QL536.Si 1 bite C.1 G368249 University of Ghana http://ugspace.ug.edu.gh INVESTIGATIONS INTO THE ECOLOGICAL DETERMINANTS OF DISTRIBUTION OF ANOPHELES GAMBIAE S.S. (DIPTERA: CULICIDAE) LARVAL POPULATIONS IN URBAN ACCRA, GHANA BY ISAIE SIBOMANA (B. Sc. Agronomy, Engineer. Agronomy) A thesis presented in partial fulfilment of the requirements for the degree of M. Phil. Entomology of the University of Ghana Insect Science Programme* University of Ghana Legon August 2002 * Joint interfaculty international programme fo r the training o f the entomologists in West Africa. Collaborating Departments: Zoology (Faculty o f Science) and Crop Science (Faculty o f Agriculture) University of Ghana http://ugspace.ug.edu.gh DECLARATION I do hereby declare that the experimental work described in this thesis was carried out by me with the exception o f references to other people’s works that have been duly acknowledged. This thesis, either in whole or in part has not been presented elsewhere for any other degree. Isaie Sibomana (Candidate) Prof. Michael D. Wilson (Supervisor) (Supervisor) University of Ghana http://ugspace.ug.edu.gh DEDICATION To my wife, Sylvie INSHUTI; my son, Brice Ulrich ISHIMWE my brother in-law, Aristide BAKUNZI and to the families Francois Xavier Karamage and Madeleine Hakizimana University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT I wish to express my sincere gratitude to my supervisors, Prof. Michael D. Wilson and Dr. Daniel A. Boakye both of the Parasitology unit of Noguchi Memorial Institute for Medical Research (NMIMR) for their expert guidance, treasured advice, patience and support throughout this work. I am very grateful to Professor D. Ofon-Adjei, Director of I the NMIMR for granting me permission to use the facilities of the Institute. Special thanks go to Mr. C A Brown for his great contribution and brilliant ideas towards the laboratory work and data analysis. I greatly appreciate his personal support, guidance and useful teaching. My heartfelt thanks also go to Dr. K.A. Koram and Dr. Bill Roger for their invaluable help with the data analysis. I will like to acknowledge the goodwill and the cheerfulness shown by all the staff of Parasitology Unit. I also appreciate the tremendous help and friendly support received from the workers and student colleagues of room 133 of the Institute I am particularly grateful to Mrs Anita Ghansah, Mrs Bridget Mariam Ogoe, Ms Nancy Duah, Mr. Evans D. Glah, Ms Adwoa Asantewa Poku, Mrs Shirley Coffie, Ms Helena Baidoo, Ms Naiki Populampu, Frederick Anokye-Danso, Fred Aboagye-Antwi and Benedicta Anumu. I want to thank the Transport officer and drivers who spared their time when the moment came for field collection. I am also grateful to the staff of water laboratory at CSIR for their immense help and encouragement in water analyses. I acknowledge the essential role played by Professor Jonathan N. Ayertey, coordinator of the African Regional Postgraduate Programme in Insect Science (ARPPIS-West IV University of Ghana http://ugspace.ug.edu.gh Africa) in making this course a success. Thanks to all my classmates. I appreciate the companionship and useful discussions with my classmate Raphael Abanja Ndondo throughout our training. My utmost gratitude goes to Dr. Dona Dakouo, Rice Research Programme Leader, EMERA, Research Station of Farako-ba, (Bobo-Dioulasso, Burkina Faso) for his advice and maximum co-operation, which contributed greatly to the final result o f this work. Finally, my warmly thanks are expressed to my lovely wife, son and brother in-law for their prayers while I was away from my house working hard on this thesis. I can never thank you enough for your love and attention throughout the time I was doing the project, leaving the house very early in the morning and coming back very late in the night seven days a week. May God richly bless you My studies at University of Ghana, Legon were funded by the Deutscher Akademischer Austauschdienst e.v. (DAAD) in Germany University of Ghana http://ugspace.ug.edu.gh DEDICATION................................................................................................................................. 111 ACKNOWLEDGEMENT..............................................................................................................IV I LIST OF ILLUSTRATIONS................................ ........................................................................ 1X LIST OF TABLES............................................................................................................................x LIST OF PLATES.......................................................................................................................... Xli LIST OF APPENDICES.............................................................................................................. xiii LIST OF ABREVIATIONS........................................................................................................xiv ABSTRACT................................................................................................................................ xviii CHAPTER ONE.................................................................................. l GENERAL INTRODUCTION................................................................................................... 1 1.1 Introduction........................................................................................................................1 TABLE OF CONTENTS D E C L A R A T IO N ....................................................................................................................................................... 11 1.2 Objectives...................................................,........................................................................ 8 CHAPTER TWO .................................................................................................................9 LITERATURE REVIEW............................................................................................................ 9 2. 1 Malaria: The Disease and Symptoms................. 9 2.2 Global Distribution of Malaria.......................................................................................11 2.3 Socio-economic Impact of Malaria............................................................................... 13 2.4 The Life Cycle and Transmission of Human Plasmodium Parasites..................... 15 2.5 The Life Cycle o f Anopheles Malaria Vectors............................................................] 8 2.6 Environmental Factors Influencing the Abundance of Malaria Vectors.............. 21 2.7 Control of Malaria.................................................................................................. 25 2.7.1 Historical background to malaria control programmes.................................25 2.7.2 Chemotherapy....................................................... '............... 28 2.7.3 Vector control............................................................... 34 2.8 Species Identification and Relevance to Vector Control Programmes................. 40 2.8.1 Anopheles gatnbiae complex and its distribution........................................ 41 VI University of Ghana http://ugspace.ug.edu.gh implications...................................................................................................................... 2.8.3 Characterization of An. gambiae s.s. populations by microsatellite DNA analysis............................................................... .............................................................. CHAPTER THREE........................ 48 MATERIALS AND M ETH O D S ............................................................................................. 48 3.1 The Study Sites..................................................... 48 3.1.1 Description of larval and pupal habitats ........................................................ 49 3.2 Methods.............................................................. ^0I 3 .2.1 Field sample collections of pre-adult mosquitoes and w a te r ..................... 50 3.2.2 Laboratory rearing of mosquitoes...................................................................... 50 3 .2.3 Morphological identification of adult Anopheles mosquitoes ....................51 3 2 4 Molecular biology studies.................................................................................. 52 3.2 4.1 The isolation of genomic DNA o f Anopheles gambiae s.l.................52 3.2.4 2 PCR identification of species of Anopheles gambiae complex 52 3.2 4 3 Microsatellite DNA analysis....................................................................55 3 .2.4 4 Analysis of PCR products................................ 57 3 2 4 4 .1 Agarose gel electrophoresis...............................................................57 3.2 4.4.2 Polyacrylamide gel electropholesis...............................................57 i 3.2.5 Physico-chemical Analyses of Water Samples ............ 58 3.2.5.1 Measurement o f physico-chemical parameters of water samples... 58 3.2.6 Data analysis.................................................................................. 69 3.2.6.1 Test of equality o f m eans ..........................................................................69 3 2.6.2 Discriminant function analysis.................................................................70 3 2 6 3 Multiple regression analysis..................................................................... 72 3.2 .6 .4 Hierarchical cluster analysis and phylogeny tree.................................74 3.2.6.5 Microsatellite data...................................... ’............................................... 75 CHAPTER FOUR............................................................................. 77 RESULTS .............................................................................................. 77 4 1 Molecular Identification of Anopheles gambiae Species Complex.......................77 2.8 2 Cryptic taxa within An. gambiae s.s. and their epidemiological 42 vii University of Ghana http://ugspace.ug.edu.gh 4.2 Genetic Structure of Anopheles gambiae s.s. Populations.......................................78 4 .2 .1 Genotype frequency distributions...................................................................... 78 4.2.2 Population differentiation index (Fst)................................................................79 4.2.3 Estimate of heterozygote deficiency and excess (Fis)................................... 79 4.3 The Physico-chemical Parameters of Breeding Habita ts .........................................82 4.3,1 Parameters predicting the presence/absence of An. gambiae s.s............... 84 4 3 2 Parameters predicting variation in distribution of An. gambiae s.s. populations........................................................................................................................ 84 4 4 Association between larval habitat types and An. gambiae s.s. populations 90 CHAPTER FIVE ...............................................................................................................93 DISCUSSION AND CONCLUSION......................................................................................93 REFERENCES........................................................................................................................... 100 I APPENDICES ........................................................................ 121 University of Ghana http://ugspace.ug.edu.gh LIST OF ILLUSTRATIONS Figure 1: Figure 2: Figure 3: Figure 4: Figure 5: I Figure 6a-tl: Figure 7a-b: Map showing the global distribution of malaria The life cycle of Plasmodium species Schematic illustration of the life cycle o f Anopheles malaria vectors Ethidium bromide stained 2% agarose gel electrophoregram of PCR products obtained from the amplification o f An. gambiae DNA for species identification Polyacrylamide gel electrophoregram of PCR products obtained from the amplification of An. gambiae s.s. microsatellite DNA with primer set AGXH7 Example of plots showing relationships between the significantly environmental parameters and proportions o f Anopheles gambiae s.s. in mosquito habitats. Molecular phylogenetic tree of An. gambiae s.s. populations only and dendrogram obtained by hierarchical clustering of their physico-chemical parameters in their habitats University of Ghana http://ugspace.ug.edu.gh Tabic I: Table 2: Table 3: Tab)e 4: Table 5: Table 6a: Table 6b:I Tabic 7: DNA sequence details of the synthetic oligonucleotide primers used for the PCR-based method for the identification of/in . gambiae s.I. species and their melting temperatures (Scott el al., 1993) DNA sequences of the oligonucleotide primer set AGXH7 used for the PCR-based amplification of microsatellite sequences in An. gambiae s.s. Frequencies of the most common alleles with the 100 % o f An.gambiae s.s. population group, 1 to 99 % o f An. gambiae s.s. population group Estimates of differentiation (Fst) and heterozygosity (Fis) within populations of An. gambiae s.s. Comparison of water parameters of habitats with An. gambiae s.s larval populations and those with other species only Summary of results of stepwise discriminant function analysis performed on dataset of the significant parameters to select the most discriminatory parameters separating habitats of An. gambiae s.s. and those of other mosquito species only Derived classification function coefficients by stepwise discriminant analysis for the separation of presence/absence of An. gambiae s.s. habitats Summary of results of multiple regression analysis (p value to enter = 0 05) using proportion of An. gambiae s.s. as dependent variable and the significantly correlated parameters as independents variables LIST OF TABLES University of Ghana http://ugspace.ug.edu.gh Table 8: Table 9a: Table 9b: Table 9c: Table 10: Summary of results of multiple regression analysis (p value to enter = 0.05) using proportion of An. gambiae s.s. as dependent variable and all the measured parameters as the independents variables Summary of results o f multiple regression analysis (p value to enter = 0,05) using turbidity as dependent variable and all the measured parameters as the independents variables Summary of results of multiple regression analysis (p value to enter = 0.05) using pH as dependent variable and all the measured parameters as the independents variables Summary of results of multiple regression analysis (p value to enter = 0.05) using calcium as dependent variable and all the measured parameters as the independents variables Environmental parameters for the group o f 100 % o f An. gambiae s.s. populations significantly correlated to the proportion of An. gambiae s.s. University of Ghana http://ugspace.ug.edu.gh Plates la-c: Hxamples o f An. gambiae s.s. larval population breeding sites Plates 2a-b: Different kinds of water associated with habitats of An. gambiae s.s. larval populations Plate 3 : Polystyrene trays used in the laboratory rearing o f Anopheles larvae Plate 4 : Wooden cages for holding emerging adult mosquitoes LIST OF PLATES University of Ghana http://ugspace.ug.edu.gh LIST OF APPENDICES Appendix I: Appendix II: Appendix III: Appendix IV: I Appendix V: Sampling sites, dates of sampling and number o f mosquitoes collected . Standard solutions used in molecular biology study Standard buffers and solutions vised in water analysis An example input data format for the population genetics analysis software POPGENE Version 1.31 Physico-chemical water parameters measured University of Ghana http://ugspace.ug.edu.gh LIST OF ABREVIATIONS AgNO. silver nitrate APS ammonium persulphate bp 1 base pair BOD biochemical oxygen demand C6H 5CH3 toluene CaCOj . calcium carbonate CHCb chloroform CO' carbonate COD chemical oxygen demand conc H2SO4 concentrated sulphuric acid C11SO4.5H2O 1 copper sulphate dATP deoxyradenosine triphosphate ddw distilled de-ionised water dCTP deoxycytidine triphosphate dGTP deoxyguanosine triphosphate DNA deoxyribonucleic acid DO dissolved oxygen DTTP deoxythymidine triphosphate EDTA disodium ethylene diamine tetraacetate. 2H ;0 EtBr ethidium bromide EtOH ethanol FAS ferrous ammonium sulphate University of Ghana http://ugspace.ug.edu.gh Fe(NH.i)2(S0 .i)2.6H2 0 : ferrous ammonium sulphate FeSOj.7HO.iO: ferrous sulphate septahydrate I GPS global positioning system H2C20.,.H20 oxalic acid h 2o water H1SO .1 sulphuric acid HCI hydrochloric acid HCOi carbonate HgCI . mercuric chloride HNO', nitric acid KjCrO.i potassium chromate K.2C r207 potassium dichromate KAc potassium acetate Kb kilobase K1 potassium iodide KOH potassium hydroxide KnaCiHjQ(i.4 l l2 0 : potassium sodium tetrate tetrahydrate L.DF linear discriminant function M molar . MnSO-i manganese sulphate MnS0.|.4H20 manganous sulphate tetrahydrate Mw molecular weight Na.