BIOLOGICAL CONIROL OF HUMAN SCHIS TOSOMIASIS USING SNAIL EATING FISH

by

Sosten Staphiel Chiotha

Dissertation subimitted to the Faculty of Graduate School of The University of Maryland in partial fulfillment of the req _rents for the degree cf Doctor of Philosophy 1990

Advisory ccxmittee: Professor Kenneth MoKaye, Chaiman/Advisor Professor Sam Joseph Associate Professor J. Edward Gates Associate Professor Raymond Morgan Professor Jay Stauffer

IIi ''APR 5 " C Copyright by Sosten Staphiel Chiotha 1990 ACKNOlqLEDGMEW~

I am grateful to the Goverrments of Malawi and United States of

America for the sponsorship to study for a Ph.D at the University of

Maryland, College Park. The fellowship was administered by the staff

of The African American Institute and I comend them for the

expertise in handling academic and welfare issues affecting students

in the programme. I am grateful to Prof. K. R. McKaye for playing a

major role as an advisor and for devoting considerable tire to the study. I am also grateful to the dissertatinn canittee nmbers,

Prof. Sam Joseph, Ass. Prof. J. Edward Gates, Ass. Prof. R. Morgan

II, and Prof. J. Stauffer for their assistance during planning and

execution of research and during writing of the dissertation. In the

University of Malawi, I wish to thank Dr. John Dubbey (Vice

Chancellor), Dr. Z. D. Kadzamira (Principal Chancellor College),

Prof. B. B. Chimphamba (Principal Bunda college), Dr. D. C.

Munthali, Dr. E. B. Khonga, Dr. E. Y. Sambo, Dr. T. Makhambera, Dr.

Phoya, Mr. J. Likongwe, and others for their assistance in working out the logistics of research in Malawi, Africa. I am particularly grateful to the Malawi Fisheries Department for kindly allowing the project to take place and the Malawi Bilharzia control Unit

(Ministry of Health) for helping in many ways. I thank Dr. E.

Van den berghe, Dr. S. Seagle, Dr. P. LaChance, Dr. R. Raesly, Dr.

G. Nelson, John Ypande, and Dave Brinker for their valuable ccmments on various drafts of the dissertation. I am especially indebted to

Prof. K. Fuller, A. Allen (Ms), P. Orndorff (Ms), Kirk, E. (Ms)

ii and other staff at the Appalachian Environmental Lab. for a number of services that created ai excellent atmosphere for Graduate study.

I am grateful to Dr. R. nzer, Dr. I. Strand, Mary-Ellen Atkinson

(Ms) and J. Wildberger (Ms) for an excellent job in co-ordinating studies in the NEES program. I am also grateful to Joan McKearnan for assistance in using the ccmputer and Dr. Lester Chitsulo for assistance in literature search. I wish to pay special tribute to

James J. Msosa (Late) who assisted me in transporting fish. I am particularly grateful to my wife Chiwa and daughter Mtisunge, 1y parents, brothers and sisters, in-laws, and W. W. Kanjo for their moral support. I wish to thank C. Jenya, S. Mponda, S. Bvalani, T.

Lovullo, A. Matambo and others who assisted in both the field and laboratory work. I wish, also, to thank J. Balarin and Dr. B.

Costa.-Pierce of ICLRARM (Malawi), D. Malikebu and the fish farmers in Zrba for assistance in several respects.

The research was funded by grants fran Agency for International Development, The Rockefeller Foundation and the International

Foundation for Science. At the recammendation of the Dissertation ccmiittee, information from a project Funded by the ICLARM-GTZ aquaculture project in Malawi has been incorporated in the dissertation.

iii TABLE OF CONTENTS Section Page

List of tables vi

List of figures viii

CHAPTER 1 NDLLUSCICIDES USED TO CONTROL SCHISIOSQIAS IS: REVIEW Introduction 1 Types of nolluscicides 2 Mode of action 5 Factors that influence toxicity 8 Toxicity to humans and other non-target organisms 11 Conclusion 13

CHAPTER 2 BIOLOGICAL CONTROL OF SCHISTOSCMIASIS USING SNAIL EATING FISH: REVIEW. Introduction 15 The parasites 17 Justification for control 21 Methods of control 24 Fish as biological control agents 25 Advantages of fish as biocontrol agents 27 Ecological constraints and caveats in the use of 32 fish

CHAPTER 3 PREY HANDLING IN Cyrtocara placodon: A SNAIL EATING FISH FROM MALAWI. Introduction 39 Materials and methods 40 Results 42 Discussion 45

CHAPTER 4 FOCD SELECTION IN TW SNAIL EATING FISH (C.placodon and Astatotilapia calliptera) FROM MALAWI AND ITS RELEVANCE TO BIOLOGICAL COIROL OF SCHIS ESOMASIS Introduction 58 Materials and methods 60 Results 63 Discussion 70

CHAPTER 5 THE INFLUENCE OF WEEDS AND SNAIL LOCATION ON THE POTENTIAL OF SNAIL EATING FISH (C. placodon and A. calliptera) FOR BIOCONTROL OF SCHISTOSOMIASIS Introduction 83 Materials and methods 84 Results 86 Discussion 90

iv CHAPTER 6 SCHISTOSOME VECTOR SNAIL CONTROL IN FISH PONDS IN MALAWI USING FISH Introduction 96 Materials and methods 97 Results 102 Discussion 117

CHAPTER 7 SUMMARY AND CONCLUDING REMARKS Water resource projects and regional expansion of schistoscniasis 121 Schistosomiasis control in inpounded waters of Malawi 125

APPENDIX 1 SNAM COLLECTION AND TRANSPORTATION 127

APPENDIX 2 FISH COLLECTION AND TRANSPORTATION 131

APPENDIX 3 ESTIMATION OF CERCARIAL DENSITY 134

REFEREhNES 135

v LIST OF TABLES

Table 1. Tim (minutes) taken for 10 molluscivores

(C.placodon) to start feeding on snails. 55

Table 2. Average nurber of snails eaten by C. placodon,

given a choice of two snail species. 66

Table 3. Average number of small and large snails

eaten by two molluscivores, C. placodon and

A. calliptera. 67

Table 4. Prey selection by A. calliptera and

C. placodon based on presence and absence

of snail shell for B. globosus. 68

Table 5. Average number of cercariae per liter in

aquaria stocked with C. placodon and A. calliptera. 69

Table 6. Frequency of C. placodon and A. calliptera

under two treatment levels of weed presence,

snail presence and snail location. 88

Table 7. The effect of weeds in aquaria on the

number of snailA consumed by C. placodon. 89 Table 8. Effect on snails of stocking experimental

ponds with C. anaphyrmis (molluscivore). 106 Table 9. Survival of C. placodon and 0. shiranus

(non-molluscivore) in experimental ponds. 107

Table 10. Survival of 0. shiranus in experimental

ponds, with and without C. placodon. 108

vi Table 11. Initial and final weight of 0. shiranus

in experimntal ponds . ill

Table 12. Weight gain by C. placodon and 0. shiranus

in experimental ponds. 114

Table 13. Initial and final snail population in rural fazners' ponds, with and without C. placodon. 115

Table 14. Pilot breeding trial of C. placodon in

experimental ponds. 116

vii LIST OF FIGURES .5 Fig. 1. Life cycle of schistosciasis 37 Fig. 2. External morphology of three snail species: 1.) Bulinus tropicus, 2.) Bulinus

globosus and 3.) Bicmphalaria pfeifferi

Fig. 3. Point load crusher for estimating static crushing resistance of snails 51

Fig. 4. Handling time in two C. placodon for B. globosus snails of different sizes 53

Fig. 5. Static crushing resistance of B. globosus, B. pfeifferi and B. tropicus

snails 56

Fig. 6. Final snail population in experimental ponds with and without C. placodon 104

Fig. 7. Initial and final ueight of C. placodon in experimental ponds 109

Fig. 8. Weight increase (%) of 0. shiranus in

experinental ponds with and without

C. placodon 112 129 Fig. 9. Map of Malawi

viii ABSTRACT Title of Dissertation: BIOLOGICAL CONTROL OF }KRAN SCHISTOSOMIASIS USING SNAIL-EATING FISH

Sosten Staphiel Chiotha, Doctor of Philosophy, 1990

Dissertation directed by: Kenneth R. McKaye, Professor, Marine Estuarine and Environmental Studies

Schistosomiasis, a debilitating disease affecting millions of

people in the tropics, is caused by flatworms (Schistosomna spp.).

Transmission involves water contact in habitats infested by vector

snails. Disease control by snail destruction with rolluscicides has

been attempted with variable success. Because nolluscicides are

costly and some are toxic to non-target organisms, particularly

fish, alternative control techniques are needed. This study was

conducted to test three cichlid fishes native to Malawi, Africa, for

biological control of schistosomiasis. The cichlids, Cyrtocara

placodon, Cytocara anaphyrmis and Astatotilapia calliptera, were

chosen based on earlier reports of snail-eating habits.

In laboratory experiments, Bulinus globosus, Biomphalaria

pfeifferi and Bulinus tropicus snails were used as prey; the first

two being vectors of humaj schistosciiasis. Choice experiments of

providing molluscivores with two snail species suggested no prey

species preference. Given single snail species of two size classes,

the molluscivores preferred small to large snails. Presented with

normal and crushed snails, the rolluscivores consumed more of the

latter. When introduced in aquaria under two treatments of

weeds/no weeds or snails/no snails, the molluscivores spent more

time in sections with weeds and snails respectively. Fewr snails

were consumed in aquaria with than without weeds. When snails were buried in sediment, the molluscivores did not detect the snails.

Alternative foods such as corn-bran and ccmercial aquarium cichlid food, except cercariae, were consumed by the molluscivores. Prey handling time in the molluscivores and static crushing resistance of snail shells, increased with snail size.

Field trials demonstrated dramatic snail reduction from e.,:perimental ponds and rural farmers' ponds stocked with C. placodon. C. placodon survived well in ponds, gained weight and appeared to co-exist with Tilapia rendalli, a tilapine aquaculture species. No snail reduction was observed with C. anaphyrmis in ponds and survival was poor. Implications of these results for biological control are discussed. C. placodon is a promising biocontrol agent. CHAPTER 1 ICIDES USED TO O1NTROL SCHIS'IOSO4IASIS

INnwrr"icN

Molluscicides are chemicals designed specifically to kill various types of molluscs (Cremlyn, 1978; Ritchie, 1973). Mollu:zcs affect man in several indirect ways. Some molluscs, namely snails and slugs, cause considerable crop damage by feeding on seedlings and mature plant leaves (Cremlyn, 1978). Such crops as autumn-sawn wheat

(Hartley and West, 1969) and tobacco in Malawi (Meredith, 1983) suffer severe attack by snails and slugs. The most serious problem from molluscs is the fact that several amphibious and aquatic snails are intermediate hosts of human and animal infections including schistosomiasis.

Schistoscmiasis is a debilitating disease (Rosenfield, 1979) caused by blood-dwelling trematodes of the genus Schistosoma

(Stirewalt, 1973). The vector snails for this disease belong to three genera, Bulinus, Biomphalaria and Oncomelania (Schmidt and

Roberts, 1985). Man is infected when the parasite penetrates the skin during contact with water.

One control strategy of schistoscmiasis involves depriving the parasites of the essential snail intermediate hosts (Farooq, 1973) primarily by applicatin of molluscicides (Bell et al., 1966). Snail control by molluscicides is a rapid and effective mans of reducing transmission (Thomas, 1973; WHO, 1980). Successful control with molluscicides has been achieved in several countries (Ritchie, 1973).

In Japan, application of molluscicides in the 1950's reduced the

1 infection rate from 73% to 2% (Hunter et al., 1962) and today schistosomiasis is of little concern in most areas of Japan (Ritchie,

1973). Similar success was also reported from a small-scale pilot control program in Malawi where significant reduction in prevalence was achieved by use of molluscicides and mass chemotherapy (Chitsulo,

1984). However, in a recent report, Teesdale (1986) is less optimistic on the feasibility of molluscicides and chemotherapy for large scale control in Malawi because of budgetary constraints and other limitations.

TYPES OF IJL=ICIDES

The number of compounds tested for molluscicidal activity is immense. In the period 1946-1955 about 7,000 campounds were screened in the laboratory and others were evaluated under field conditions

(Ritchie, 1973) and by 1966, the number exceeded 10,000 (Warren and

Weisberger, 1966). The classification of molluscicides varies widely in the literature. For example, the classifications may be according to chemical family (such as phenols), individual compounds (such as copper sulphate), availability, etc. Details about these categories are discussed fully in Ritchie (1973), WHO (1961, 1965, 1973, 1980), and Malek (1980). Only a few key groups and individual molluscicides are mentioned here. Copper ca~unds

These compounds were among the first to be used for controlling schistosomiasis (Ritchie, 1973). For example, copper sulphate has been used as a molluscicide since 1920, although its use is now

2 restricted to a few countries (Malek, 1980). Other copper compounds are copper carbonate and cuprous oxide. The active component of these campounds is the Cu II ion (WHO, 1973; Paulini, 1974).

Phenolic comdounds This group constitutes a number of compounds with varying degrees of molluscicidal activity. After studying over 7,000 phenols, Nolan et al. (1953) were able to establish a relationship between the chemical structure of these ccopounds and molluscicidal activity. For example, they showed that the most active compounds were the highly halogenated phenols, the dinitroalkyl phenols and the aryl phenols. However, the only compound that was ever used in actual control wrk was pentachlorophenol (in the salt fonrulation as sodium pentachlorophenate or NaPCP) (Nolan et al., 1953; WHO, 1961; Malek, 1980).

Carba tes

The interest in this group partly sterred fram the fact that it contains a number of effective insecticides. Hence, there was a possibility of using these compounds in ccmbined control of insect

-vectors for diseases like malaria and onchocerciasis as well as the snail vectors (Ritchie, 1973). Carbamates tried for snail control are Sevin, Zectran, Cuprobam and Rhodiacid, but they have not been used much in control wrk (Ritchie, 1973). argano-tis

Several organo-tins have been kn to be effective agricultural

fungicides and are major coiponents of marine and antifoulant paints,

stabilizers for food packaging, and anthelminthic (Malek, 1980). A

3 few are known to have molluscicidal properties. These are bis (N­ tributyl-tin) oxide, tri-n-butyltin acetate, and tri-n-propylin oxide

(Malek, 1980). Due to insufficient research on this class of ccn-ounds, their use is not recommended any more (WHO, 1980). Molluscicides of plant origin

Some plants have chemical adaptations for resisting attack by insects and other herbivores (Thomas, 1973). Examples of such chemical compounds are alkaloids, tannins, cyanoganic glycosides, saponins and others (Harbourne, 1972). Many of these chemicals have successfully been develope; to control insects. Pyrethrm is one example (Thomas, 1973). Attempts are now being made to search for molluscicides of plant origin. Pranising results have been reported on extracts frao the fruits of Phytolacca dodecandra (Lemma, 1970), and from root bark of Xerophis obovata and Psorosperum febrifuqum

(Chiotha and sonthi, 1986). The active principal in P. dodecandra is a triterpene glycoside (WHO, 1973). It is now believed that a

ccaplex ester (molecular weight of 360) isolated from pond water and

shown to be toxic to snails (Berrie and Viser, 1963) was probably of

plant origin (Thomas, 197i3). -A screening process is going on in

several countries (Malek, 1980), but it will take time before plant

molluscicides become available for wide field application.

Others Some of the commonly used molluscicides are Yurimin (3, 5­

dibromo-4 '-nitrobenzene), Frescon (N-tritylmorpholine), and

Bayluscide (5, 2" dichloro-4-nitrosalicylanilide). These are

discussed fully in Ritchie (1973), WHO (1973), and Malek (1980).

4 MODE OF ACrICN

Molluscicides were applied to snail habitats before the mechanisms by which they killed snails were known (Berry et al. 1950;

Cheng and Sullivan, 1974). Like other pesticides, much of the early research was devoted to solving practical problems associated with the formulation and application (Wilkinson, 1976). However, increasing demand for more and better pesticides as well as concern over pollution may have encouraged research into their mode of action. Some of the mechanisms by which olluscicides affect snails are as follows: uncoupling of oxidative phosphorylation, inhibition of oxidative phosphrylation, inhibition of acetylcholinesterase, inhibition of protein synthesis, hemorrhage, desiccation and irritation.

Uncoupling of oxidative phosprylation

Under normal conditions, the mitochondrial respiratory chain is linked to and controlled by the reactions associated with oxidative phosphorylation (Fukami, 1976). Studies have shown that the addition

of a respiratory substrate to an intact ("coupled") initochondrial­ preparation increases the rate of respiration (state 3) until the

supply of ADP and Pi is exhausted through conversion to ATP (Fukami,

1976). Thus, oxidative phosphorylation is controlled by supply and

demand in that when the energy requirements of the cell decrease,

fewer molecules of ATP are used, fewer molecules of ADP become

available, and electron flow is decreased (Iehninger, 1978). In the

presence of uncoupling agents, the close coupling between the

respiratory chain and phosphorylation is broken (Fukami, 1976).

5 Oxidation continues because electron transport along the chain still occurs but at a maximal rate and without formation of KTP (Lehninger,

1978). Thus the energy, made available by oxidation is not converted into active phosphate, but is expended as heat (Finkel, 1982).

Therefore, the uncoupling agents cause respiratory control to be lost.

Generally, uncoupling agents contain both anionic and lipophylic groups in their structure as in some nitrophenol analogues, halogenated compounds, antibiotics, and unsaturated fatty acids

(Fukami, 1976). Of the nitrophenols, two important molluscicides are dinitro-o-cresol (DNOC) and dinitro-cyclohexyl phenol (Dimex), but their use is limited (Cremlyn, 1978). For the halogenated uncouplers, the best known is NaPCP, a common and effective molluscicide (Olivier and Haskins, 1960; Ritchie, 1973). Its uncoupling properties are only at concentration below 20 mg/l

(Olivier and Haskins, 1960). Inhibitory effects

Moliuscicides could act by inhibiting several processes.

Oxidative phosphorylation is one process. Here the stimulation of

oxygen consumption by ADP and phosphorylation of ADP to ATP are both

prevented (Lehnxiger, 1978). This has been demonstrated with high

concentrations (' 20 ppm) of NaPCP (Weinbach and Nolan, 1956).

Similarly, Cheng and Sullivan (1974) demonstrated that copper

molluscicides inhibited oxygen consumption and also lcered heart

rate. However, Cheng and Sullivan (1974) were not able to determine

6 if the snails died from asphyxiation. These researchers also demonstrated disruption of the surface of rectal ridge by copper.

A second process that suffers inhibition is protein synthesis which has been demonstrated with chloramphenicol (also an

antibiotic). This compound preven-ts attachment of messenger RNA to

the ribosome, the intracellular site of protein synthesis (Weisberger

and Wolfe, 1964). This was confirmed by Warren and Weisberger (1966)

who demonstrated that chloramphenicol inhibited growth of snails and

caused significant reduction in egg laying. These researchers also

demonstrated an added advantage of this molluscicide in that it had a

direct effect on the parasite by inhibiting its development in the

snail.

Another inhibitory effect due to molluscicides concerns

acetylcholinesterase. Acetylcholinesterase (AchE) is one of the

enzymes (hydrolases) which split subtrates by the introduction of the

elements oi water. Its function is to catalyze the split of

acetylcholine to acetate and choline. Carbamate iolluscicides such

as Zectran react with AChE (O'Brien, 1976). The mechanism is

analogous to that of true substrates except that the low K3 rates of

carbamates tie up so much of the enzyme, thereby preventing

hydrolysis of true substrates such as acetylcholine (O'Brien, 1976). Exactly how this inhibition hanns snails is not known.

Desiccation, hemorrhage and irritation

When snails and slugs are exposed to compounds such as

metaldehyde, increased secretion of mucus is observed (Cremylyn,

1978). The result is inobilization and eventual loss of water;

-7 death is considered to be due to the loss of water (Cremylyn, 1978).

How the chemical acts to induce the above changes is not known, but reports suggest that molluscicidal activity is specific to the metaldehyde (1)while the trimer (paraldehyde with a six membered ring) is not active and neither is the monomer (acetaldehyde)

(Hartley and West, 1969). The compound is particularly effective

against slugs, but not as good against -vectorsof schistoscmiasis.

A readily observable effect of saoe molluscicides is that they

cause hemorrhage and heforrhagic inflammation (Visser, 1964). Both

synthetic and plant molluscicides affect snails in this manner (Chiotha and Nsonthi, 1986).

Som molluscicides appear to irritate the snails and c9use them

to leave the poisoned water. For example when NaPCP was applied to

some streams, the snails were seen to leave the water by climbing

on to vegetation (Ritchie, 1973; WHO, 1961). Similar observations

were made 'hen snails were exposed to plant extracts of T. vogelli.

Here the snails climbed out of the test solution and stopped on the

edge of the beaker (Chiotha and Msonthi, 1986). The significance of irritation to toxicity is not known, but it may probably induce

hemorrhage and excess mucus production and eventually lead to death as discussed above.

FACTORS = INFLMUENE TXICITY

A number of factors related to the chemical nature of the

mlluscicide are important determinants of toxicity. The structure

is one of them. Small changes in the structure of the chemical may

8 For example the alter its toxicity (Rand and Petrocelli, 1985). a phenol ring leads to introduction of a second hydroxyl group into activity (Nolan et al., an increase or decrease in molluscicidal (where activity is 1953). These results may be due to synergism reduced) (Nolan et al., increased) or antagonism (where activity is

1958; Rand and Petrocelli, 1985). species (Rand and Toxic effects will also depend on the target Bulinus globosus are less Petrocelli, 1985). For example, adult mg/l) while Biomphalaria susceptible to Niclosamide (LC 50 - 0.27 (IC 50 - 0.15 mg/l) pfeifferi are more susceptible to this chemical adult (WHO, 1973). Differences have also been observed between the molluscicide NaPCP is snails, juveniles and eggs. For example, 1973; WHO, highly toxic to snail eggs but less so to adults (Ritchie, are equally 1980). Other molluscicides such as N-tritymorpholine

effective against all stages (Malek, 1980). and Winkler (1962) An interesting anomaly appeared when Chi sulphate to eggs of compared the toxicity of NaPCP and copper The results indicated Onccrmlania at different stages of incubation. to newly laid eggs that NaPCP was more toxic than copper sulphate the eggs of while for older eggs the reverse was true. Apparently, of mucus and mud at Onccmelania became enclosed in a capsule composed the entry of the a later stage. Probably this capsule prevented molecules of copper large molecules of NaPCP while the smaller

sulphate ware not (Chi and Winkler, 1962). differential If uptake of a chemical is successful, then the molluscicide. The toxicity may depend on the fate of

9 increases or decreases nolluscicide could be altered in a way that excreted or accumulated toxicity. The molluscicides could be rapidly NaPCP, for example, at (Wilkinson, 1976; Brown, 1978). does not kill the adult snails but concentrations lower than 0.1 ppm, number of eggs laid and the has long term effects by reducing the and Haskins, 1960). viability of the few eggs laid (Olivier was removed after a few However, in studies where the chemical recovery of egg production and days' exposure, there was some the suppression of fecundity improved egg viability, indicating that low concentrations. is not permanent in all snails at these toxicity. For example, Several environmental factors affect spp.) use molluscicides as a some aquatic bacteria (Pseudomonas the toxicity of the source of nitrogen and this reduces Bayluscide, Niclosamide and N­ molluscicides. Molluscicides like detoxification (Bell et al., tritylmorphorine suffer this kind of have also been found in the 1966; WHO, 1973). These bacteria does not appear to be any digestive tract of the snails, but there

symbiotic relationship (Bell et al., 1966). Trifenmorph is Exposure to light detxifies some molluscicides. methanol which is not dehydrolysed in water to triphenyl hydrolysis is accelerated by molluscicidal (Amin et al., 1976). This in the field is 25-30 hours sunlight. The half life of the chenical laboratory (70 hours) (WHO, which is less than that measured in the life is also affected by 1973). Apart from sunlight, the half

suspended silt (Malek, 1980).

