The biology of Biomphalaria choanomphala and B. sudanica in relation to their role in the transmission of Schistosoma mansoni in Lake Victoria at Mwanza, Tanzania

Bull. Org. mond. Sante| 1972, 47, 331-342 Bull. Wld Hithu Org. The biology of Biomphalaria choanomphala and B. sudanica in relation to their role in the transmission of Schistosoma mansoni in Lake Victoria at Mwanza, Tanzania MARGARET MAGENDANTZ 1 A study of the intermediate snail hosts of Schistosoma mansoni in Lake Victoria at Mwanza, Tanzania, was begun in October 1969, the main aims being to investigate the distribution and seasonal variations in population densities ofBiomphalaria choanom- phala and B. sudanica in relation to the nature of the lake bottom and the biological features of the lake shore, the factors influencing variations in the intensity of S. mansoni transmission along the Mwanza shoreline, and the age structure of populations of B. choanomphala. Field surveys were made at 70 sites near Mwanza and in nearby bays, B. choanomphala being collectedfrom the lake bottom by means of a wire-mesh dredge. Variations in the distribution and population density of B. choanomphala were correlated with the nature of the bottom and its depth profiles at depths of0.5-6.0 m. Approximately 1-20 snails/iM were found on mixed sand and mud but only about 1 snail/iM on the pre- dominantly muddy bottom farther out from the shore. Seasonal variations in the age structure and fluctuations in the population densities of B. choanomphala of as much as 10-13-fold were observed. A large and a small form of B. choanomphala, possibly eco- phenotypes, were found. S. mansoni infection rates in B. choanomphala ranged from 0.2 % to 3.3 %, suggesting a tendency to higher infection rates in mature snails. For the past 12 years the East African Institute for Medical Research at Mwanza, Tanzania, has conducted intensive research on schistosomiasis. In addition, the WHO/Tanzania Schistosomiasis Pilot Control and Training Project (Tanzania 2101) has, among its other activities, recently examined the factors influencing the transmission of intestinal schistosomiasis in Mwanza (McCullough et al., 1972; McCullough & Eyakuze, unpublished data; McCul- lough & Magendantz, unpublished data). These studies, together with those of Wijers & Munanga (1971), indicate that transmission of Schistosoma mansoni in Lake Victoria and its tributaries is an increasing public health problem, especially as towns and villages along the shores of the lake are growing rapidly. 1 Visiting Research Worker, East African Institute for Medical Research, Mwanza, Tanzania. Lake Victoria contains an endemic species of Biomphalaria of the Choanomphala group, which lives on the bottom at distances up to 150 m offshore and is highly susceptible to infection with S. mansoni. In addition, B. pfeifferi is found in streams draining into the lake, and B. sudanica in habitats near the shores of the lake. These three species were first shown to be susceptible to S. mansoni infections by Cridland (1955), while Webbe (1962) and McClelland & Jordan (1962) first described the role of B. choa- nomphala in the transmission of S. mansoni at Mwanza and Bukoba, respectively. An apparatus was designed by Prentice (1966) to collect B. choa- nomphala and to study the transmission of S. mansoni at Entebbe, Uganda (Prentice et al., 1970); more recently, Prentice (1970) has experimented with the application of molluscicides to the lake, Baalawy (1971) reported high densities of B. choanomphala at Mwanza, low densities at Musoma and Bukoba, 2904 -331- 332 M. MAGENtDANTZ and no snails at Kisumu. These studies have pro- N vided considerable information about the distribu- tion of B. choanomphala, but little on its population dynamics and natural rate of infection with S. man- soni. Webbe (1962) found infected specimens of B. suda- nica in backwaters behind the lake shore, and regarded them as an important source of S. mansoni transmission. In Uganda, Berrie (1964) related the growth and reproductive rate of this snail to seasonal changes. However, little is known about the growth of B. sudanica and the rates of infection with S. man- soni in different types of habitat. The main aims of the present study, carried out from October 1969 to August 1971, were to investi- gate the distribution and seasonal variations in den- sity of B. choanomphala and B. sudanica in relation to bottom substrata and physical and biological features of the lake shore, the factors influencing variations in intensity of S. mansoni transmission along the Mwanza shorelines, and the age structure of populations of B. choanomphala. It is hoped that these studies, together with those already mentioned, will contribute to the planning of more satisfactory control measures against S. mansoni in Mwanza. MATERIALS AND METHODS Field methods Field surveys for B. sudanica and B. choanomphala were made at 70 sites in Mwanza North and South bays, at