-COi sodium carbonate Na2SO< 5HiO sodium thiosulphate w University of Ghana http://ugspace.ug.edu.gh NaCI sodium chloride Nal sodium iodide NaN? sodium azide SlaOH sodium hydroxide NaOH-Nal sodium hydroxide-sodium iodide NHj ammonium NHjOH ammonium hydroxide NO .1 nitrite NO^ nitrate PCR polymerase chain reaction PH hydrogen-ion exponent PO, phosphate rDNA ribosomal DNA RNA ribonucleic acid Rnase ribonuclease rpm 1 revolution per miriule sddH.O sterile double distilled water SDS sodium dodecyl sulphate s.e. standard error S i0 2 silica s.l. sensu lato s.s. sensu stricto 1 SO, sulphate TAB Tris-Ace tate EOT A University of Ghana http://ugspace.ug.edu.gh TBE Tris-Borate EDTA TDS total dissolved solids TE Tris-EDTA TEMED N,N,N, ‘N'-letramelhyl ethylene diamine Tm melting temperature Tris 2-amino-2-(hydroxymethyl)-l ,3 propanediol P-l microlitre HM micromolar UPGMA Unweighted Pairgroup Method with Arithmetic Averaging ZrOCb 81-hO zirconylchlorideoclahydrate University of Ghana http://ugspace.ug.edu.gh ABSTRACT Anopheles gambiae s.s. larval habitats are important determinants o f adult distribution and abundance, which also determine the geographical pattern of malaria disease. Information on their habitat characteristics can contribute greatly to a better planning of its control strategies through environmental management. To identify the environmental parameters, which influence the distribution of An. gambiae s.s. larval populations in urban Accra, Ghana, 30 habitats with An. gambiae s.I. and 24 without but with other mosquito species were studied Wosquito larvae and pupae were collected and reared to adults Water samples were taken al the same time of mosquito collection for physico- 1 chemical analyses. The adult mosquitoes obtained were morphologically separated into An. gambiae s.l. and other species and counted. Then PCR-based methods were used on 300 mosquitoes to identify the members of An. gambiae s.l. and for microsatellite DNA analyses of An. gambiae s.s. AGXH7 locus using published oligonucleotide rnicrosateilite primers. The microsatellite DNA data generated was used to determine the population structure and construct the phylogenetic relationships between An. gambiae s.s. populations. Twenty-eight physico-chemical parameters o f each water sample were measured and those that on comparison were found to be significant were analysed further using linear discriminant function analysis and multiple regression analysis to reveal the best predictors of present o or absence, and abundance o f An. gambiae s.s. in habitats. Then hierarchical cluster analysis using these water parameters was performed to reveal similar habitats, which was then compared with molecular phylogeny obtained. University of Ghana http://ugspace.ug.edu.gh All the 300 An. gambiae s i. were Klenlified as An. gambiae s.s. Turbidity, pH and calcium wore found to be the most significant discriminatory parameters associated with the presence or absence of (he A i;i. gambiae s.s. in the habitats and were also selected as the best predictors of larval abundance. It was also observed that the molecular phylbgeny of An. gambiae s s populations that bred alone in habitats fitted exactly with the clustering obtained using their water parameters. Also, closer populations were more likely to have relatively similar values of turbidity, pH and calcium concentrations and were also more likely to have similar allelic and genotypic frequencies. University of Ghana http://ugspace.ug.edu.gh c h a p t e r o n e GENERAL INTRODUCTION LI In troduction Malaria is the most important disease transmitted by mosquitoes. However, despite tremendous progress made in the acquisition of knowledge of the biology o f the malaria parasites, the anopheline mosquitoes and the human host, and the development o f anti- 1 malaria! drugs and insecticides, the disease has proved far harder to control. At every turn when it was believed that the disease could be eradicated, either the mosquitoes and/or thg parasite have eased their way out of extinction Both, the malaria parasites and the mosquito vectors have become resistant to anti-malarial drugs and to insecticides respectively. For many reasons, we are still far from solving the malaria problem, which is actually getting worse in some pails of (he world and even returning to areas from which it has once been eradicated (Dobson, 1999). An estimated two billion people (more than 40 % of the world population) presently live in areas with malaria risk and I the global annual incidence ranges between three to five hundred million clinical cases, with a annual death toll of between two to three million (WHO, 1998). Malaria also accounts for 10 % of Africa’s disease burden, causing the greatest suffering and impoverishment among poor people, with pregnant women and children under five years of age, being the most vulnerable (Okenu, 1 999 ). University of Ghana http://ugspace.ug.edu.gh In Ghana., malaria is the most important parasitic disease, accounting for 7.8 % o f all t certified deaths. About 40 42 % of all outpatient attendance in Ghana hospitals is attributed to malaria (Ahmed, 1989) The disease also is associated with considerable economic burden, including direct cost to governments and patients for hospital admissions and outpatient consultations, cost to households for treatment sought outside the official system, and cost due to absenteeism from productive work or education (WHO, 1997). The estimated direct and indirect cost of the disease in Africa alone is estimated to be $2000 million per year (WHO, 2 0 0 0 ). ( Human malaria is normally transmitted frorn one person to another by the bite of female mosquito species infected with malaria parasites Of the thousands of described mosquito species, only a fraction of those in the genus Anopheles are known vectors. Thpse Anopheline species that do hot Iransmit malaria parasites either do not feed on humans or are not susceptible to human malaria parasites, and a number have life spans that are too short to allow the parasite 1o fully mature The vector species that pose the greatest threat are usually abundant, long-lived, commonly feed on humans and typically dwell in close proximity to humans, 'I here are some 400 species of Anopheles mosquitoes, but only about 70 species are kuovVn to Iransmit malaria About 30 are of major importance, accounting for a significant amount of all malaria cases (Teklehaimanot & Pushpa, 1991). Their role in malaria transmission depends largelv on the presence of a favorable environment for Iciival development and adult survival, and the ability to feed on humans. Transmission University of Ghana http://ugspace.ug.edu.gh also depends significantly on human habits that promote the host-vector contact (National Academy Press, 1991), characteristics of vectors, such as their abundance, susceptibi 1 ily to infection, longevity, degree of contact with humans, and the type of species involved in the transmission (Appawu et al., 2001). Finally, it depends also on I behavioral attributes of vectors, such as finding and biting o f hosts and choice o f resting and oviposition sites, that vary both within and between species (Coluzzi et al., 1979), the daily mosquito survival rale, the lime between mosquito infection and sporozoite production in the salivary glands, the vectorial competence (even if an uninfected Anopheles feeds on an infectious host, either the mosquito may not acquire the viable infections, or Plasmodium parasite may fail to replicate within the vector. Furthermore, the mosquito may not transmit the infections onwards at a subsequent mealX and some factor expressing the human recovery rate from infection (Gullan & Cranston, 1 994). I In Africa, the most important vectors are members of the An. gambiae Giles complex and An. Junes fits complex (Teklehninmnot & Pushpa, 199]). The An. gambiae complex comprises six named species. An. gambiae s.s., An. arabiensis, An. melas. An. merus, An. quadriannulalus and An. bwaiubae (Gillies and Coetzee, 1987), one unnamed species (Hunt el a /.. 1998) and seveial incipient species (Coluzzi et al., 1979). The species aie morphologically indistinguishable yet genetically and behaviorally distinct. I he distinct behavioral characteristics determine their distribution and efficacy as I vectors (I1 avia et a l . 1997) Anopheles gambiae sensu stricto and An. arabiensis are the two most effective vectors o f Plasmodium falciparum within the An. gambiae complex (WHO, 2000) Ihese vectors have proven effective in transmitting the parasite to humans across the region, in rural and urban areas alike However, all the transmission University of Ghana http://ugspace.ug.edu.gh characteristics of vectors are influenced by environmental conditions, such as climate, rainfall and vegetation. Anopheles gambiae s.s. is extremely versatile, regarding tolerance to a wide variety o f micro and macro environmental conditions, as evidenced by its wide geographical distribution (Lanzaro el a/., 1998). In this species, ecological and behavioral plasticity I has been observed to be associated with polymorphisms in the form o f paracentric chromosomal inversions, microsatellite DNA and isozyme variability within localized populations (Lanzaro el al., I 995) The spatial distribution o f certain gene arrangements also shows strong association wilh specific regional habitats and their frequencies change seasonally in places with seasonal fluctuations in weather, especially rainfall (Toure el al., 1998; Coluzzi el al., 1 985, Biyan el al., 1982; Coluzzi el al., 1 979). Clinal geographic and microspatial variation in gene frequencies and arrangements in An. gambiae s.s. populations also have been correlated with behavioral differences I (Besansky el al., 1994). There is also suggestive evidence that West African populations oi'/ln. gambiae s.s. are highly structured and that the spatial and temporal distributions of chromosomal inversion polymorphisms in West African populations of An. gambiae are non-random (Coluzzi el a/., 1979). The strong association of certain karyotypes with specific habitats has led to the description of distinct chromosomal “forms’1 o f An. gambiae s.s. called ecophenotvpes in West Africa (Lanzaro et al., 1995). Three ecophenotypes, Bamako , “Mopti” and "Savanna” , occur in sympatry at numerous sites in Mali, Burkina I'aso, Ghana (Toure. 1991; Lanzaro el al., 1995), Gambia and Senegal t (Bryan el al., 1982). University of Ghana http://ugspace.ug.edu.gh Vectorial capacity can vary tremendously even between members of the same species complex (Coluzzi, 1984) Furthermore, Toure el at. (1 983) reported that Bamako, Mopti and Savanna forms differ in their vectorial capacity, y/hich means that relevant biological differences are detected not only between species but also between incipient taxa within the same species Small-scale spatial variation and temporal heterogeneity in mosquito densities can have important consequences for disease transmission (Smith et al., 1995). The strategy of malaria control is based on breaking the chain of transmission o f the parasites between humans and mosquitoes either by controlling the parasites with drugs or breaking the contact between people and vectors. Several strategies for malaria management have been evaluated and these include habitat management, chemotherapy, vaccines and vector control using spraying, mosquito nets or future use of transgenic mosquitoes (Fontenille & Lochouarn, 1 999; WHO, 2000). However, none o fthe control strategies has been reported to achieve complete control. Moreover, human behavior including small changes in land use, wars and population movements have complicated, disrupted'and frustrated many attempts at control (Dobson, 1999). Malaria control programmes are very expensive and countries where malaria is endemic are often very poor and lack the finances and infrastructure to maintain effective campaigns (Dobson, 1999) To complicate issues further, resistance to anti-malarial drugs has appeared in recent years and it is increasingly becoming widespread, anti-malarial vaccines are eagerly awaited but the problems of developing a cheap and effective vaccine are immense and it is not likely to be available for some years to come (Fontenille & Lochouarn, I 999). University of Ghana http://ugspace.ug.edu.gh Management through drainage, filling, leveling intermittent flushing of vector breeding luibiials are cost-effective options (Rafatjah, 1988) for vector control Because of insecticide assistance and the environ mental impact of spraying, considerable resources have been devoted to search lor biological control agents. Vector control by larviciding has been implemented in some circumstances, especially when the use o f residual adulticides was not effective or was too expensive. For most malaria vectors, reducing mosquito vector population densities by means of larviciding is generally an inefficient way1 of alfeding transmission, because larval mortality among many anopheline populations may be density dependent, and population reduction at the larval stage does not affect the mean longevity of the surviving adult population. However, when a large proportion of larval habitat can be easily identified and target, larval control can be very effective (Collins & Paskewitz, 1995). Whatever me1hod(s) is to be adopted, it is critical that there should be a detailed understanding of the basic ecology of the vector, including a sound knowledge of the genetic diversities that may exist within vector species populations. Habits o f the local anopheline mosquitoes are what determine the geographical pattern o f the disease and information on the micro-ecology of vector breeding sites can contribute greatly to a better understanding of the relationship between the vector, man and the environment in disease transmission. I )espile the importance o[An. gambiae s.s. as a vector, relatively little is known about its liiival ecological preferences. There are Iwo likely reasons given for the dearth of larval University of Ghana http://ugspace.ug.edu.gh studies on llie malm in vectors. I lie main reason being that malaria control in Africa, Irudilionally, has been directed al (he adult stages, therefore studies of larval ecology have beeh thought to be irrelevant by some workers (Gemnig el al., 2001). This narrow view clearly is obsolete since an understanding of population dynamics - which includes an understanding of flucliuilions in adult populations - requires a thorough appreciation of laclors affecting larval abundance. Larval habitats are important determinants of adult (I isl ribution and abundance Although the transient habitats of An. gambiae may not be a reasonable target for a vector control, an understanding o f dynamics and productivity of larval habitats is required if efforts to model and predict adult abundance are to succeed. I hi' second reason is that until recently such studies would have been impossible to carry out because the tools for characterizing different populations of An. gambiae s.s. did not exist Ilowevei, microsalellite analyses, which are PCR-based (Lehman et al., 1006; Lanzaro ct 37.5°C), shivering, mild chills, worsening headache, malaise and loss of appetite. I'lie periodicity of fever also depends on the type o f parasite species that the patient harbours. If infection is left untreated, the fever in P. vivax and P. ovale infections regularises to a 2-day cycle and for P. malariae the fever occurs every 3 days. I"oi l \ falciparum however, the fever remains erratic and may not regularize to tertian i pattern (White, 1996) Plasmodium malariae and P ovale infections cause little morbidity and almost no mortality, P. vivax infections are more severe and debilitating Inil are usually self-limiting in healthy individuals. P. falciparum infections are always life-threatening in non-immune individuals (Collins & Paskewitz, 1995). WHO (1990) has defined severe malaria, which is the acute form o i falciparum malaria to include, severe anaemia, renal failure, pulmonary oedema, hypoglycaemia, circulatory collapse, bleeding, convulsions, haemoglobmuria, coma, hyperparasitaemia, jaundice I arid hyperpyrexia. In highly endemic areas for P. falciparum , where entomological inoculation rates can exceed several hundred infective bites per year, severe anaemia in inliints is the principal manifestation and major contributor to mortality (Beier et al., 1990) Cerebral malaria is the most prominent feature of severe malaria and is defined strictly as unarguable coma. This is caused by the sequestration o f infected erythrocytes m the microvasculature of the brain. University of Ghana http://ugspace.ug.edu.gh Malaria is endemic in a total o f 101 countries and territories. There are 45 countries I ^ in WHO’s African Region, 21 in WHO’s American Region, 4 in W HO ’s European Region, 14 in Eastern Mediterranean Region, 8 in W HO’s South-East Asia Region, and 9 in WHO’s Western Pacific Region (WHO, 1998) (Figure 1). Although this number is considerably less than it was in the mid-1950s (140 countries or territories) more than : \400 million o f the word’s population are still at risk (WHO, 2000). The inc idence o f malaria worldwide is estimated to be 300 - 500 million clinical cases each year, with about 90 % o f these occurring in Africa south o f the Sahara (WHO, 2000) Outside Africa two thirds of the reported cases are concentrated in only 6 l countries, namely (in decreasing order) India, Brazil, Sri Lanka, Afghanistan, Vietnam and Colombia (WHO, 1995). Malaria is thought to kill between 2 and 3 million people worldwide each year, of whom about 1 million are children under the age o f 5 years in Sub-Saharan Africa These childhood deaths, resulting mainly from cerebral malaria and anaemia, constitute nearly 25% of child mortality in Africa. Fatality rates o f 10 - 30% have been reported among children referred to the hospital with severe malaria, although these rates are even higher in rural and remote areas where patients have restricted access to adequate treatment (WHO, 2000). In addition to children under age 5, the others at greatest risk ol dying from the disease include pregnant women, people moving from malarious zones for reasons o f work, migration, refuge, war or tourism, and travellers who visit endemic countries and return home with the disease (WHO, 19%). 2.2 Global Distribution of Malaria University of Ghana http://ugspace.ug.edu.gh G >Uc u? o E rj ™ X E, 3 - u 2 o •* c u o I a i s c * n ° n — ■>{ i- W _n <9 O>oc co Ci *- ■-< o < < 2<5—c o £ I I Q A k f * I I - f o i : 9- £ j^ «• i Fi gu re 1: G lo ba l di str ib ut io n of m al ar ia (A fte r W H O , 19 97 ) University of Ghana http://ugspace.ug.edu.gh The distribution o f the disease varies from one country to another and even within countries because o f Ihe flight range o f the vector, which is thought to be about 2 miles, irrespective o f the prevailing wind (WHO, 1998). Plasmodium vivax parasite can remain endemic in some Central American populations at very low levels that are difficult to measure in active prevalence surveys, while in some parts o f Africa, the prevalence o f P. falciparum approaches 100% in young children and malaria- specific infant mortality may exceed 20% (Spielman etal., 1993). University of Ghana http://ugspace.ug.edu.gh I ho enormous loll ol lives and ol clays of l&bour lost, and the costs of treatment, make malaria a major social and economic burden in developing countries (WHO, 1905 ). Among nil infectious diseases, malaria continues to be one o f the biggest contributors to disease Inn dens in terms o f deaths and suffering. By undermining the health and capacity to work of hundreds o f millions o f people, it is closely linked to poverty and contributes significantly to stunting social and economic development (WHO, 1996). Malaria and poverty are intimately connected. Cross-country recess ions lor I hr 190S 1090 period confirm the relationship between malaria and economic growth (Gallup & Sachs. 