10 of The pH of the water is another factor affecting the toxicity molluscicides. For example, ICI 24223 (Isobutyl-triphenyl­ 4 nitro] methylamine) and Bayer 73 (5-chlorosalicylic acid [2 chloro, against snails aniline) were recommended for their effectiveness that they were (WHO, 1961), but suffered one major setback in (Fox et al., detoxified in waters with high acidity or alkalinity found in 1963). This creates a serious problem since snails are waters of varying pH levels. In Brazil, for example, snails inhabit Fox et waters with a pH range of 4-9 (Milward de andrade, 1954, in would be al., 1963). This rans that large amounts of molluscicide needed in these extreme pH conditions (Fox et al., 1963).

TOXICITY TO HANS AND OER NON-TARGET ORGANISMS Almost all pesticides are toxic not only to the target species, true but also to non-target organisms (Paperna, 1980). This holds same for molluscicides. WHO reports (1961; 1973) indicate that and molluscicides such as NaPCP, Yurimin, N-tritylmorphorine to Niclosamide are strongly piscicidal at the concentrations used 1964) kill snails. This was confirmed by reports from Egypt (Berg,

and Japan (Ritchie, 1973) about mass fish kills due te synthetic

molluscicides. Plant molluscicides are also piscicidal (Ransford, and this is 1948; Chiotha and Msonthi, 1986; Chiotha et al. in prep.) why some plant molluscicides have traditionally been widely used in the Malawi as a fishing technique (Seyani and Chiotha in prep). On good other hand some molluscicides are phytotoxic. This is a

property if they are applied to natural waters because the plants

11 upon which the snails depend for food, oviposition (Brown, 1978;

Thomas, 1973) and protection from predators (Louda et al., 1985) are rmoved. However, this might cause problems if the mlluscicides are applied to irrigation canals (Ritchie, 1973). As a matter of fact

NaPCP was found to destroy cultivated crops at concentrations used to kill snails during the early 1960s (Gonnert, 1961), but there have not been any similar reports since (Ritchie, 1973). This is possibly due to the fact that very few of the commonly used nolluscicides are phytotoxic (WHO, 1973).

Several invertebrates may be killed by molluscicides. In experimental work to investigate this effect on aquatic invertebrates, it was found that some molluscicides (e.g. NaPCP, copper sulphate) rapidly killed cladocera, copepods, diatoms, and insect larvae while others (e.g. N-tritylmorpholine) had no significant effect (WHO, 1973).

Interestingly, very little is known about the effects of molluscicides on humans. Early reports were contradictory. For example, people wading in water to which a plant extract (T. vogelli) had been applied ccrplained of numbness of the legs and roughness of the skin (Dalziel, 1937). However, Ransford (1948) was not able to confirm paresthesia after wading in poisoned water nor any toxic effects in four Malawian volunteers who drank 1/2 pint each of the plant extract. However, both authors were in agreement that the people who ate the fish killed by the molluscicide did not suffer any ill effects. Similarly, no ill effects were reported in Egyptians

(Berg, 1964) and Japanese (Ritchie, 1973) who ate fish killed by

12 mentioned the ynthetic molluscicides. None of these reports ossibility of long term effects. the molluscicides Recent evidence indicates that the majority of used for killing ire not toxic to humans at the concentrations the amounts to be mails. However, care is needed in formulating these chemicals are applied, realizing that the water to which There should also be applied could be used for drinking (WHO, 1973). like NaPCP dust, may :are in handling these chemicals because scme, solution (- 1%) may cause -ause skin irritation (Malek, 1980) and in lermatitis (WHO, 1961; 1973).

MUMIM is one There is general agreement that the use of nwlluscicides the transmission of effective way of quickly interrupting a major debilitating schistosomiasis (Shiff, 1961; WHO, 1973), 1979, Garfield 1985). parasitic infection in the tropics (Rosenfield there is no better Some authors even suggest that at present is far from alternative (Ritchie, 1973). But this conclusion put a strain on the satisfactory. Molluscicides are expensive, and is endemic. As a budgets of developing countries, where the problem has slowed matter of fact, research into new and better molluscicides do not see any down in recent times because pharmaceutical companies (Malek, 1980). A prospective economic returns from such investments are toxic to non­ more serious concern is the fact that molluscicides in Africa target organisms (Paperna, 1980). Mass fish kills reported (Ritchie, and Japan when olluscicides were applied are an example

13 1973; Berg, 1964). In Egypt, fishenmn refused to have their fishing grounds sprayed with mlluscicides (Weil and Kvale, 1986). The irony is that if snails get killed by molluscicides the people may be free fran schistosomiasis, but they may also suffer from malnutrition obvious due to fish distruction (debont and debont Hers 1952). Less is the toxicity of molluscicides to microscopic and submicroscopic seriously organisms in the water which may also be poisoned and thus It appears affecting the food chains (Berg, 1964; Ritchie, 1973). man has not had any serious direct poisoning from molluscicides in the (Malek, 1980). But, as it happened with malathion used there is control of malaria where over 2,000 people were poisoned, always a risk with any pesticide (WHO, 1973). Chemical impurities,

failure to take precautions, faulty fornrulation and application, and accidental exposure may have serious consequences. The good news, however, is that some molluscicides are rapidly detoxified by sce bacteria (Fox et al., 1963). Furthermore, several environmental

factors such as pH, silt, vegetation, sunlight etc. reduce toxicity

of the mlluscicides in the environment (WHO, 1973; Malek, 1980).

14 -EPTM CHAFTER 2 BIOLOGICAL CXNT1R)L OF SCHIST ISWIS 'USING S

FISH FRM IAKE MAAWI.

fl X'TION caused by blood dwelling Schistosmiasis, a debilitating disease 1979), is a major flat worms of the genus Schistoscma (Rosenfield, and the subtropics (WHO, public health problem in the tropics around the Great Lakes of 1965). This disease probably originated to other parts of Africa, West Africa, and eventually spread and Webe, 1969). Schistosomiasis Indies, and South America (Jordan among parasitic diseases, it is ni' occurs in 74 countries; and, of human sickness (Noble and second only to malaria as a cause estimates put the number of Noble, 1982; Ukoli, 1984). Current (Mott, 1984). Another 600 million people infected at 200 million as the disease continues to are at a constant risk of infection

spread to new areas (Jordan et al., 1980). in Malawi was recognized The seriousness of schistosomiasis Dye, 1924; Cullinam, 1945; over sixty years ago (Conran, 1913; (Teesdale and Chitsulo, 1983) Ransford, 1948), and recent studies is increasing. All indicate that the incidence of schistosomiasis disease have reported high past and present studies of this of Lake Malawi's shoreline. infection rates among the inhabitants of the lake are the major Apparently, the swamps on the fringes 1937). Here vector snails are sites of transmission (Gospil, the open water where habitat relatively abundant when compared to survival (Teesdale and Chitsulo, conditions are unfavorable for snail

15 1983). Similar observations have been made in the Lake Chilwa area, southern Malawi (Morgan, 1979; Cantrell, 1981). Snails occurring in weed beds appear to have a refuge from predators (Louda et al.,

1985).

The presence of snail eating fish (molluscivores) in Malawi has been reported in the past by several wrkers (Bertram et al.,

1942; Jackson et al. 1963; Fryer and Iles, 1972). More recently these molluscivorous cichlids have been hypothesized to be responsible for preventing the thin-shelled bilharzia vector snail, Bulinus globosus (Morelet) from successfully invading the open regions of Lake Malawi (McKaye et al.,1986). A test of this hypothesis was made by excluding molluscivores from open sand habitat and providing snails with a caged refuge. Under such

experimental conditions the dpnsity of snails increased, between 40­

60% within a week. Furthermore, the Lake Malawi molluscivores

disproportionately consumed snails of the genus Bulinus relative to

the more heavily armored genus, Melanoides spp. When given a

choice of snails under laboratory conditions, the Lake Malawi

cichlid Cyrtocara lacodon (Regan), selectively fed upon the thin

shelled bilharzia vector, Bulinus globosus (McKaye et al., 1986).

These results could have applicability beyond a better

understanding of the ecology of Lake Malawi. If these fishes prove

to be effective in eliminating bilharzia vectors within a natural

ecosystem, they could be successfully utilized within aquaculture

ponds to eliminate the vector snails that often invade these

ponds. With over twenty native fish species available that feed on

16 possibility of utilizing these snails, research exploring the disease in man-made water bodies Malawian fishes to control this promise. Sam of the snail eating seems practical and may hold much Clarias ellncji-, Cyrtocara fish from Malawi are: Barbus crystmus, Cytoara a na is (Bertram incola, Cytocara mola, Cytocara kirk, 1963; Fryer and Iles, 1972). et al., 1942; Jackson et al.,

TE PARASITES of harboring trematodes of the Schistosomiasis is a condition name bilharzia (bilharziasis) is genus Schistosma. The other popular the parasite in 1851 (Mott, named after Bilharz who discovered medical concern are Schistosoma 1984). The important species of (Bilharz), and Schistoscma mansoni (Sambon), Schistosama haenatobium 1986). Malawi is endemic for both japonicum (Katsurada) (Garfield, the former is the most dcinant S. mansoni and S. hainatobium, and of of less medical importance or (Ransford, 1948). For species Belding (1965), WHO (1967), and veterinary importance, refer to

Manson-Bahr and Apted (1983). though hermaphroditism has Adult schistosomes are dioecious, in 1924). Females are often found been reported occasionally (Faust, where copulation takes place the gynaecophoric canals of males normally inhabit the victim's (Muller, 1975). The paired worms bladder (S. haematobium) or the veins that drain blood from the and S. japonicum) (Choudhry and intestinal lining (S. mansoni

Teesdale, 1982).

17 Life cycle

The different species of schistoscres that infect humans have similar life cycles (Barbour, 1982) (Fig. 1). Each adult female produces between 100 and 3,500 eggs per day, with S. mansoni producing the fewest (Stirewalt, i973). A large number of the eggs leave the host either through urine (S. haematobium) or feces

(S. mansoni and S. japonicum). When eggs are deposited in a suitable environment, they hatch into swimming larvae known as miracidia (Jordan and Webbe, 1969; Stirewalt, 1973). The miracidia are positively phototropic and negatively geotropic. Such responses bring them to the surface where a large proportion of snail intexnediate hosts are found (Jordan and Webbe, 1969). Nevertheless, miracidia can also infect snails that live at the bottom (Brown,

1980). In this case, the miracidia irny be attracted by chemicals produced by the snail host (YcInnis, 1965; Brown, 1980).

Although the miracidia can home in on the snails, the attraction is not strong enough and, therefore, large numbers of snails are needed to ensure that the miracidia get close enough to snails to be attracted (Shiff and Kriel, 1970; Clarke, 1977). In the snail, the narasites undergo two asexual multiplication stages (first and second generation sporocysts) giving rise to cercariae which are shed into the water. Human infection takes place upon contact with the water and cercariae subsequently penetrate the skin. In about a month after cercarial penetration, the worms start laying eggs.

Individual worms can survive up to twenty years, although the

18 and Jordan, average life span is between 3 and 5 years (Goddard

1980).

Snail intermediate hosts hosts of Particular species of snails serve as intermediate the vector schistosomes. For both S. mansoni and S. haematobium, Planorbidae (WHO, snailr inhabit fresh water and belong to the family to two subgenera, 1965). Those transmitting S. haematobiun belong 1984). Bulinus (Physopsis) and Bulinus (Bulinus) (Ukoli, in Africa and Approximately 300 species of Bulinus are known to exist few transmit the islands in the Indian Ocean, but only a for example, S. schistosomiasis (Brown, 1980). In Malawi, (Physopsis) haematobium is exclusively transmitted by Bulinus 1983). The globosus (Morelet) (Fig. 2) (Teesdale and Chitsulo, nodules characteristic shell of B. globosus shows spirally arranged of 22.5 x 14 or irregular corrugations and reaches a maximum size mm (Brown, 1980). one genus, The intermediate hosts of S. mansoni, belong to known (Brown, Bicrphalaria (Garfield, 1986), of which 4 species are x 5.2 mm, (Fig. 1980). Biciphalaria pfeifferi (Krauss), averaging 13 (Teesdale 2) is the most important intenrediate host in Malawi

and Chitsulo, 1983) and in Africa generally (Brown, 1980). family For S. japonicum, the intermediate hosts belong to the or potential) Hydrobidae, subfamily Pomastinae. All known (actual single genus, intermediate hosts of S. japonicum belong to a (Weil and Oncomelania (Wright, 1973). These snails are amphibious dry season. Kvale, 1985) and are not found anywhere with a distinct

19 are aquatic and prefer Both Bulinus and Biomphalaria snails for food, shelter, shallow water, owing to the favorable conditions (WHO, 1965). The span of their and oviposition near the surface from the humid tropics to distribution is quite broad, ranging Despite this general desert oases (Weil and Kvale, 1985). have been observed. For similarity, fine ecological variations prefers streams, seepages, and a example, Biomphalaria pfeifferi irrigation channels, and variety of man-made water bodies (including in seasonal pools or large swanps swinming pools). It is not found 1980). Within genus variation favored by Bulinus globosus (Brown, the vector snails. For example, has also been demonstrated among waters with rich aquatic B. globosus is ccomon in slowly flowing (Krauss) lives in a variety of vegetation, while Bulinus africanus plants (Brown, 1980). water bodies with or without higher by the snail vectors have These restrictive habitat preferences pattern of schistosomiasis in probably influenced the geographical in a region may create Malawi. However, ecological changes never existed before. As conditions favorable for a vector that a waterproject is initiated, a Smith (1977) has argued, whenever Furthermore such changes potential bilharzia problem is created. vector sril at the expense of could tip the balance in favor of one Delta where irrigation projects another. This occurred in the Nile the expense of Bulinus snails favored Biomphalaria snails at

(Garfield, 1986).

20 JUSTIFICATION FR CxIFM5 of the disease Morbidity and possible regional expansion high. The prevalence of this disease in sam endemic areas is "must rank as Surveys of Malawi in the 1940's reported that Malawi world for the infection" one of the worst countries in the is little changed (Ransford, 1948) and today the situation to new (Teesdale and Chitsulo , 1983). Also the disease is spreading widespread than the areas, because the snail vectors are more these snails occur is a parasites. Each disease free area in which infected migrants arrive potential new focus of schistosomiasis, when

(Weil and Kvale, 1985). (Lovett-Campbell, Schistosamiasis is a debilitating disease damage caused by the 1948; Rosenfield, 1979), with the greatest host (Manson-Bahr and proportion of eggs that fail to leave the infections, the Apted, 1983). In the case of S. haematobium resulting in increased kidneys, ureters, and bladder are affected at night (nocturition) or rate of urination (micturition) especially and even renal failure total loss of bladder control (incontinence), between schistosmiasis (Choudhry and Tesdale,1982). An association Bullough (1975), but and infertility in Malawi was reported by causes of infertility Wright et al. (1982) have argued that other

may be ore important. countries, most In Egypt, as well as other Middle Eastern cell carcinoma and bladder cancer is of the type known as squamous have sevre and repeated occurs almost exclusively in patients who 1985). How this comes urinary schistosomiasis (Sherrif and Ibrahim,

21 about is not known but neoplastic transformation of the urothelium in high risk areas could conceivably be initiated by low doses of a local environmental or endogenous carcinogen (Hicks, et al. 1977).

The transformation could be promoted to tumor growth by the irritant The effect of schistoscme ova on the bladder (Hicks et al. 1977). to be high prevalence of S. haematobium in Malawi is believed responsible for higher rates of bladder cancer compared to other parts of East and Central Africa (Lucas, 1982; Wright et al.,

1982).

Infection with S. mansoni and S. japonicum usually causes damage to the intestinal lining and also enlargement of the liver and spleen (Muller, 1975; Noble and Noble, 1982; Manson-Bahr and

Apted, 1983). Ransford (1948) attributed cases of splencniegaly in Northern Malawi to high intensity of infection and frequency of

reinfection with schistosomiasis. Today, such symptcms are still

encountered in some patients, but it is also realized that the

splenomegaly may be due to both schistosomiasis and malaria

(Molyneux et al., 1979).

The parasitic effects ary, such that, those infected might not

show symptas (Rosenfield, 1979). In Malawi, the infected often

realized that they had not been well only after being treated

(Chitsulo, 1984). Schistosomiasis, nevertheless, can cause

considerable pathological changes, in a comparatively large section

of the population, but only a fraction of those infected die of

the disease (Jordan and Webbe, 1969; Foster, 1967). Similarly,

gynecological lesions observed in same Malawian patients infected

22 with schistosomiasis are unlikely to cause death directly, but definitely add considerably to the overall morbidity of schistosmiasis (Wright et al., 1982).

Economic importarce

The most direct cost to society fram schistoscrniasis is the money spent on treatment and other control inputs. In Malawi, for example, the annual direct expenditure was estimated at $0.75 per infected person per annum (Teesdale and Chitsulo, 1983), but now the cost is likely to be much higher than the 1983 estimate. The indirect effects on a country's economy can be even more costly because weakened manpower, in a community with high disease incidence, must be of socioeconomic significance (Stockard, 1978).

A study conducted in Sudan, for exanple, indicated that the blood haemoglobin levels of the infected had been reduced to the extent that oxygen flow to the muscles and the brain was limited, thereby impairing physical activity (Rosenfield, 1979). Above all, a 16­

18% work reduction capacity was found in inf cted Sudanese workers

(El Karim et al., 1980). While there was no direct difference in productivity in East Africa, between infected and uninfected workers on an irrigation scheme, the infected had a higher rate of absenteeism (Foster, 1967).

Many other indirect effects are probable, but can be difficult to measure and quantify. One example is the manner in which

schistosamiasis affects scholastic ability (Castle et al., 1948).

Measuring directly the effect of schistosomiasis on academic

performance and adequately controlling for other factors is not

23 easy. In his report of 1950, the Director of medical services in

Malawi suggested that schistosomiasis had a tremendous negative effect on the educational achievent .of infected children (Blair,

1956). This conclusion was possibly arrived at from intuitive observation rather than from controlled experimentation. Nevertheless, a study in Tanzania demonstrated that pupils improved their class performance after treatment against schistosomiasis (Jordan and Randall, 1962).

METHODS OF CONTROL

In prehistoric times, long before the mode of infection was known, various preventive measures were proposed. For example, the ancient Egyptians and the Zulus of South Africa believed that the parasite entered through the penile orifice and hence reccrmended that a penile sheath be worn before bathing (Farooq, 1973). In

Malawi, the inhabitants of Licm Island associated the disease with lack of circumcision (Anon, 1904). Hciever, in the early twentieth century it was demonstrated that the infection was through the skin

(Fujinam and Nakamura, 1909) and that the parasites used snails as intermediate hosts (Miyairi and Suzuki, 1914; Leiper, 1915; Itube and

Gonzales, 1917),all cited by Farooq (1973) and Warren (1984).

Henceforth, more practical control measures could be developed.

The aim of schistoscniasis control is to achieve a substantial reduction of prevalence and intensity of infection. Eradication of the disease is the ideal, but it is acceptable if prevalence and incidence are lowered to a level where the disease was of little

24 significance in relation to other endemic diseases (Hoffman et al.,

1979). The methods of choice are: mollusciciding, agricultural engineering, imrproved water supply, sanitation, health education, chemotherapy, and biological control agents. Detailed treatment of these methods are given by Michelson (1957), Berg (1964 and 1.973),

WHO (1965), Jordan and Webbe (1969), Fergusson (1978), Kuris (1973),

Ansari (1973), Bradley and Webbe (1978) and Rosenfield (1979). Limitations of various kinds have made it necessary to focus research on biological control of schistosomiasis (Fergusson, 1978) and the most common target for such research is on a variety of organisms capable of removing the vector snails (Barnish and

Prentice, 1982; Frandsen and Madsen, 1979).

FISH AS BIOCIN1IM AGENTS

Early attaq ts

Research on the use of fish to control snail hosts of

schistosomes has been confined to relatively few countries such as Cameroon, Kenya, Zaire and Brazil (Brown, 1980; Jordan et al.,

1980; McCullough and Malek, ms) but has not progressed far.

Nevertheless, preliminary trials of fish as control agents in fish

ponds and reservoirs has been attempted. Debont-Hers (1952, 1956)

and debont (1956 a, b) were among the first to report success using

a cichlid fish Serranochromis (Sargoch s) mellandi (Boulenger) in

fish ponds in Zaire. Another cichlid fish, Astatoreochronis alluadi

(Boulenger), has been successfully cultured with tilapine fishes in

both the Cameroons (Bard and Fbvogo, 1963; Movogo and Bard, 1964;

25 et al., 1977)...... In et al., 1964), and in Kenya (4Mahon Gamet s brasiliensis the cichlids, Astronotus ocel (Cuvier),

bimaculatus (Gill), and S. mellandi (Quoy and Gaimard), Hemicshrms percularis (Linnaeus) (family along with the jewel fish, Mac a variety Belontiidae), all appeared to control snail vectors under Malek, ins). M.re recently, studies of circumstances (McCullough and reduction in small concrete in Sudan demonstrated significant snail (Daffala et al., (protpterus annect ) ponds by the lung fish capacity in 1985). The lung fish exhibited amazing snail eating snails fish eating as many as 200 the laboratory with a large

in a day (Daffala et al., 1985). limited. eating fish in Malawi is The literature on snail to report among the earliest researchers Bertram et al. (1942) are noted report, Bertram et al. (1942) on snail eating fish. In their were as Serranoc s thurmbergi, that some cichlid species, such fed on water snails. No with heavy grinding teeth and equipped of the potential of S. thunmbergi mention is made in the above report

Later, in schistosomiasis control. or other snail eating fish large added Cyrtocara placodon and however, Jackson et al. (1963) the list of calliptera (pellagrin) to specimens of Astatotilapia use lu that these species could !e snail eating fish and indicated

in schistosomiasis control. for eating habit of C. placodon Attempts to study the snail successful to control snails was less incorporating in fish culture Malawi the move inland from Lake as only one specimen survived that the Pruginin (1976) reported (Pruginin., 1976). Nevertheless

26 snail eater. ther than... single specimen was a "voracious" on fishes suitable for culture report , which mainly concentrates no further there appears to have been rather than snail control, preference work of 1986 demonstrated the work until ycKaye et al.'s of C. placodon for B. globosus.

