of tillers on farm = 53000 

Number of tillers sampled = 1478

Number of tillers infested = 135

V = (1478 - 135~>
135 

= 9.95 
K = 1

56

University of Ghana          http://ugspace.ug.edu.gh



D = 0.3 (ie 70% confidence level)

n _ K2 Nv2 
ND2 + K2 V2

By substitution,
n = 53000 x 98.964

53000 x .09 + 98.964

simple size = 1077 tillers 

Incidence of “deadheart”: The results for incidence of'deadheart’ by shootborers in Farm 

B have been illustrated in Fig. 12 (see also appendix 9); those of Farm A in Fig. 13 (see also 

Appendix 8). In Farm B, where sampling spanned the period February to December 1991, 

percentage infestation by lepidopterous stemborers showed two possible peaks: a questionable 

peak around February and a distinct peak in October. That by the shootflies showed similar 

trend in the same months. Combined infestation was similar to that by the shootflies because 

the infestation by the shootflies was generally higher. In Farm A, where sampling spanned 

the period January to June 1991, the infestation levels for the lepidopterous stemborers peaked 

in February and dropped thereafter. That for the shootflies peaked in March and dropped to 

a low in May. However, it was beginning to rise in June when work was abandoned because 

of the excessive flooding of the field. Combined infestation also followed the trend by the 

shootflies. By comparing data from farms A and B, it appears there are two peaks in 

infestation. Analysis of variance following square root transformation of the mean 

percentage infestations showed no significant difference between infestation by the shootflies 

and stemborers (P > 0.05). Nevertheless, damage in Farm A by both groups of insects during

57

University of Ghana          http://ugspace.ug.edu.gh



the same period was higher than in Farm B. For instance, from February to June, the mean 

percentage incidence of'deadheart' by shootflies in Farm A was 0.65% and that of Farm B 

was 0.44%. Similarly, that by stemborers in A was 0.29% compared to 0.19% in Farm B. 

Generally, infestations were low in both farms (Figs. 12 and 13).

4.3.2 Rainfall, Temperature, R.H and Evaporation

The following results have been illustrated in Fig. 14 (see also Appendix 10). Rains 

started late in February at Osofoiaman and the wettest month was May. February and 

December were the driest months. Temperature did not vary much within the months. Mean 

monthly temperatures ranged between 25.5°C in August and 29.1°C in February and March.

The r.h. too did not vary much within the year; the lowest was 80% and the highest 91.5%. 

Mean evaporation had two peaks, 2.3 mm in March and 1.9 mm in December. There were 

weak linear relationships between 'deadheart' by both groups of shootborers and rainfall, and 

between deadheart’ caused by shootflies and other climatic factors of evaporation, 

temperature and relative humidity. A regression analysis of percentage deadheart' by 

lepidopterous stemborers and rainfall gave a relationship of y = -0.0017x + 0.6298 (Fig 15: 

Appendix 11). The relationship of shootflies and the climatic factors were as follows: rainfall, 

y = -0.0024x + 1.3413 (Fig. 16: Appendix 12); evaporation, y = -0.8958x + 2.2639 (Fig. 17: 

Appendix 13); temperature, y = -0.5826x + 17.15 (Fig. 18: Appendix 14); and relative 

humidity, y = 0.1292x -9.8968 (Fig. 19: appendix 15); where y = percentage 'deadheart' and 

x = climatic factor.

58

University of Ghana          http://ugspace.ug.edu.gh



P»
rt

«n
t*

a*
 

*
d

«f
td

h
t«

n

59

fig. 12: Percentage 'deadheart’ caused by shootflies and lepidopterous stemborers, 

Osofoiaman farm B, 1991.

University of Ghana          http://ugspace.ug.edu.gh



ten Fob Vivch Apal U«v JUf«
Month

Percentage 'deadheart9 caused by shootflies and lepidopterous stemborers, 

Osofoiaman farm A, January - June, 1991.

University of Ghana          http://ugspace.ug.edu.gh



61

Fig. 14: Mean montly rainfall ̂ evaporation, temperature, relative humidity and percentage 'deadheart’ 

by shootflies and stemborers, Osofoiaman farm B, 1991.

University of Ghana          http://ugspace.ug.edu.gh



62

Fig. 15 Relationship between percentage 'deadheart’ by stemborers and rainfall.