2001). Taking into account initial poverty, econom ic . policy, tropical location and life expectancy, among other factors, countries with intensive malaria transmission grew 1.3% less per person per year, and 10% reduction in malaria is associated with 0.3% higher growth (Gallup & Sachs, 200 I ) The disease cause 10.6% of lost disability adjusted life years (DALYs), second only to 11IV/AIDS I'hc cost o f malaria in economic terms is also high. The cost o f a case from society’s view point is $ US 9 84 or 12 days equivalent of productivity (WHO, l (->)9). Treatment ranges in cost between $ US 0.80 and $ US 5.30 depending on lo< al drug resistance (NIAID, 2000) and the total cost in Africa alone is estimated to have reached US 2,000 million per year (WHO, 2000). In many parts ol the world, malaria is becoming an even greater problem than before, hpidemics aie recuiring in areas where transmission had been interrupted and are generally associated with deteriorating social and economic conditions (WHO. 2.S Socio-economic Impact of Malaria University of Ghana http://ugspace.ug.edu.gh 1996). Even when non-fatal, malaria produces considerable impact on the health of die African children mostly during their first five years o f life, increasing susceptibility to other infections and hampering their development It is also a very seiious problem, when associated with pregnancy, contributing to maternal and neonatal doatli and low birth weight (WHO, 1992). The rural communities are the most affected because the rainy season is often a time o f intense agricultural activity a time for poor families to earn most o f their annual income. The disease can make lluse families even poorer due to the loss o f labour (WHO, 1998). Workers suffering a bout can be incapacitated for 5 - 2 0 days. It is estimated that malaria stricken family spends an average o f one-quarter o f its income on malaria treatment, and prevention and such a family can only harvest 40% o f crops harvested by a healthy liitnily (WHO, 1990) Malaria also leads to the chronic school absenteeism in children and there can be impairment of learning ability (WHO, 1998). University of Ghana http://ugspace.ug.edu.gh 2A The Life Cycle and Transmission of Human Plasmodium Parasites Human malaria is normally transmitted from one person to another by the bite o f a female Anopheles mosquito infected with malaria parasites. The life cycle o f the parasite involve a vertebrate host and an insect vector (Figure 2). There are three pluses of development in the life cycle o f the parasite namely, exo-eiythrocytic stages in the liver, erythrocytic schizogony in erythrocytes and sexual processes in tin mosquilo (Smith, 1996). The malarial cycle commences with an infected female ■iiiopheles mosquilo taking a blood meal from a human host. As it feeds, it injects saliva contaminated wilh sporozoite stage of Plasmodium directly into the blood stream Some sporozoites are killed but the survivors after 30 minutes o f their entry into the blood stieam migrate to the liver (Smith, 1996). All sporozoites leave the peripheral blood circulation within 45 minutes of injection into the host. Once in the liver, pre-(or exo-) erythrocytic schizogonous cycle takes place in the parenchyma tells This leads to the formation o f a large schizont, which depending on the Plasnmdiiim species may contain from 2,000 to 40,000 merozoites (Gullan & ( mnslon, I‘)'M). The prepatent period of infection that started with an infective bite ends when the merozoites are released and either infect further liver cells o ren te r th e bloodstream to invade erythrocytes. Invasion occurs when the erythrocytes mvaginate and engulf the merozoite, which subsequently feeds as a trophozoite within a vacuole. The first and several subsequent erythrocytes schizogonous cycles produce a liophozoite that becomes a schizont, which releases from 6 to 16 m< rozoiles The duration o f the erythrocyte schizogonous cycle is the duration o f the interval between attacks 48 hours for certain malaria and 12 hours for quartan University of Ghana http://ugspace.ug.edu.gh malaria (Gulltm A Cranston, 1994). Within the erythrocytes, merozoites fuel their activities by consuming haemoglobin and develop into ring-shaped trophozoites. There is another round o f asexual production within the erythrocytes, which takes 48 hours and this time when the merozoites are released some invade other erythrocytes whilst others change into sexual forms (micro and macrogametocytes) within 1 0 - 12 days (Mons, 1985). The synchronous release o f merozoites from the erythrocytes liberates paiasite products that stimulate the host’s cells to release cytokines (a class of immunological mediators), which provoke the fever and other symptoms associated with malaria attack. Normally, a variable number o f asexual cycles occur before any gamelocytes are produced (Carter & Gwadz, 1980). The gametocytes have no further activity in the human host. These circulate in the blood stream until they are picked up by another mosquito as it takes blood from human In the gut o f the mosquito, there is exflagellation of the microgametocytes and subsequent fertilization of the macrogametocytes. The zygote, which is called ookinete, penetrates the wall of the midgut and develops into an oocyst. Sporogony within the oocyst produces many threadlike sporozoites, which are released as the oocyst ruptures. The spoiozoites develop and become up to 1,000 times more infective than when in the oocyst (Smith, 1996). They then migrate to the salivary glands where they reside until injection into another human host. I With the exception o f a few cases o f transplacental and blood transfusion associated transmission, malaria parasites are transmitted exclusively by mosquitoes o f the genus Anopheles. Nearly 20% of the almost 400 described species o f anopheline mosquitoes have been implicated as vectors, but most o f these are probably o f minor or incidental importance. With the exception o f Southeast Asia, which may have ten University of Ghana http://ugspace.ug.edu.gh or more important malaria vectors, most malarious regions o f the world have only Lhiee or four major vectors (Collins & Paskewitz, 1995). In sub-Saharan Africa, where 90% o f the world’s malaria-infected people are found, most transmission is caused by three anopheline species, An. gambiae s.s., An. arabiensis and An. fimes/us, with An. gambiae s.s. being the most important vector (Collins & Paskewitz, 1995, Cirimnig eta!., 2001). University of Ghana http://ugspace.ug.edu.gh Late trophozoite Penetration ot red blood cells / Merozoites Erflagellating m icrogametoeyte Penetration o f parenchymal cells Ookinete Sporozoites injected by mosquito Oocysl containing Oocyst on stomach wall sporozoites LIFE CYCLE of PLASMODIUM Spp. A d a p ted a n d redrawn from NCDC Figure 2: Life cycle o f Plasmodium species University of Ghana http://ugspace.ug.edu.gh 2.5 The Life Cycle of Anopheles Malaria Vectors Anopheles mosquitoes have four distinct stages in their life history - the egg, the larva, the pupa and the adult (Figure 3). The first three stages occur in water while the adult is an active Hying insect. They have a large range o f breeding sites bu t the most common are the shallow open sun lit pools (Service, 1993). After mating and a blood meal, gravid Anopheles lays some 50 - 200 small (1 mm long) brown or blackish boat-shaped eggs on the water surface (Service, 1980). Each blood meal provides enough nutrition for a female mosquito to lay a batch o f eggs (White, 1998). In most Anopheles there is a pair o f conspicuous lateral air-filled chambers called the floats on the eggs, which in few species completely extend round the egg. These floats help maintain eggs floating on the water surface. Anopheles eggs cannot withstand dessication and in tropical countries they hatch after 2 to 3 days, but in colder temperate climates hatching may not occur until about 2 to 3 weeks, the duration depending on temperature (Service, 1980). All mosquito larvae are aquatic and metapneustic and pass through four larval instars (Service, 1993). The larvae are filter feeders and unless disturbed remain at the water surface feeding on bacteria, yeasts, protozoa and other microorganisms and also breathing in air through their spiracles (Service, 1980). Rates of larval growth are influenced by environmental factors such as temperature, photoperiodicity, food supply, degrees of overcrowding and the species (Service, 1993). Pupation occurs at the water surface and the whole process takes about 3 to 5 minutes. Pupae are non-feeding and normally spend most ol their time at the water surface breathing through the paired trumpets (While, 1998). Pupal duration is mainly determined by temperature. In tropical.countries it is usually 2 to 3 days, but can be as short as 26 hours at 30°C University of Ghana http://ugspace.ug.edu.gh (Service, 1993). The duration also varies according to species. Male larvae generally pupate before females and in most species male pupal life is shorter than female pupal life When ibis process is complete, the fully formed adult emerges from the pupal case. Emergence usually takes 12 - 15 minutes and within minutes afterwards the newly emerged adult can make very short flights. The teneral adult usually seeks shelter amongst vegetation until ready for mating, which in the case o f the female usually occurs a few hours after emergence or sometimes much sooner (Service, 1993). Males do not mate until their genitalia have rotated through 180°C, a process that take between 20 - 24 hours but in some species as little as 6 - 12 hours (Service, 1903). Mating is often preceded or accompanied by swarming Most male mosquitoes usually die after mating. The female require a blood meal for ovarian development, followed by the maturation and oviposition o f a batch o f eggs. There is a usually heavy mortality, especially among larvae due to predators, disease, drought etc. Larval loss due to predation is one o f the factors that reduce the numbers o f larvae that develop into adults. It is recognized that predation o f larvae in esiablished pools is an important factor in limiting their number. For example, Culex hgripes sometimes colonises the same pools as An. gambiae , causing a dramatic reduction in An. gambiae larval density (Haddow, 1942). It may be noted that the agility displayed by An. gambiae larvae in contrast to species such as An. funestus tend to increase their vulnerability to attack by predators (Service, 1980). Adult females Anopheles can live for several weeks to several months, dependent upon the prevailing environmental conditions. University of Ghana http://ugspace.ug.edu.gh Adult Anopheles usually rests with the body at an angle to the surface with proboscis and abdomen in a straight line In some species, they rest at almost right angles to the surface, whereas in others such as An. culicifacies the angle are much smaller (Kettle, 1992). Most Anopheles are crepuscular or nocturnal in their activities, thus emergence from the pupae, mating, blood feeding and ovipisition normally occur in the evenings, at night or nearly in the morning around sunrise. Some species such us An. albimantis, a malaria vector in Central and South America, bite man mainly outdoors (exophagic) from about sunset to 21 hours whereas others will rest outside (exophilic) in a variety o f natural shelters, such as amongst vegetation, in rodent burrows, cracks and crevices in trees, under bridges, in termite mounds, and other cracks in the ground. Most Anopheles species are not exclusively exophagic or endophagic exophilic or endophilic, but exhibit a mixture o f these extremes behaviour. Similarly, Anopheles do not feed exclusively on either man or animal, most feed on both man and animals but the degree o f anthropophily and zoophily varies according to species (Coluzzi et al., 1999). University of Ghana http://ugspace.ug.edu.gh MOSQUITO LIFE CYCLE Figure 3: Schematic illustration o f the life cycle o f Anopheles vectors o f malaria (After Service, 1980). University of Ghana http://ugspace.ug.edu.gh 2.6 Environmental Factors Influencing the Abundance of Malaria Vectors [ he major malaria vectors are characterized by some degree o f association i.e. anihropophily with humans, which presumably evolved from primatophilic ancestors and/or as a by-product o f adaptation to human environments (Constantini et al., 11)‘)7 ). Humans acl as a constant evolutionary challenge, as they provide a source of environmental change and heterogeneity to which anthropophilic Anopheles have to respond with a highly vector-host relationship (Coluzzi, 1992). Humans, through activities such as clearing forested areas for agro-related activities either create ideal conditions for the vectors or often exacerbate malaria situation by their moving into areas heavily populated by vectors (Rahman et al., 1993). Malaria is governed by a large number of environmental factors, which affect its distribution, seasonality and transmission (Snow et al., 1999). Any beneficial impact of development in significantly reducing in transmission potential o f Plasmodium has only been observed with vector species that have complex larval habitats vulnerable lo pollution and/or land exploitation. This is the case for the malaria vectors in the Mediterranean area, namely An. sacharovi, An. labranchiae and An. superpictus (Bruce-Chwartt & Zulueta, 1980). The opposite trend is however observed in various tropical vectors and particularly in the most anthropophilic species o f the An gamluae complex, which usually exploit man-made larval habitats such as bare-edged, temporary, freshwater pools etc, exposed to sunlight. The spread of these breeding sites is clearly favoured by agricultural practices associated with deforestation, desalination o f coastal areas and irrigation of arid savannas The result is an increase in recent decades of malaria transmission potential in many areas of 21 University of Ghana http://ugspace.ug.edu.gh Africa south o f Sahara because of a progressive adjustment o f the main vectors in the An. gambiae complex to man-made ecological changes (Coluzzi, 1984). Changes in environmental conditions are important determinants o f the ecology of living organisms, which adapt to the different ecosystems leading to an evolution of distinct genetic variabilities within species. More obvious modifications o f the vectorial system have been reported in relation to man-made ecological changes in the forest zone o f Southern Nigeria and in the lagoon area o f Cotonou (Coluzzi, 1992). Both instances were the direct consequences o f urbanization. Studies in southern Nigeria revealed unexpected concentrations o f An. arabiensis in urban and periurban situations within the ecological zone formed by the West African forest belt (Coluzzi, 1992) This zone is normally occupied solely by the Forest form o f An. gambiae s.s which is characterized by the standard chromosome -2 arrangement. This taxon was however, found breeding in small forest villages around the urban periphery and its absence from the central part of towns was thought to be due to either competitive exclusion by An. arabiensis or the possibility that any genetic adaptation to the urban environment was continuously disrupted by gene flow from foiest-adapted genotypes. In the lagoon area o f Cotonou, An. gambiae s.s., polymorphic for the Savanna chromosomal arrangements 2Rb-2La, replaces the less effective vector^/?, me las in lagoon zones where pile-dwelling traditional villages have been converted into unplanned urban settlements (Coluzzi, 1992). Other obseivations o f adult populations o f An. gambiae s.s. and An. arabiensis indicate a spatial or temporal separation o f these species. In Mali, Tour te ta l . (1998) found that the savanna form of An. gambiae s.s. predominated in relatively humid 22 University of Ghana http://ugspace.ug.edu.gh areas with larval production occurring almost exclusively during the rainy periods, whereas An. arabiensis prevailed in more arid areas and reproduced throughout the year. In Tanzania, An. arabiensis was common during the short rains and just before the- long rain, whereas An. gambiae s.s. predominated during and just after the long rains (White et a/., 1972). Studies carried out in Nigeria, indicated th a t^« . gambiae s s is common during the long rains, with populations o f An. arabiensis increasing as the rains receded (White and Rosen, 1973). In Kenya (at Kisumu) Haddow (1942) reported higher populations o f An. gambiae s.s. with at least five inches o f rain per month than with less than lhat amount Anopheles gambiae exists only in frost-free regions (Grilles & de Meillion, 1968), or where the minimum temperature in winter remains above 5°C (Leeson, 1931). The overall relationship between mosquito abundance and rainfall has been demonstrated on several occasions (Molineaux & Gramiccia, 1980, Le Sueur & Sharp, 1988). Rainfall provides breeding sites for mosquitoes and increases the humidity, which enhances their survival. Temperature affects the transmission cycle o f P. falciparum in many different ways, but the effects on the duration o f the sporogonic cycle o f the parasite and the vector survival are particularly important (Snow et a l , 1999). Temperatures below 22°C are the determining factors of the number o f mosquitoes surviving parasite’s incubation period, which takes 55 days at 18°C and ceases around 16°C (Detinova, 1962). After 55 days the proportion of a cohort o f mosquitoes lhat survives is only 0.003 (Martens, 1997). I he breeding habitats of Anopheles malaria vectors vary from large and usually permanent collections o f water, such as fresh water swamps, marshes, rice fields and University of Ghana http://ugspace.ug.edu.gh borrow pits to smaller collections o f temporary water such as small pools, puddles, water filled car tracks, ditches drains, galleys, hoof prints, etc. The most common breeding sites are the shallow open sun lit pools (Service, 1993). Larvae occur in wells and man-made container habitats such as clay pots, motor vehicle tyres, water storage jars and tin cans (Chinery, 1984). Most observations o f the larval habitats of An. gambiae s.I have noted a preference for temporary, sunlit pools (Gillies & DeMeillon, 1968; Gillies & Coetzee, 1987). However, not all suitable-appearing, sunlit pools examined in the Freetown, Sierra Leone area, were positive for An. gambiae. Generally, many larvae would be found in one pool whereas other similar- appearing pools were completely negative for larvae (Muirhead-Thompson, 1945). University of Ghana http://ugspace.ug.edu.gh 2.7 C on tro l o f M iliar ia In principle, (here are two main approaches to the control o f malaria, chemotherapy and vector control. Chemotherapy using antimalarial drugs has been the main method for parasite control and for clinical management o f the disease. Vector control aims in eliminating vector or reducing human vector contact using various methods including insecticide, environment management, bednets etc. Various attempts through national programmes have been made to control the disease worldwidely, using these approaches. 