AGEMS ADVAMS OF FISH AS BIOODRM) Multiple use new in the tropics have created Development of water projects of has been increased transmission snail habitats. The result over half 1986). In Sudan for example schistosomiasis (Garfield, are infected Gezira irrigation scheme of the residents on the Kariba and man-made lakes such as (Gaddal, 1984). Similarly, and and Malingunde in Malawi (Teesdale Volta (Jordan et al., 1980) The iniportant sites of transmission. Chitsulo, 1983) have become endemic area additional water body in an earlier belief that an by Forsyth any worse was challenged does not make transmission pathological demonstrated an increase in and Bradley (1964). They and from man-made water bodies changes in children infected children expectancy of the infected suggested that the life bear in mind Also it is important to would be reduced by 15-20%. of the quantitative and the extent that schistosomiasis is to the intensity of infection disease is directly related the snail Therefore, the more numerous (wormload) (Clarke, 1977). infection. likelihood of repeated habitats, the greater the continues of water development projects Nevertheless, the expansion

27 irrigation, and or a variety of reasons: protein production, ydroelectric power. harvested from an acre More protein in the form of fish can be or beans frcm an acre of )f pond than in the form of beef, corn, 1985). In many African :he finest farmland (Schmidt and Roberts, of providing the :egions, fish fanning is the only possibility and deBont-Hers, 1952). people with animal protein (deBont be a secondary result of Furthermore, protein production can lakes designed for stocking fish in reservoirs and man-made and Stauffer, 1983). irrigation and or power generation (lcKaye are an economic Although the water development projects health risks should not necessity (Garfield, 1986), the resulting for a new and low cost be underestimated. The situation calls snail feeding fish into innovative approach. The introduction of in the design these man-made bodies of water should be considered of all such projects. snails and control Theoretically, these fish would feed on the could still serve the the disease while at the same tine the water eaten for food since other functions. The fishjcould also be the fish. For example, schistosoniasis is not transmitted from that Serranochromis deBont and deBont-Hers (1952) reported snails and was a macrocephalus (Boulenger) successfully controlled Hora (1952) in good sportfish and excellent to eat. Similarly, Pangasius pangasius Slootweg (1985) reported on the cat fish fish not only provided (Hamilton) which fed on snails. This used as grit for poultry. protein, but left shells intact to be

28 For example, Zakaria Fish may also have secondary effects. not only cleared a canal of Bulinus (1963) reported a cat fish that feet, and he argued that apart in Iran but also stung people's it wuld also discourage people from controlling the vector snails, infected water. It is, however, fran getting in contact with been found that can sting people unfortunate that no creature has or feces (Lie et al., 1968). who contaminate water with urine Lanoval might benefit fish farming. Apart fron disease control, snail such as Bellamya bengalensis, This is so because some snails, fish to the extent of causing compete for food with cultured 1986). Hence snail removal malnutrition among the fish (Raut,

woDuld reduce such food competition.

Toxic free control of the snail host is one Parasitologists agree that control infection rate (Berg, 1964). This effective means of reducing the to apply molluscicides to view prompted public health workers been varying degrees of success snail habitats. The results have all effective molluscicides are (Rosenfield, 1979). However, nearly organisms (WHO, 1965; Jordan toxic in varying degrees to non-target hence the growing resistance to and Webbe, 1969; Paperna, 1980), deBont-Hers, 1952; Berrie, 1966; their use in fish ponds (deBont and nolluscicides evaluated in Malawi McCullough and Malek, ms). Plant killed fish at much lower (Kamwendo et al., 1985) actually the snails (Chiotha and Msonthi, concentration as compared to and South Africa, rapid amphibious 1986). Similarly, in Zimbabwe molluscicide application and the and fish kills wre noticed after

29 1985). Worse dams had to be restocked with fish (Appleton, were believed still, some fish killed in the South 'African case, and hence the to keep nosquitoes down by preying on their larvae threatening major mass fish kills allowed build up of mosquitoes, outbreaks of malaria (Appleton, 1985). medically The need to develop new methods for the control of of the US-Japan ixportant snails was highlighted in a report "Opposition to Medical Programme cited by Berg (1973) as follows: will the dissemination of toxic compounds such as molluscicides of their use". For undoubtedly result, eventually, in restriction have objected example, in upper Egypt, near Aswan, local fishermen Kvale, 1985). It is to the use of toxic molluscicides (Weil and control be thus inportant that other methods such as biological

explored.

Cost effectiveness method of The use of fish could be the most cost effective no further control because once introduced and properly managed, trained personnel, inputs should be required. The need for highly should not be sophisticated equipment or frequent application fulfill one of the necessary. The use of molluscivores would which requires control strategies proposed by Maoshou-Pai (1984) a control policy that the disease be fought repeatedly. Such measures would have long term effects and minimal need for future

(Kuris, 1973). (1977) Encouraging results were reported by McMahon et al. term effects that the cichlid fish A. alluadi had significant long

30 Kenya. Similar results on snail populations in artificial ponds in bimaculatus, were also obtained with another cichlid fish, H. the snails which, after introduction in a pond in Brazil, el.iminated 20 fish at the in 6 months and their number increased from ms). As a result, beginning to more than 500 (McCullough and Malek, of selected fish McCullough and Malek (ms) suggested that use in large man-made species in integrated control of schistosomiasis control or lakes (such as Kariba and Volta) where chemical should be environmental control are costly, difficult and repetitive,

thoroughly explored. health In the past, control was solely the job of public now viewed to be workers. However, community participation is 1984; Dazc, crucial for successful control programs (Chitsulo, fish should win 1984). In this respect, measures that utilize prove cheaper in popular participation by the local cmmunnity and would capitalize the long run. In Malawi, such a control strategy in increasing on the already existing interest by the government 1971; McKaye and fish output through fish culture (Pruginin, that lead to the Stauffer, 1983) and local self-help structure (personal success of the rural piped water supply project

observation).

31 ECOLDGICAL ONSTRAINTS AND CAVEATS IN THE USE OF FISH

All biological control programs are operationally of an snail ecological nature (Lie et al., 1968). Hence the control of intermediate hosts using fish must be considered in the general context of ecological principles (WHO, 1984). A consideration of some of the ecological constraints are given below.

Prey preference

The fish utilized should be capable of feeding on the adult and or juvenile snail stages. However the degree to which molluscs serve as food is variable. Molluscs may constitute as low as 1%, of the diet as in the yellow perch (Perca flavescens, Mitchell) or

100%, as in the freshwater drum (Aplodinotus qrunniens, Rafinesque)

(Michelson 1957). Which type of fish should be chosen?

A species whose diet consists entirely of snails would be a good

candidate since the probability is high that it would consume all due to snails in a pond. However if the snails disappeared

predation or seasonal changes, such a specialized fish would starve

and would have to be continuously reintroduced when the snails

return. Therefore, the diet for a fish of choice should consist

mainly of snails but also other food stuff in the absence of snails

(MCullough and Malek, ms.; WHO, 1984).

Nevertheless, possible problems exist with such a generalized to feeder. For example, at what snail density do they switch

another type of food? Predators tend to be polyphagous and avoid

exhausting their potential food supplies (Slobodkin, 1968).

Therefore, some infected snails might still renain. Furthermore, the

32 that some fish is carplicated by the fact problem of prey preference are another if more than one type will prefer one vector snail to to A. alluadi, preferred Biamphalaria found together. For exanmple, the two and might not be useful where Bulinus (btMahon et al., 1977) Also, the species preference of vector types are found together. on the size of the snail and the molluscivores will vary, depending

the size of the fish. plasticity and habitat Problems of co-evolution, morphological

diversity shell appears to be directly The morphology and thickness of the (Zipser and Vermeij, 1978). related to the intensity of predation that increased shell Venreij and Covich (1978) have hypothesized is a co-evolved adaptation to thickness in the genus Melanoides lakes of Africa. This view is reduce predation by cichlids in the who demonstrated that C. placodon supported by McKaye, et al. (1986) the larger sizes of thick shelled consumes a lower proportion of thinner shelled B. globosus snails. snail species, and preferred the bone and teeth used to crush the Also, the form of the pharyngeal of is fed soft food (as in the case snails may atrophy if the fish 1965). Such a response could A. alluadi) and not snails (Greenwood, snails are initially removed and be a problem in fish ponds if the

were to later reappear. consume snails that are in Furthermore, fish may fail to et al. (1983) suggested that protective habitats. For example, Louda gastropods below 4 m was due the drop in the density of Lake Malawi In the shallower water, where to increased predation by cichlids. 33 snails, beds provide a refuge for vascular macrophyte (weed) efficiency in capturing them densities were high because predator in which molluscivores were was reduced. Recent experiments sand 60% increase in snails over open excluded by cages, resulted in jond the physical structure of a (McKaye et al., 1986). Therefore in the for the snails' persistence or reservoir may still allow beds to be more effective, all weed presence of molluscivores, and As Huffaker (1958) has indicated, ,uld probably have to be rmoved. of prey is facilitated by availability coexistence of predator and

refuges.

Co-existence with other fish hence alluadi are good eating and Same of the fish such as A. they do Such fish will be favored if can be accepted in ponds, of fish that people prefer. not interfere with the raising fish this requirement as well. A AstatoreochrmiUs alluadi fulfills it d a l l i might appear suitable, because like Tilapia ten (Boulenger) and their eggs. However, it may feeds on weeds, adult Bianphalaria the fish ponds because it reduces not be acceptable in farmers' (Fergusson, 1978). Therefore, growth rate of other flishes several species of fish are required experiments in polyculture of of fishes to introduce into these to determine the optium ctmminity

man-made systens.

34 Fig. 1. Life cycle of schistosaniasis.

35 a"A aditoooms

HJAAN ACTvM

w ah 0

~OMIRAMDA CERCARL&E

BbPD" L~rm oc*come&a

VECTORS

36 Fig. 2. External morphology of three snail species: A.) Bulinus tropicus,

B.) Bulinus globosus C. ) Bicrphalaria pfeifferi.

37 38 A sNAIL-EATING FISH CHAPTER 3 PREY HANDLING IN gvrtocara placodon: FRCM MAIAWI.

I1717DUC OCN several Exploitation of gastropods for food has been reported in fishes (Slootweg, animal groups such as snakes, birds, crabs, and significant predators 1987) but the latter two are the most pose special problems (Palmer, 1979). As a food resource, gastropods and Iles, 1972) is because the highly nutritious flesh (Fryer (Hoogerhoud, 1987). enclosed in a hard, dense, indigestible shell adaptations to However, molluscivores have developed -morphological 1965; Fryer and Iles, gain access to the gastropod flesh (Greenwood, fishes for example, use 1972; Liem, 1973). Shell-breaking crabs and 1979; Meyer, 1989) chelae (Palmer, 1979) and pharyngeal jaws(Palme-r, respectively, to crush snails. behavior in This study was conducted to observe foraging of Lake Malawi. Cyrtocara placodon, a native cichlid molluscivore a biological control This species had earlier been suggested as al., 1963; Pruginin, agent of human schistosomiasis (Jackson, et 1986). Therefore, 1976; McKaye et al., 1986; Chiotha and McKaye, snail eating fish is the understanding the foraging behavior of this reduce vector snail first step in predicting its ability to populations (Slootweg, 1987). relationship of Experiments were also conducted to test the size, using a point load snail shell crushing resistance to snail Bulinus tropicus and crusher. Three snail species, Bulinus globosus,

39 with the exception Biomphalaria pfeifferi were used. These species, in Malawi of B. tropicus, are vectors of human schistosomiasis significant (Chitsulo, 1984). Previous studies have demonstrated certain snail differences in static crushing resistance between choice of these species (Stein et al., 1984; Hoogerhoud, 1987). The vector snails fran snails allowed us to test this hypothesis on the in static Malawi and one non-vector snail species. Differences shell shape, shell crushing resistance are mainly due to variation in sculpture thickness, degree of whorly overlap and other external in shape; Bulinus (Palmer, 1979). The snails in this study differed (Brown, 1980). species are globose while Bicaphalaria are discoidal

MATERIALS AND MN S General observations on feeding behavior 0.76) were Ten molluscivores (mean SL = 12.86 cm; SD = to acclimate introduced into a glass aquarium (100 1) and allowed were then provided for 24 hours without food. The molluscivores initiation of with B. tropicus snails and observations made on and the feeding, prey capture t~chniques, prey processing, of the feeding interaction of the molluscivores during feeding. Same analysis at activities were recorded on video tape to permit detailed fish were placed a later stage. Observations were also made when the

individually in tanks.

40 of molluscivores Evaluation of prey handling tine and capacity an) were held individually Two molluscivores (SL= 13.2 an, 13.8 '3. globosus snails, and had in 100 1 glass aquaria, fed individual determined. Handling time was their handling time per snail snail capture to last ejection determined as the time interval frcm 1987; Meyer, 1989) or when of shell fragments (Hoogerhoud, 1987). The size range of respiration became normal (Hoogerhoud, mm, but they were presented snails offered to the fish was 4-13 The size above which the fish individually in a random fashion. represented the upper limit of failed to capture and consume snails were continued for two weeks handling capacity. The investigations for satiation. Between each with the same molluscivore to control 24 hours. trial the molluscivores were starved of snail shells Estimation of static crushing resistance of three snail species, B. Determination of crushing resistance was done using a point load globosus, B. tpicus and B. pfeifferi, the techniques employed by crusher following, with modifications, (1987); see fig. 3. A snail of Stein et al. (1984) and Hoogerhoud end of the piston and water known size was placed at the bottom on the piston platform until the poured slowly into a beaker placed water was noted and the procedure shell was crushed. The volume of applied to the snails against repeated for other snails. Force was logical considering that same their minimal dimension. This seems manipulate snails between molluscivores, such as Lepomis macrolophus, their miniml dimension (Stein pharyngeal plates to crush them across tropicus the shells were placed et al. 1984). For B. globosus and B.

41 pfeifferi with the opening with the opening facing down, and for B. and piston assembly on their sides. The weight of the water constituted the force required to crush the shells.

RESULTS

Prey capture and processing prey and sucking it Prey capture involved swinning taards the both for snails on at close proximity. This technique was employed were attached to objects sediment or suspended in water. If snails molluscivores rammed the such as glass, plants, or other snails, the consumed them as before. snails several tines to dislodge them and the jaws and ingested. Large snails (>7 mm) were picked up with between the lips before Soetimes the large snails were manipulated

ingestion or abandoned if too large to be ingested. crushed between the When successfully ingested, the snails were a high pitch crunching pharyngeal bones, a process that produces a series of quick jaw sound. During crushing, the molluscivores made of shell fragments and gill cover movements, followed by ejection jaw and gill cover through the mouth and the gills. The rapid periods of slow or no movements were interspaced with one or several

nm nnts. and appeared to The ejected shell fragments were small (< 0.5mm) when examined under a have very little flesh attached to them pieces of shell fragments dissecting microscope. Occasionally, large These lost fragments are (> lmm) with flesh are lost from the mouth. or another one. usually reconsuned by the sane molluscivore,

42 mlluscivore interaction was not always The initiation of feeding by C. placodon snails as soon as predictable. Sometimes the molluscivores consumed times they took they were dropped into the water, but at other tank longer. The average initiation time for 10 molluscivores in one the molluscivores was 14.2 minutes (SD = 14.18); see table 1. When greater than were held individually in separate tanks, initiation was day. When one 1 hour and some did not feed until the next went for the same snail, molluscivore approached a snail, others tank. Most often, despite the presence of other snails in the close to one molluscivores without prey, positioned themselves snatch the captured processing a snail, making several attempts to lost from the prey or waiting to consume whole snails or fragments

mouth.

Evaluation of handling time and capacity for C. pl placodon with The results of feeding the same specimens of C. increases with different sizes of snails suggests that handling time models for increasing snail size; see fig. 4. The appropriate

handling were found to be quadratic:

2 2 = Y = 117 - 38.40 x + 4.04 x (n = 25, r 0.90)

2 = 2 = 0.95). Y = 48.97 - 18.37 x + 2.48 x (n 24, r could sometimes not For small snails (less than 4 nun), handling time fast and few if be accurately detennined because the processing was

any shell fragments were released. limit of For C. placodon (SL, 13.2 cm; 13.8 cm), the upper greater than 10 handling (i.e. handling capacity) is for snail sizes 43 rm. None of the two fish or others used in the first investigation

(initiation of feeding) successfully ate snails larger than 10 mu and same of the snails between 9 and 10 mm were consumed only after several attenpts.

Estimation of crushing resistance of snail shells using a point load crusher

Crushing resistance increased with snail size for all the three snail specie. (fig. 5). Interaction between force and snail size was significant (P< 0.01, ANCOVA ). Using stepise regression analysis, the best models for static crushing resistance were found to be quadratic:

2 2 Y = -15.06 + 4.13 X ­ 0.09 x (r = 0.82, n = 61)

Y = 6.99 - 0.04 x + 0.15 x2 (r 2 = 0.61, n = 43)

Y = -0.01 + 1.45 x + 0.02 x2 (r2 = 0.78, n = 30),

for B. globosus, B. tropicus and B. pfeifferi respectively. The

difference in static crushing resistance among the snails was tested

using Duncans' multiple range test on residuals from a common line.

The differences were significant for the three species (P>0.01,

ANOVA).

Visual inspection of the graphs (Fig. 5) and a plot of residuals

against snail size suggest increased variability in static crushing

resistance eaLtong the large snails. However, the hypothesis of

hcmogenous scatter tested, using split-sample analysis of residuals

(Younger, 1985), could not be rejected.

44 DISCUSSION

General feeding behavior Prey capture in fishes has been a subject of several studies capture (Alexander, 1967; Nyberg, 1971). The dominant mode of prey motion in bony fish consists of a combination of suction and forward pattern for (Leewen and Muller, 1984) and C. placodon fits this The strategy snails suspended in the water and lying on sediment. glass was different for snails sticking to objects such as aquarium Under or plants or when several snails are sticking to each other. times such circumstances, the molluscivores rammed the snails several

to dislodge them before ingestion. This strategy is employed. strong probably because the suction pressure from the mouth is not possible enough to free the snails attached to objects. Another

explanation for separating snails from each other is mouth gape

limitation (Osenberg and Mittelbach, 1989). The behavior of snails sticking to each other or other objects

might reduce vulnerability to predation. If any snails were uneaten to from the aquarium experiments, the majority were found sticking not have the glass walls and other objects. The molluscivores did these any difficulty in consuming snails on sediment probably because of the snails could not maintain a tight grip on the loose grains

sediment.

C. placodon is a pharyngeal crusher and ingested snails are both moved to the pharynx where they are crushed by massive teeth on The the lower and upper pharyngeal bones (Fryer and Iles, 1972). to allow advantage of crushing snails, especially large ones, is

49 (Hoogerhoud, 1987). However, them to pass through the inte3tines Experimental evidence from crushing should also facilitate digestion. a tilapine fish species observing Oreochrcms spilurus spilurus, snails, demonstrated that all which opportunistically ingests some uncrushed snails survived crushed snails were digested, whereas 1981). the digestive process (Canpbell, C. placodon makes a We noticed that during snail crushing, mov ents with intervals of slow series of quick jaw and gill cover have been made on A. alluadi or no movemrent. Similar observations and L. microlophus (Stein et al. (Hoogerhoud, 1987; Hoogerhoud, ms) resting periods (Hoogerhoud, 1984). These intervals are probably is swallowing snail flesh (Stein et 1987) or when the molluscivore

al., 1984). significance. Studies on A. Shell ejection also has biological increases the amount of flesh alluadi suggest that this process 1987). Furthermore, shell stored in the stomach (Hoogerhoud, of carrying around the dense ejection reduces energy costs vertical migration due to the indigestible shell material and of be needed to compensate for extra swiibladder volume 4hich would

ingested shell (Hoogerhoud, 1987). feeding suggests that C. The interaction of nolluscivores during on group members to locate food. placodon forage in groups, relying on C. placodon in L. This is consistent with insitu observations fishes (Keenleyside, 1979). Malawi (McKaye pers. comm) and other of watching others digging in Scme fishes have developed a strategy feed on the exposed prey the substrate and darting in to

46 demonstrated for the goatfish Keenleyside, 1979). This has been a variety of other species, all Mullidae) which are followed by of the goatfish (Hobson, Eeeding on the prey disturbed by the digging spined sticklebacks (Gasteresteus 1968). Similarly, foraging three search for food around one individual aculeatus) quickly converge and prey position towards newly discovered that pitches into a head dG.m a yellow perch (perca flavescens), (Keenleyside, 1955). Among the and repeatedly by different individuals single prey may be struck at 1973). finally eaten by one (Nursall, effective in restricting prey Group feeding may also be more do However, considering that snails that flee (Keenleyside, 1979). have molluscivore, group feeding ay not flee from an approaching feeding One possibility is that group evolved for other functions. single solitary feeding, in areas where wuld be more effective, than fish aggressive territory-holding fish are excluded from by more contact with mimbers of the group (Keenleyside, 1979). Also, keeping snails. This is particularly might be more effective in locating populations are not uniformly important considering that snail (EL-Gindy, 1962). Furthermore, distributed in natural habitats be from other molluscivores may taking snails or fragments lost reducing time of crushing. The advantageous in saving energy and called from other individuals is behavior of stealing food The reported among certain birds. kleptoparasitism and has also been demonstrated this habit among work of Hatch (1970), for example, (fish) from terns. laughing gulls which stole prey

47 was not always The initiation of snail feeding by C. faCOdon the day there are certain times of predictable (Tablo '-). rdrnaps Such at its peak for these nolluscivores. when feeding activity is carp (Cyprinus carpio), which behavior has been reported in the any time at dusk and night, but feed exhibit high endogenous activity et al. 1986). The presence of of the day when conditioned (Sibbing of is probably conducive to initiation other molluscivores in a tank in a molluscivores that fed frequently the feeding repertoire because for social when isolated. This need group became less active common among cichlid molluscivores. facilitation is probably Lake snail eating Haplochrines fram Hoogerhoud (1987) reported that could see feed normally when the fish Vitoria held in separate tanks would to test if other fish species others. It would be interesting other C. molluscivores in the same way affect the feeding behavior of p1acodon did. and capacity for C. placodon Evaluation of harfdling tine with same specim-ens of C. Rlacodon The results of feeding the with that handling time increases different sizes of snails suggests 13.8 am), the C. 2:acodon (SL = 13.2 an, increasing snail size. For tropicus handling capacity) is for B. upper limit of handling (i.e. others used 10 nm. None of the two fish or snail sizes greater than than 10 mm successfully ate snails larger in the first investigation after 9 and 10 mm were consumed only and some of the snails between

several atteopts.

48 using a point load :stimation of crushing resistance of snail shells xusher for all the three Crushing resistance increased with snail size that small snails ;nail species. This data supports the suggestion and Veineij, 3980; Stein Lre easier to crush than large ones (Zipser and Mittelbach, 1989). ,tal., 1984; Hoogerhoud, 1987; Osenberg handling time data (fig. 4) klso, because of the parallel between the (fig. 5), it is clear that 3nd the static crushing resistance data factor influencing handling static crushing resistance is a major time in C. placodon. in static crushing Previous studies have shown that differences are due to shape, shell resistance between species of snails and other types of external thickness, degree of whorly overlap, is important because shell sculpture (Palmer, 1979). Shell thickness solids (Wainwright et al. resistance to breaking, like all other (Palmer, 1979). 1976), increases as the third power of thickness

Implications for biocontrol of schistosniasis use in controlling Cyrtocara placodon clearly has potential for study has confirmed earlier snail vectors of schistosomiasis. This snail eater even in reports that this molluscivore is a %voracious' Pruginin's case (1976) captivity (Pruginin, 1976). However, unlike were able to observe where only one molluscivore survived, we feeding. We have interaction of several molluscivores during may be less vulnerable to observed that snails attached to objects would probably be more predation and hence the molluscivores providing objects effective in simple habitats than complex habitats 49 seem to indicate for firm attachment of snails. The results also molluscivores are that better results would be achieved if several need for social used in any control attempts because of the handling time with facilitation during foraging. The increase in molluscivores in this increasing snail size and the failure of the raises important experiment to consume snails larger than 10 mm (SL <13.8 concerns regarding the effectiveness of small molluscivores the significant cm) in controlling large vector snails. Finally, snail species difference in crushing resistance of the two vector would prefer the raises the questions on whether the molluscivores These issues species that is easier to crush in natural situations. the following were investigated further and are reported in chapters.

50 static crushing Fig. 3. Point load crusher for estimating resistance of snails.