one site in Nyegezi Bay, and at two in Pasiansi Bay. B. choanomphala was collected with a 2-mm wire mesh dredge measuring 75 cm deep and 55 cm wide at the opening, supported on a triangular frame. The net was dropped from a boat at distances of 20, 40, and 60 m from the shore, pulled for a measured distance, and then lifted out of the water. This simple method of dredging yielded hundreds of snails in a relatively short time (about 1 hour). The samples were washed through a 1-2-mm mesh sieve to remove snails and egg masses, which were then spread out in shallow basins and covered with water so that the smaller snails could be more easily seen. The average number of B. choanomphala per haul was calculated by dividing the total number collected by the number of hauls. In order to assess the sensitivity of the dredge in collecting snails, this method was compared with an exhaustive stationary sampling method, in which -iM2 plots on the lake bottom were marked off with a circular frame. The bottom material inside the frame was dug out to a depth of 4 cm and sieved as described above. Altogether, 10 such samples and 2 30-m dredge hauls (starting 30 m from the shore and passing as close as possible to the stationary sampling stations) yielded the following data for B. choanomphala. 1-M2 samples Distance from shore and substratum No. ofsnails/m' 10 m; sand 0 10 m; sand 0 15 m; sand 0 15 m; sand 0 20 m; sand & mud 4 20 m; sand & mud 3 25 m; sand & mud 10 25 m; sand & mud 16 30 m; sand & mud 18 30 m; sand & mud 11 Dredge hauls 96 snails/haul = 12.8 snails/M2 a 176 snails/haul = 23.4 snails/m2 a These data suggest that a single dredge haul for a short distance on a sandy-muddy bottom collects at least 70% of the snails in a transect. Each site of roughly 1 500 m2 was transected by an average of 3-4 dredge hauls on each collecting day. It was found that the dredging method was particularly suited to shallow bays with sandy-muddy bottoms and no submerged vegetation, whereas an Eckman grab was ineffective at shallow depths on hard sandy bottoms. Variations in population structure and infection rates of B. choanomphala over a period of 1.5-2 years were followed monthly at 5 sites in Mwanza North and South bays and on 5-12 occasions at 6 sites in these bays and in Nyegezi and Pasiansi bays. All the other sites were surveyed on 1-4 occasions. Sites for repeated collections were selected on the basis of their similarities and contrasts with respect to substrata, depth profiles, human activities, and proximity to housing. Some examples of the dif- ferent substrata and depth profiles are given in Table 1. Various types of human activity and hous- ing condition were represented. Six of the sites were probably used to a greater degree than any of the others by 30-50 or more different individuals in a single day, for bathing, washing, and fishing. la The calculations were made as follows: between the distances of 30 m and 15 m from the shore each dredge haul traversed an area of 7.5 m'. Dividing the number of snails in each of the 2 hauls by 7.5 yields the average number of snails per m. BIOLOGY OF BIOMPHALARIA IN LAKE VICTORIA AT MWANZA o) o Lo 0 0N4 LO U) co O co c'i 4 C_ ~_ ~ C c ~ ~~~c U) E) N U) U o o _ - - 0 'U 'L Lo C C) C Ca CO ' LO LO C .U' 'U o N 0 c c _ N co % 0 _ U) C') o E E 0 0 C') U) '_ '0 M0 '0 ' co E E E E VI _- LooC N X w c N N C' le U) a C Uo c (A m c c c_'0 '0 co co 0 0 0~ UP r"~ U) C 0 _ N C' le (I u 2u 2o --,aiX nc C 'D r. 5 w to E w E E 0 I- 0 N U) N:V:l w00E w0E - -,w co E~ co E CD E2c2 0 0 0 0 It co 0 0 In the Mwanza bays there are virtually no sites with access to the lake that are not frequented, except those near the Capri Point water supply inlet for Mwanza where it is forbidden to bathe and settle. B. sudanica was collected with a scoop net in emergent vegetation, irrigation furrows, and grassy seepages along the Mwanza North Bay and South Bay shores. Seasonal variations in the population structure of B. sudanica were followed monthly at two sites. When these sites began to dry out, snails could only be collected with forceps. Laboratory methods E 0co co 0 Co *I coE .0E co co C .C 0.zE C._ 0as E Co (U(N Cw C B. sudanica and B. choanomphala collected in both Mwanza bays were screened for mammalian schisto- some infections by placing snails in glass vials con- taining 15 ml of filtered lake water under a lamp for 12 hours at a temperature of 25-26°C. Albino mice were exposed to mammalian-type schistosome cercariae pooled from 3 or more B. sudanica or B. choanomphala by partially submerging them in 50 ml of filtered lake water containing 150-200 cer- cariae. The identification of human schistosome cercariae was confirmed by the recovery of S. man- soni ova from the livers ofmice autopsied 2-3 months after exposure to the cercariae. B. choanomphala snails collected from the lake bottom did not readily lay egg masses in aquaria. However, during the first few days after being brought