2.7.1 Historical background to malaria control programmes Systematic control o f malaria begin after the discovery o f the causative parasite by Alphonse Laveran in 1880, the demonstration by Ronald Ross o f its complete cycle in 1887 and its specific dependence on Anopheles mosquitoes by the Italian malariologistes, Giovanni Battista Grassi, Amico Brignami and Giuseppe Bastianelli ( I'eklehaimanot & Bosman, 1999). With the discovery o f the organochloride insecticide, dichlorodiphenyltrichloroethane (DDT) during World War II, it was thought eradication o f malaria globally through vector control was possible (WHO, 1998), therefore the WHO Global Malaria Eradication Programme was created. During this period, Africa hosted several pilot pre-eradication projects, but was never included in the programme, in spite o f burden of the disease on the continent (WHO, 1998). The justifications were related to the general under-development o f health services, communications and infrastructures (Gramiccia & Beales, 1988). In the end, the major donors withdrew their commitment to malaria eradication and the pilot projects were terminated and countries had to depend entirely on their own untrained technical stall', their limited infrastructure and meagre financial resources. 25 University of Ghana http://ugspace.ug.edu.gh With the Declaration of Alma Ata in 1978, WHO promoted the idea o f malaria control in the con text o f primary health care, but this vision was not translated into sticnglhening malaria control efforts in Africa (Teklehaimanot & Bosman, 1999). Because o f the resurgence o f malaria in various parts o f the world and malaria epidemics in Africa, the World Health Assembly in 1990 requested the Director General to reorganize the programme to increase its human resource and develop an appropriate strategy lo address the prevailing situation. As a result, the Global Malaria Control Strategy (GMCS) was developed through partnership and consultation involving all major development and bilateral organizations and UN agencies. This started with a meeting on "malaria waiting for the vaccine” at the London School of Hygiene and Tropical Medicine (U.K.), to debate what could be done at that time with the tools available (Targett, 1991). Then the control strategy adopted was further discussed and developed, and was finally endorsed in 1992 by over 92 Ministers o f Health from endemic countries and the major international partners during [he Ministerial Malaria Conference in Amsterdam, The Netherlands (WHO, 1993). It was subsequently endorsed by the World Health Assembly, the Economic and Social Council o f the United Nations, the UN General Assembly and by the Organization of African Unity (Teklehaimanot & Bosman, 1999). The four technical elements of this global strategy were to provide early diagnosis and prompt liealment, to plan and implement selective and suitable preventive measures including vector control, to detect early, contain or prevent epidemics and to strengthen local capacities in basic and applied research to permit and promote the regular assessment of a country’s malaria situation, in particular the ecological, social and economic determinants of the disease (WHO, 1998). University of Ghana http://ugspace.ug.edu.gh Be I ween 1992 and 1996, there was a gap between commitments and real financial support. Initial preparatory activities were completed to support the development of national plans o f action and training of programme managers. However, the limited support given did not produce any impact on the malaria situation (Teklehaimanot & Bosnian, 1999). Because o f resurgence of malaria and the slow progress made to combat the problem, the World Health Assembly requested the Director General to explore the possibility o f establishing a special programme on malaria and to mobilize resources to support malaria control activities at country level. Thus, during 1997-8 US$ 20 million were allocated to intensify malaria control in 34 African countries. Then the Accelerated Implementation o f Malaria Control was conceived with the aim o f establishing sustainable foundations for malaria control in Africa and capacity building o f National Malaria Control Programme (NMCP) was given the highest priority with a locus on disease management, involving health professionals from the general health services (WHO, 1998). A significant development was the declaration on malaria by Organization o f Africa Unity in 1997, whereby the Heads o f State committed themselves to malaria control in Africa, followed by establishment o f the African Initiative for Malaria Control which includes the African countries and major partners such as World Bank, UNICEF, USAID, among others. In 1998, the G 8 Summit in Birmingham discussed also the need for higher commitment to malaria control, particularly in Africa. During the same year, the Director General ol WHO established Roll Back Malaria initiative, as one o f the priority projects for the Organization, to be implemented as a global partnership within the context ofhcalth sector development (Teklehaimanot & Bosman, 1999). University of Ghana http://ugspace.ug.edu.gh 2.7.2 Chemotherapy Anlimalarial drugs however can be grouped into four categories according to the stage in the life cycle of the malaria parasite that they attack. These are tissue and blood schizonticides, gametocytocides and sporonticidals. Drugs that destroy all exo- erythrocytic forms are referred to as tissue schizonticides. Blood schizonticides act on the asexual erythrocytic parasites and the gametocytocidals act on gametocytes, particularly the immature forms. The antimalarials when taken up in blood meal inhibit the development o f oocyst in the mosquito and thereby prevent the production of sporozoites and these are referred to as sporonticidals. Quinine is a bitter tasting blood schizonticide and in case o fP vivax and P. malariae also acts as a gametocytocide. It was originally extracted from the bark o f Cinchona ledge liana (Cinchona tree) and remains the only drug, which over a long period o f lime has remained largely effective for treating the disease (Davis, 2 0 0 1 ). Quinine was previously widely used but has undesirous side effects including acute massive inlravascular haemolysis and haemoglobinuria (black water fever). It is now used only for emergency treatment of P. falciparum malaria, when response to Chloroquine and most antimalarials have failed In the 1940’s, an intensive research programme to find alternatives to quinine lead to the manufacture of Chloroquine and other chemical compounds that are more cllective and less toxic (NIA1D, 2000). Chloroquine became the most successful anlimalarial drug ever synthesised, because of its safety, affordability, ease o f use and its great efficacy (WHO, 1967; Ginsburg et al., 1999; Djimde et al., 2001). It was introduced in 1945 and is an excellent blood schizonticide. It is also a University of Ghana http://ugspace.ug.edu.gh gametocytocide agai 11 si P. vivax, P ovale and P. m alariae. Chloroquine assumed a major role in primary heallh care because of the rationale o f preventing mortality and curbing morbidity and suffering o f afflicted persons. However, the efficacy o f Chloroquine in malaria chemotherapy has been compromised with the development of the resistance to the drug by the malaria parasites. The first documented cases o f / ’ falciparum resistance to Chloroquine were in South America in the 1950s (Peter, 1970), and confirmed in Thailand in 1959 (Harinasuta et al., 1962). Since then, Chloroquine resistance has spread to almost all the malarious areas o f the world with heterogenic distribution both in frequency and degree (Bjorkman and Phili-Howard, 1990). Chloroquine now no longer offers protection against South East Asian P. falciparum and increasingly in other regions because of resistance (Navy Med. Dept., 200 I). In Africa, resistant strains o f P. falciparum were first recorded in Kenya and Sudan in 1978 (WHO, 1986) and more recently in West Africa (Cheesbough, 1991). In many parts of Africa, the drug is no longer used alone for therapy (Brasseur et a l, 1998). While effective in suppressing the falciparum in some parts o f Africa and most strains of P vivax, resistant forms o f P. vivax are appearing and have been reported in Papua New Guinea, Indonesia, Thailand and India (WHO, 2000). Many studies are underway to discover the cause of resistance in the parasite although how Chloroquine works at the molecular leyel, the biochemical and molecular mechanisms o f drug resistance in P. falciparum remain unknown (Wellems, 1992; Wellems et a l, 1997; Martiney et al., 1999). In the 1980s, Halofantrin which belongs to the class o f compounds called phenanthrene-methanols and not related to quinine was introduced as a blood schizonticidal against the erythrocytic stages o f Chloroquine resistant P. falciparum 29 University of Ghana http://ugspace.ug.edu.gh and also the erythrocytic stages o f P. vivctx but not the hypnozoites. The drug is used to treat acute form of uncomplicated and multi-resistant P. falciparum malaria and as a “stand by" drug if when chemoprophylaxis fails and there is no medical aid available. Its short half-life of 1 to 2 days however does not make it suitable as a prophylactic (Davis el a/., 2001). Halofantrin has been associated with neuropsychiatric disturbances and also it is contra-indicated during pregnancy and not advised for women who are breastfeeding (WHO, 1998). Proguanil was first synthesised in 1946 and introduced in 1948 under the trade name Biguanide. It destroys the early tissue stages, especially P. falciparum and hence is used as causal prophylactics. It also acts as blood schizonticide working more slowly than chloroquine and as a sporontocide (Navy Medical Dept., 2001) and it is free from unpleasant side effects. Proguanil is combined with chloroquine to prevent transmission by killing gametocytes. It is still used as prophylactics in some countries. Pyrimethamine was introduced in 1952 as a diaminopyrimidine with similar activities as Proguanil. It is a daig that affects the synthesis and utilisation o f folate. Pyrimethamine (2,4, diaminopyrimidine) acts by inhibiting the dihydrofolate reductase necessary for synthesis o f tetrahydrofolate, a precursor in the parasite DNA synthesis (Navy Medical Dept., 2001). Some of side effects are skin rashes and higher doses will affect human dihydrofolate releases leading to megaloblastic anaemia, holale supplement is usually given in pregnancy to reduce the efficacy of the drug (Nwanynwu el al., 1996; Basco & Rignald, 1998). Resistance to Pyrimethamine has been reported in Africa (WHO, 1990). 30 University of Ghana http://ugspace.ug.edu.gh Fansidar is a blood schizonticide and it is a combination o f Pyrimethamine and Sulphadoxine that was pressed into service as the first line o f defence against Chloroquine resistant P. falciparum parasite in South America and South East Asia (Plowe tit al., 1997). It is used for uncomplicated malaria that cannot be cured by chloroquine, or as first line treatment where Chloroquine has been dropped from use (Krogstad, 1996). Pyrimethamine-dapsone (Fansidar) is combined with Chloroquine to prevent transmission by killing gametocytes. In Southeast Asia where parasites are often resistant to both Pyrimethamine and Sulphadoxine, Fansidar is clinically useless (Winstanley, 1996). Resistance to Fansidar is now widespread and serious side effects have also been reported (WHO, 1998). The infusion of qinghoo Artemisia annua has been used in China for at least the last 2000 years (WHO, 1999) and its active ingredient “ qinghaosu” also known as Artemisinin has recently been identified. The two most widely used derivates o f Artemisinin are Aitesunate and Artemether and both are blood schizonticides. While they are widely used in Southeast Asia, they are not licensed in much o f the “western world” including Australia. A high rate of treatment failures has been reported and it is now combined with Mefloquine for the treatment o f P. falciparum malaria (WHO, 1998). Mefloquine was first introduced in 1971 and is a quinoline methanol derivative structurally related to quinine. It is a blood schizonticide and is also effective in killing hypnozoites if given as a combination treatment with Primaquine. The compound was found to be effective against malaria that was resistant to other drugs 31 University of Ghana http://ugspace.ug.edu.gh and because o f its long half-life it is also a good prophylactic. There have been reports of acute brain syndrome, which is estimated to occur in 1 in 10 ,000 to 1 in 20.000 o f the people taking this drug (WHO, 1998). Multi-drug resistance is an acute problem in some areas, particularly in Southeast Asia, where P. falciparum exhibits resistance to virtually all available antimalarials, including mefloquine (Chids et al., 1991; Peters, 1985). Widespread resistance has now developed to Mefloquine and this together with undesirable side effects has resulted in a decline in its use (Fernandez, 2001). A combination o f Proguanil and Atouaquone known as Malarone was first introduced in 1998 in Australia. Atouaquone was introduced in 1992 and it was used with success for the treatment o f Pneumocystis carrinii. When combined with Proguanil there is a synergistic effect and the combination is at present a very effective anti malarial treatment. Malarone has undergone several large-scale clinical trials and has been found to be 95 % effective in drug resistant P. falciparum malaria (WHO, 1998) It has been claimed to be largely free from undesirable side effects even though Proguanil is an Antifolate Mopacrine that was introduced in 1935 and used as a prophylactic on a large scale during the World War II (WHO, 1998) is now considered absolute. This drug, which is an effective blood schizonticide has disadvantages such as lading down in the skin and the recipient turns bright yellow It is now considered to have too many undesirable side effects and is no longer used. 32 University of Ghana http://ugspace.ug.edu.gh Chemoprophylaxis as a control strategy has been attempted but it still remains a subject o f debate. Prophylaxis is generally considered necessary for pregnant women but that for children is debated because of the risk o f long-term side effects and of selection for resistant parasite strains (Carnevale & Mouchet, 1987). Countiywide prophylaxis is expensive and also requires strong organization. Maintaining the involvement o f the community is also difficult, because at any one time only a proportion o f the community participates in drug distribution programs. In addition, drug administration can prevent natural immunity from occurring and may simply delay disease until children are older (Carnevale & Mouchet, 1987). 