51 52 for snails (B. Fig. 4. Handling time (sec.) in two C. placodon appropriate models globosus) of different sizes. The for handling are quadratic:

2 2 0.90) Y = 117 - 38.40 x + 4.04 x (n = 25, r =

2 x 2 (n = 24, r = 0.95). y = 48.97 - 18.37 x + 2.48

53 160* 140 120 * ' 0100

~80 * S60

40

4 5 6 7 8 9 10 Size (mm) 10 moAluscivores (C. pla.codon) Table 1. Timre taken (minutes) for were to start feeding on snails. All the rolluscivores time for kept in one glass aquarium. Mean initiation

feeding is 14.20 min. (SD = 14.14).

starts Trial Time feeding(Min.)

1 28.00 2 16.00 3 19.00 4 43.00 5 22.00 6 1.00 7 12.00 8 0.08 9 3.00 10 0.08

55 Fig. 5. Static crushing resistance (Newtons) of three snail species. Reg.ression equations are quadratic as follows:

2 2 Y = -15.06 + 4.13 X - 0.09 x (r = 0.82, n = 61)

Y = 6.99 - 0.04 x + 0.15 x 2 (r 2 = 0.61, n = 43)

2 2 = = Y = -0.01 + 1.45 x + 0.02 x (r 0.78, n 30),

for B. globosus (1), B. tropicus (2) and

B. pfeifferi (3) respectively; where Y is force (N) and

X is snail size (ram).

56 30 1

25 ­

..ao020

0 3 6 9 12 15 18 Size (mm) FISH (gyrt... CHAPTER 4 FOOD SELECTION IN TWO SNAIL EATING placodon and Astatotilapia calliptera) FRFM MALAWI AND ITS

RELEVANcE TO BIOLOGICAL Comm'DL OF SCHIsTOC1AIASIS

as intermediate hosts of A number of fresh water snails serve infection caused by human schistosamiasis (bilharzia), a parasitic (Manson-Bahr and blood dwelling flukes of the genus Schistosoma is the fourth major cause Apted, 1983; Bisseru, 1984). This disease malaria, malnutrition, of morbidity in the tropics and subtropics; 1983). Most experts on and tuberculosis rank higher (Dzik, is the most effective single schistosmiasis agree that snail control disease (El-Gindy, 1962; measure for control of this debilitating

Dazo, et al., 1966; McMullen, 1973). applied to snail Chemical molluscicides have been extensively of success (Rosenfield, 1979; infested habitats with varying degrees costs and toxicity to Chiotha and Morgan, 1987). However, rising (Appleton, 1985; Teesdale, non-target organisms, particularly fish and more environmentally 1986), clearly suggest that cheaper This realization has made acceptable control measures are required. disease attractive. The research on biological control agents of this usual target fc,': attack with snail intermediate host is also the 1982). biological control agents (Barnish and Prentice, toward testing the The present research was directed from Malawi, Africa, for suitability of two snail-eating fish species There are several possible use in the control of schistoscmiasis.

58 and les, 1972), but in tis study snail eating fish in Malawi (Fryer two cichlid fishes, C2rtocara placodon and Astatotilapia calliptera, of such fish was reported from wre chosen. Although the existence in et al., 1942), their potential early studies in Malawi (Bertram, until much later by schistosomiasis control was not popularised et (1963), Pruginin (1976), McKaye wc.kers such as Jackson et al. (1986). Malawi has one of the al. (1986), and Chiotha and McKaye schistoscmiasis in the world (Teesdale highest prevalence rates of" a search for alternative control and Chitsulo, 1983) and hence Snail-eating fish are particularly measures is worth exploring. increase in the number of snail­ attractive because of the current and a result of fish farming (Chiotha infested habiiats in Malawi as ponds might be mre practical than Jenya in prep); their use in these of The target snails for control snail control by chemicals. and Bulinus (Physopsi.s) globosus schistosomiasis in Malawi are these two species are the confirmed Biomphalaria pfeifferi, because is and Teesdale, 1982). The former vectors in the country (Choudhry caused by Schistoscma haematobium a vector of urinary schistosafiasis S. bowel schistosomiasis caused by while the latter is a vector of

can be used as biocontrol Before any of the snail-eating fishes ecological questions needing agents, there are a number of question is about food investigation. One such fundamental documented the existence of food selectivity. Several studies have (Ivlev, 1961; Emlem, 1966; Bohl, selectivity by foraging animals Because of the implications of 1982; Schmitt and Holbrook, 1984). 59 of biological control such selectivity on the success or failure fish species in using a predator (WHO, 1984), the two snail-eating in the laboratory. this study were studied for selectivity to snail size, snail species, Selectivity was studied in relation other than snails. presence and absence of snail shell and food

MERIAIS AND MMEDDS

Prey selection based on snail species. rectangular fiberglass The experiments were conducted in ten sand at the bottom and tanks (1 m x 1 m x 1 m), each with 3 cm of water. Pond water was not half filled with dechlorinated tap or well observations of feeding used because of its murky condition making ten similar size snails activity difficult. To each fiberglass tank, were introduced in of two species, B. globosus and B. pfeifferi, for 24 hours and equal proportio.-n. Snails were allowed to acclimate introduced into each then a single mrolluscivore (C. placodon) was 24 hours prior to use. tank. All molluscivores had been starved for was noted, but without The number of snails eaten every day after 4 days. The replacement. The experiment was terminated with B. tropicus, a procedure was repeated with B. qlobosus paired were tested non-vector snail species. A total of 20 molluscivores

against each treatment.

60 Prey selection based on snail size with the same number of Fiberglass tanks were set up as above ten snails were of one species snails in each tank. Hcever, all the classes, large and small. The but in equal proportions of two size 6-7 mm and 8-9 m for small size range in each class was initially experiments were altered to and large, respectively, but replicate between the two size classes. increase the average size difference and 8-10 mm for small and Hence, later size classes were 4-7 mm of snails eaten in each size large snails, respectively. The number

class was recorded as before. of snail shell Prey selection based on presence or absence by presence or absence To test how prey choice could be affected was followed with some of snail shell, the protocol above snails of one species, B. modifications. For example, small size was presented with equal globosus, were used.. Each molluscivore shells) and those whose proportions of nonral snails (with intact This investigation was shells had been crushed and removed by hand. calliptera and resuxlts read conducted using both C. placodon and A. as in the other choice experiments above. of B. globosus and C. placodon was also presented with a choice For the purposes of B. pfeifferi, both with shells removed. to non-molluscivores to comparison, snail flesh was also presented species of non-molluscivores see if they would eat the flesh. The and Tilapia rendalli were Pseudotropheus zebra, Oreochrcmis shiranus 1983). These fishes had (Fryer and Iles, 1972; McKaye and Marsh,

61 been kept in aquaria for over 6 months and had never taken snails with intact shells as food. Consumption of food material other than snails by molluscivores An attempt was made to test if the molluscivores would take alternative sources of food in the presence and absence of snails.

Three food types were tested and these were Tetra Cichlid Fish Food

(cmmercial fish food), "madeya" (corn bran) and "nsima" (corn-meal which is a human staple food in Malawi). The latter two were chosen because fanrers in Malawi use them as supplementary fish food in aquaculture (Chiotha and Jenya in prep.). The nsima used is left over fran meals, while the adeya is a by-product of corn processing.

Corn meal was tried rather than corn-flour because of the suggestion that starch products are utilised better by fish when cooked (Anon, 1988).

Investigations were also done to test cercariophagic potential

of the molluscivores. At least ten wild infected snails were

introduced into a 10 1 aquarium filled with dechlorinated tap water.

To enhance shedding of cercariae, the aquarium was illuminated for

two ours with a bench lamp (100 Watt). Two hours was long enough

for a reasonable number of cercariae to be shed and hence the snails

were removed and kept in a holding aquarium for later use. Next,

four molluscivores were introduced into the aquarium containing the

cercariae. Two of the nolluscivores were A. calliptera and the other

two were C. placodon and had both been starved for 24 hoars prior to

their use. The investigation was run for 6 hours after which the

molluscivores were killed, dissected, and the alimentary canal

62 examined for cercariae under low power dissecting microscope. The experiment was repeated until 24 molluscivores had been tested. The number of cercaxiae in the aquarium was estimated at introduction and removal of molluscivores by modification of techniques described by

Pellegrino et al. (1962, 1966) and Choudhry and Teesdale (1982); see appendix 3.

RESULTS

Prey selection based on snail species Given a choice of B. tropicus and B. globosus snails, C. placodon appeared to show no special preference to either species and the average numbers of these snails eaten in four days were not significantly different for the 20 molluscivores tested (P> 0.05,

Mann-Whitney U). Similarly, there was no significant difference in average number of snails eaten, when C. placodon was given a choice of B. globosus and B. pfeifferi (P>0.05, Mann-Whitney U). The results are summarized in Table 2.

Pre selection based on snail size

Given a choice of relatively small and large size snails , C. placodon preyed more on the small size snails. When the size range

for small was 6-7 mm and large, 8-9 mm, the difference in average number of snails taken was not sigiificant (P> 0.05, Mann-Whitney U).

However, by increasing size difference between the two snail

classes, the number of small size snails eaten by C. placodon was

significantly greater than for the large size class (P<0.01, Mann­

Whitney U, Table 3). Similar results were obtained with the second

63 the two molluscivore, A. calliptera (Table 3). Considering fewer snails (P< 0.01, molluscivores, A. calliptera ate significantly of comparable size. When Mann-Whitney U, Table 3) than C. placodon were mouthed for a while and ingested snails were too large, they were abandoned from then ejected whole. Sometimes the large snails the lips without being ingested. shell Prey selection based on presence or absence of snail more Both C. placodon and A. calliptera consumed significantly nonral snails of the sane B. globosus snails with shells removed than 4). However, there was no species (P<0.05, Mann-Whitney U, Table globosus and B. pfeifferi, significant difference in choice of B. U, Table 2). An both with shells removed (P>0.05 Minn-Whitney to three non-molluscivores, attempt to feed snails with shells were unsuccessful. namely, P. zebra, T. renialli, and 0. shiranus, as food, after the shells However, these fishes accepted snail flesh had been removed. Consumption of food material other than .,,nails by nolluscivores fed readily on The nolluscivores, A. calliptera and C. placodon, to snails. However, C. corn bran and commercial food in addition food only, for four placodon that had been fed commercial fish of starving. While weeks, never accepted snails' even after 72 hour- C. placodon did the other non-molluscivores would eat corn-meal,

not. revealed no Examination of the gut in the molluscivores and A. calliptera cercariae. This seems to suggest that C. placodon comes from the are probably not cercariophagic. Further evidence

64 that the number of cercariae was water analysis in which it was clear (P>O.05, T-test, Table unaffected by the presence of Irolluscivores

5).

65 Table 2. The average number of snails eaten by C. placodon (n=20)

when the nolluscivore was presented with a choice of two

snail species. Snails in experiment 3 had shells removed. The differences were not significant in all three

experiments (P>0.05, Mann-Whitney U). The snails used were

B. tr2icus, B. globosus and B. pfeifferi.

Trial Average no. of snails eaten no. Snail species in 4 days S.D.

1 B. globosus (vector) 2.50 (1.10) B. tropicus (non-vector) 2.15 (1.14) 2 B. globosus (vector) 2.55 (1.19) B. pfeifferi (vector) 3.10 (1.12)

3 B. globosus (shell removed) 2.85 (1.53) B. pfeifferi (shell removed) 2.65 (1.87)

66 and large) Table 3. The average number of two size snails (small eaten by two mlluscivores, C. placodon and A. calliptera, in in fiberglass tanks in four days. The differences

snails eaten between the two size groups -were significant

(P<0.01, Mann-Whitney U). The snails used were B. tropicus

and B. pfeifferi.

Average number of snails eaten in 4 days per molluscivore

Molluscivore Snail species sp&cies Small snails: Large: 4-7 mn (S.D). 8-10rm (S.D).

C. placodon Bulinus 2.90 (1.85) 1.55 (1.81) (n=40) tropicus

C. placodon Bicnphalaria 3.35 (1.53) 1.50 (1.05) (n=20) pfeifferi

A. calliptera B. tropicus 0.30 (0.47) 0.00 (0.00) (n=20)

67 Table 4. Prey selection by C. placodon and A. calliptera based on presence and absence of snail shell for B. globosus. The difference in the average number of snails eaten betwen the two groups of snails was significant (P<0.05, Mann- Whitney U)

Average number of snails eaten Molluscivore spp. Shell Shell present S.D. absent S.D.

C. placodon (n=20) 1.45 (1.47) 3.00 (1.41)

A. calliptera (n=20) 0.55 (0.89) 2.75 (1.48)

68 liter in aquaria stocked Table 5. Average number of cercariae per A. calliptera). The with molluscivores (C. placodon and before and 6 hours difference in the number of cercariae

after molluscivore introduction is not significant

(T test, P>0.05).

Average Average number of cercariae/l Molluscivore standard spp. length (cm) (SD) Initial (SD) Final (SD)

9.18 (0.70) C. placodon 172.22 (55.18) (-n=12) 188.89 (79.29) A. calliptera 9.75 (0.26) (n=12)

69 prey selection based on snail species and B. pfeifferi have The choice experiments with B. globosus no relative preference 'or clearly demnstrated that C. placodon has has potential for controlling aither vector. C. placodon, therefore, together. These results are both vector species even when both occur species can sometimes occupy encouraging because the two vector snail for example, we were able to the same habitats. During this study, dams at Makoka and Thondwe collect both snail types from reservoir fig.9, Appendix 1). in Zomba district (see map of of Malawi, distribution in Similary, a country-wide report of schistosomiasis suggests that in other Malawi (Teesdale and Chitsulo, 1983; 1985) Mzixrba, and Nkhota-kota, the areas of the country such as Chitipa, though not necessarily two vector snails are commonly found together, also have applicability in equal abundance. These results could in countries such as beyond the frontiers of Malawi. For instance, and Frank 1964), Kenya Souch Africa, Transvaal region (Schutte 1969), both vector snails (McMahon et al., 1977) and Zambia (Hira have been reported together in sare habitats. prey preference The fact that C. placodon did not show distinct (non-vector) may have between B. globosus and B. tropicus control using this implications on the overall success of disease that there was no preference molluscivore. It is an advantage that where B. tropicus is in favor of the non-vector because in habitats should still effect a found together with the vectors, C. placodon Furthermore, by being less reduction in the vector population.

70 array of prey items to choose selective, C. placodon will have a wide on a long term basis in from and should thus be easier to maintain absence of the vector snails. snail-infested habitats, even in the that aquatic habitats having This is particularly crucial considering by vector snails at a future no vector snails now, may be invaded date. to effectively reduce Also, if C. placodon can be shown the field, then C. placodon populations of the two vector snails in tried as biological would be superior to some snail eating-fishes example would be A. alluadi, control agents of schistosomiasis. An This fish has reportedly the snail-eating fish from East Africa. of its significant reduction been a pramising control agent because its weakness has been the lack of of Biomphalaria vector snails, but vector snail, Bulinus significant reduction in numbers of a second (Slootweg, 1985). consumption of non- The primary disadvantage of indiscriminate the role non-vectors play vectors and vectors is the possible loss of transmission. Of the many in reducing the extent of schistosmiasis dynamics o parasitic processes which influence the population to host is one of the most organisms, transmission from host snails can interfere important (Carter, et al. 1982). Non-vector rates in vector snails with this process by reducing the infection snails intercept miracidia through their action as decoys. Decoy the miracidia from (See life cycle, fig. 1), thus preventing as intermediate hosts reaching snails that normally serve miracidia infect the decoy (Laracuente, et al. 1979). When the

71 w-uld snails, they fail to develop into cercariae. This situation population density was probably be important only when non-vector because by chance alone higher than the vector population to feed on the most common prey indiscriminate fish predators tend

(Horn, 1983). the natural interactions in We do not have information regarding the non-vector snails in Malawi. the habitat between the vector and non-vector snails controlled A point of concern would be if the Any predation or other vector population through competition. non-vector might then allow pressure reducing the influence of the where the molluscivores lose vector numbers to build up to a level that molluscivores not their effectiveness. Thus it is desirable vector snail populations or interfere with natural factors limiting their infection rates.

Prey selection based on snail size C. placodon, preyed The to molluscivoret, A. calliptera and (<7 rm) than large snails (>7 significantly more on smaller snails of small size individuals to rm). The increased vulnerability in nature. Studies have shan predation appears ',- be wide spread at night to avoid visually that same small gastropod species feed species forage with relative feeding fish while large.gastropod et al., i983). impunity throughout the day (Bertness, to explain how prey size Several mechanisms have been proposed predators (Schmitt and selectivity is achieved by individual tested in this study, the Holbrook, 1984). For the molluscivores snails are easier to crush first possible explanation is that smaller

72 than larger ones. This phenomenon was clearly demnstrated in our snail measurements of static crushing resistance for the three of species (chapter 3 of thesis) and is consistent with the reports snails Zipser and Vermeij (1980) who have demnstrated that juvenile and other are indeed more easily crushed lethally by crabs (chapter predators than are adults. Also fram handling time studies quickly 3 of thesis), C. placodon processed small snails much more easier to than large ones, another indication that small snails were crush. fish. The second possible explanation is gape size of the

According to Hartman (1958), Lawrence (1957), Schmitt and Holbrook buccal (1984), and Osenberg and Mittelbach (1989), mouth gape and or ingested cavity determine the maximum size of an iten that can be it whole. This was supported by our studies with C. placodon where that were was occasionally observed mouthing large snails manipulated before successful swallowing or abandoned when too large. Furthermore, for predators that crush snails, an increase in the effective diameter of a snail shell, decreases the mechanical advantage of a predator's crushing apparatus (Palmer, 1979). This the was probably the reason why the molluscivore A. alluadi, used in more East African studies, ate few large Bulinus snails which are rotund than small ones (M-Mahon et al. 1977). Our hypothesis is that

gape size was probably a major limiting factor in the selection of the prey in our investigations. Nevertheless, saime large snails that had been successfully ingested were ejected without being crushed.

73 We have suggested (chapter 5 of thesis) that for C. placodon, visual perception appears to be an important cue in prey location.

According to theories advanced for visual predators, the size of potential prey varies with distance of prey from the predator

(Holling, 1966; Werner and Hall, 1974). For predators that pursue their prey over a certain distance, the ccmposition of prey itens in the diet as a function of distance will consist of small sizes at short distances while at long distances the diet will be biased towards larger prey items (Maiorana, 1981).

C. placodon sometimes selected prey fram a distance and this probably affected the size of some snails taken. As indicated before, the molluscivores sometimes selected snails that w too large to be swallowed. Hence, the fact that more small snails than large snails were eaten by the nmlluscivores is probably a reflection of the overall outcome and tells us little how often large snails were tried and abandoned.

The preference for small size snails might be a disadvantage for biological control. This is so because several studies have shown that the majority of infected snails tend to be large ones. Donnelly and Appleton (198 ), for example, reported that 98% of infected snails in a sample of B. pfeifferi had shell diameters greater than 9 nm. Similar-results were reported by Sturrock (1973) who found at least 60% of infected snails (B. clabrata) in the largest size group. By showing preference for smaller snails, C. plact.ion might actually miss the group that is active in transmitting the disease.

74 is the Another disadvantage of preference for smaller snails when it suggestion that a biological control agent is most effective major acts on developmental stages that have already survived most aquatic density dependent limiting factors (WHO, 1984). For at the snails, the dominant density dependent process is mortality confirmed by egg stage or very young stage (WHO, 1984). This is snail species studies on Lymnaea elodes which demonstrated that this 1966; had a survivor&hip to maturity of only about 2% (Eisenberg, would, under Bron and DeVries, 1985). Our candidate molluscivore of its preference this scenario, he considered less effective because for smaller snalls. snails One differential motality factor between young and older such as in natural habitats could be infection with parasites, than schistosomes. Infected snails have higher mortalities appears uninfected snails (Oliver et al. 1954; Sturrock, 1966) and it (Sturrock, that these differential rates are highest among juveniles Food is 1973; Anderson and May, 1979; Donnelly and Appleton, 1985). on L. another factor that would affect snail populations. In studies elodes, focd limited ju -nile survival and growth of this aquatic snail (Eisenberg, 1970). might be An added complication resulting from infected snails the extent to which thea parasites affect snail growth. Schistoscaes modify groth rate of snails (Najarian, 1961; Pan, 1963; Sturrock, aperture 1966, 1967) and also induce distortion of snail shell both the (Sturrock and Sturrock 1971). This has been observed in snails infected laboratory and the field on Biomphalaria glabrata

75 with S. mansoni where infected snails grew faster at first and slowed down later when compared to uninfected snails (Sturrock and Sturrock 1970; 1971). The inplications of this for biological control are not obvious, but if this phencrenon is common to all vector snails, then infected snails would be at an advantage in escaling predation by quickly leaping into a size range that is not preferred by the molluscivores. What is not reported in Sturrock and Sturrock's work is whether or not shell strength oi the infected snails is affected as well. This also has further implications on snail predation by the fishes. Despite the above possible disadvantages, we are still confident that C. placodon has great potential in schistosomiasis control. First of all, although food limitation contributes towards juvenile snail mortality, it may not be an important regulating factor in all habitats. Several people have argued that in unpredictable habitats food may be a more important limiting factor than in predictable habitats (Eisenberg, 1966; Hunter, 1975; Brown et al. 1985). We consider the fish ponds and other small impound d waters in which C. placodon is targeted for use to be predictable habitats. This is so because aquaculture practices require that certain conditions such as water level and pond productivity be maintained as long as fish are kept and this could cover a greater part of the ye& if not the whole year. Similary, impoundments for agriculture and domestic water supply allow persistence of water in some habitats which would be dry otherwise.