University of Ghana          http://ugspace.ug.edu.gh



63

Fig 16 Relationship between percentage deadheart’ by shootllies and rainfall.

University of Ghana          http://ugspace.ug.edu.gh



Pa
rc

an
ta

ga
 

~ 
da

ad
ha

ar
t 

~ 
by 

dl
pt

ar
ou

i 
■

h
a

ot
ijo

ra
n

64

Evaporates ( raa]

Fig 17: Relationship between percentage deadheart’ by shootflies and evaporation.

University of Ghana          http://ugspace.ug.edu.gh



iu
s 

sh
Q

ot
fc

ar
an

65

Tmnvmrtmrm ( C ]

Fig 18: Relationship between percentage deadheart’ by shootflies and temperature.

University of Ghana          http://ugspace.ug.edu.gh



Pa
re

an
ta

ga
 

'd
a

a
d

h
a

a
rl

' 
by 

dl
pt

ar
au

i 
m

o
o

tt
o

ra
n

&er

R«Urilv« honddy (X)

Fig. 19: Relationship between percentage 'deadheart’ by shootflies and relative 

humidity.

University of Ghana          http://ugspace.ug.edu.gh



Incidence of deadheart' by shootflies and lepidopterous stemborers

Results of borer incidence in Farm A were not complete and therefore may not give 

much information but the fact that work had to be abandoned as a result of flooding, shows 

the unsuitability of marshy areas for the cultivation of citronella.

In Farm B, the results showed no significant difference (P > 0.05) between the means 

of infestation rates by the shootflies (1.2%) and stemborers (0.6%). Therefore, as far as insect 

attack of the plant is concerned, both groups of insects are equally important, even though 

percentage infestation by the shootflies was observed to be higher.

That infestation by the stemborers has two very distinct peaks is similar to the 

observation made by Endrody - Younga (1968) of infestation by E. saccharina and Sesamia 

botanephaga Tams and Bowden (Lepidoptera: Noctuidae). However, while Endrody -  

Younga’s (1968) peaks occurred in July and October, the peaks in this study occurred in 

February and October. The two peaks imply two distinct generations of the insects in a year 

as suggested by Endrody - Younga (1968). By the same reasoning, the shootflies show three 

generations in a year (Fig. 12), an observation which agrees with the number of generations 

a year of other Chloropidae reported by Cunlife (1925), Sabrosky and Wilbur (1936) and 

Simmonds (1952).

The simultaneous rising and falling of infestation levels makes it possible for both 

shootflies and lepidopterous stemborers to be controlled together, assuming both react equally 

to same control measures. It also suggests that both groups of insects respond similarly to the

67

4.4 Discussion

University of Ghana          http://ugspace.ug.edu.gh



Relation between incidence of'deadheart' and Rainfall, Evaporation, Temperature and 

Relative Humidity.

The linear relationships between the infestation rates and the above-mentioned climatic 

factors are rather weak, implying weak influences of these factors on infestation. 

Nevertheless, rainfall appears to be the dominant factor impacting on both groups of insects. 

Thus, infestations are always low in high rainfall and vice versa (Fig. 14).

The low infestations from April to July indicate that low populations of the larvae 

attack the crop during periods of high precipitation. In the case of the lepidopterous 

stemborers, this agrees with the observation of Jerath (1965) in Nigeria, who reported very low 

larval density of R. saccharina during the wet season. Also low infestation in the period above 

supports the suggestion by Easwaramoorthy and Jayaraj, (1988) that high amounts o f rain 

might be directly responsible for less borer incidence. In this work, high level of surface run­

off occurred in both Farms A and B during high precipitation and since the exit/entry holes 

were so low above the ground surface, the tunnels became susceptible to flooding, thus 

rendering the larvae vulnerable during heavy and continuous rains. This probably contributed 

to the absence of damage by the stemborers during the rains (Fig. 14). The gradual build up 

of infestation by the larvae of the chloropids from August could imply corresponding build 

up of adult population from the end of the rainy period in July (Fig. 14), which would agree 

with the conclusion by Sabrosky and Wilbur (1936) that adult chloropids emerge after periods 

of heavy rains. It is, however, at variance with the findings of Cunlife (1925), and Simmonds

68

factors influen