33 University of Ghana http://ugspace.ug.edu.gh 2.7.3 Vector control Mularia eradication was achieved in southern European countries in the early 1950s using insecticide spraying (Coluzzi, 1992). Successes with vector control have also been reported in Malaysia, Brazil, Egypt, Cuba and Panama. However, in many other areas attempts to eliminate the vector or reduce transmission have met with limited success and frequent failures (Harrison, 1978; Bynum & Fantini, 1994, 1998). Insecticide formulations with long-term residual activities are sprayed on surfaces, such as walls of houses that the potentially infectious mosquito is likely to encounter. This strategy targets the weak link in the transmission cycle, i.e. the requirement for long-term survival of the infected mosquito for complete parasite development (Collins & Paskewitz, 1995). Prior to the Second World War, control was based mainly on antilarval measures, including source reduction, while pyrethrum spray as an adulticide was tried on a small scale in certain areas with variable results (Zahar, 1984). During 1940s, with availability of DDT, malaria control by house spraying was initiated on small scale in certain countries and on a larger scale in others such as Madagascar, Mauritius, South Africa (Natal and Transvaal), Swaziland, and Zimbabwe, formerly Southern Rhodesia (Bruce-Chwatt, 1963). DDT became widely used because o f its cheapness per unit weight and its durability, which enabled spraying to be carried out twice a ye;ir, or only once in areas with a short annual malaria mosquito season (Curtis, 1996). Early studies demonstrated that DDT was also a repellent, irritant and has toxic effects on malaria vectors (WHO, 1998), hi 1960s and 1970s there were claims about supposed elfects of DDT on human health such as residues in human breast milk, which was usually attributed to contaminated food (Curtis, 1994). Resistance to 34 University of Ghana http://ugspace.ug.edu.gh DDT developed in the insects, which rendered it ineffective for vector control. DDT resistance in An. gam biae s i. has also been reported in areas in W est Africa where it was used in house spraying (Ivoira Cano and Bakri, 1975 unpublished report to WHO). The use o f other pesticides in agriculture coupled with the extensive use of DDT in cotton production areas were believed to have contributed to the selection pressure for resistance (Zahar, 1984) DDT was then replaced with other classes o f insecticides and later on they were also reported to be ineffective to the resistant mosquito vectors. Larval control with paris green combined with pyrethrum house spraying was extremely successful in elim inating/4/7. gambiae from Brazil (Soper& W ilson, 1943; Soper, 1949; Gratz & Pal, 1988), and paris green was the only agent used in the eradication ol 'An. gambiae from a focus in Egypt in 1944-1945 (Soper, 1949; Gratz & Pal, 1988; Russell, 1995). In general, organophosphates (especially temephos) are mostly widely used larvicide, but DDT, dieldrin, larvicidal oils, arsenical compounds, and development inhibitors have all been used with varying degrees o f success (Russell, 1995). Other measures, such as bleeding habitat modification and biological control agents, have been effective in limited settings, at elim inating or drastically reducing malaria prevalence in several regions. Synthetic repellents are usually used by residents and visitors to endemic regions Plants as natural products or synthetic repellents are used in many areas such as (he use o f Citronella products in India against anopheline mosquitoes (Curtis el al, I 990). In Tanzania, smoke from burning plants are used to protect against mosquito bites. However, the effectiveness o f these methods is 35 University of Ghana http://ugspace.ug.edu.gh probably limited and will depend on both the biology o f the local vectors and the intensity o f transmission (Collins & Paskewitz, 1995). Large-scale use o f insecticide impregnated bed nets (ITBNs) in endem ic countries is being promoted as a means to control malaria. Studies carried out in Senegal (Alonso cl al., 1991) and China (Cheng el al., 1995) initially demonstrated the efficiency o f IlBNs in reducing infant mortality. These findings were subsequently confirmed by a large-scale multicenter study in six countries across Africa including Ghana (Binka d al., 1996, Lengeler et at., 1996; Nevill et al., 1996). Nevertheless, acquisition o f new infections still occurred at a very high rate during the high transm ission season. For example, it was estimated that in Kenya 100 % o f the children should have been infected with P falciparum with i 3 6 weeks in the bed net villages compared with 10 6 weeks for the controls (Beach el a l, 1993). Also in Burkina Faso, where transmission levels are also high, the use o f bed nets was found not to have impacted on P. falciparum malaria incidence (Curtis el a l, 1990). The behaviour o f the vectors largely affects the success o f the control method. During the high transmission season, substantial members o f vectors may be feeding outdoors, often during the early evening. Thus, bed nets may be most useful in areas where transm ission is less sLable, seasonal or o f low intensity. Problems include non-compliance in proper use of the bed nets and failure to maintain the insecticide dose, which can be reduced substantially during net washing (Curtis et a]., 1990). Another notable possibility is introducing malaria refractory genes into natural populations of vectors (James, 1992; Crampton et a l, 1994). Progress is being made towards identifying refractory mechanisms and their underlying genetics (Vem ick et 36 University of Ghana http://ugspace.ug.edu.gh a l, 1995) and in developing the technology to introduce genes into the mosquito genome (M iller el a l, 1987). However, strategies for integrating selected refractory genes into field populations are lacking. The possibility o f population sub structuring is an important complication as well as the dynamics o f the rate to fixation o f a gene introduced into a population that consists o f two or more subpopulations maintained by mating barriers, which will be very different than the dynamics in a population o f randomly mating individuals (Lanzaro e1 a l, 1995). Research has also turned toward the vector, with the main aim o f using recombinant DNA technology to replace a highly competent vector population with an identical population engineered to be an incompatible host for the malaria parasite (Besansky el a l, 1992; Coluzzi, 1993; Crampton et a l, 1990; James, 1992; Eggleston, 1991). Several genetic control research projects targeting Anopheles mosquitoes have been carried out, none with any particular success (Collins & Paskewitz, 1995). The basic idea is to drive a genetic construct into the existing vector population. This construct ca iT je s a parasite-inhibiting gene or genes and will not significantly affect the vector population’s fitness In short, this control strategy targets the parasite w ithin the mosquito rather than the mosquito itself (Graves & Curtis, 1982). This control strategy is unlikely to bear fruit for decades, and the malaria control tools that do emerge from this effort may be entirely unlike what is imagined today (Collins & Paskewitz, 1995). Considerable resources have been devoted to the search for other alternative methods of controlling malaria vectors. For example, biological control agents have been in use and to date only larvivorous fish have been used successfully in malaria control 37 University of Ghana http://ugspace.ug.edu.gh projects and theses cases are few. Hackett (1931), Hadjinicolaou & Betzios (1973) and W ickramasinghe & Costa (1986) have all reported that the use o f the North American fish Gambnsia affinis successfully reduced malaria incidence in Italy and Greece, where malaria transmission was unstable. R ishikesh et al. (1988) have summarized efforts to identify useful pathogens and parasites, including virus, fungi, nematodes and protozoa The main pathogens under study include the fungi ( "olcomomyces spp ., Lagenidium spp ., Ciilicinomyces clavosporus, andMetarhizium anisopliae, the protozoan Nosema algerae and the neam tode Romanomermis cnlicivorax. None o f these agents have shown any promise for malaria control, having proven difficult to rear and store, as well as being unstable or inefficient in the field The bacterial endospore toxins produced by various strains o f Bacillus sphaeriais and B. ihnringiensis israe/ensis have been used as larvicidal agents in some situations (de Barjac & Sutherland, 1989). Unfortunately, the Bacillus toxins are stiil relatively expensive and because they have no residual activity, they either require frequent application or are only suitable for environments where a one-time control measure produces a valuable outcome (Collins & Paskewitz, 1995). Malaria vectors exhibit a wide variety o f life history strategies therefore, there is no simple and universal applicable form o f vector control. In those parts o f the world where malaria has been eradicated the previous important malaria vectors still remains - in many cases, in numbers as great as during the periods when malaria was endemic (Collins & Paskewitz, 1995). Although vector analysis is part o f the process o f malaria control, this has not always been acknowledged as shown by the recent hi story of malaria control which has been characterized by the uncritical simplification o f success stories that are m isleading to the general application o f 38 University of Ghana http://ugspace.ug.edu.gh single control tools (Coluzzi, 1992). The proliferation o f man—made-malaria , which accompanied the push for economic development in most of the endemic countries, spurred the need for the control interventions and while great successes were obtained in many specific projects, the general campaigns proposed by the enthusiasts o f vector control faced increasing difficulties in their practical implementation in the field (Najera, 2001). The value and relevance o f malaria vector control have not been clearly recognized and its effectiveness has declined for reasons including poor use o f available alternative control tools, inappropriate use o f insecticides and reduced effectiveness due to vector resistance and lack o f an epidemiological basis for vector control interventions. The. problems o f malaria control are further aggravated by changing environmental conditions in areas in which exploitation o f natural resources and development activities are taking place. Expending agriculture, the cleaning o f forest, or the building of dams and irrigation schemes, and unplanned urban development provide mosquitoes with new breeding grounds, while at the same time bringing more people into contact with them. The emergence o f resistant vectors has comprom ised the use of most applicable insecticides and newer generation replacements are consistently more costly and are often less effective or exhibit greater toxicity on non-target fauna. 39 University of Ghana http://ugspace.ug.edu.gh 2.8 Species Identification and Relevance to Vector Control Programmes Species are considered as biological units, defined by intrinsic m echanism s o f reproductive isolation and characterized by some kind o f discrete genetic difference that is not necessarily expressed at the morphological level (Mayr, 1996). A lthough morphological features appear to be in many cases very useful tools for the characterization of species, the genetic analysis o f various groups o f closely related species or species complexes shows that reproductive isolation may be acquired without or before morphological divergence, resulting in speciation that is undetectable through morphological observation alone (Mayr, 1996). Species complexes containing morphologically cryptic species that vary in their behaviour and vectorial capacity present a very real problem to malaria control programme managers (Coetzee ei al., 2000). The identification o f sibling species is o f major mahiriological importance (Coluzzi el a l, 1979). Morphological similarity generally implies close phylogenetic relationships and recent speciation processes, but it does not imply similarity or identity o f bionomics when dealing with sympatric taxa. Such differences are important in the epidemiology o f malaria. Failure to recognize sibling species o f Anopheles may result in the failure to distinguish between a vector and a non vector species o f malaria as was the case with the Anopheles maculipennis complex (Hackett, 1937; Bates, 1940). A Lhorough understanding o f Lhe malaria problem and risk, sound know ledge o f molecular biology of the vector, human host and environment are prerequisites for effective planning and targeting of vector control interventions (WHO, 2000), 40 University of Ghana http://ugspace.ug.edu.gh The Anopheles gambiae complex was initially considered as a single species until 1944 and presently, six formally named species, as well as forms within them. There are morphologically indistinguishable yet genetically and behaviourally distinct mosquito species that vary dramatically in their importance as vectors o f malaria and in distribution in Africa (Coluzzi et al., 1979; Service, 1985). Members o f An. gambiae complex have a wide geographical distribution and have been reported from the most African countries and adjacent islands including Madagascar, as well as Saudi Arabia and Yemen (Coetzee et al., 2000). The two most widespread are An. gambiae and An. arabiensis. These species breed in temporary water often associated with human disturbances, from the southern limits o f the Sahara Desert south to most parts o f the continent including Madagascar They are largely sympatric in the narrowest sense; both are found as larvae in the same pools and as adults in the same huts (Powell et al., 1999). Anopheles gambiae predominates in forest and humid savanna zones whereas An. arabiensis is more successful in arid savannas and steppes, including those o f southwestern part o f Arabian Peninsula (Coetzee et al., 2000). Anopheles arabiensis is recorded more often than An. gambiae in areas where rainfall is less than 1000 mm and the reverse is observed where rainfall is grater than 1000 mm (Gillies & Coetzee, 1987; Hunt et al., 1998). Anopheles quadriannulatus has a narrower distribution in southeast Africa, Ethiopia and Zanzibar. This geographical discontinuity has apparently produced allopatric speciation, as shown by Hunt et al. (1998) through crossing experiments. Both An. 2.8.1 Anopheles gambiae complex and its distribution 41 University of Ghana http://ugspace.ug.edu.gh quadriatmulcitus species are generally sympatric with An. arabiensis and less frequently with An. gambiae s.s. • 4 Anopheles merits and An. me (as axe salt and blackish water breeders confined to the East and west coasts respectively o f Africa (Powell et al., 2000), although An. melas has been found breeding in fresh water streams in the Gambia (Chinery, 1984). Because o f their ecological differentiation into salt water, neither An. merus nor An. melas are sympatric as larvae with the other members, although adults o f these species may encounter both An. gambiae and An. arabiensis adults in certain situations (Hunt el al., 1998). Anopheles melas has a short dispersal range from preferred breeding sites and adults-are usually not found at distances more than 3 km from the saline environment (Bryan, 1987). Anopheles bwambae is known only from the Semliki forest in the Rift valley near the Zaire border where it breeds in geothermal mineral' springs (Coluzzi et a l, 1979, White, 1985). 2.8.2 Cryptic taxa within An. gam biae s.s. and their epidemiological implications In the early 1980s studies showed that in West Africa further taxonomic complexity may exist within An. gambiae s.s. Subsequent genetic and behavioural variations observed within this species were found to be associated with different cytological forms (Forest, Savanna, Bamako, Mopti and Bissau)which showed restricted or no inter-breeding in the field and whose distribution depended on environmental factors such as climate, breeding sites, etc. (Toure et a l, 1994). The evidence for cryptic taxa within An. gambiae s.s. is the observation that the various gene arrangements of the 2nd chromosome (differing by inversions) were far from a Hardy-Weinberg equilibrium at certain times o f the year in some areas o f West Africa (Coluzzi et a l , 42 University of Ghana http://ugspace.ug.edu.gh 1999). The strong association o f certain karyotypes with specific habitats led to the description o f distinct chromosomal forms o f An. gambiae known as eco-phenotypes. Three ecopbenotypes; Bamako. Mopti and Savanna occur in sympatry at numerous sites in Mali, West Africa. Studies o f inversion karyotype frequencies at these sites also revealed a deficit o f heterokaryotypes relative to Hardy Weinberg expectations (Toure, 1991). The relative frequencies o f these ecophenotypes also vary ecologically and seasonally (Toure et a/., 1983, 1994, 1998; Coluzzi et al. , 1985). The Savanna form is typically found away from major rivers and flooded or irrigated areas and is fully dependent on rainfall for larval breeding, and therefore, this form usually is absent during the dry season. The Bamako form is closely associated with riverine basins and breeds monthly from mid- to late rainy season. Its distribution coincides with the occurrence o f the larval habitats, which are quite unusual for An. gambiae s.s. The habitats are mainly slowly moving water and residual pools along the laterite edges of riverbeds. The Mopti form is closely associated with flooded plains and irrigated fields. It is the only form o f An. gambiae s.s. which is more arid-adapted than An. arabiensis, which it could competitively displace (Toure et al., 1998). Two other forms o f gambiae s.s. namely Forest and Bissau have also been recently proposed. The Forest form refers to forest-breeding An. gambiae s.s. which is nearly fixed for the 2R/2L standard chromosomal arrangements. The Bissau form has been recorded in Gambia and Senegal where it is associated with rice fields along the Gambia River (Bryan et a/., 1982), 43 University of Ghana http://ugspace.ug.edu.gh An extremely important implication is that populations o f An. gambiae s.s. are not homogeneous entities as so often assumed in epidemiological models and in planning of control measures (Coluzzi et al., 1999). The fact that often the vector population is an heterogeneous mix o f units, ranging from intraspecific polymorphism to cryptic taxa, affects both the efficiency o f disease transmission and the relative value o f the control measures (Coluzzi, 1984; ,1992). For example, the evolution form o f An. gambiae that breed through the dry season i.e. the Mopti form can produce year- round transmission when ordinarily it is interrupted in the dry season (Toure et al., 1996). As environments vary spatially and temporally, different genotypes come to predominate, such that the total population fitness is buffered from such perturbances. The fact that different karyotypes have behaviours that differentially place them in contact with humans clearly has important medical implications (Coluzzi et a l, 1999). It has also been demonstrated that different karyotypes display differences in frequency o f having human or animal blood meals (Petrarca & Beier, 1992). This leads to. different karyotypes being differentially infected with Plasmodium falciparum (Toure et a l, 1986; Petrarca & Beier, 1992). In addition to differences in blood meal choice, these behavioral differences also affect the efficacy of insecticide spraying (Coluzzi et a l , 1999). The adaptive flexibility shown by An. gambiae s.s. has important epidemiological implications. The adaptability is clearly exhibited in the species ability to rapidly adapt to man-made environments. While some populations still breed in ancestral habitats, the vast majority o f populations o f both An, gambiae s.s and An. arabiensis breed in human disturbed environments. This has meant that as humans clear more forest for agriculture, desalinate more coastal regions, and also irrigate more arid 44 University of Ghana http://ugspace.ug.edu.gh savanna, they are expanding the geographic range o f these highly efficient vectors o f malaria (Powell et al., 1999). There has been much discussion recently o f genetically manipulating vector populations to make them less efficient vectors o f the disease. If species are sharing genes, even if only occasionally, clearly this will have important consequences for any genetic replacement program. The target species could relatively quickly regain the ability to transmit by acquiring the gene or genes from another species. This is especially true if manipulated gene or genes are at a selective disadvantage compared to the wildtype gene they replace. 