76 Furthermore, while trematode larval infections might limit snail population through castration (Wright 1966), the effect is not epidemiologically significant in transmission foci (Kuris, 1973). Also, environmental manipulations for water development projects create conditions for low species diversity, allowing snail populations to rise above levels that can be restricted by trematode infections (Kuris 1973). The survey of ponds in Malawi (Chiotha and Jenya in prep.) which showed that fish ponds were heavily infested with vector snails is further indication that natural limiting factors may not prevent build up of large snail population numbers in these irpounded waters. Hence, should C. placodon put sufficient pressure on the younger snails, the population of snails should eventually diminish due to reduced recruitment of snails into the larger and reproductively active group. This is probably what happened in the pond experimnts in which C. placodon dramatically reduced snail numbers (chapter 6 of thesis). For A. alluadi in Kenya, some reduction in Bulinus population may have occurred by a process of attrition because the fish were able to eat the younGer and less bulbous snails (McMahon et al., 1977). Brown and DeVries (1985) have also argued that fish predators are capable of lowering juvenile snail densities below levels expected on the basis of productivity. Similarly, a dramatic illustration of fish in controlling disease is the disappearance of human echinostomiasis, from Lindu valley, Indonesia (Carney, et al. 1980). This disease, whose transmission involves the mollusc Corbicula, is believed to have disappeared due

77 mossambica) that either to an introduced cichlid fish (Tilapia stage of canpeted with the mollusc for food or preyed on the velliger the mollusc (Carney, et al. 1980). smaller snails Although our volluscivore specimens preferred snails as large as when given a choice, they were also able to handle size range of snails 10 mm (Chapter 3 of thesis) and hence the the size of sm infected consumed overlaps to a certain extent with size range handled by C. snails. An additional advantage of the This is important placodon is that it includes medium size snails. are considered to be the because redim-sized, sexually mature snails been confirmed in the most resistant to desiccation. This has suggestions by laboratory (Sturrock, 1970) and supporteu earlier age group is one best Webbe (1962) and Cridland (1957) that this it refills after a suited to repopulate an aquatic environment when for our biological dry spell. The applicability of this finding medium size vector control agent is that successful predation on that would survive stages should help in removing the snail stage stocking periods. desiccation when fish ponds are drained between snail being infected Also, as indicated before, the chance of a pressure on juvenile increases with age (Sturrock, 1973). Predation snails fran living snails by C. placodon should therefore prevent case, long enough to get infected and shed cercariae. In any life expectancy as schistosome-infected snails have a lower average West Indies, ccmpared to uninfected snails. Studies in St. Lucia, over one week demonstrated an average life expectancy of a little and Webbe, 1971). after the onset of cerca:-ial .shedding (Sturrock 78 the start therefore only be a proble at Large infected snails should improve with C. and the situation should of a control programme An added advantage is that placodon exerting its influence. success suffer loss in reproductive schistosome-infected snails if they (Hairston, 1973). Hence, even compared to uninfected snails to the next generation of snails escape predation, their contribution should be small. the basis for food selection Although we have tried to explain in that our are aware of our limitation in our molluscivores, we of 10 lab.ratory at a total density experiments were done :.n the used was Also the size of molluscivores snails for each experimiet. We do not in the wild (< 16 cm, SL). limited to what we could get this size range, or when provided know how molluscivores outside would behave. with different snail densities, fish by other researchers, we Based or studies on snail-eating should molluscivores are used they strongly believe that if larger example, without any difficulty. For handle even the largest snails small (1952) stocked ponds with when debont and debont Hers spp.), the 4 eating fish (Serranocromis specimens(<1 cm, SL) of snail s-urvived. snai-l and some large snails zolluscivores ?-;e mostly small were used, all the snails including However, when large mlluscivores that large A complication with our hope the large ones disappeared. some large snails is the fact that molluscivores would handle Holbrook, dietary shifts (Schmitt and predators exhibit ontogenic C. placodon on ontogenic dietary shift in 1984). There is no data this test. to put the volluscivore to and plans are already underway

'9 Prey selection based on presence and absence of snail shell barrier to The shell of snails is regarded as a formidable in our study predation (Fryer and Iles, 1972). This is confinmed shwed stronger preference because both C. placodon and A. calliptera calliptera, this would for crushed snails than intact snails. For A. flesh from the be expected because it feeds on snails by extracting A. calliptera shell rather than crushing. This may be because Hence, it would lacks the jaw apparatus for crushing snail shells. advantage of the appear logical that A. calliptera should take Also we did not availability of snail flesh already extracted. chemical defence detect any evidence o! injured snails secreting were not avoided. against predation by fish since the crushed snails on snails The crucial question is why C. placodon, which feeds shell, and has in by successfully crushing and separating flesh fran show such strong fact the right jaw structure for crushing snails, is that shell preference for snails without shells? One possibility (Stein et al. crushing is costly in terms of energy spent crushing 1984). shells against Further evidence of t14 deterrent effect of snail in this predation comes frd the finding that three non-molluscivores The three non­ study also.accept.d snail flesh as food. do not normally molluscivores, T. rendalli, 0. shiranus and P. zebra, Iles, 1972; feed on snails in their natural environment (Fryer and did not consume Pruginin and Arad, 1977; W.Kaye and Marsh, 1983) and 18 months of this any intact snails in aquaria throughout the from Nigeria, investigation. Similar observations have been reered

80 feed on snails in the ihere 10 riverine species, that also did not (Reid, 1984). In latural envirorment, ate snails crushed'by hand America are also ;imilar manner, studies on cichlids of Central to Barlow (1974), the :onsistent with the above results. According broad spectrum of feeding "'entral American cichlids exhibit a are not profound because adaptations, but these adaptations when suitable prey herbivorous fish were able to feed as carnivores the most carnivorous fish were presented and, above all, even food. species readily took inanimate laboratory by nolluscivores Consumption of food material other than snails other food material The demonstration that C. placodon consumes corn-bran is a food such as corn-bran is encouraging because ponds in Malawi. supplementation for tilapine fishes in aquaculture of molluscivores on a This should make it easier to maintain a stock any food inrt specific long tezr basis in aquaculture ponds without should be able to to the -olluscivores. Also the molluscivores snails are in low numbers survive on this food supplementation when kept on commercial or absent. Hoever, the failure of molluscivores further" confixnation food for four weeks to consume snails needs eating habit of this because it may suggest that the snail We also failed molluscivore is plastic under certain ci.cumtances. in the molluscivores. to demonstrate any consumption of cercariae the guppies, Lebistes Such feeding habits have been demonstrated in the molluscivores reticulatus (Pellegrino, et al. 1966), and probably guppies. It would be lack the filtering mechanisms found in the fish. interesting to repeat this experiment with planktivorous

81 were not exhaustive Admittedly our studies on the molluscivores the suitability of the as our main focus was towards testing Nevertheless, the results have molluscivores for disease control. just focusing on the feeding opened up new areas for further research example, several authors have behavior of these molluscivores. For affect food selection. Thompson addressed a number of factors which individual differences toward food (1965) demonstrated the effect of have been total food selection. Other factors in literature food, and predator satiation abundance, spatial distribution of times of different food (Ivlev, 1961); caloric values and consumption be advanced for our molluscivores (Emlem, 1966). These areas would These models describe the by students of optimal foraging models. particular behaviors and, in costs and benefits associated with cost/benefit function addition, solutions which minimize a postulated are derived (Mittelbach, 1981).

82 POIENTAL S AND Sf ICD=C CN 7M HPTER 5 7HE MINCE OF WE: (Cyrtocara placodon. and OF SNAIL-EATING FISH FOR BIOCONTROL OF AstatotilaPia calliptera) sSISc14IASIS nURODUCTIO involves contact with water Schis'-osomiasis transmission 1966). The snails (Farooq and Mallah, harboring infected vector disease cycle the weakest link in the snail vector is considered at the most control efforts are directed (Weil and Kvale, 1985) and knowledge of For effective snail control, snail vector (WH, 1965). and their distribution, survival, ecological conditions controlling the 1962). Furthermore, when abundance is essential (El-Gindy, of using predators, the relationship method of choice is biological, agent to their environment and bhth the target snails and the control the two populations need thorough th_; interrelationships between understanding (Berg, 1964). interaction of two snail eating This study was conducted to test in aquarium conditions simulating fish species with aquatic snails, The and complex habitat (with weeds). simple habitat (without weeds) with that snails tend to be associated rationale for this work is and hence Hira, 1969; cMillen, 1973) aquatic plants (El-Gindy, 1962; whether or not our candidate it is important to investigate habitats. affinity for the complex biocontrol agents have similar and snail habitats are diverse (Bradley Furthermore, considering that in testing the effect of complex Webbe, 1978), we were interested

83 The final objective habitats on snail capture by the molluscivores. buried in sediment was to test an extreme situation of snails being successfully find to see if the candidate biocontrol agents would the prey.

MATERIAIS AND MHS

The effect of weeds on habitat choice by roliuscivores aquaria (100 1), Experiments were conducted in clear glass of habitat complexity having 3 am of sand at the bottom. Two levels placing ten plastic weeds (with and without weeds) were created by of the aquarium. The (approximately 10 cm apart) in one half only aqua scapers, which cam plastic weeds were of the type called Meta 2 cm wide. The leaves in in a bunch of 13 leaves, each leaf about to 30 cm. A vertical line each bunch vary in length fran about 15 cm indicate the boundary was marked on the side of the aquarium to molluscivore was between the two sections. A 24 hour starved acclimated for one introduced at the center of the two sections, to the two treatments hour and thereafter its position in relation Both C. placodon and recorded at 30 second intvals for 75 minutes. Assignment of A. calliptera were tested in this investigation. was random in replicate treatments to the 1.:o aquarium sections experiments.

84 The response of molluscivores to different locations of snails in the habitat protocol, The investigations were designed following the above cage (10 cm x 5 cm x 2 cm except the weeds. One small rectangular each half of the aquarium. high) was placed on the sediment in wrapped around a wire (2 m Each cage consisted of nylon netting while the second was thick) frame. One cage contained ten snails, The snails were visible without snails to control for cage effect. (2 nm) was too small for through the netting, but the mesh size prepared as before was snails to get out. A molluscivore and the frequency of stay introduced at the center of the aquarium C. placodon was tested in in each half recorded as before. Only with the two cages this investigation. The procedure was repeated completely covered with sediment. of wes Predatory efficiency in the presence and absence (50 1), half of which Experiments were perfoned in ten aquaria weeds. The weed contained plastic weeds and the rest were without distributed. To each density was twenty per aquarium, evenly snails (4-7mm) were aquarium, ten similar size B. globosus Next a nolluscivore, introduced and left to acclimate for 24 hours. to each aquarium and prepared by starving as before, was introduced Small size snails the number of snails eaten in four days recorded. (chapter 4 of thesis). were used to control for size limitation

85 RESULTS The effect of weeds on habitat choice by molluscivores sections of the When given a choice between open and weedy SL = 10.70 cmI, SD = 1.56) aquarium, both A. calliptera (n = 10, mean cm, SD = 1.53) moved back and C. placodon (n = 15, mean SL = 11.71 on the average the and forth, between the two sections. However, higher than in frequency of stay among the weeds was significantly 6). the open sections (P>0.01, T-test, Table locations of snails in The response of molluscivores to different the habitat. of C. placodon With the cages placed on sediment, the frequency was significantly higher in (n = 10, mean SL = 13.14 an, SD = 1.46) than in the section with an the aquarium section with a snail cage On several occasions, empfy c;.ge (P<0.01, T-test, Table 6). attempting to take molluscivores approached the cage with snails, buried in sediment the the snails. However, when the cages were (n = 10, mean SL = 11.39 difference in the tine spent by C. placodon was not significant cm, SD = 1.23) in each of the two sections sometimes took (P>0.05, T-test, Table 6). The molluscivores seconds, but this was not mouthfuls of sand and ejected it in few

restricted to the section with buried snails. of weeds predatory efficiency in the presence and absence 20, mean SL = 11.68 Fewer snails were eaten by C. placodon (n = C. placodon (n = 20, mean an, SD = 1.92) in aquaria with weeds than (Mann-Whitney U, SL = 11.84 cm, SD = 1.56) in aquaria without weeds in the two treatments (200 P<0.01, Table 7). The size of snails used

86 different (P>0.05, Ma5n­ snails per treatment) were not significantly most of the snails were attached Whitney U). In aquaria with weeds, randamly without weeds, the snails were to the weeds while in those walls, as well as floating freely distributed on the sediment, glass

on the water surface.

87 Table 6. The average number of thes molluscivores (C. placodon and A. calliptera) were observed in aquaria sections under two

treatment levels of eed presence, snail presence and snail

location. The asterisk (*) indicated the results were not

significant (P>O.05, T test), two asterisks (**) indicate results were significant (P<0.Cl, T test).

Molluscivore Aquarium Average frequency under species (n) condition (treatient) each treatment level

C. placodon (13) Weeds prescnt 110.53 (24.16) ** (15) Weeds absent 39.47 (24.16)

C. placcdon (10) Snails present (on sedinent) 106.00 (20.94) ** (10) Snails absent 44.00 (20.94)

C. placodon (10) Snails present 78.20 (19.32) * (10) Snails absent 71.80 (19.32)

A. calliptera(IO) Weeds present 125.80 (21.87) ** (10) Weeds absent 24.20 (21.87)

88 Table 7. The effect of plastic weeds in aquaria on the number of snails consumed by C. placodon in four days. The differences in the average number of snails eaten from aquaria with and without weeds were significant (P<0.01, Mann-Whitney u)

Average number of snails eaten in four days by each rolluscivore Treatment SD

Weeds present 2.55 (1.79) Weeds absent 8.50 (2.23)

89 DISCUSSION The effect of weeds on habitat choice by mlluscioes molluscjvOres in this study for The strong preference shown by common non-weedy habitats is probably weedy habitats as ccp-ared to the affinity for cover and among fishes. Several authors confirm 1967; Fraser and Cerri, 1982). Affinity structure by fishes ( Lewis, to increased density levels of for structure could be attributed case 1957). This is definitely the prey in such habitats (Gerking, to follow the pattern of for snails whose distribution in ponds tends bottom, floating or emergent (WHO vegetation whether living, dead, 1988 ). 1965; Dazo et al., 1966; Anon, exhibit a similar If the mlluscivores in our investigation for then it would be advantageous behavior pattern in ponds, high molluscivores into areas of biological control to place the the seem to be congruent with snail density. This would the al. (1983) that fishes have demonstration by Werner et in based on relative resource levels flexibility to shift habitats the habitats. for the weeds may not have However, the strog preference encountering large numbers of prey. resulted from the likelihood of 1977; Mittelbach, 1981) have noted Some workers (Hall and Werner, fish to be restricted to weed beds that the general tendency of young of to conform to the predictions in natural lakes does not appear Small bluegills (<100 mm, models specifying optimal habitat use. in feeding rates in open water than SL), for exarnple, achieve higher mostly in vegetation (Mittelbach, vegetation and yet they forage 90 additional dimension to the 1981). These observations suggest an to habitat choice. observed behavior of fishes in relation and Mittelbach (1984) on The work of Werner et al. (1983) evidence that the concentration bluegills has provided experimental in the vegetation of natural lakes of the young of many fish species risk. Hence it appears is a behavioral response to predation in our investigation probably conceivable that the molluscivores made foraging among the weeds displayed a fright response which appear safe. and the fishes natural Further studies in both the laboratory underlying mechanisms for the environment would elucidate the in habitats of different behavioral responses of the molluscivores known concerning predation upon levels of complexities. Little is Nevertheless, C. placodon C. placodon in the natural envirorment. larger fishes of prey such as are possible prey for some of the (Fryer and Iles, 1972). Bagrus meridionalis and predatory cichlids locations of snails in the The response of molluscivores to different

habitat of snails in cages have The results of different positions to detect snails within the dFonstrated that C. placodon is able that the sae is true when water. However, we did not get evidence results suggest that prey the snails are buried in sediment. These a greater extent on visual detection is probably dependent to finding is in agreement with perception rather than chemical. This perception is perhaps the most Stenson (1978) who argued that visual fish. However, important fon of prey detection by most predatory

91 under certain circustance Dther senses could be more important poor, smell shown that when visibility is Studies on the Carp have error towards food and, in bottam guides this fish by trial and of food (Sibbing et al, 1986). feeding, taste is the main detector scooping and spiting back out Although, C. placodon are observed dig it appears that they do not mouthfuls of sand during foraging, contrast a few millimeters. This is in into the sediment more than capable Lake Malawi cichlid, that is to Letrinops furcifer, another of (up to eye level) in search of digging deeper into sediment a 1972). HcwJer, L. furcifer has invertebrate prey (Fryer and Iles, a (Fryer and Iles, 1972) suggesting sharper snout than C. 2lacodon deeper failure of C. placodon digging morphological limitation on the into sediment. sediment might be a disadvantage The lack of prey detection in vector because there is evidence that for schistosomiasis control buried in sediment during their snails became partialy or copletely (WHO, or due to deposition of silt ordinary movement on soft mud 1978). These snails would 1965) or when they estivate (Fergusson, I back into the water. Hver, escape predation and make their way little sediment is probably of the presence of snails in is First, use of molluscivores epidemiological significance. on snail and continuous pressure expected to put long term should be vulnerable to predation populations and snails in sediment from sediment. The temporary by the molluscivores when exposed chemical of greater significance for refuge in sediment should be come out diluted at the tine the snails molluscicides which might be

92 sediment is Df the sediment. Second, the proportion of snails in closer to the likely to be small because the majority tend to be to water surface (Jordan and Webbe, 1969) and should be exposed predation by the molluscivore. Predatory efficiency in the presence and absence of weeds rich growth of Aquatic snails are abundant in waters containing show that suitable water plants (WHO 1965). Our studies eaten from aquaria significantly greater numbers of snails were results suggest without weeds as ccxpared to those with weeds. These efficiency in that the molluscivore (C. placodon) has reduced of this is capturing snails in weedy conditions. The implication vector snails that the effectiveness of C. placodon in controlling plants. wuLld be reduced in habitats with a lot of aquatic with The results from these aquarium experiments are consistent habitat for C. some field studies in Lake Malawi, the natural noted a drop in placodon. Louda et al. (1983), for example, that ths gastropod density in Lake Malawi below 4 m and suggested waters. In the was dui to increased predation in the deeper densities were high shallower waters of Lake Malawi, however, snail against predation, because the macrophyte weed beds provided refuge beds meant that while in the deeper waters absence of these weed (.ouda et al., snails were expq.osed to predation by cichlid predators 1983). was The hypothesis that weeds reduced predatory efficiency to exclude tested by McKaye et al. (1986) by cage experiments the hypothesis molluscivores in open habitats. The results confirmed 93 was observed because an increase of 60% in snail population density by the cages. over open sand wl-re molluscivores had been excluded in Africa. In There have also been similar reports from elsewhere in dense patches of Kenya, greater numbers of Bulinus were found. with A. alluadi, a water weed than open sections of ponds stczked the difficulty snail-eating fish (MacMahon et al., 1977). Similary, (MacMahon et in capturing snails among weeds by fish predators for the large al., 1977) also appears to be a valid explanation and floating number of Bulinus snails found in mats of Ceratophyllum in Ghana recorded by vegetation islands in the man-made Volta Lake Paperna (1969). could be as The basis for reduced vulnerability of the snails random encounters of Crowder and Cooper (1982) have suggested, that In addition, the prey by the predator are reduced by weeds. and Stein, 1982). behavior of the predator could be altered (Savino a significant In their study, Savino and Stein (1982) observed bass (Micropterus altered predatory behavior of the largemouth macrochirus). salmoides) towards its prey, the blue gill (Lepomis captures in dense The predators did not compensate for reduced the bluegills. vegetation by increased searching or following of investigation did Hence it is probable that the molluscivores in 6ur weeds. It is also not increase search effort in the presence of the maintain a tight grip possible that snails sticking to plants be interesting requiring several attacks to be dislodged. What would have affected the to know is how different weed densities would the predatory predatory efficiency of C. placodon. The study on 94 of the behavior of the larguth bass suggest.s that the alteration vegetation predator's behavior is not affected by small changes in Stein, 1982). density but by changes of large magnitude (Savino and C. Despite the reduced effectiveness in weed environments, of schistoscmiasis placodon still has great potential in the control 6 of thesis). as was shown in the pond experiments (chapter aquatic plants (El­ Furthermore, although snails are associated with Pip, 1978), such Gindy, 1962; Calow, 1973; Pip and Stewart, 1976; snail habitats (WHO, vegetation is not the only essential feature of snails appear to be 1965; McMullen, 1973). Also, even though bulinid the occurrence of slightly more dependent than planorbids upon in the absence aquatic vegetation, both types of snails can subsist is available of aquatic vegetation, provided other food material survey of snails in (McMullen, 1973). This is probably why in a were found even in rural farmers' ponds in Zomba, Malawi, snails in prep). Also, ponds that were properly weeded (Chiotha and Jenya C. placodon are since the target of water bodies to be stocked with condition is not fish ponds and sinall water iirnundments, the weed ponds, as would be going to be a big problem even in poorly managed 1981). the case i.n some natural swamps in Malawi (Cantrell, study was probably Furthermore, the aquarium weed density in this deliberately higher than one wuld ever get in fish ponds, but we eme conditions chose this level to test the molluscivores under ext: of weed density.

95 CHAPTER 6 SCHISTOSONE VEW SNAL OWUL IN FISH P IN3 MIMI

USWI FISH

Snails are an important caoponent of the fauna of both temporarY In Africa and other and permanent aquatic habitats (Cridland, 1957). many of these snail species tropical and subtropical countries, and debont hers, 1952; Berrie, transmit human schistosmiasis (debont habitats are used for 1966; McMahon et al., 1977). When aquatic water supply, some of the aquaculture, agriculture, and domestic of molluscicides conventional control methods, such as application practical. Application of and prolonged draining are not because of toxicity to fish ard molluscicides would not be acceptable is also not consistent with other non-target organisms and, draining in many countries where national and local development needs Appleton, 1985; Teesdale, schistoscoiasis is endemic (Berg, 1964; control might be a 1986; Chiotha and Morgan, 1987). Biological in such aquatic habitats better alternative form of snail control (Chiotha and M:Yaye, 1986). infection rates of Malawi has a long history of high Chitsulo, 1983) and the schistoscmiasis (Cull-iierm, 1945; Teesdale and water resource projects in present expansion of aquaculture and other habitats (Teesdale and the country has increased the number of snail ). The earliest attempt Chitsulo, 1985; Chiotha and Jenya in prep. schistosomiasis in Malawi to use biological control agents against crayfish fron Madagascar probably occurred in the late 1930's when

96 were introduced in Mulunguzi and Domasi rivers in Zomba district (E.A.R., 1938). This venture, however, was unsuccessful because no,. there was no effect on snails and the survival of cray-fish could then be verified after several sampling efforts. Attention has since shifted to other biocontrol agents, particularly fish (McCullough, several 1981). With over twenty snail-eating fishes in Malawi, offer a researchers have suggested that these fishes might possibly et al., 1963, means for biocontrol of schistosomiasis (Jackson Pruginin, 1976; McKaye, et al., 1986; Chiotha and McKaye, 1986). The present study was conducted in fish ponds to asses: 1) the inpact of two Lake Malawi snail-eating native fishes, Cyrtocara 2) if these placodon and Cyrtoc anaphymis, on snail populations, two species could coexist with a tilapine fish (Oreochrcmis shiranus) if commonly used for aquaculture in Malawi (Pruginin, 1971), and 3) C. placodon would reproduce in ponds. The successful production and utilization of these snail eating fishes in ponds could allow food production to proceed with reduced or no health risks to the human population from schistosomiasis.

ATERIS AND MEMH!S Expzrental pond studies

Studies were conducted at Bunda College, Malawi in 12 cement p.nds. The size of each pond was 0.0015 ha (5 m x 3 m x 1 m deep). Previous aquaculture experiments in these ponds had created favorable conditions for snails, which were found in large numbers. The most cammon snail species in these ponds was Bulinus tropicus, but Bulinus

97 globosus (vector of human schistoscmiasis) and e spp. were also to get an found. The study began with a snail survey in all ponds from estimate of population size. Five random samples were collected description in WHO each pond using a scoop designed following the square (30 (1965) and Choudhry and Teesdale (1982). The scoop was Previous cn x 30 cm) with a long light metal handle (1.5 meters). provided a studies in these ponds had established that five samples The scoop fair estimate of snail population (Stauffer, pers. ccmm). but in is not practical in aquatic habitats with dense vegetation, were our study the ponds were almost free of macrophytes. Two people by involved in the sampling and the technique was standardized size of the holding the scoop at arm's length. Because of the small The ponds, sampling was done along the edges as well as the center. where the individual samples fran each pond were placed in a sieve, then mud was washed away and the snails counted. The snails were for returned to the same pond after counting. Further suggestions given in improving accuracy of sampling using these techniques are sampler Browr and Zar (1984), such as calibration with a cylinder (also called stove pipe sampler or Wilding sampler). with The experimental designI was that of pair block design random assignment of treatments in each block. The two treatment levels were presence and absence of molluscivores. For treatment five of ponds (presence of mnlluscivores), each pond was stocked with (absence of C. placodon and twenty of 0. shiranus. For control ponds 25 per nolluscivores), each one was stocked with 0. shiranus only, at was within pond. This stocking rate was approximately 17,000/ha and

98 as reported by Pruginin the range of stocking densities in Malawi and Arad (1977). the effect of the This procedure allowed us to test both coexistence with molluscivores (C.placodon) on pond snails and the Weight and standard tilapine non-molluscivores (0. shiranus). before introduction length of the fish were individually determined supplementation in the in the ponds. Each pond received a daily food amount provided an form of corn-bran at a rate of 10 g/day. This as suggested by average feeding rate of about 5% of body weight, (Ann. Fish. Rep. Pruginin and Arad (1977). Maize bean is 12% protein the nutritional needs 1972) and probably does not adequately meet derive additional of the fishes. HuAver, the molluscivore should from the algae in the nutrition fran the snails and the tilapine fish but introduced at pond. The food supplenentation was not broadcast, technique is that two or three points. The advantage of this feeding (Anon, 1988). The it maximizes food intake and reduces food wastage snail survey experiment was terminated in four weeks and another survival and growth conducted. The fishes were then harvested, and times, but in one determined. The experiments were replicated three stocked with C. investigation, some of the treatment ponds were ponds were numbered, anaphyrmis as the molluscivore. Although the ponds were stocked the people sampling snails did not know which with the molluscivore.

99 Experiments in rural fanrers' ponds farmers in This was a pilot study involving three rural fig. 9). Only Namikango area of Zatba district (see map of Malawi, were within a two ponds of each farmer were selected and these ponds simple design, radius of approximately 2 km. All the ponds were of in the being shallow (less than 1 m deep) earth ponds, excavated or from ground and relying on ground water supply from within had sparse external ground wells via small canals. The ponds snail type marginal emergent vegetation, mostly grasses. The common spp. were also in the ponds was B. globosus, but Lymnaea than the encountered. These ponds were about ten times larger done as in the cement experimental ponds above. Snail sampling was larger experimental ponds above. Because the rural farm ponds were from each than the experimental ponds, twenty samples were collected of snails pond, along the edges. In these fish ponds, the majority is not are along the edges of the fish ponds, but distribution uniform (Chiotha and Jenya in prep.). were Treatment levels (presence and absence of mlluscivore) The randomly assigned to the two ponds of each farmer as before. into the farmers were each provided with 30 C. placodon for stocking treatment ponds. All the fish ponds had been stocked with tilapine introduction of C. fishes (0. shiranus and T. rendalli) prior to end of placodon. Snail surveys were conducted at the beginning and was done investigation as before. Stocking and management of ponds by the fanrers to ensure that C. placodon was tested under conditions ran for in which the rural fanrers manage their fish. The experiment

100 survival of C. five months (February-June, 1989) and at harvest by placodon was determined. This experiment was entirely controlled the farmers the fanrers' programme and terminated at five months when harvested the ponds. Breeding trials of C. placodon in experimental ponds newly (1988) This study was done at Bunda College in the x 1 m deep). The constructed brick and cement ponds (1.5 m x 2 m (6 cm thick) to substrate in each pond consisted of a layer of sand sites in Lake mimic the spawning substrate in their natural breeding Malawi. Malawi and C. placodon has lek-based breeding system in Lake cKaye, 1988; several females spawn with a single male (Stauffer and with male YcKaye in press). In this respect the ponds were stocked Only two and female C. placodon in the ratio of 1 to 5 respectively. was seven. ponds were stocked and the total number of fish per pond of increased The advantage of having ore females is the possibility collected production of ova. The olluscivores were fed on snails snails were from nearby earthen ponds. No specific number of were provided but enough were made available to ensure that snails corn-bran (4 g present in the ponds at all tines. Small amounts of The per day) were also supplied to each pond as food supplement. in May experiment was initiated in February (1989) and terminated (1989).

101 RES ETS Experimental pord study

The average number of snails per unit area (per m2) in ponds stocked with C. placodon dropped dramatically in four weeks and approached zero in some ponds. This dramatic drop in snail numbers in the treatment ponds was highly significant as compared to control ponds (Mann-Whitney U, P<0.01, Fig. 6). However, there was no

significant difference in snail numbe.-; between treatment and control

ponds with C. anaphynmis as the molluscivore (P>0.05, Mann-Whitney

U, table 8).

Both 0. shiranus and C. placodon had excellent survival rates

in the fish ponds, averaging above 70% (Table 9). No significant differences in survival were detected between 0. shiranus in

treatment and control ponds (P>0.05, Mann-Whitney U, Table 10).

Survival of C. anaphynmis was poor and at the end of 4 weeks there was not a single specimen alive in the experimental ponds.

Both C. placodon and 0. shiranus gained weight during the four weeks in fish ponds. The average weight gain (24%) for C. placodon was significant (P>0.01, Sig Test, Fig. 7). Similary, 0. shiranus exhibited significant growth (51%) (P<0.05, Mann-Whitney U, Table

11). The difference in the growth rates in the two fish species was

significant (P<0.05, Mann-Whitney U, Table 12). A comparison of weight gain between 0. shi.ranus from control and treatment ponds revealed no significant differences (P>0.05, Mann-Whitney U, Fig.

9).

102 Experiments in rural faners' ponds After 18 months, every pond with C. placodon had significantly fewer snails than any pond without the nolluscivores (P<0. 01, Mann- Whitney U, Table 13). Survival of C. placodon in the three ponds was 40%, 40% and 100% respectively. Breeing trials for C. placodon in cment ponds Fry were noticed swimring close to the water surface after 15 weeks. In pond 1, fifteen fry were collected and from pond 2, seventeen fry were collected. Survival of the parental stock was 100%. The results are sunmarized in Table 14.

103 Fig. 6. Final snail population in experimental ponds at Bunda College, Malawi, after a four week period. Every pond with C. placodon had fewer snails than any without the

molluscivore. The results were highly significant (P<0.01,

Mann-Whitney U)

104 50 45 Mmolluscivore present EZmolluscivore absent u( 4040- 35

-~30- oA N 25­ E 25 Cf) 15 C10- Ix0j it l --I 12 3 4 5 6 7 8 9 1011 121314

Pond pair number with C. anaphyrmis Table 8. The effect of stocking ponds Malawi (molluscivore) on snails in ponds at Bunda College, between initial after a four week period. The difference (P>0. 05, Mann- and final snail numbers was not significant Whitney U).

2 Average number of snails/m (n=5) Treatment final (sd) Pond Initial (sd) no.

10.7 (6.80) 2 12.0 (6.10) 21.3 (8.26) 4 11.6 (6.23) Molluscivore 10.4 (4.04) 12.9 (5.59) 6 16.7 (7.65) 7 present 13.6 (5.72) 17.3 (6.80) 1 15.3 (8.01) Fjlluscivore 12.7 (8.85) 10.9 (3.56) 3 9.1 (6.38) 5 absent 6.2 (5.34) 5.3 (3.35) 8 8.0 (5.58)

106 kept together Table 9. Survival of C. placodon and 0. shiranus in ponds for four weeks at Bunda College, Malawi.

Fish species C. placodon 0. shiranus

260 No. of fish at stocking 70 216 No. of fish at 4 weeks 54 83.08 % Survival 77.14

107 of Q. shiranus in ponds with and Table 10. A comparison of survival without molluscivore (C. placodon). The results indicate in survival of 0. there was no significant difference of C. placodon shiranus in the presence and absence (P>0.05, Mann-Whitney U)

No. of 0.shiranus %Survival Treatment Initial Final 83.08 C. placodon 260 216 present 275 83.33 C. placodon 330 absent

108 4 weeks) of C. Fig. 7. Initial weight and final weight (after 0. placodon (molluscivore). The ponds also contained shiranus

109 60 55 - initial 50 1 final Uj) 45 N 40 + 35 -

I-. c- 30 O C) 25 " 20 ~15­ c 10 ci) 5-

Pond Number Table 11. Weight of 0. shiranus at the beginning of experiment and four weeks later. Ponds with asterisk (*) were also

stocked with C. placodon. The difference between initial

and final weight of 0. shiranus was significant (Mann- Whitney U, P<0.05).

Average weight of 0. shiranus (g) %weight Pond Initial sd Final sd increase

2* 14.50 (5.50) 22.73 (8.92) 56.7596 3* 11.24 (3.57) 15.98 (2.80) 42.1708 5* 11.58 (1.86) 18.68 (2.61) 61.3126 8* 10.89 (2.17) 18.00 (3.90) 65.2893 9* 9.54 (1.16) 13.85 (2.34) 45.1782 12* 7.56 (1.08) 12.91 (3.25) 70.7672 14* 24.81 (8.48) 32.97 (9.84) 32.8900 16* 13.81 (2.48) 24.82 (6.10) 79.7248 18* 12.92 (3.06) 17.65 (4.54) 36.6099 19* 17.43 (3.96) 23.92 (7.02) 37.2347 22* 10.24 (2.12) 18.86 (2.89) 84.1797 23* 11.33 (2.24) 14.96 (2.66) 32.0388 25* 8.66 (1.62) 13.48 (1.78) 55.6582 28* 10.72 (2.78) 17.06 (3.07) 59.1418 1 17.70 (5.08) 23.56 (7.68) 33.1073 4 11.57 (2.72) 15.86 (3.77) 37.0787 6 9.23 (1.07) 15.68 (2.95) 69.8808 7 14.88 (4.87) 18.57 (6.29) 24.7984 10 8.94 (0.91) 13.14 (2.22) 46.9799 11 8.42 (1.41) 13.26 (5.33) 57.4822 13 22.67 (6.23) 32.12 (7.50) 41.6850 15 15.13 (4.63) 22.10 (7.97) 46.0674 17 12.54 (2.22) 18.10 (3.98) 44.3381 20 8.72 (1.70) 15.03 (2.75) 72.3624 21 21.36 (5.10) 34.10 (8.13) 59.6442 24 12.81 (2.71) 23.94 (3.74) 86.8852 26 11.24 (3.10) 18.65 (3.99) 65.9253 27 9.37 (2.20) 13.14 (2.90) 40.2348

ill Fig. 8. weight increase (%) of 0. shiranus from ponds with and without C. placodon in experimental ponds at Bunda College, Malawi. No significant difference was detected between the two groups of 0. shiranus (Mann-Whitney U, P

112 100 90 with C. placodon D8 Ejwithcut C. placodon

70­ o 60 C 4-j 50­ ..-" 40 " 300 0"R20­ 0 10 0 12 3 4 5 6 7 8 9 1011 1213 14 Pond pair number C. placodon and o. Table 12. A comparison of weight gain between shiranus in experimental ponds at Bunda College in four

weeks. Weight gain for 0. shiranus was significant (Mann­ (Mann- Whitney U, P<0.05) but not for the C. placodon

Whitney U, P>0.05)

%weight increase in 4 weeks 0. shiranus Pond C. placodon

56.7586 2 31.3839 42.1708 3 21.0292 17.9389 61.3126 5 65.2883 8 8.5323 24.2518 45.1782 9 70.7672 12 19.4890 32.8900 14 33.1087 79.7248 16 23.5989 6.9492 36.6099 18 37.2347 19 7.4506 26.4721 84.1797 22 32.0388 23 55.5345 55.6582 25 14.0000 59.1418 28 46.2654 33.1073 1 37.0787 4 6 69.8808 24.7984 7 46.9799 10 57.4822 11 41.6850 13 46.0674 15 44.3381 17 72.3624 20 59.6442 21 86.8852 24 65.9253 26 40.2348 27

114 Table 13. The effect of C. placodon (molluscivore) on snails in rural

farmers' ponds at Namikango, Malawi, after an 18 week

period. Every pond with C. placodon had significantly

fewer snails at 18 weeks than control ponds (P<0.01, Mann- Whitney U)

Average number of snails/m2 (n=5) Treatment Pond Initial sd Final sd No.

2 olluscivore 4.5( 4.40) 0.0 (0.00) 4 present 24.3(12.91) 1.1 (1.11) 5 7.0( 5.30) 0.3 (0.47)

1 Mlluscivore 4.3 (4.14) 8.5 (6.67) 3 absent 13.2 (7.19) 10.8 (6.14) 6 7.7 (7.15) 13.45(7.64)

115 Table 14. Results of pilot breeding trial of C. placodon in ponds at

Bunda College, Malawi. Survival of the parental stock was

100%.

No. of C. placodon Parental No. of juveniles produced Pond stock in 15 weeks (Feb.-May)

Male Female

1 2 5 17

2 2 5 15

116 DISCUSSION for This study demonstrates that C. placodon has great potential ponds and rural schistosomiasis control. In both the experimental ponds was fanTers' ponds, the reduction of snail numbers in trtment were also significant as compared to control ponds. The results had fewer snails consistent because every pond with molluscivores fact that the than ponds without. Even more encouraging was the had been only ponds from which we did not detect snail presence stocked with C. placodon­ that The average survival above 70% is remarkable considering Lake Malawi, the the conditions in the ponds are different from important natural habitat for this molluscivore. Another whole growing consideration is that this molluscivore survived the ponds and only a season of tilapine fishes in the rural farmers' tilapine fishes single introduction for each growing season of the This should be should be necessary to get the desired effect. additional economical because there should be no need for McCullough and expenditures or labor after introduction (Berg, 1964; is an added Malek ms). Also a 24% growth rate for C. placodon by humans. advantage because the molluscivores can also be consumed to C. The higher growth rate for 0. shiranus (51%) as compared not of the placodon is probably because the two species of fish were similar age. should The re!,ults in this study also suggest that C. placodon to detect readily be accepted in fish ponds because we were not able The tilapine any harmful effect on tilapine aquaculture species.

117 species used in our experimental ponds actually grew substantially and there was no detectable difference between those harvested from treatment and control ponds. If properly managed, the system should because the tilapine result in net gai '. - public health (Berg, 1964) fish should provide the much needed protein nutrition while the molluscivore should reduce the risk of schistosamiasis infection. The successful trial with rural farmers indicates that the molluscivore can survive under ralistic aquaculture conditions. Furthenrore, since molluscivores spawned in earthen holding ponds and in the pilot spawning trial, pond production of fry for distribution to fanmers is possible. Despite the above advantages of using C. placodon, we consider the results preliminary, awaiting further trials to confirm the results and answer broader practical questions. For example, the lack of coaplete elimination of snails in some of the treatment ponds might raise questions about the epidemiological significance of the few snails that survive predation. Some studies have shown that chances of vector snails getting infected might improve at low snail densities (Ewers, 1964; Carter, et al., 1982; Appleton 1983). However, the snail reduction achieved in our investigations, exceeding 50%, should probably be below the optimal level for snail infections by miracidia. Another consideration regarding few snails that survive a control programme is that they can generate population recovery. However, even when molluscicides are used, hardy snails can persevere after most of the snail population has been destroyed (Dzik, 1983).

118 that of We believe the goal for use of C. placodon should be with vector reduction. This goal is realistic and we are in agreement snails will McCullough (1981) that total elimination of vector Above all, probably never be achieved with biological control. that all reduction is realistic because there is no way of telling the available snails had been eliminated fram a natural habitat with have argued, techniques and equipment. As Brower and Zar (1984) obtain in accurate estimates of absolute density are difficult to useful in aquatic environments. Nevertheless, these techniques ure snails. demonstrating trends in population levels of the pond health What does the goal of snail reduction mean in terms of is generally benefits? For insect pest control programmes, it at least keep accepted that a useful biological control agent should the level of the pest population at equilibrium density that is below The same can causing economic damage (Flanders, 1966, Clarke, 1970). necessary to be true for snail control because it does not seem Schayck, exterminate the snail species to control transmission (Van rates 1986). Several field studies have demonstrated that prevalence low of infected snails in endemic areas are characteristically (Farooq and Hairston, 1966; Sturrock, 1973; inchella and Lorverde, been 1983) and in sme areas infection rates of less than 0.3% have the high human recorded (Van Schayck, 1986). This suggests that rates but infection rates can not be explained by these low infection (McDonald, rather by the high population numbers of vector snails in 1965; Van Schayck, 1986). Hence snail reduction should, This principle, give rise to reduced transmission of the disease.

119 of has been demonstrated on Vieques Island where transmission with schistosomiasis was controlled by keeping vector population low to chemicals (Fergusson et al., 1965). It could also be advantageous such combine control using fish with other environmental techniques seasons. as weed re=n-al and draining the ponds in between stocking health Because schistosomiasis is closely linked to human behavior, disease education should also be part of control programs for this (Teesdale, 1986). Although we have shown that C. placodon can breed in ponds, the enough number of fry produced was too few to be practical. Producing goal. f.-y for distribution to farmers should be the ultimate be too Catching these fishes from the lake by scuba divers would C. expensive and impractical. Also from this study it appears that anaphyrmis would not be suitable for use by the farmrs, basically of because of poor survival in fish ponds. The lack of reduction a direct snails in ponds stocked with this molluscivore is probably from a result of poor survival. Poor survival could have resulted number of factors. One possibility is disease because C. anaphynnis days became easily infected with fungi such as Saproleqnia spp. a few may be a of being moved from Lake halawi. However, poor survival result of failure to adjust to conditions outside Lake Malawi. Furthermore, being an obligate nolluscivore (Stauffer, pers. cam.), C. anaphynmis may have failed to utilize the supplementary feed supplied to the fish ponds.

120 CHAPTER 7 SUMMARY AND COUDING REMARKS

Water resource projects and regional expansion of schistosaniasis Many tropical countries are developing their water resource development projects for a number of uses. The primary development objectives have been generation of hydroelectric power, increased agricultural production, fish farming and water supply for doaestic and industrial use (Bisseru, 1984; Derban, 1983; WHO, 1985). Most of these projects have fulfilled the underlying objectives, but unfortunately the projects have also increased health risks. Schistosomiasis is one of the major health risks fram water -esource projects (WHO, 1985). Since these impoundments create new aquatic habitats, they are easily invaded by schistosome vector snails. These impounded waters also become centers of human activities. Chances of water contamination by feces and urine and frequency of water contact are therefore increased significantly. Few carparisons have been made between disease distribution and intensity before and after creation of aquatic habitats. However, available studies are from major water impoundments and &realmost exclusively concerned with schistosamiasis (WHO, 1985). For ex _:ple, major increases in prevalence have been reported on the Aswan dam in Egypt (2-75%), Sennar dam in Sudan (5-86%) and Akascinbo dam in Ghana (5-90%) (Derban, 1983; WHO, 1985). However, camprehensive data is

lacking for small impoundments (WHO, 1985). In Malawi, increased concern over detrimental effects to health from man-made aquatic habitats led to the formation of the

121 schistosomiasis control schemes in the late 1960's (Teesdale and Chitsulo, 1985). There are no reports of prevalence before and after creation of these habitats in Malawi, but other epidemiological Major studies provide evidence of increased risk of schistosamiasis. evidence for increased transmission is based on the demonstration of higher prevalence of schistosomiasis in villages sited near water impoundments than those further away fram the impoundments (Teesdale and Chitsulo, 1985). In one study, prevalence was 37% in villages near a man-made lake and 0% in villages further away fram the lake (Teesdale and Chitsulo, 1985). Evidence of additional risk is based on the distribution of schistosomiasis in the country. As indicated S. before (chapter 2 of thesis), Malawi is endemic for S. mansoni and haematobium, transmitted by B. pfeifferi and B. globosus respectively. Creation of man-made lakes in same areas of Malawi has resulted in introduction of a second fonn of schistosomiasis where only one existed before. In Chileka area of Blantyre district (See map of Malawi, Fig. 9), transmission of S. mansoni is focal, associated only with man-made lakes, while the rest of the natural habitats transmit S. haematobium (Teesdale and Chitsulo, 1985). A different pattern has emerged in Lilongwe district of Malawi, where S. mansoni transmission is associated with natural habitats but focal transmission of S. haematobium takes place solely in man-made lakes (Teesdale and Chitsulo, 1985). The existence of B. globosus and B. pfeifferi in water impoundments in Lilongwe and Blantyre respectively, suggests that the water impoundments create favorable conditions for natural

122 invasion of the waters by a second snail vector. A similar situation has been reported in Egypt where the irrigation projects on the Nile Delta led to increased abundance of Biomphalaria spp. in areas where Bulinus spp. were predominant prior to the project (Garfield, 1986). Furthermore, the movent of fish from one region of Malawi to another may have introduced eggs of new vector snail into man­ made lakes where they did not exist before (Chitsulo, pers caomm.). Similary, the introduction of Biamphalaria straminea vector snails into Australia has been connected with consignments of aquarium fish from Hong Kong (Dudgeon, et al., 1983). Further evidence of increased risk of transmission of small imoundments in Malawi is based on a study conducted in early 1989, sponsored by Malawi-ICLARM-GTZ aquaculture project (Chiotha and Jenya, in prep.). In this study 26 fish ponds in the Namikango area of Zomba district were surveyed for vector snails and the results showed that over 70% of the ponds harbored B. globosus. This confirms the presence of vectors in the majority of the ponds. In some of the areas where the ponds have been constructed, natural pools of water dried completely during the dry season. The fish ponds, however, have allowed water to persist through out the year. This persistence of large number of aquatic habitats critically affects the rate of reinfection in populations treated against schistosomiasis. For example, reinfection rates in Malawi are naturally low during years of low rainfall (Pugh and Teesdale, 1984). For ponds constructed in swampy areas of Namikango, significantly more snails were found in fish ponds than from the

123 swamps surrounding the ponds. This suggests, that in most areas, the ponds have amplified snail numbers. This amplification may be explained by the fact that conditions suitable for fish culture, such as increased primary productivity, are also ideal for snails (Raut, 1986). A test of snail infection by exposing 30 randomly selected snails from each pond to light (See appendix 3) and microscope did examination of crushed snails when the light test was negative, not detect infection in all but one snail. This suggests that infection rates among the snails are probably low. However, infection rates in most endemic areas are low (Farooq and Hairston, 1966; Sturrock, 1973; Minchella andLorverde, 1983). More frequent sampling and larger samples than was possible in our study, are required to detect infection. In any case, infection rates are difficult to interpret in terms of their epidemiological significance (Farooq and Hairston, 1966) and the critical test will be measuring infection rates in the human population at risk . We also noticed that sore of the fish ponds in Namikango were also constructed within 10 meters of houses and that pond water was used for domstic jobs not related to the keeping of fish. This proximity of fish ponds to dwelling units and the dependence on the water for dcmestic uses will increase chances of contamination and infection.

124 schistoscuiBss control in water inpoundments in Malawi the risk of Water impoundments in Malawi have clearly increased associated with schistoscmiasis. Increases in disease prevalence across the greater use of water have been simultaneously reported water development whole continent of Africa in countries engaged in these man-made projects (WHO, 1985). Control of transmission from functions aquatic habitats is probleatic. Because of the important damestic water served by these impoundments, such as fish production, if anything, supply and irrigation, they can not be eliminated and, of molluscicides their numbers will continue to grow. Application high cost of would not be acceptable because of fish kills and the limitations can purchase. Teesdale (1986) has argued that these two health budget not be overlooked in a country like Malawi, with a low water and where fisheries is an important industry. Furthermore, in prep. )and contact is inevitable during fishing (Chiotha and Jenya other dcrestic activities (Teesdale, 1986). due to Because of the possible spread of schistosomiasis to increased pond construction, the present research was conducted Cyrtocara study snail-eating habits of Astatotilapia calliptera, fron Malawi, placodon, and Cyrtoc anaphyrmis, three rolluscivores The snail­ for possible use as biocontrol agent of schistosamiasis. eating habit of these species and potential for schistosomiasis et al., 1963; control had been suggested in earlier studies (Jackson, Chiotha Fryer and Iles, 1972; Pruginin, 1976; NcKaye et al., 1986; and YzKaye, 1986). A. calliptera and C. placodon are facultative

125 molluscivores while C. anaphyrmis is an obligate molluscivore

(Stauffer, pers.cam.). the species The results of this investigation suggest that, of controlling studied, only C. placodon has potential for placodon schistosomiasis in impounded water. This is because C. ponds caused significant reduction in snail numbers from experimental placodon and in 3 rural farm ponds. Other useful attributes of C. with 0. included significant growth in fish ponds and coexistence

shiranus, an important aquaculture species in Malawi. placodon These results clearly demonstrate that wild caught C. pond-reared C. will consume snail vectors. Now, further trials with this species placodon are required to determine the full potential of

as a biological control agent.