2.8.3 Characterization of An. gam biae s.s. populations by m icrosatellite DNA analysis Microsatellites ar-e defined’ as simple tandemly repeated DNA sequence elements • • * • • usually as a dinucleotide or a trinucleotide (e.g. [GT]n, [GAC]n, etc. ) found abundance in the genomes of just about every known organism and organelle (Zheng, 1997; Chambers & MacAvoy, 2000). Some researchers defined them as 2 - 8 bp repeats (e.g.. Armour et a l, 1999) and others as 1 - 6 bp repeats (e.g. Goldstein & Pollock, 1997) or even 1 - 5 bp repeats (e.g. Schlotterer, 1998), M icrosatellite loci have been described as “ideal” markers to measure population level phenomena (e.g. population structure) due to their high polymorphism, codominance, abundant presence throughout the genome and relative ease in scoring (Bughanan et a l , 1994; Scribner et al, 1994; Lanz^ro et al, 1995). They have come into prominence over the last decade because scientists have found them to be remarkably versatile molecular tools and applications range from their use as highly accessible genetic markers for chromosome segments identification o f individuals to tracking the biological history o f populations. They are usually highly polymorphic in length due University of Ghana http://ugspace.ug.edu.gh to a variation in the number o f repeats within a given microsatellite locus as a result o f uneven crossover (Jeffrey el al., 1985) or slippage o f the DNA polymerase during replication (Tautz, 1989). The high polymorphism o f microsatellite results from high mutation rates, estimated to range from 10 2 to 10 5 locus/gamete/generation (Dallas, 1992; Weber & Wong, 1993) with most estimate being between 10"3 and 10"4 (Lehmann et al., 1996). This polymorphism also leads to an increased probability o f finding heterozygous individuals and the microsatellite markers can be mapped by recombination since they are generally inherited in a co-dominant Mendelian fashion (Weissenbach et al., 1992). The high degree o f polymorphism has also made microsatellite DNA useful markers for evolutionary and phylogenetic studies o f organisms and in the study o f the origins o f different human populations (Santos et al., 1997). Several markers have been used in the characterization o f An. gambiae s.s. populations from both East and West Africa (Lanzaro et a l 1995; Kamau et a l, 1999; Lehmann et al., 1998). The use o f microsatellites in An. gambiae s.s. studies is now extensive, and these types o f markers are increasingly being recognized as valuable for studies on other vectors (Wang et a l, 1999). They are becoming the markers o f choice for high-density genome mapping for An. gambiae (Zheng et al., 1993). Lanzaro et al. (1995) have also demonstrated high polymorphism, codominance, abundance throughout the genome o f An. gambiae and the relative ease in scoring made the authors conclude that microsatellite loci are superior to allozymes for population studies. 46 University of Ghana http://ugspace.ug.edu.gh Polymorphisms in microsatellite are being used to generate a genetic map o f the An. gambiae s.s. Microsatellites have all potential as tool for studying the population genetics o f this malaria vector. Microsatellite polymorphisms provide a more sensitive measure o f divergence and therefore can potentially distinguish more effectively populations that may have recently diverged (Lanzaro et al., 1995). The large number o f loci, high degree o f polymorphism, and abundance o f low and intermediate frequency alleles also suggest that microsatellite will provide a superior tool for estimating gene flow (Wright, 1951) and determining allele distribution (Slatkin, 1985). . Microsatellite DNA analysis may be a PCR-based method that uses primers and scoring the different band sizes o f the amplified products for subsequent analysis. Several computer-based algorithms including POPGENE (Population Genetic • / Analysis) etc, are currently available for analysing micro satellite DNA data. The output from POPGENE includes genotype and allele frequencies, effective allele number, polymorphic loci, genetic distance, expected homozygous, expected heterozygous, differentiation indices (Fst), heterozygosity deficiency or excess (Fis), Gene flow, etc which aid in characterizing different populations University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE MATERIALS AND METHODS 3.1 The Study Sites Mosquito larvae arid pupae together with water from their habitats were collected at fifty-four breeding sites in the Greater Accra Region and in the Apam District in Central Region. The Greater Accra and Central Regions are adjacent to each other and both are located in the southern part o f Ghana and all sites are located in the coastal savanna ecological zone, which is characterized by dry climatic conditions receiving the least amounts o f rainfall in Ghana (Dickson & Benneh, 1988). It has two rainfall peaks. The first occurring April to June and the second from September to October, with the total amount o f rainfall ranging between 740 and 890mm a year. The lowest mean monthly temperature (about 26°C) is recorded during August and the highest (about 30°C) between March and April. The relative humidity throughout the year ranges between 65 and 75% in the afternoons. The vegetation consists mainly o f grass with isolated patches o f scrub and sparse trees. University of Ghana http://ugspace.ug.edu.gh The study area was divided into zones and each zone was surveyed on foot to locate breeding sites. Samples were obtained from a variety o f breeding sites, including gutters, marshes, ponds, small pools o f stagnant water, muddy water, borrow pits, runoff from bathrooms and irrigated rice fields (Plates la-c). W ater samples o f different qualities were also obtained, but most contained organic debris with filamentous green algae floating on the surface (Plates 2a-b). In addition, most o f the breeding sites were also shallow, temporary and exposed directly to sunshine. A global positioning system (GARMIN GPS 40™, VBA) was used to determine the geographical coordinates o f each site (Appendix I) 3.1.1 Description of larval and pupal habitats University of Ghana http://ugspace.ug.edu.gh Plate la: Sample collection site at Achimota with a narrow stretched pool o f water flowing from a leaking water supply pipe. University of Ghana http://ugspace.ug.edu.gh Plate lb: Sample collection site at Madina with a shallow pool o f stagnant water flowing from a residential washing area. University of Ghana http://ugspace.ug.edu.gh Plate lc: Sample collection site at Adenta with an open gutter containing stagnant water and a mixture of both Anopheles and Culex larvae. University of Ghana http://ugspace.ug.edu.gh Plate 2a: Examples o f type o f water from habitats where An. gam biae s.s. was found breeding University of Ghana http://ugspace.ug.edu.gh Plate 2b: Examples o f type o f water from habitats where An. gam biae s.s. was found breeding University of Ghana http://ugspace.ug.edu.gh 3.2 Methods 3.2.1 Field sample collections of pre-adult mosquitoes and water Anopheles larvae identified from their horizontal position on the surface o f water and other species identified by their angular position were carefully collected with a 350­ ml dipper and transferred into small plastic containers which were loosely capped. Water samples wer? also collected thereafter into a 1.5 litre plastic bottles for physico-chemical analyses. For the estimation o f dissolved oxygen (DO) a 300-ml glass-stoppered bottle covered with aluminium foil to shield light was filled with the water and 2 ml o f manganese sulphate solution, and then 2 ml o f alkali-iodide-azide solution added to fix the oxygen. The temperature o f each habitat was also recorded The samples were then transported in iceboxes to the laboratory, making sure that the humidity inside the box was high and that the larvae and pupae did not suffer from excessive heat stress. 3.2.2 Laboratory rearing of mosquitoes Once in the laboratory, the water with pre-adult mosquitoes were poured into trays to a depth o f approximately 2 cm and each tray labelled to indicate the site and date o f collection (Plate 3). The trays containing mosquitoes were kept at 27 - 30°C and 76 ± 2% relative humidity with a 12h:12h light and dark cycle. Every two days, about 200 mg o f ground Nutrafin goldfish food (Rolf Hagen, USA) were used to feed the larvae in the tr3 to 1.0 ml o f EDTA titrant at the calcium indicator end point), which is related to the volume o f standardized EDTA titrant was calculated using the equation (5). C B = .............................— (Equation 5) Volume ofEDTA Where, C = volume o f standard calcium solution used to standardize the EDTA The calcium hardness (mg o f CaCC>3 /l) was determined using the equation (6 ). Concentration o f Ca2+ (Equation 4) Calcium hardness = mg/1 (Equation 6 ) 0.4 Where, 0 4 = atomic weight o f Ca / molecular weight o f CaCO^ Magnesium hardness was determined using the equation (7). Mg hardness (mg/1) = Total hardness (Eq. 3) - Ca hardness (Eq. 6 ) (Equation 7) University of Ghana http://ugspace.ug.edu.gh Magnesium ion concentration was determined using the equation (8 ). Mg2' = [Total hardness (Eq. 3) - Ca hardness (Eq. 6 )] x 0.243 mg/1 (Equation 8 ) Where, 0.243 = atomic weight o f Mg /molecular weight o f C aC 0 3 Chloride ion (Cl ) concentration was determined using the argentometric method. The pH o f 100 ml o f the water sample was first adjusted between 7 and 10 using conc.H2S0 4 , then 1ml o f potassium chromate (K2C r0 4) indicator solution was added followed by titration with .standard AgN03 until a pinkish yellow colour endpoint was reached. A blank value was also determined by titration method. The chloride ion (Cl ) concentration was determined using the equation (9). (A-B) x M x 35,450 Cl = ---------------- - -------------mg/1 Equation 9 Volume o f sample Where, A = Volume o f AgN03 used for sample titration, B = Volume o f AgNC>3 used for blank titration, M = Molarity o f AgNCh Potassium (K+) and Sodium (N ah) concentrations (mg/1) were determined using flame photometric method. Samples were introduced in a Flame Analyzer (Gallenkamp model FGA - 350L, UK). The calculations o f potassium or sodium were given directly by the flame photometer at wavelengths o f 768 and 589nm respectively. In instances when either potassium or sodium concentration was very high the water samples were diluted and the concentrations were calculated as follows: K or Na ) = (K^or Na' in aliquot) x D Where, D = dilution factor, which is determined using equation (10). Volume of sample + distilled water used for dilution D = — (Equation 10) Volume o f sample University of Ghana http://ugspace.ug.edu.gh Ammonia-nitrogen ’ (NH4-N) amounts (mg/1) were determined by direct Nesslerization. For the calibration curve, 10 ml o f the stock ammonium chloride solution was diluted to 100 ml with distilled water. From this intermediate solution, each o f 1, 2, 3 , 4 and 5 ml was diluted to 100 ml with deionised water. 5 drops o f Rochelle salt were added to 50 ml o f each concentration into 100 ml conical flask, mixed well and followed by addition o f 2 ml Nessler’s reagent. They were then allowed to stand for 10 minutes for colour development The water sample was left to settle and an aliquot of 50 ml o f the supernatant was pipetted into a fresh 100 ml conical flask Then, 1 to 5 ml of sample was pipetted into a fresh flask and ammonia- free water added until it reached 50 ml mark For very turbid samples, the water was filtered and the filtrate was used for the analyses. Two drops o f Rochelle salt were added to the diluted sample or 5 drops in the case o f undiluted sample, mixed well and 2 ml o f Nessler’s reagent added They were then allowed to stand for 10 minutes for colour development and the absorbance for samples and blank determined using a UV/V1S spectrophotometer (Philips model PU 8625B, The Netherlands) at a wavelength o f 410nm using a 1 cm light path cuvette. The concentration o f ammonia-nitrogen in the unknown sample was determined by extrapolating from a calibration curve. Diazotization method was used to determine the concentration (mg/1) o f nitrite- nitrogen (N 0 2-N). A series o f standards (blank o f 0, 0.5, 1, 1.5, 2, 3, 4, 5 and 10 ml o f working nitrite solution) was diluted to 50 ml with distilled water. Then, 2 ml o f buffer-colour reagent were added to 50 ml o f sample in a Nessler tube, mixed well and left for at least 15. minutes to allow colour to develop. The absorbance was University of Ghana http://ugspace.ug.edu.gh measured in the UV/VIS spectrophotometer at 540nm against the blank. The concentration o f nitrite-nitrogen (mg/1) was determined by extrapolating from the calibration curve. To determine the concentration (mg/1) o f nitrate (NO3-N), the hydrazine reduction method was used. NO 3 calibration standards were prepared by diluting to 10 ml with distilled water the following intermediate nitrate solutions: 0, 2 , 4 , 6 „ 8 and 10 ml. In a test-tube, 1 ml o f 0.3M NaOH was added to 10ml o f the water sample and mixed gently, followed by the addition o f 1 ml o f reducing mixture and mixed gently. The mixture was heated at 60°C for 10 minutes in water bath, cooled to room temperature and 1 ml o f colour developing reagent was added, and mixed well by shaking. The absorbance at 520nm was read and the nitrate concentration was directly computed from a calibration curve. The value obtained however is due to both NO 3 -N and NO 2 -N, therefore the concentration o f NO 3 -N was derived by subtracting the concentration o f NO 2 -N above from the value obtained. Phosphate (PO4-P) concentration (mg/1) was determined using the stannous chloride method Standard phosphate solutions o f known concentrations ranging from 0.1 mg/1 to 1.0 mg/1 were prepared and treated as samples. To a 100 ml sample free from colour and turbidity was added 0.05 ml (approximately 1 drop) o f phenolphthalein indicator. To the sample turned pink, a strong acid solution was added dropwise to discharge the colour. Where more than 0.25 ml (5 drops) was required, a small volume of sample was diluted to 100 ml with de-ionised water then a drop o f phenolphthalein indicator was added, discharged if the sample turned the pink colour with the acid. A volume o f 0 4 ml molybdate reagent I was added with thorough University of Ghana http://ugspace.ug.edu.gh mixing after each addition and was followed by 0.5ml (10 drops) o f stannous chloride reagent. The absorbance at 690 nm was measured after 10 minutes (but before 12 minutes) using a UV/VIS spectrophotometer. The phosphate concentrations in the samples were determined from the calibration curve. Silica (S i0 2) concentration (mg/1) was determined using the molybdolicate method. From the silica intermediate standard solution, 5, 10, 15, 20, 25 and 30 mg/1 concentrations were prepared followed by addition o f the colour development reagents as the samples were treated. 1 ml o f HC1 and 2 ml o f ammonium molybdate reagent were added in rapid succession to 50 ml o f sample, mixed by inverting at least six times and left to stand for 5 to 10 minutes. This was followed by addition o f 2 ml o f oxalic acid solution and then mixed thoroughly. Colour was read after 2 minutes (but before 15 minutes after the addition o f oxalic acid) at a wavelength o f 410 nm on a spectrophotometer. The silica concentration was determined directly from the calibration curve o f known standards. Sulphate (S 0 4) concentration (mg/1) was determined using a turbidimetric method. From the sulphate solution, each o f 5, 10, 15, 20, 25 and 30 ml was diluted to 100 ml with distilled water and used as standard solution in calibration. 5 ml conditioning reagent were added to 100 ml sample in a 250 ml Erlenmeyer flask and mixed by stirring. Thereafter a spoonful o f barium chloride crystals was added still stirring at a constant speed for 60 seconds, After stirring and within 5 minutes the absorbance at 420 nm was measured. The concentration o f sulphate was extrapolated from the calibration curve o f known standard concentration. University of Ghana http://ugspace.ug.edu.gh To determine amounts o f fluoride (mg/I) the SPADNS method was used. Standard concentrations o f fluoride (0.2, 0 .4 , 0 .6 , 0 .8 , 1, 1.2 and 1.4 mg/1) were prepared from standard fluoride solution (0.1, 0.2, 0.3, 0.4, 0.5, 0.6, and 0.7 ml o f standard fluoride solution diluted to 50 ml with distilled water). 