126 APPENDIX 1 SNAIL OCUBCTION AND TRA1 I(OFlCN These Three species of snails were used in the investigations. haematobium), Biomphalaria were Bulinus qlobosus (vector of S. and Bulinus troicus (not a vector pfeifferi (vector of S. mansoni) College and the main of human schistosames). The fish ponds at Bunda B. tppicus, especially the reservoir dam are heavily infested with manure for raising ponds that are fertilized with pig and chicken were collected from a tilapine fishes. Biomphalaria pfeifferi and Thondwe Dam (Zomba reservoir dam at Nkhom (Lilongwe district) was never found district). The third snail species, Bulinus qlobosus Dzmasi Fish Farm and at or near Bunda College and was collected from Malawi (fig. 9 ) for the Lake Chilwa (Zomba district). See map of districts mentioned in this dissertation. were used in Scoops as described in Choudhry and Teesdale (1982) that some of the collection of snails. Because of the possibility with schistosomiasis, all the vector snails could be infected were taken during possible precautions to avoid getting infected of rubber gloves and collection. These precautions included wearing vector snails or water wading boots to avoid direct contact with the (30 am long) were used to from the snail habitats. Long tweezers pick snails from the scoops. buckets with The snails were transported in 15 liter plastic liters). Leaves of enough water to keep the snails afloat (about 2 habitat, were also aquatic plants, mostly grasses from the snail as possible during ismersed in the water, to keep the water as stable substrate for transportation and also to provide additional

127 attachment of snails. This method of moving snails was found convenient and hundreds of snails were transported with no mortality at all. Upon reaching the laboratory, the snails were transferred into glass aquaria. The water was aerated and the snails fed with lettuce and cabbage. After one week of acclimation, the snails were ready for use in the feeding experiments.

128 Fig. 9. Map of Malawi. The districts shown are: 1. Karonga

2. MRmphi

3. Nkhatabay

4. Mzinba 5. Nkhota-kota

6. Kasungu

7. Lilongwe 8. Mangochi

9. Zrba

10. Blantyre

129 I Io

- - - " - a a~. • I APPENDIX 2 FISH OELECTICN AND TRANPORICN Four species of fish were used in the investigations. These were Cyrtocara placodon, Cyrtocara anaphyrmis, Astatotilapia calliptera and Oreochromis shiranus. The first two species were collected fra the southern end of Lake Malawi. All C. placodon were collected mostly fron Cape Maclear and occasionally from Chirombo Bay, both near Monkey Bay (Mangochi district). The choice of Cape Maclear for C. anaphyxmis was based on increased availability of this species in Cape Maclear as compared to Chirombo Bay. The third mlluscivore, A. calliptera and 0. shiranus, were collected from the main reservoir dam at Bunda College and Domasi Fish Farm. For the lake species, gill nets were used as block nets. Being deep water fish (5-15 m), the fish were chased into the nets by scuba divers. The mesh size was such that the fish did not go through or get caught by the gills. The fish were sorted out into the right species under the water and then put in holding cages made of nylon netting. These holding cages were kept in the water for at least 24 hours to eliminate weak ones in preparation for transportation. Survival during transportation was better for fish held overnight than those transported soon after catching. Two techniques were used for transporting the fish from the lake. The first method involved carrying the fish in a 200 liter drum. With this method, an average of 60 fish were transported without any difficulty. The drum was filled to about a quarter with water and air was supplied by an aerator connected to the car battery. The trip to Bunda took an average of 4 hours. To reduce

131 morning or late mortality, the fish were transported in the early the fish were afternoon hours when it was cooler. At Bunda College, together with transferred to smaller and shallow (20 1) containers Gradually the enough lake water to keep the fish fully submerged. holding ponds, shallow containers were filled with water from cerent the fish a process lasting approximately 20 minutes and thereafter gradual mixing were transferred into the cement holding ponds. This the fish to sudden of pond and lake water was done to avoid exposing the lake water was temperature change which could be lethal. Also, which might not poured into the ponds to avoid transferring parasites have moved into the water from the fish during transportation. gem which In the cement holding tanks, the fish were fed maize 1972). Highest contains about 12% protein (Annual Pep. Fish. Dept., the mortalities were seen within the first 24 hours of transporting that the fish when we lost about 5 to 10%. Later it was discovered fish did not cement holding ponds were not satisfactory because the At first it survive more than a month, but gradually disappeared. (Scopus was felt that the many carnivorous birds such as, hamerkop them because umbretta) and grey heron fAnda dinerea), were eating was there was no trace of dead fish. Control of the loss of fish The outcome attempted by covering the holding ponds with netting. us to of this action was not a reduction in mortality, -but it enabled that the see the dead fish under the nets. Hence, we suspected the day and cement tanks probably became too hot for the fish during were dead and that the carnivorous birds were picking off fish that to the floating on the surface, or stressed ones swimming close

132 surface. We then started holding the fish in one of the earthen ponds and survival was much better. Sam fish survived for nore than 6 months and spawned. The second method of transporting the fish made use of polyethylene bags. These bags could hold a maximum of 20 liters of water but were only filled with about a quarter of that volume. On the average, 5 fish were put in each bag and then filled with oxygen. Using rubber bands, the ends of the bags were tightly closed. Double bagging was done to prevent the inner bag from accidental puncture and deflation. The bags were packed in styloform boxes, or cooler boxes to reduce temerature fluctuations. When it was too hot, a little ice was placed at the bottom of the box. Ideally, temperature should be between 24 C and 28 C for optimum diffusion of oxygen into the water (Guerrero 1987). Collection of A. calliptera and 0. shiranus was done using seine nets since the ponds at Bunda and Domasi are shallow (maximum depth 1 M).

133 APPENDIX 3 EStIMATIO OF CERCARIAL DENSITY were taken from the Ten random water samples (1 ml each) and transferred to a 20 aquarium containing cercariae using pipettes up with 5 ml of hot water to ml test tube. The test tube was topped left to stand for 5 hours stun the cercariae. The test tube was then the bottom, after which most of to allow the cercariae to settle at ml. To the remaining water, the water was decanted, leaving about 2 to stain the cercariae. By 3 drops of 2% Lugols iodine were added of cercariae to microscope transferring the stained suspension counted under a low power slides, the number of cercariae was fitted with paraffin wax dissecting microscope. Concavity slides, to allow for greater rings, were used for the microscope examination than would be possible volume of cercarial suspension to be analyzed on Choudhry and Teesdale with ordinary slides. This method was based (1982) and Pellegrino et al. (1962).

134 REFEREES of the Alexander, R. McN. 1967. The functions and mechanism Zool. protrusile upper jaws of some acanthopterygian fish. J. Lond. 151: 43-64. and A. N. Amin, M. A., A. Fenwick, J. M. Osgerby, A. P. Warley with Wright. 1976. A large scale snail control trial Wld. trifenmorph in the Gezira irrigation scheme, Sudan. Bull. Hlth. Org. 54: 573-588. Anderson, R.M. & May, R.M. 1979. Prevalence of schistosome infections within molluscan populations: observed patterns and theoretical predictions. Parasitology. 79: 63-94. Annual Report 1972. Fisheries Department: Ministry of Agriculture and Natural Resources. Government Printer, Zomba, Malawi. Anon, 1904. Medical and surgical reports, 1899-1900. Likoma Diocese. Malawi National Archives. NA/ 145/ DC/ 4/ 1. Anon, 1988. Fish farming in ponds. Ministry of Agricultural Education Dept. The Hague, Netherlands. Ansari, N. 1973. Epidemiology and control of schistosomiasis (bilharziasis). University Park Place, Baltimore. Appleton, C. C. 1983. Wetlands and public health. J. Limnol. Soc. Sth. Afr. 9(2):117-122. Appleton, C.C. 1985. Molluscicides in bilharziasis control - the South African experience. Sth. Afr. J. Sc. 81: 356-360.. Barbour, A.D. 1982. Schistosomiasis. In R.M. Anderson (ed.) Population dynamics of infectious diseases. Theory and application. springer-Verlag, New York.

135 Bard, J. and Movongo, 1963. Note d'information sur

l'Astatoreochromis alluadi poisson molluscophage utilisable

dans la prophylaxie de la bilharziose. Bull. Soc. Path. Exot. 56: 119-120.

Barlcw, G. W. 1974. Contrasts in social behavior between Central American Cichlid fishes and coral-reef surgeon fishes. Amer. Zool. 14:9-34.

Barnish, G. and Prentice, M. A. 1982. Predation of the snail

Biomphalaria glabrata by freshwater shrimps in St. Lucia, West

Indies. Ann. Trop. Med. Parasit. 76(1): 117-120.

Belding, D.L. 1965. Textbook of parasitology. Appleton-Century- Crofts, New York.

Bell, E. J., F. J. Etgets and L. J. Jennelle. 1966. The identity

and source of a molluscicide-degrading bacterium. Am. J. Trop.

Med. Hyg. 15: 539-543.

Berg, C. 0. 1964. Snail control in trematode diseases: the possible

value of schizcuyzid larvae snail-killing diptera. Adv. Parasit. 2: 259-309.

Berg, C.O. 1973. Biological control of snail borne diseases. Rev. Exp. Parasit. 33: 318-330.

Berrie, A. D. 1966. Fish ponds in relation to the transmission of bilharziasis in East Africa. E. Afr. Agric. Forest. J. 31 (3): 276-282.

Berrie, A. D. and S. A. Viser. 1963. Investigation of a growth­ inhibiting substance affecting a natural population of fresh water snails. Physiol. Zool. 36 :167-173.

136 Berry, E. G., M. 0. Nolan and 0. Gonzales. 1950. Field tests of molluscicides against A. glabratus in endemic areas of

schistosomiasis in Puerto Rico. Publ. Hlth. Rep. 65: 939-950.

Bertness, M. D., Garrity, S. C., and Leving, S. C. 1981. Predation

pressure and gastropod foraging: a tropical-temperate

comparison. Evolution 33: 417-431.

Bertram, C. K. R., Borley, H. U. H. and Trewavas, E. 1942. Report on the fish and fisheries of Lake Nyasa. London: Crown Agents for the Colonies.

Bisseru, B. 1984. Epidemiology and control of Schistosoiasis. Post graduate Doctor: 75-82.

Blair, D. M. 1956. Bilharziasis survey in British West and East

Africa, Nyasaland and the Rhodesias. Bull. Wld. Hlth. Org. 15: 203-273.

Bohl, E. 1982. Food supply and prey selection in planktivorous

cyprinidae. Oecologia. 53: 134-138.

Bradley, D. J. and Webbe, G. 1978. Ecological and habitat methods

in the schistosomiasis control. Proc. Int. conf. Schist.:

691-706.

Brower, J. E. and Zar, J. H. 1984. Field and laboratory methods for

general ecology. a. C. Brown Publishers. Dubuque, Iowa.

Brown, A. W. A. 1978. Ecology of pesticides. John Wiley and Sons. New York.

Brown, K. M. 1979. The adaptive demography of four fresh water pulmonate snails. Evolution 33:417-431.

137 Africa and their Medical Brown, D.S. 1980. Freshwater snails of Importance. Tylor and Francis Ltd. I-ondon. and the Brown, K. M. and DeVries, D. R. 1985. Predation Oecologia distribution and abundance of a pulmonate snail. 66:93-99. of female Bullough, C.H.W. 1976. Infertility and bilharziasis 83: genital tract. British J. Obstetrics and Gynaecology 819-822. Malham Tarn, Calw, P. 1973. Gastropod association within Yorkshire. Freshw. Biol. 3: 521-534. (Pisces, Campbell, K.L.I. 1981. Fishes of the genus Sarotherodon Kenya. J. Cichlidae) of springs along the Northern Oaso Ngiro, East. African Nat. Hist. Soc. 173, 12 pp. level fluctuation Cantrell, M. A. 1981. L11harzia snails and water in a tropical swamp. Oikos 36: 226-232. a Carney, W. P., Sudomo, M. and Purnomo. 1980. Echinostomiasis: disease that disappeared. Trop. Geogr. Med. 32: 101-105. R. A. 1982. Carter, N. P., Anderson, R. M., and Wilson, to snail: Transmission of Schis-osoma mansoni from man miracidial Laboratory studies on the influence of snail and densities on transmission success. Parasit. 85: 361-372. 1948. The Castle, W.M., Clarke, V. de V., & Hendtikz, E., in effects of subclinical bilharziasis on mental ability school children. South Afr. Med.J.48: 2035-2038.

138 ode of entry, action, and Cheng, T. C. and Sullivan, J. T. 1974. In: Cheng, T. C. (Ed.), toxicity of copper molluscicides. pp. 89-154. Academic Molluscicides in Schistosomiasis control.

Press Inc. New York. 1962. The effects of copper Chi, L. W. and L. R. Wrinkler. sulfonated hydrocarbon sulphate, sodium pentachlorophenate and a Trop. Med. Hyg. 11: 851-854. on the eggs of Oncmlania. Am. J. 1986. Screening of indigenous Chiotha, S. S. and J. D. Msonthi. bilharzia transmitting plants for possible use in controlling 193-197. snails in Malawi. Fitoterapia LVII(3): 1986. Possible biological control Chiotha, S. S. and McKaye, K. R. by Lake Malawi molluscivores of schistosomiasis (bilharzia) 11-24. Luso: J. Sc. Tech. (Malawi). 7(1): 1987. IMlluscicides used to control Chiotha, S. S. and Morgan, R. P. (Malawi). 8 (1&2): 11-22. schistosomiasis. Luso: J. Sc. Techn. The potential of fish ponds in Chiotha, S. S. and Jenya, C. In prep. transmission. Malawi­ Malawi in bilharzia (schistosomiasis)

ICIARM-GTZ aquaculture project report. E. In prep. The possible Chiotha, S. S., Seyani, J.H. and Fabiano, and managemnt problem use of local plants in solving health Malawi. Malawi-ICLARM-GTZ associated with fish farming in

aquaculture project report. years. World Health. WHO Chitsulo, L. A. 1984. Three successful magazine, pp. 18-19. 1982. Bilharzia. A manual for Choudhry, A. W. and C. H. Teesdale. Ltd., Nairobi. health workers in Malawi. Ciba-Greg

139 situation through the life Clarke, L. R. 1970. Analysis of pest L. Rabb and F. E. Guthrie, systems approach. pp. 45-57. In R. Carolina University Press, concepts of pest manageent. North

Raleigh, N.C. in of transmission of bilharziasis Clarke, V.deV. 1977. Patterns 2-5). Rhodesia. C. Afr. J. Med. 23(11): in the North Nyasa Conran, P.G. 1913. A report of ankylostomiasis pp68. District. Ann. Med. Rep. Nyasaland, and mode of action. John Cremlyn, R. 1978. Pesticides: preparation

Wiley and Sons. New York. factors affecting the numbers of Cridland, C. C. 1957. Ecological J. Trap. Med. Hyg. 60: snails in temporary bodies of water. 287-293. E. 1982. Habitat structural Crowder, L. B., and Cooper, W. between bluegills and their prey. complexity and the interaction

Ecol. 63(6): 1802-1813. disorders in East Africa. Trans. Cullinam, E. R. 1945. Medical (5): 353-372. Roy. Soc. Trop. Med. Hyg. XXXIX Amin, M.A. 1985. The lungfish, Daffala, A.A., Ellias, E.E. and a biocontrol agent against Protopterus annectans (Owen) as Med. 88(2): 131-134. schistosom vector snails. J. Trop. useful plants of West Africa. Crown Dalziel, J. M. 1937. The agents. London. Wld. Hlth. pp 14-15. Dazo, B.C. 1984. Outlook bright.

140 Dazo, B. C., Hairston, N. G., and Dawood, I. K. 1966. The Ecology alexandrina and its of Bulinus truncatus and Biomphalaria 4 9 Bilharziasis in the Egypt- implications for the Control of 35: 339-356. Project area. Bull. Wld. Hith. Org. M.J. 1952. Mollusc control and deBont, A.F. and deBont Hers, Nature London. 170: 323-324. fish fanning in central Africa. control of freshwater molluscs deBont, A.F. 1956 a. Biological them. Annls. Soc. Belg. Med. and of diseases transmitted by Trop. 36: 667-672. les mollusques dans les eaux deBont, A.F. 1956 b. Lutte contre Belge 47: 337-380. africaines. Bull. Agricole du Congo M.J. 1956. Haplochromis mellandi deBont, A.F. and deBont Hers, Cichlidae). Rev. Zool. Bot. Blgr. poisson malacophage (Faro. Afr. 53: 370-376. and bilharzia: lessons from Derban, L. K. A. 1983. Man-made lakes the Volta. February/March: 24-25. field C. 1985. Observations on the Donnelly, A. F. and Appleton, C. mansoni and S. matheei in transmission dynamics of Schistoscma 91: 281-290. southern Natal, South Africa. Parasitology report on the gastropod fauna of Dugeon, D., Yipp, M.W. 1983. A with special reference to aquarium fish farms in HongKong Malacol. Rev. 16(1-2): introduced schistostosm host species. 530 splenaregaly in Central Africa. Dye, W. H. 1924. Schistosomiasis and J. Roy. Amy med. Corps. 43(3): 161-181.

141 Dzik, A. J. 1983. Snails, Schistosomiasis and irrigation in the tropics. Public Health (London) 97(4): 214-217.

E.A.R. 1938. cray-fish fish and bilharzia. East Africa and

Rhodesia, 17th Feb., 1938. Malawi National Archives S1/284/37. Eisenberg, R. M. 1966. The regulation of density in a natural

population of the pond snail, Lymnaea elodes. Ecology 47: 889­ 906.

Eisenberg, R. M. 1970. The role of food in the regulation of pond snail, Lymnaea elodes. Ecol. 51: 680-684.

El-Gindy, H. I. 1962. Ecology of snail vectors of bilharziasis.

Proc. 1st Int. Symposium on bilharziasis Cairo Govt. Print, pp. 305-318.

El Karim, M. A. A, Collins, K. J., Brotherhood, J. R., Dore, C.,

Weimer, J. S., Sukkar, M.Y., Omer, A. H. S. and Amin, M. A.

1980. Quantitative egg excretion and work capacity in a

Gezira population infected with Schistosoma mansoni. Am. J. Trop. Med. Hyg. 29: 54-61.

Emlem, J. M. 1966. The role of time and energy in food preference. Am. Nat. 100: 603-609.

Ewers, W. H. 1964. The influence of the density of snails on the

incidence of larval t reatodes. Parasit. 54: 579-583.

Farooq, M. and Hairston, N. G. 1966. The epidemiology of

Schistosoma haematobiun and S. mansoni infections in the Egypt­

49 Project area. Bull. Wld. Hlth. Org. 35: 331-338.

142 social Farooq, M. and Mallah, M. B. 1966. The behavioral pattern of and religious water-contact activities in the Egypt-49

Bilharziasis project area. Bull. Wld. Hlth. Org. 35: 377-387.

Farooq, M. 1973. Historical development. pp. 1-4. In: Ansari (Ed.), Epidemiology and control of schistosomiasis

(Bilharziasis). haematobium Faust, E.C. 1924. The reaction of S. japonicum and S. in the presence of their intermediate hosts. J. Parasit. 10 : 199-204.

Fergusson, F. F., Dawood, J. K. and Blondeau, R. 1965. Preliminary

field trials of Acrolein in the Sudan. Bull. Wld. Hlth. Org.. 32: 243-248.

Fergusson, F.F. 1978. The role of biological agents in the control of schistoscre bearing snails. U.S. Dept. of Health, Education and Welfare/ Public Health Service. CDC/ Bureau of Laboratories/ Atlanta, Georgia 30333. pp.107. Finkel, A. J. 1982. Industrial toxicology. John Wright. Boston.

Flanders, S. E. 1966. The circumstances of parasitic replacement among parasitic Hymenoptera. Canadian Entomologist 98: 1009­ 1024.

Forsyth, D.M. and Bradley, D.J. 1964. Irreversible damage by

Schistoscma haematobium in school children. Lancet. 2 : 169-171.

Foster, R. 1967. Schistosomiasis on an irrigated estate in East Africa. III. Effects of asymptomatic infection on health and industrial efficiency. J. Trop. Med. Hyg. 70: 185-195.

143 Fox, I., L. A. Berrios-Duran, L. P. Frick and L. S. Ritchie. 1963. Detoxification of ICI 24223 by free chlorine and by certain pH

levels. J. Parasit. 49: 648-652.

Frandsen, F. and Madsen, H. 1979. A synopsis of Helisrma durvi in biological control. Acta tropica. 36: 67-84.

Fraser, D. F. and Cerri, D. R. 1982. Experimntal evaluation of predator-prey relationships in a patchy environment: consequences for habitat use patterns in minnows. Ecol. 63(2): 307-313.

Fryer, G. and Iles, T. D. 1972. The Cichlid fishes of the Great

Lakes of Africa. Olivier and Boyd, Edinburgh.

Fukami, J. 1976. Insecticides as inhibitors of respiration. pp.

353-396. In: Wilkinson, C. F. (Ed.), Insect biochemistry and physiology. Plenum Press. New York.

Gaddal, A.A. 1984. Blue Nile, Half time report. Wld. Hlth. pp 20-21.

Garnet, A., Brottes, H. and Movogo, L. 1964. Premiers essais de

luthe contre les vecteurs des bilharzioses dans les etangs

d'une station de pilciculture au Cameroun. Bull. Soc. Path. Exot. 67: 118-120.

Gariield, E. 1986. Schistosomiasis: the scourge of third world. Part 1. Etiology. Current contents. 109): 3-7.

Gerking, S. D. 1957. A method of sampling the littoral macrofauna and its application. Ecol. 38: 219-225.

144 Goddard, x. and Jordan, P. 1980. On the longevity of Schistosoma mansoni in man in St. Lucia, West Indies. Trans. Roy. Soc. Trop.

Med. Hyg. 74: 185-191. the Gonnert, R. 1961. Results of laboratory and field trials with

nolluscicide Bayer 73. Bul. Wld. Hlth. Org. 25: 533-542.

Gospil, W. L. 1937. Some observations on schistosomiasis in the

North Nyasa District. Ann. Ned. Rep. Nyasaland. pp 395-396.

Greenwood, P.H. 1965. Environmental effect on the pharyngeal

mill of the cichlid fish, Astatoreochromis alluadi, and

their taxonomic implications. Proc. Linn. Soc. Lond. 176: 1­

10.

Guerrero III, R.D. 1987. Tilapia farming in the Philippines. Technology and Livelihood Resource Center.

Hairston, N. G. 1973. The dynamics of transmission. In:

Epidemiology and Control of Schistosomiasis (Bilharziasis).

Ansari, N. (Ed.). University Park Place, Baltimore.

Hall, D. J. and Werner, E. E. 1977. Seasonal distribution and

abundance of fishes in littoral zone of a Michigan Lake. Trans.

Am. Fish. Soc. 106: 545-555.

Harbourne, J. B. 1972. Phytochenical ecology. Academic press, New York.

Hartley, C. F. and T. F. West. 1969. Chemicals for pest control. Pergamon press, Oxford.

Hartman, G. F. 195n. Mouth size and food size in young rainbow

trout, Salmo airdneri. Copeia: 233-234.

145 Hatch, Jeremy. 1970. Predation and piracy by gulls a ternery in Maine. The Auk. 87: 244-254. Hicks, R. M.; James, C.; Webbe, G. and Nelson, G. S. 1977. Schistoscaa haematobin and bladder cancer. Laboratory Meeting on Schistosomiasis Demonstrations. Roy. Soc.TrOp. Med. Hyg.: 288. Hira, P. R. 1969. Transmission of schistosamiasis in Lake Kariba, Zambia. Nature 224: 670-672. Hobson, E. S. 1968. Predatory behavior of some shore fishes in the Gulf of California. Res. Rep. U.S. Fish Wildl. Serv. 73: 1-92. Hoffman, D. B., Lehman, J. S., Scott, V. C., Warren, K. S. and Webbe, G. 1979. Control of schistosomiasis - Report of a Workshop. Am. J. Trop. Hyg. 2: 249-259. Holling, C. S. 1966. The functional response of invertebrate predators to prey density. Mon. Enttowl. Soc. Can. 48: 1-86. Hoogerhoud, Ms. Optimal prey processing in molluscivorous cichlids Part 1. Leiden, The Netherlands. Hoogerhoud, R.J.C. 1987. The adverse effects of shell ingestion for molluscivorous cichlids, a constructional morphological approach. Netherlands Journal of Zoology 37 (3-4): 277-300. Horn, M. H. 1983. Optimal diets in ccoplex environments: Feeding strategies of two herbivorous fishes from a temperate rocky intertidal zone. Oecologia. 58: 345-350.