5 ml each o f SPADNS solution and o f Zirconyl-acid reagent were added to 50 ml o f the water sample and mixed well. The absorbance was then read at 570 nm. If the absorbance was beyond the range o f the standard curve, then water samples were diluted with deionised water to 50 ml into a conical flask and the procedure repeated. The concentration o f Fluoride in the sample was determined using the calibration curve. Total suspended solids were determined by gravimetric method. Filter was wetted with 10 ml o f deionised water followed by a transfer o f suitable volume o f sample to the funnel (to yield not more than 200 mg o f residue) after the bottle had been vigorously shaken. Filter was washed with three successive 10 ml volumes o f distilled water and suction was continued for 3 minutes after filtration was complete. Filter was removed from the filter holder, transferred into a dish and dried for at least one hour at 105°C in an oven. Filter was weighted after cooling in a desiccator. The drying cycle was.repeated until a constant weight was obtained. Total suspended solids (mg/1) were determined using the equation ( 11) ( A - B ) x 106 Total suspended solids = ----------------------------- mg/1 (Equation 11) Volume o f sample filtered (ml) Where; A = weight o f filter + dish + residue (g) B = weight o f filter + dish (g) Total dissolved solids were also determined using gravimetric method. The water sample was shaked and any volume, which was enough to yield between 10 and 200 University of Ghana http://ugspace.ug.edu.gh mg dried residue was taken. Sample was filtered through the glass fibre filter and a vacuum was applied for 3 minutes after filtration to ensure that as much water as possible was removed. After washing with three successive 10 ml volume of deionised water and continuous suction for 3 minutes, total filtrate (with washings) was transferred to an evaporating dish and evaporated to dryness on a water bath. Evaporated sample was dried at 105°C in an oven for at least 1 hour. The filter was removed and placed in a desiccator to cool and then weighed. The drying cycle was repeated until a constant weight was obtained. The total dissolved solids (expressed in mg/1) were determined using the equation ( 12) ( A - B ) Total dissolved solids = ------------------------------------ mg/1 (Equation 12) • Volume o f sample filtered Where, A = weight o f dried residue and dish (g) B = weight o f dish (g) Chemical oxygen demand (COD) was determined by the closed tube reflux method. 3 ml o f standard potassium dichromate solution (0.0167 M) and 7 ml o f H 2SO4 reagent (silver sulphate in sulphuric acid) were added to a sample o f 5 ml into a culture tube and mixed well by shaking. The tubes were placed in a COD Heating Digester (VELP Scientifica) preheated to 150°C and refluxed for 2 hours. After cooling to room temperature, 1 to 2 drops o f ferroin indicator were added to the products and titrated with 0.1M standard ferrous ammonium sulphate solution (FAS) to change colour from blue-green to reddish brown or wine end point. In the same manner, a blank containing the reagents and a volume o f deionised water equal to that o f sample was refluxed and titrated COD as mg o f 02/1 was calculated as follows: University of Ghana http://ugspace.ug.edu.gh ( A - B ) x M x 8000 COD = --------------------- mg/1 Volume o f sample Where: A = volume o f FAS used for blank B = volume o f FAS used for sample M = Molarity o f FAS 8000 = milliequivalent weight o f oxygen X 1000 ml/litre Determination o f dissolved oxygen (DO) was done using azide modification o f W inkler's method. 2 ml conc. H2SO4 was added to a sample o f 250 - 300 ml in a BOD bottle, shaken till dissolution was complete. From that solution, 100 ml were taken and titrated with M/80 sodium thiosulphate solution to a straw yellow colour, followed by addition o f 1 - 2 ml starch solution and titration continued until the blue colour turned to colourless. Dissolved oxygen as mg/1 O2 was calculated as follows: Volume of M/80 thiosulphate x 101.6 0 2 =----------------------------------------------------------mg/1 Volume o f water sample Biochemical oxygen demand (BOD) was determined by dilution method. 2 ml o f MnS04 were added,to the diluted 250 - 300 ml sample in BOD bottle followed by 2 ml of alkaline-iodide-azide and corked carefully to exclude bubbles. After the precipitate has settled, 2 ml o f conc.H2S0 4 were added, corked and the bottle inverted several times to dissolve the precipitate which gave intense yellow colour. The solution was then titrated with M/80 sodium thiosulphate until a pale yellow colour was obtained. Then 1 ml of starch as indicator was added and the titration was continued until t}ie appearance o f blue colour. The rest o f the water sample was incubated at 20°C for 5 days after which it was treated likewise. The difference in dissolved oxygen (DO) o f days 1 and 5 was used to calculate the BOD as follows: University of Ghana http://ugspace.ug.edu.gh BOD = (D1 - D2) x P Where: D 1 (mg/1) = DO o f diluted sample immediately after preparation D2 (mg/1) = DO of diluted sample after 5 days incubation at 20°C P = decimal volumetric fraction o f sample used University of Ghana http://ugspace.ug.edu.gh Statistical analyses 'of water parameters were performed using SPSS 10.0 software (SPSS Inc., USA). Samples were firstly divided into two groups: group with Anopheles species and that without but with other mosquito species. A t test was used to compare these groups o f observations. Water parameters whose means were significantly different were submitted to a linear discriminant function analysis to identify the best discriminatory parameters between the two groups. To determine the relationship between ecological parameters and the observed variation in the proportions o f An. gambiae larval populations, correlation and multiple regression analyses were performed The parameters were first analysed using univariate statistical methods. Then, those which were significantly correlated (Pearson correlation coefficient) to the proportion o f An. gambiae were included in multiple regression analysis that yielded a regression model containing the best combination o f parameters significantly contributing to the variation. To assess the similarity in An. gambiae larval habitats whole samples from An. gambiae were first analysed together and later divided into two groups o f 100% An. gambiae and mixed populations. The results..of the hierarchical cluster analysis o f water parameters was visually compared to the phylogeny tree generated by POPGENE software using the data from microsatellite DNA- analysis o f An. gambiae s.s. for similarities. 3.2.6.1 Test of equality of means This test deals with comparing groups o f observations with respect to continuous data, starting with the simplest case where we wish to compare a single group o f observations with some pre-specified value, and moving through to the case where 3.2.6 Data analysis University of Ghana http://ugspace.ug.edu.gh we have several sets o f observations on each a group o f individuals. The t test is appropriate for this type o f analyses. The most common statistical analyses are those for comparing two independent groups o f observations. With these groups, we are interested in the mean difference between the groups, but the variability between subjects becomes important. The two sample t test is based on the assumption that each set o f observations is sampled from a population with a normal distribution, and that the variances o f the two populations are the same The F test or variance ratio test was used in situations when the differencies in variations o f the two groups were not known. Variance ratio is the ratio o f the sample variances or the square o f the ratio o f the sample standard deviations. The variance ratio observed in the sample is calculated by taking the larger standard deviation divided by the smaller and looks up the square o f this value in the table o f F distribution. The distribution o f the F statistic has two values o f degrees o f freedom, one corresponding to each variance. Because we take the ratio o f the larger variance to the smaller we consider only the upper tail o f the F distribution. The t test for equality o f means was used to compare the means o f parameters two independent groups of mosquito habitats; habitats with An. gambiae s.s. and those without but with other mosquito species. Confidence limit was set at 95 %. 3.2.6.2 Discriminant function analysis The linear discriminant function analysis (LDF) is useful where you want to build a predictive model o f group membership based on observed characteristics o f each University of Ghana http://ugspace.ug.edu.gh group. The procedure generates a discriminant function (or, for more than two groups, a set o f discriminant functions) based on linear combinations o f the predictor variables that provide the best discrimination between the groups. There are several purposes for LDF: i) To investigate differences between groups ii) To determine the most parsimonious way to distinguish between groups iii) To discard variables, which are little related to group distinctions iv) To classify cases into groups v) To test theory by observing whether cases are classified as predicted Basically, a discriminant function score is predicted from the sum o f the series o f the variables, each weighted by a coefficient. Therefore, there is one set o f discriminant function coefficients for the first discriminant function, a second set o f coefficients for the second discriminant function and so forth. Cases get separate discriminant function scores for each discriminant function when their scores on variables are inserted into the equation: D j = di]Xi + dj2X2 + d i3X3 + dinxn Where D is the standardised score on the ith discriminant function, x is the standardised score on each variable and dj, is the discriminant coefficient. Just as D j can be calculated for each specimen, a mean value o f D can be calculated for each group. Thus, members o f each group considered together have a mean score on a University of Ghana http://ugspace.ug.edu.gh discriminant function that is the distance o f the group in standard deviation units from the zero mean of the discriminant function. This is analogous to multiple regression, but the dj’s are discriminant coefficients which maximize the distance between the means o f the criterion (dependent) variable. Cases are also assigned into groups using the basic classification equation: Yj = bjo + bjiZ] + bj2Z2 + bjnZn Where Yj, is the score on the classification function for the group j, Zi is the raw score o f the variable and b, the associated classification function o f the variable, bjo, a constant This method then can be used to predict to which subgroup a new individual is likely to belong Discriminant function analysis is also used to select the variables most useful in predicting group membership if performed in a stepwise manner The discriminant function analysis was used to select the parameters, which were most useful in distinguishing the habitats o f An. gambiae and o f no An. gambiae Parameters were initially screened by univariate analysis. Those that were significant at a = 0.05 were included in multivariate analysis. The probability o f F to enter the model was 0 .05 and that o f removal was 0.051. 3.2.6.3 Multiple regression analysis The general purpose, o f multiple regression is to learn more about the relationship between several independent or predictor variables and a dependent or criterion variable. Multiple regression yields a regression model in which the dependent (or University of Ghana http://ugspace.ug.edu.gh outcome) variable is expressed as a combination o f the explanatory variables The statistical significance o f each variable in the multiple regression model is obtained simply by calculating the ratio o f the regression coefficient to its standard error and relating thi.s value to the t distribution with n-k-1 degrees o f freedom, where n is the sample size and k is the number o f variables in the model. The t statistic, which is calculated as b/std error, where b is the regression o f the coefficient, and is equal to the square root o f the F statistic for the extra variability explained by the present model in comparison with the model excluding that particular variable. The principle involves testing the significance o f particular regression coefficients and then applying logical approach to select the best set o f independent variables, where the best is to be interpreted as including all variables, which really affect the dependent variable and only those remain the same Thus, for three variables, xj, X2, and xj, the significance o f the effect jc7, is tested by comparing the residual sum o f squares for the regression on all-three variables and for the regression on X2 and X3 only. The difference between the residual mean square for the full model and the F-test is used. 4 Three methods can be employed in this analysis. The forward entry method selects the parameter with the .highest level o f significance first before any other The stepwise selection method enters the parameters in their order o f significance with the parameter having the highest level o f significance being entered first before any other. The backward elimination method removes the parameter with the least level o f significance completely out o f the model and does not consider it at all in any compilations. University of Ghana http://ugspace.ug.edu.gh For this study, multiple regression analysis was used to identify which combination of the parameters w^s significantly contributing to the observed variation in the proportion o f An. gambiae s.s. larval populations in the various habitats. For this analysis parameters that were significantly correlated (Pearson coefficient) to the proportion o f An. gambiae s.s. (oc = 0.05) were included in the multivariate analysis. 3.2.6.4 Hierarchical cluster analysis and phylogeny tree Hierarchical cluster analysis is a statistical method for finding relatively homogeneous clusters o f cases based on measured characteristics. It starts with each case in a separate cluster and then combines the clusters sequentially, reducing the number o f clusters at each step until only one cluster is left. When there are N cases, this involves N -l clustering steps, or fusions. This hierarchical clustering process can be represented as a tree, or dendrogram, where each step in the clustering process is illustrated by a join o f the tree Distance or similarity measures are generated by the Proximities procedure. Our choice o f clustering method or clustering criterion will determine the way in which the proximity between two clusters is measured. For example, using single linkage, the proximity between two clusters is the highest 'similarity (or smallest distance) between any two cases, one from each o f the clusters. It's their nearest neighbours. By contrast, with average linkage, the similarity between two clusters is the average o f the proximities between all pairs o f cases, one from each o f the two clusters. It is often recommended to optimize the Euclidean Sum of Squares which involves finding the mean o f each cluster and the distance from each case contained in each cluster and its mean, then squaring these distances and summing the squared distances for all the cases in all the clusters. It is University of Ghana http://ugspace.ug.edu.gh a measure o f the within-cluster variance, or the diversity o f a particular classification or cluster model. Phylogeny is the study o f the evolution o f life forms. It helps to understand the life through time— not just at one time in the past or present, but over long periods of past time. Before we can attempt to reconstruct the forms, functions, and lives o f once-living organisms, we have to place these organisms in context. The context o f evolutionary biology is phylogeny, the connections between all groups o f organisms as understood by ancestor/descendant relationships. Evolutionary history is typically represented by a phylogeny tree, a tree o f species with the root being the oldest common ancestor and the children o f a node being the species that evolved directly from that node. Each species in a set is represented by a set o f traits or character values _ For this study, hierarchical cluster analysis was used to assess the similarity between An. gambiae larval habitats. Phylogeny tree was used to assess the ancestor/descendant relationship which connects the An. gambiae s.s. populations collected from different habitats. Samples were firstly analysed together and after divided into group o f 100% An. gambiae and that o f the mixture with other species. Visual comparison between hierarchical cluster analysis and phylogeny tree was done 3.2.6.5 Microsatellite data The banding profile o f alleles o f each sample as observed on the gel was scored for the absence or presence o f alleles (bands) depending on DNA fragment sizes. University of Ghana http://ugspace.ug.edu.gh Alphabetic capital letters were used to denote different DNA fragment sizes Single bands were scored as homozygote and recorded as AA whereas bands appearing in pairs or in fours were scored as heterozygote and recorded as AB (Appendix IV). Missing bands were denoted by two dots ” Population genetics analysis software, » POPGENE, Version 1.31 for Microsoft Windows™ (Yeh ei a l 1999) was used to analyse the alphabetic diploid and co-dominant data. The output statistics o f this analysis included the genotypic frequency, differentiation indices (Fst), and heterozygosity deficiency or excess (F js), and allele frequency. F^ is an estimate o f inter-population genetic differentiation and was considered to be low if the differentiation index (Fst) was below 0.05 and high when it was greater than 0.05 (Wright, 1978). Positive F;s values indicate heterozygote deficiency while negative values indicate heterozygote excess (Lehmann eta l., 1997). University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR RESULTS Out o f 54 samples collected An. gambiae s.l. larvae were recorded alone at 5 (9.3%) sites whereas mixtures o f An. gambiae s.l. with other mosquito species were obtained in 25 (46.3%) sites. Habitats o f other species w i th o u t^ , gambiae s.l. were recorded at 24 (44.4%) sites. About 95 % of the larvae and pupae o f both Anopheles and other species that were reared in the insectary completed their development to adults. However, it was observed -that the development from the larval to the pupal stage for Anopheles species lasted 3 - 5 days (depending on the larval stage when collected) and adult emergence from pupae lasted 1 - 2 days. It was also observed that in the first batch of Anopheles mosquitoes that emerged there were predominantly more males than the subsequent batches, which had more females. A total o f 3,652 mosquitoes that were studied were morphologically identified as 2,735 (74.9%) An. gambiae s.l. and 917 (25 .1%) as other species. 