Huffaker, C.B. 1958. Experimental studies on predation : dispersion factors and predator-prey oscillations. Hilgardia 27: 343-383.

146 Hunter, R. D. 1975. Growth, fecundity, and biogenetics in three populations of Lymnaea palustris in upstate New York. E. Col. 56: 50-63. Hunter, G. W., K. Okabe, J. C. Burke and J. Williams. 1962. Aspects of schistosomiasis control and eradication with special reference to Japan. Ann. Tcp. med. and Parasit. 56: 302-313. Ivlev, V. S. 1961. Experinental Ecology of the feeding of fishes. Yale University Press, New Haven. Jackson, P.B.N., Iles, T.D., Harding, D., and Fryer, G. 1963. Report on the survey of Northern Lake Nyasa 1945-1955. Joint Fisheries Research Organization. Gov. Printer, Zomba, Nyasaland. Jordan, P. and Randall, K. 1962. Bilharziasis in Tanganyika: Observations on its effects and the effects of treatment in school children. J. Trop. Med. Hyg. 65: 1-6. Jordan, P. and Webbe, G. 1969. Human schistosomiasis. Charles Thomas. Springfield. Jordan, P., Christie, J. D. and Unrau, G. 0. 1980. Schistosomiasis transmission with particular reference to ecological and biological methods of control. Acta Tropica. 37: 95-135. Kansndo, W. Y., Chiotha, S. S. and Msonthi, J. D. 1985. Screening of plants used traditionally to control schistosomiasis in Malawi. Fitoterapia. LVI (4): 229-232.

Keenleyside, M. H. A. 1955. Sae aspects of the schooling behavior of fishes. Behavior 8: 183-248.

147 Keenleyside, M. H. A. 1979. Diversity and adaptation in fish

behavior. Springer-Verlag. New York.

Kuris, A. M. 1973. Biological control: Implications for the analogy between the trophic interaction of insect pest­ parasitoid and snail trematode systems. Exp. Parasit. 33: 365-379.

Laracuente, A., Brown, R. A., and Jobin, W. 1979. Comparison of four species of snails as potential decoys to intercept

Schistosomemiracidia. Am. J. Trop. Med. Hyg. 28(1): 99-105.

Lawrence, J. M. 1957. Estimated sizes of various forage fishes

largemuth bass can swallc. In: Proc. S. E. Ass. Game Fish.

Cann. 11: 220-226.

Leewen, J. Van and Muller, M. 1984. Optimum sucking techniques for

predatory fish. Trans. Zool. Soc. Iond. 37: 137-169. lehninger, A. L. 1978. Biochemistry. Worth. Publishers, Inc. Ifena, A. 1970. Laboratory and field evaluation of the

mnlluscicidal properties of Phytolacca dodecandra. Bul. Wld. Hlth. Org. 42: 597-612.

Lewis, S. L. 1967. Physical factors influencing fish populations in pools of a trout stream. Trans. Am. Fish. Soc. 98: 14-19. Lie, K. J., Basch, P. F., Heyneman, A. J. 1968. Implications for trematode control of interspecific larval antagonism within snail hosts. Trans. Roy. Soc. Trop. med. Hyg. 62: 299-319.

Liem, K.F. 1973. Evolutionary strategies and morphological

innovations: cichlid pharyngeal jaws. Syst. Zool. 22: 425­ 441.

148 Louda, S. M., W. N. Gray, K. R. McKaye and J. M. Mhone. 1983. Distribution of gastropod genera over a vertical gradient at

Cape Y4Clear, Lake Malawi. Velliger 25(4): 387-391.

Iovett-Campbell, A. C. 1948. A note on bilharziasis in West African troops. Trans. Roy. Soc. Med. Hyg. 41: 821-822.

Lucas, S. B. 1982. Squamous cell carcinoma of the bladder and schistoscniasis. E. Afr. Med. J. 54: 345-351.

Maiorana, V. C. 1981. Prey selection by sight: random or econcmic. Am. Nat. 118: 450-451. Malek, E. A. 1979. Control of snail hosts of Schistoscmiasis. In Pathways in Malacology, Pp. 251-277; van der Spoel, S., Van Bruggen, A. C. and Lever, J. (Eds.). Utrecht: Bohn, Scheltema and Holhma. Malek, E. A. 1980. Snail-transmitted parasitic diseases. Vol. 1. CRC Press, Inc., Boca Raton, Florida. Manson-Bahr, Apted. 1983. Manson's tropical diseases. Baillere Tindal. London. Mao Shou-pai. 1984. No secret weapon. Wld. Hlth.: 12-13.

McCullough, F.S. and Malek, E.A. Ms. The use of fish for biological control of snail transmitted diseases. McCullough, F. S. 1981. Appraisal of the potential of the use of fish for control of disease vectors other than mosquito ­ TDR/BCV/IC.81.2/WP.22, Geneva, Switzerland. McDonald, G. 1965. The dynamics of helminth infection with special reference to Schistosomes. Trans. Roy. Soc. Trop. Med. Hyg. 59: 489-506.

149 McInnis, A. J. 1965. Responses of Schistosoma mansoni miracidia to chemical attractants. J. Parasit. 51 (5): 731-746. MYKaye, K. R. Sexual selection and the evolution of the cichlid fishes of Lake Malawi, Africa. In: Keenleyside (ed.). Behavior, ecology and evolution of cichlid fishes. Chapman, Hall, London. 32pp. In press. MHKaye, K. R. and Stauffer, J. R. 1983. Malawi Fisheries: An assessment and overview. AID project No. 698-0135-19. Lilongwe, Malawi. 72 pp. IcKaye, K.R. and Stauffer, J.R. 1988. Seasonality, depth and habitat, distribution of breeding males of Oreochromis spp, chambo in Lake Malawi National Park. J. fish Biol. 33: 825-834. MYKaye, K. R. and Marsh, A. 1983. Food switching by two specialized algae-scraping cichlid fishes in Lake Malawi, Africa. Oecologia. 56: 245-248.

c!Kaye, K.R., Stauffer, J.R., and Louda, S.M. 1986. Fish predation as a factor in the distribution of Lake Malawi gastropods. McMahon, J.P., Highton, RIB. and Marshal, T.F. de. 1977. Studies on biological control of intermediate hosts of intemediate hosts of schistosamiasis in Western Kenya. Env. Conserv. 4: 285­ 289. HcMullen, D. B. 1973. Biological and environmental control of snails. In: Ansari (ed.), Epidemiology and control of Schistoscmiasis, pp. 533-.591. Karger, Basel and University Park Place, Baltimore.

150 Meredith, Ii. M. 1983. An introduction to slugs and land snails, with

a key to genera of the larger species known to occur in Malawi. Nyala. 9(1/2): 23-34. specialization: feeding Meyer, A. 1989. Cost of morphological plymYrphic cichlid perfonnance of two morphs in the trophically 80: 431-436. fish, Cichlascma citrinellum. Oecologia control of Michelson, E.H. 1957. Studies on the biological parasites of fresh schistosome-bearing snails. Predators and Parasit. 47: 413­ water molluscs: A review of the literature. 426. Laboratory caparison of Minchella, D. J. and Loverde, P. T. 1983. stocks which are the relative success of Bicmhalaria qlabrata with Schistosoma susceptible and insusceptible to infection mansoni. Parasit. 86: 335-344. and body size; A study Mittelbach, G. G. 1981. Foraging efficiency gills. Ecol. 62: 1370­ of optimal diet and habitat use by blue 1386. partitioning in two Mittelbach, G. G. 1984. Predation and resource sunfishes (Centrarchidae). Ecol. 65 (2): 499-513.F and Bartlett, A. 1979. Molyneux, M.E., Hutt, M.S.R., Clear, A.S. and schistosomal Serum inuunoglobulin concentrations, malarial in Malawi. J. antibodies in patients with massive splenmegaly Trop. Med. Hyg. 82: 183-187.

151 Morgan, P.R. 1979. The Influence of fluctuations in lake level on Bilharzia Transmission. In: Kalk, M., McLachlan, A. J. and Howard-Williams, C. (eds.), Lake Chilwa : studies of change in a tropical ecosystem, Monograph Biologicae 35: 380-384. Junk Publishers, London. Mott, K.E. 1984. Schistosmiasis: New goals. Wld. Hlth.: 3-4.

Movogo, L. and Bard, J. 1964. Seconde note d'information sur

l'Astatoreochrcnis alluadi poisson malacophage utilisable dans la priphylaxie de la bilharziose. Bull. Soc. Path. Exot. 57: 21­ 23.

Muller, R. 1975. Worms and disease. A manual of medical

helminthology. William Heinman medical books, Ltd. London.

Najarian, H. H. 1961. Egg-laying capacity of the snail Bulinus

trancatus in relation to infection with Schistoscar haematobium. Tex. Rep. Biol. Med. 19: 327-331.

Noble, E. R. and G. A. Noble. 1982. Parasitology. The biology of animal parasites. Lea and Febiger. Philadelphia. Nolan, M. 0., Bond and Mann. 1953. Screening tests for molluscicidal activity. Am. J. Med. Hyg. 2: 716-751.

Nursall, J. R. 1973. Teitory in redlip blennies (Ophioblennius

atlanticus - Pisces: Blenniidae). J. Zool. Lond. 182: 205-223.

Nyberg, D. W. 1971. Preyr capture in the lexgemouth bass. Am. Midl. Nat. 86: 128-144. O'Brien, R. D. 1976. Acetylcholinesterase and its inhibition. pp. 271-296. In: Wilkinson (Ed.), Insect biochemistry and physiology. Plenum Press, New York.

152 Olivier, L. J., Earbosa, F. S. and Coelho, M. G. 1954. The influence of infection with Schistosoma mansoni on survival of A. glabrata. Publ. Av. do Cent. Peq. Inst. Agg. Magal. 3: 63­ 71. Olivier, L. and W. T. Haskins. 1960. Effects of NaPCP on A. glabratus eggs. Am. J. Trop. Med. Hyg. 9: 199-205.

Osenberg, C.W. and Mittelbach, G.G. 1989. Effects of body size on predator-prey interaction between pumpkinseed sunfish and gastropods. Ecol. Mon. 59(4): 405-432). Palmer, A. R. 1979. Fish predation and the evolution of gastropod shell sculpture: Experimental and geographic evidence. Evolution 33: 697-713.

Pan, C. 1963. Some biochemical and immnological aspects of host­ parasite relationships: generalized and local tissue responses in the snail Australorbis glabratus infected with Schistoscma mansoni. Ann. N.Y. Acad. Sci. 113: 475-485. Paperna, I. 1969. Aquatic weeds, snails and transmission of Bilharzia in the new manmade Volta Lake in Ghana. Bull. Inst. Fond. Afr. Noire, Ser. A. 31 (2): 487-499. Paperna, I. 1980. Parasites, infections, and diseases of fish in Africa. CIFA Techn. Paper No. 7. FAD, Rcme. Paulini, E. 1974. Copper molluscicides: Research and goals. In: Cheng, T. C. (Ed.), Molluscicides in Schistosamiasis control. Pp. 155-170. Academic Press Inc. New York.

153 Pellegrino, J., Olivier, C. A., Faria, J., and Cunha, A. S. 1962. New approach to the screening of drugs in experimental

Schistosomiasis mansoni in mice. Am. J. Trop. Med. Hyg. 11:201­

215.

Pellegrino, J., DeMaria, M., DeYbura, M. F. 1966. Observations on the predatory activities of tebistes reticulatus (Peters, 185a) on Cercariae of Schistosoma mansoni. Am. J. Trop. Ned. Hyg. 15 (3): 337-341.

Pip, E. 1978. A survey of the ecology and composition of submerged

aquatic snail-plant communities. Can. J. Zool. 56: 2263-2279.

Pip, E. and Stewart, J. M. 1976. The dynamics of two aquatic plant­ snail associations. Can. J. Zool. 54: 1192-1205.

Pruginin, Y. 1971. The development of fish fanning in Malawi. FAD, Rtome. FI: SF/ MLW 7.

Pruginin, Y. 1976. Fish fanning in Malawi. FAD Report, Rome. FI:

DP/ MLW/ 71/ 516/ 8.

Pruginin, Y. and Arad, A. 1977. Fish farming in Malawi. FAO Report, Rcme. FI: DP/ MLW/ 73./ 516/ 10.

Pugh, R. N. and Teesdale, C. H. 1984. Long-term efficacy of single­ dose oral treatment in schistosomiasis haematobium. Trans. R. Soc. Trop. Med. Hyg. 78: 55-59.

Pullin, R. S. V. and Shehadeh, Z. H. 1980. Integrated agriculture­ aquaculture fanning systes. ICLARM-SEARCA.

Rand, G. M. and S. R. Petrocelli. 1985. Fundamentals of aquatic

toxicology. Hemisphere Publishing Corporation. Washington,

D.C.

154 Ransford, 0. N. 1948. Schistosomiasis in the Kota-Kota district of Nyasaland. Trans. Roy. Soc. Trop. med. Hyg. 41 (5): 21-23.

Raut, S. K. 1986. Inhibition of fish growth by the fresh water snail Bellamya Bengalensis. Environ. Ecol. 4(2): 332-333. Reid, G. McG. 1984. Prospects for Bilharzia control in Northern Nigeria in relation to the ecology of snail vectors in the Sokoto-Rima river (Niger systen). The Nigerian Field. 49: 3­ 19. Ritchie, L. S. 1973. Chemical control of snails. pp. 458-532. In: Ansari (Ed.), Epidemiology and control of schistosomiasis (Bilharziasis). University Park Place, Baltimore. Rosenfield, P. L. 1979. The management of schistosomiasis. Resources for the future. Washington, D. C. Savino, J. F., and Stein, R. A. 1982. Predator-prey interaction between largemouth bass and blue gills as influenced by Simulated Submersed vegetation. Trans. Am. Fish. Soc. 106: 596­ 501. Schimdt, G. D. and L. S. Roberts. 1985. Foundations of parasitology. Times irror/Mosby College publishing. St.

Louis. Schmitt, R. J. and Holbrook, S. J. 1984. Ontogeny of prey selection by black superperch, Embiotoca jacksoni (Pisces: Embiotocidae): the role of fish morphology, foraging behavior, and patch selection. Mar. Ecol. 18: 225-239.

155 Schutte, C.H.J., and Frank, G. H. 1964. Observation on the distribution of freshwater mollusca and chemistry of the natural waters of the South-Eastern Transvaal and adjacent Northern Swaziland. Bull. Wld. Hlth. Org. 30: 389-400. Seyani, J. H. In prep. A survey of traditional fishing methods using plant poisons in Malawi. Malawi-ICLARM-GTZ aquaculture report. Sherrif, M. and Ibrahim, A.I. 1985. A preventable cancer. Wld. Hlth.: 26-27.

Shiff, C. J. 1961. Trials with a new molluscicide, Bayer 73, in Southern Rhodesia. Bull. Wld. Hlth. Org. 25: 559-562. Shiff, C. J. and Kriel, R. L. 1970. A water soluble product of Bulinus globosus attractive to Schistosoma haematobium miracidia. J. Parasit. 56: 281-286.

Sibbing, F. A., Osse, J. W. M. and Terlouw, A. 1986. Food handling in the carp (Cyprinus carpio L.): its movement patterns, mechanism and limitation. J. Zool. Lond. 210: 161-203.

Slobdkin, L.B. 1968. How to be a predator. Amer. Zool. 8: 43­ 45.

Slootweg, R. 1985. Biolgical control of snail intenrediate hosts of Schistosoma spp. by fish: A summary of literature. Zool. Lab. Dept. of Ed. Morphology. Leiden University. Netherlands.

Slootweg, R. 1987. Prey selection by molluscivorous cichlids foraging on a Schistosomiasis vector snail, Biomphalaria qlabrata. Oecologia 74: 193-202.

Smith, E. B. 1977. Schistoscmiasis. C. Afr. J. Med. 23 (11): 1

156 Stein, R. A., Goodman, C. G., and Marschall, E. A. 1984. Using time and energetic measures of cost in estimating prey value for fish

predators. Ecology 65: 702-715.

Stenson, J. A. E. 1978. Differential predation on two species of chaoborus (D,.ptera, Chaoboridae). Oikos. 31: 98-101.

Stirewalt, M. A. .973. The important features of schistosomes, pp.

17-31. In: Epidemiology and control of schistosomiasis

(bilharziasis), Ansari (Ed.). University Park Press, Baltimore.

Stockard, J. L. 1978. Economic justification for schistosomiasis

control. Proc. Int. Conf. Schistoscmiasis, Cairo. 1: 3-11.

Sturrock, R. F. 1966. The influence of infection with Schistoscma mansoni on the growth Pnd reproduction of Biomphalaria

pfeifferi. Ann. Trop. Ned. Parasit. 60: 187-197.

Sturrock, R. F. 1967. The effect of infection with Schistosoma

bulinus (Physopsis) nasutus products. Ibiden. 61: 321-325.

Sturrock, R. F., and Sturrock, B. M. 1971. Shell abnormalities in

Biomphalaria glabrata injected with Schistosoma mansoni and

their significance in field transmission studies. J. Helminth.

XLV (213): 201-210. Sturrock, B. M. and Sturrock, R. F. 1970. Laboratory studies of the host parasite relationship of Schistosoma mansoni and

Biomphalaria glabrata from St. Lucia, West Indies. Ibidem 64: 357.

Sturrock, R. F. and Webbe, G. 1971. The application of catalytic models to schistosomiasis in snails. J. Helm. 45: 189-200.

157 Sturrock, R. F. 1970. An investigation of same factors influencing the survival of Biomphalaria qlabrata deprived of water. Ann. Trop. Med. Parasit. 64: 365-371.

Sturrock, R. F. 1973. Field studies on the transmission of

Schistosma mansoni and on the bionomics of its intermediate host, Bicmphalaria glabrata, on St. Lucia, West Indies. Int. J. Parasit. 3: 175-194.

Teesdale, C.H. :986. The role of health education to reduce

transmission of schistosaniasis. Trop. Med. Parasit. 37: 184­ 185.

Teesdale, C. H. and Chitsulo, L.A. 1983. Schistosomiasis in

Malawi - A review. Malawi Epidemiological Quarterly. 4: 10­ 36.

Teesdale, C. H. and Chitsulo, L. A. 1985. Schistosaniasis in Malawi

- a review. Trop. Med. Parasit. 36: 1-6. Thanas, J. D. 1973. Schistosomiasis and the control of rolluscan hosts of human schistosomes with particular reference to possible self regulatory mechanisms. Adv. Parasit. 11: 307-395.

Thcrpson, D. 0. 1965. Food preferences of the meadow vole in

relation to habitat Lffinities. Am. Midi. Vatur. 74: 76-86. tkli, F. M. A. 1984. Introduction to parasitology in tropical

Africa. John Wiley and Sons., Ltd. New York.

Van Schayck, C. P. 1986. The effect of several methods of aquatic

plant control on two bilharzia-bearing snail species. Aquat. Bot. 24: 303-309.

158 Venneij, G. J. and Covich, A. P. 1978. Coevolution of freshwater gastropods and their predators . Amer. Nat. 112: 833-843.

Venreij, G. J., Zipser, E. J. and Dudley, E. C. 1980. Predation in timre and space: breakage and drilling in terebrid gastropods. Paleobiology 6 (3): 352-364. Visser, S.A. 1964. Yolluscicidal activity of surface active agents on the snail Biomphalaria sudanica, a vector of bilharzia. Nature 204: 292. Wainwright, S. A., Biggs, W. D., Currey, J. D., and Gosline, J. M. 1976. Mechanical design in organisms. Halstead Press, New York, 423 p.

Warren, K.S. 1984. Water-poison disease. Wld. Hlth.: 5-6. Warren, K. S. and A. S. Weisberger. 1966. The treatment of molluscan Schistosomiasis mansoni with chloramphenicol. Am. J. Trop. Med. Hyg. 15 (3): 342-350. Webbe, G. 1962. The transmission of Schistosoma haematobium in an area of Lake Province, Tanganyika. Bull. Wld. Hlth. Org. 27: 59-85. Weil, C. and M. Kvale. 1985. Current research on geographical aspects of schistoscniasis. Geographical Rev. 75 (2): 186-216. Weinbach, E. C. and M. 0. Nolan. 1956. The effect of pentachlorophenol on the metabolism of Astralobis glabratus. Exp. Parasit. V: 276-284.

Weisberger, A. S. and S. Wolfe. 1964. Effect of chloramphenicol on protein synthesis. Fed. Proc. 23: 976-983.

159 Werner, E. E. and Hall, D. J. 1974. Optimal foraging and the size selection of prey by blue gill sunfish (Lepcmis microchirus). Ecol. 55: 1042-1052.

Werner, E. E., Gilliam, J. F., Hall, D. J., Mittelbach, G. G. 1983.

An experimental test of the effects of predation risk on habitat use in fish. Ecol. 64: 1540-1548.

WHO. 1961. Molluscicides. Report of a WHO Expert Cammittee. Techn.

Rep. Wld. Hlth. Org. 214: 1-50.

WHO. 1965. Snail control in the prevention of bilharziasis. Mon. Ser. No. 50. Geneva.

WHO. 1967. Epidemiology and control of schistosomiasis. Wld. Hlth.

Org. Techn. Rep. Ser. 378.

WHO. 1973. Schistosomiasis. Report of the WHO Expert Committee.

Wld. Hlth. Rep. Ser. No. 515. WHO. 1980. Epidemiology and control of schistosomiasis. Techn. Rep. Ser. 643.

WHO. 1984. Report of an informal consultation on research on the biological control of snail intermediate hosts. WHO, Geneva, 22­ 27 January 1984.

WHO. 1985. The control of schistoscniasis. Techn. Rep. Ser. No. 728.

Wilkson, C. F. 1976. Insect biochemistry and physiology. Plenum Press. New York.

Wright, C. A. 1966. The pathogenesis of helminths in the mollusca.

Helm. Abstr. 35: 207-224.

160 Wright, W. H. 1973. Geographical distribution of schistosomes and their intermediate hosts.: 32-24'i. In Ansari N. (Ed.), Epidemiology and control of schistosomiasis (bilharziasis). University Park Press. Baltimore. Wright, E. D., Chimphangwi, J. and Hutt, M. S. R. 1982. Schistosomiasis of the female genital tract. A histopathological study of 176 cases from Malawi. Trans. Roy. Soc. Trop. Med. Hyg. 76 (6): 822-829.

Younger, M. S. 1985. A first course in linear regression. Second edition. Duxbury Press, Boston. Zakaria, H. 1963. Heteropneustes fossilis, a possible agent for the biological control of the snail hosts of schistoscmes. Ann. Trop. Yed. Parasit. 57 (2): 157-160. Zipser, E. and Vermeij, G.J. 1978. Crushing behavior of tropical and temperate crabs. J. Exp. Mar. Biol. Ecol. 155-172. Zipser, E. and Venrmeij, G. J. 1980. Survival after nonlethal shell diunage in the gastropod Conus sponsalis. Micronesica 16 (2): 229-234.

161