4.1 Molecular Identification of Anopheles gambiae Species Complex Ten morphologically identified An. gambiae s.l. mosquitoes per site were studied and PCR identification (Scott et a l 1993) o f a total 300 specimens was carried out. PCR amplifications were successful in all cases and revealed only An. gambiae s.s. indicated by the diagnostic size o f the amplified DNA fragment which is 390 bp (figure 4). University of Ghana http://ugspace.ug.edu.gh 1 2 3 4 5 6 7 8 M + Figure 4: Ethidium bromide stained 2 % agarose gel electrophoresis o f PCR products obtained from the amplification o f An. gambiae DNA for species identification. Lanes 1 to 8 = An. gambiae s.s. diagnostic band size; M = 100 bp Molecular weight marker; negative control; “+”= positive control University of Ghana http://ugspace.ug.edu.gh 1 2 3 4 5 M 6 7 8 9 10 < — 10Obp Figure 5: Polyacrylamide gel electrophoresis of PCR products obtained from the amplification of An. gambiae s.s. microsatellite DNA with primer AGXH7. Lanes 1, 6 and 7 = Heterozygous set of four bands (AB ), Lanes 2, 5 and 8 = No bands Lanes 3. 4 and 9 = Heterozygous double band (AB), Lane 10 = Homozygous single band (AA), Lane M = 100 bp Molecular weight marker University of Ghana http://ugspace.ug.edu.gh A total o f 270 An. gambiae s.s. mosquitoes were studied by means o f microsatellite DNA analysis, using the AGXH7 oligonucleotide primer set designed to amplify a locus on the X chromosome (figure 5). Out o f the 270 specimens, 260 (96.3%) were successfully amplified on first attempt, 3 (1.1%) were successful on re-amplification and 7 (2.6%) that failed even after re-amplification was taken to be due to their possession o f null alleles. DNA bands that differed in sizes were taken to be distinct alleles. The microsatellite locus AGXH7 studied was found to be polymorphic. Ten loci were recorded for the overall populations and a total o f 12 alleles were also recorded and were not uniformly distributed in all the populations. The loci in 100% o f An. gambiae s.s. populations were found to be 70% polymorphic, whereas those from habitats with 1 to 99% of An. gambaie s.s. were 100% polymorphic. A total o f 10 alleles were recorded with the 100% of An. gambiae s.s population group whereas 12 alleles were recorded with the 1 to 99% o f An. gambiae s.s. population group. Three (6.0%) out o f the overall total o f 50 mosquitoes studied from the first group were heterozygous for 99 bp, 150 bp, 340 bp and 430 bp alleles, whilst 49 (22.3%) out o f overall total o f 220 mosquitoes studied from the second group were heterozygous for different band sizes. In addition 5 mosquitoes.were found to be heterozygous for the 99 bp allele. 4.2.1 Genotype frequency distributions As shown in Table 3, ten alleles were the most common for both groups. However, with the 100% An. gambiae s.s. population group, allele 1 was the most dominant 4.2 Genetic Structure of Anopheles gambiae s.s. Populations University of Ghana http://ugspace.ug.edu.gh (0 .433 ) and alleles 5 , 6 , 7 , 8 , and 9 were found to have the same allele frequency (0.056). With the 1 to 99% o f An. gambiae s.s. population group, alleles 1 (0.206) and 2 (0.208) were relatively dominant and were present in 17 out o f the 22 populations. In the pooled populations, alleles 1 (0.241) and 2 (0.195) were relatively dominant and were present in 21 out o f 27 populations. 4.2.2 Population differentiation index (Fst) The estimate o f inter-population genetic differentiation (F^) indicated a considerable differentiation between populations for both, the group o f 100% o f An. gambiae s.s. % populations and that o f 1 to 99% of An. gambiae s.s. (Table 4). Fs, was significant, being higher than 0.05 in both cases. Estimates o f Fa were high for the group o f 100% o f An. gambiae s.s. populations (Fst = 0.9385) than for the other group (Fst = .0.9132). A h igh, differentiation between populations also was observed when the analysis was done with the pooled populations (Fa = 0.8714). 4.2.3 Estimate of heterozygote deficiency and excess (F^) The h'eterozygote deficiency or excess which was calculated to determine heterozygote deficiency (positive F;s value) or excess (negative F is) for both groups indicated that all these groups had negative Fis values (Table 4). The estimated heterozygote excess for the two groups were - 0.2725 and - 0.8081 respectively. The estimated heterozygote excess for the pooled populations was - 0 .7 5 7 4 . University of Ghana http://ugspace.ug.edu.gh Ta bl e 3: Fr eq ue nc ie s of the m os t co m m on all ele s wi th the 10 0 % of An .g am bi ae s.s . po pu la tio n gr ou p, 1 to 99 % of An . ga m bi ae s.s . po pu la tio n University of Ghana http://ugspace.ug.edu.gh Table 4: Estimates o f differentiation (Fst) and heterozygosity (Fis) within populations o f An. gambiae s.s. Population Fst Fk 100 % o f An. gambiae s.s. 0.9385 - 0.2725 1 to,99 % o f An. gambiae s.s. 0.9132 - 0.8081 All populations 0.8714 - 0.7574 University of Ghana http://ugspace.ug.edu.gh 4.3 The Physico-chemical Parameters of Breeding Habitats The comparison o f the mean values o f the parameters o f habitats with Anopheles and non Anopheles revealed that 12 parameters; turbidity, pH, conductivity, suspended solids (SS), total dissolved solids (TDS), sodium, calcium, potassium, chloride, sulphate, total hardness and magnesium hardness significantly differed between the two groups. The values o f all these parameters were higher for the water o f An. gambiae s.s. habitats as shown in Table 5. University of Ghana http://ugspace.ug.edu.gh University of Ghana http://ugspace.ug.edu.gh 4.3.1 Parameters predicting the presence/absence of An. gam biae s.s. The results o f the stepwise linear discriminant function analysis revealed that pH, calcium and turbidity . were the best discriminatory parameters between presence/absence o f An. gambiae s.s. larval habitats (Table 6a). pH was first selected followed by calcium and then turbidity in that order. The derived classification function coefficients (Fisher’s linear discriminant functions) obtained are shown in Table 6b. 4.3.2 Parameters predicting variation in distribution of ,4m. gam biae s.s. populations The bivariate correlation procedure computed using Pearson’s correlation coefficient revealed that the proportions o f An. gambiae s.s. in larval habitats were significantly and positively correlated to turbidity (r = 0.515, p = 0.000) , pH (r = 0.429, p = 0.001), conductivity (r = 0.380, p = 0.005), total suspended solids (r = 0.384, p = 0.004), total dissolved solids (r = 0.443, p = 0.001), sodium (r = 0.310, p = 0.023), calcium (0.256, p = 0.048), magnesium (r = 0.272, p = 0.046), chloride (r = 0.363, p = 0.007), sulphate (r = 0.299, p = 0.028), total hardness (r = 0.269, p = 0.050) and magnesium hardness (r = 0.287, p = 0.036) (figure 6a-d). When multiple regression analysis was performed on the dataset o f significantly correlated parameters to select the best predictors o f proportion o f An. gambiae s.s. it revealed turbidity, pH and calcium as the best predictors in the model. These together accounted for 42.9% of the total variation observed (Table 7). All the three models that were applied selected turbidity, pH and calcium in that order. However, when University of Ghana http://ugspace.ug.edu.gh multiple regression analysis was performed using 28 parameters, only turbidity and dissolved oxygen were selected as best predictors associated with proportions o f An. gambiae s.s. (Table 8). The two parameters however accounted for only 36.7% o f the total variation. To determine which o f the non-selected parameters were either counting to or were associated w ith turbidity, pH and calcium,' multiple regression analyses were performed using these as dependent variables. The results obtained are shown in Tables 9a - 9c. The analysis revealed that suspended solids, sulphate, total dissolve solids, magnesium hardness, ammonium and biological oxygen demand were the best variables, together accounting for approximately 79% o f the variation existing in turbidity. Furthermore, all were significantly correlated to turbidity (P < 0.05 in all cases) However, .dissolved oxygen, total dissolved solids, silica and calcium were the best, variables associated with pH, together accounting for 47% o f the variation existing in pH whereas total hardness and suspended solids were the only variables ' r associated with calcium and accounting for 42% o f the observed variation. University of Ghana http://ugspace.ug.edu.gh Table 6a: Summary o f results o f stepwise discriminant function analysis performed on dataset set on the significant parameters to select the most discriminatory parameters separating habitats o f An. gambiae s.s. and those o f other mosquito species 6nly Step Number of parameters Exact F statistic W ilks’Lambda P value . 1 l a 14.752 0.779 0.000 2 2b 14.558 0.637 0.000 3. 3C 14.888 0.528 0.000 a pH ' b. pH and calcium c. pH,' calcium and turbidity Table 6b: Derived classification function coefficients by stepwise discriminant analysis for the separation o f presence/absence o f An. gambiae s.s. habitats (p value to enter'= 0.05) Habitats Parameters 'With A n. gambiae s.s. O ther species only . ' pH 28.353 25.906 Calcium 6.166E‘02 4 9 13e-°2 , Turbidity 4.61 7e"03 1.048e‘03 Constant - 118.962 - 97.966 University of Ghana http://ugspace.ug.edu.gh o o o o o o o O O =30 0 0 o o o o o o \ o i n m in - (l/Sui) spqos pO/V|OSSip |tnoi Fi gu re 6a -d : Ex am pl e of pl ot s sh ow in g re la tio ns hi ps be tw ee n the sig ni fic an tly en vi ro nm en ta l pa ra m et er s an d pr op or tio ns of An op he le s ga m bi ae s.s . in m os qu ito ha bi ta ts . Th e fit ted lin e is dr ow n in bo ld University of Ghana http://ugspace.ug.edu.gh Table 7: Summary o f results o f multiple regression analysis (p value to en ter — 0.05) using proportion o f An, gambiae s.s. as dependent variable and the significant correlated parameters as independents variables Model Correlation R • Adjusted Standard F P square R square error value value 1 0.515“ 0.265 0.251 37.4293 18.743 0.000 2 •0.597b 0.357 0.331 35.3619 14.128 0.000 ->J 0.6 5 5C 0 429 0.394 33.6536 12.502 0.000 a. Predictors: (constant), turbidity b Predictors: (constant), turbidity, pH c. Predictors:.(constant), turbidity, pH, and calcium Table 8: Summary of-results o f multiple regression analysis (p value to enter = 0.05) using proportion o f An. gambiae s.s. as dependent variable and all the measured parameters as the independents variables Model Correlation R Adjusted Standard square R square error F value P value l a 0.515 0.265 0.251 37.4293 18.743 0.000 2b Q.606 0.367 0.342 35.0817 14.764 0.000 a. Predictors; (constant), turbidity b Predictors: (constant), turbidity and dissolved oxygen University of Ghana http://ugspace.ug.edu.gh Table 9a: Summary o f results o f multiple regression analysis (p value to en ter = 0.05) using turbidity as dependent variable and all the m easured param eters as the independents variables Model Correlation R square Adjusted R square Standard error F value P value r . 0.728, 0.531 0.521 245.2238 58.757 0.000 2b ' 0.778 0.605 0.590 227.0838 39.079 0.000 3C 0.806 0.650 0.629 215.8864 30.968 0.000 4d 0.854 0.730 0.708 191.6289 33.093 0.000 5e 0.872 0.760 0.735 182.5416 30.376 0.000 6f 0.888 0.788 0.761 173.4506 29.064 0.000 a. Predictors: (constant), SS b. Predictors: (constant), SS and sulphate c. Predictors: (constant), SS, sulphate and TDS d. Predictors: (constant), SS, sulphate, TDS and Mg hardness e. Predictors: (constant), SS, sulphate, TDS, Mg hardness and ammonium f. Predictors: (constant)' SS,-sulphate, TDS, Mg hardness, ammonium and BOD University of Ghana http://ugspace.ug.edu.gh Table 9b: Summary o f results o f multiple regression analysis (p value to en ter - 0 .05) using pH as dependent variable and all the measured param eters as the independents variables Model Correlation R Adjusted Standard F P square R square error value value f ‘ 0.401 0.161 0.145 0.5878 9.964 0.003 2h . 0.549 0.301 0 273 0.5417 10.973 0.000 3C I 0/624 0.390 0.353 0.5111 10.647 0.000 4 d _ 0.685 ‘ 0.470 0.426 0.4814 10.842 0.000 a. Predictors: (constant), DO b. Predictors: (constant), DO and TDS c. Predictors: (constant), DO, TDS, and silica d. Predictors: (constant), DO, TDS, Silica, and Calcium Table 9c: Summary of results of multiple regression analysis (p value to enter using calcium as dependent variable and all the measured parameters independents variables = 0.05) as the Model Correlation R Adjusted Standard F P square R square error value value l a 0.568. 0.322 0.309 91.5114 24.717 0.000 2b . 0.650 0.423 0.400 85.2630 18.696 0.000 a. Predictors: (constant), total hardness b. Predictors: (constant), total hardnfess and suspended solids University of Ghana http://ugspace.ug.edu.gh 4.4 Association between larval habitat types and An. gambiae s.s. populations To determine if the characterised An. gambiae s.s populations were specifically associated with certain larval habitats, the phylogeny o f the different populations was constructed using the microsatellite DNA data and POPGENE software. Then, a phenogram was constructed to cluster together similar habitat types, using hierarchical cluster procedure in SPSS software version 10.0 (SPSS inc., USA) and environmental parameters that were significantly correlated to proportions o f An. gambiae s.s. Visual assessment o f the phylogenetic tree and the dendrogram did not i reveal any associatioa Then the two analyses were again performed using microsatellite DNA and environmental parameters o f habitats that harboured only An. gambiae s.s. In this case there was a perfect match (Figures 7a-b) and in both cases, the An. gambiae s.s. proportions and the habitat at one site, Legon village B was more distant (standing alone) from the rest. Table 10 shows the environmental parameters for the five sites where 100% of An: gambiae s.s. populations occurred and it shows that the closer values o f environmental parameters were clustering together especially, turbidity, pH and calcium as observed at Madina estate A and B sites. The values o f the environmental parameters' for Legon village B which was standing alone on both trees were lower when compared to those o f the rest o f this group. University of Ghana http://ugspace.ug.edu.gh Madina estate A ________________________________________Madina estate B ----------------------------------Korle-bu vegetable farm E ------------------------------------------------------------ Achimota ------------------------------------------------------------ Legon village B H Figure 7a Madina estate B Madina estate A Korle-bu vegetable farm E Achimota Legon village B Figure 7b Figure 7: Molecular phylogenetic trees obtained with microsatellite DNA o f An. gambiae s.s. o f pure populations (a) and hierarchical clustering (b) o f their habitats University of Ghana http://ugspace.ug.edu.gh a X O c: o C o D­ O uu-o O c cd O l5‘5OA C .2 J3 3 CL o o . C/i C/5 8 XC4— o Npox O o <4- O Q.. a o CD 13 £ cdu< cdCl 3C3 H •I *S B: Sc M 'P S « * H 0 H ■oUcd j= -w08 —3 3to oj *V• M _© 2 U OX) C3 U z c/r O H c/3 C/J L i ' 3 - a e o U a a £ ; a 15 3 H - C/i in CO in co ^ r oTf VO r - 00in o00 3 V = Volume (ml) o f NajCCb used A = Volume (ml) o f acid used t Standardization of 0.0141 M AgNC>3 with NaCl 2 drops o f potassium chromate indicator were added to 10ml o f standard sodium chloride (0.0141 M) into a flask titrated with the AgN03 solution to a pinkish yellow end point M = (0.014 x 10)/V Where, M = Molarity o f AgNCb V = ml o f AgN03 used in tritration Stannous chloride reagent 2.5g fresh SnClj.fibO were dissolved in 100ml glycerol, heated in water bath and stirred with a glass rod to hasten dissolution Starch solution 6g o f soluble starch were added to a small quantity o f distilled water and mixed. This mixture was added to 1000ml boiling water and allowed to boil for 5 minutes. The solution was left to stand overnight. The solution was preserved with 2 drops o f toluene c 6h 5c h 3). The solution was titrated with the acid to an orange end-point. The normality o f acid was obtained from (NxV )/A where: University of Ghana http://ugspace.ug.edu.gh Strong acid solution 300ml c o n c e n t r a te d H 2S 0 4 were slowly added to about 600ml o f distilled water. The solution was cooled and 4ml Nitric acid (H N 03) added and the volume made up to 1 litre. Sulphate solution (1000 mg/1 SO4) 1.47929g anhydrous Na2S 0 4 was dissolved in distilled water and diluted 1 litre mark into a volumetric flask. Sulphuric acid (H2SO4) 2.8 ml concentrated sulphuric acid, density 1.84 was diluted to 1 litre with deionised water and standardized against 0.1N sodium carbonate using methyl orange indicator. This solution was further diluted to 0.02N solution by taking 200ml aliquot o f the stock solution and topping to 1 litre The solution was stored in a covered glass bottle. Sulphuric acid reagent (silver sulphate in sulphuric acid) 11 g silver sulphate were dissolved to 1 litre conc.H2S 0 4 Zirconyl-acid reagent 133mg of Zirconylchlorideoctahydrate (Z r0C l2.8H20 ), were dissolved in about 25ml distilled water Then 350ml conc.HCl was added to the solution before making up the volume to 500ml with distilled water. University of Ghana http://ugspace.ug.edu.gh APPENDIX IV Input data format for the population genetics analysis software POPGENE Version 1.31 /*Diploid data o f 27 populations each with 10 loci */ Number o f populations = 27 Number o f loci = 1 0 Locus name: LI L2 L3 L4 L5 L6 L7 L 8 L9 L10 NAME = MADINA HANNAH STREET GG ..BC DE FF BC DE FF BC DE FF AB DE FF BC DE FF BB DD AB DE DD BC DD HH II JJ KK LL JJ 137 University of Ghana http://ugspace.ug.edu.gh APPENDIX V PHYSICO-CHEMICAL WATER PARAMETERS MEASURED Parameter Formula Units Value Notes WHO guideline Temperature °C Turbidity NTU 5 pH pH units 6 5 - 8 . 5 Conductivity . M-S/cm Suspended solids (SS) mg/1 Total dissolved solids (TDS) mg /1 1000 Sodium Na mg/1 20 0 Potassium K mg/1 30 Calcium Ca mg/1 20 0 Magnesium Mg mg/1 150 Total iron Fe mg/1 0.3 Ammonium NH 4-N mg/1 Chloride Cl mg/1 250 Sulphate S 0 4 mg/1 400 Phosphate P 0 4-P mg/1 Silica S i0 2 mg/1 Nitrite NO2-N mg/1 Nitrate NO3-N mg/1 10 Total hardness as CaCOi mg/1 500 Total alkalinity as CaCO? mg/1 Calcium hardness as CaCO , mg/1 Magnesium hardness as CaCO? mg/1 Fluoride F mg/1 1.5 Bicarbonate HCO^ mg/1 Carbonate CO , mg/1 Dissolved oxygen (DO) mg/1 dUD Pnn —---- ------------- mg/1 LUJJ mg/1 138 University of Ghana http://ugspace.ug.edu.gh