into the laboratory they deposited egg masses in glass vials. These egg masses were care- fully removed to initiate a laboratory colony of B. choanomphala in the following way. The egg masses were placed in linen cloths suspended in aerated lake water in basins or aquaria. Aeration was necessary to prevent the egg masses from quickly becoming covered with bacterial and fungal slime. For a few weeks before use the cloths were allowed to become coated with green and blue-green algae on which the hatchling snails could feed. After the snails emerged they were fed a mixture containing boiled lettuce, calcium carbonate, ground- up mouse pellets, and sand. Once a week the cloths were rinsed to remove faecal matter and bacteria. The young snails grew to maturity and laid eggs abundantly on the cloths. RESULTS Biomphalaria choanomphala: distribution and varia- tion in population density in relation to bottom substrata 333 Q Co ._ co CL 0~ z m 0 CU N C a CU 0 co CU N Cu CO I.: (D c CU 0e 0_ ,o QU U) n cn 4 CL DE co N Ca E 4 CL cn .0 3 LO~ 4 C Z s E co 0CL E ,l E z 0 co C ,E 4 C ,E E *n0 (1)1 4 D C'N- U) 0 334 M. MAGENDANTZ From the distribution of B. choanomphala and B. sudanica, and the variations in population den- sity of the former species, in Mwanza South and North bays, it appears that B. choanomphala is mainly a bottom-dwelling species, although Webbe (1962) reported finding it on submerged vegetation as well as on the lake bottom. In the present study, B. choanomphala was found on the bottom, but never on emergent vegetation or on the shoreline, except when living snails were washed up after a storm. At some sites there were many empty shells of B. choanomphala, but at other sites with more steeply sloping shores there were few or no shells, even though high densities of snails were found on the adjacent lake bottom. The lack of shells on the shore could therefore not be used to indicate the absence of snails at any particular site. On the lake bottom, variations in snail density were closely cor- related with various types of substratum. No snails, or very few (1-5 snails per 40-m dredge haul), were found on mud bottom situated offshore from exten- sive areas of emergent grasses, sedges, and papyrus and containing much silt and decaying vegetation. Extensive emergent vegetation bordered approxi- mately 5 km in a stretch of 10 km of the South Bay and 1 km in a stretch of 5 km of the North Bay, but there was none bordering the 10 km of Pasiansi Bay. Moderately high to high densities of B. choanom- phala (>15 snails per 40-m haul) were found on sand and mud bottom offshore from a small zone of emergent vegetation or from an exposed shore. Roughly one-half of both the North Bay and South Bay sites had this type of bottom and these snail densities. At both sites in Pasiansi Bay, which had the same type of bottom, there were high densities of B. choanomphala. In order to obtain a rough idea of the number of B. choanomphala on the lake bottom in certain areas, population densities can be roughly estimated from the following equation: total no. of snails total no. of snails collected within a certain collected within a certain distance from the shore a distance from the shore a area of lake bottom 0.55 m b x distance dredged dredged x no. of dredge hauls For example, the density of B. choanomphala at site 46 in Mwanza South Bay on 23 March 1970, a 20, 40, and 60 m. b Width of the dredge. based on data in Table 2, is calculated from the above equation as follows: 205 (total no. of snails collected in one 40-m haul) = 9.3 snails/mr2. 0.55 x 40 x 1 At this site at the time of dredging maximum snail densities were found within a distance of 10-40 m from the shore. Thus, taking a section of lake bottom 100 m long and 30 m wide within a distance of 10-40 m from the shore, the total area would be 3 000 M2. For a population density of 9.3/m2, a rough estimate of the total number of snails along a 100-m stretch of beach within 40 m from the shoreline is 3 000 x 9.3 = 27 000 snails per 3 000 M2. Such calculations of the total number of snails at a particular site indicates the high densities of B. choanomphala in some areas. There was such marked variation in the substrata, depth profiles, and snail densities in Mwanza North and South bays and Pasiansi Bay that the bays could not readily be characterized. Table 1 shows these variations for several sites in each bay. Densi- ties of B. choanomphala declined to low levels (1-5 snails per 40-m haul) at various distances from the shore where the bottom was predominantly mud. In Mwanza South Bay the highest snail densities (15-300 snails per 40-m haul) occurred within 60 m from shore on mixed sand/mud bottom at depths of 0.75-3.0 m, whereas beyond a distance of 60 m densities declined to 1-5 snails/40-m haul on pre- dominantly muddy bottom. In contrast, at some Mwanza North Bay sites and in Pasiansi Bay, where mixed sand/mud bottom extended beyond 100 m from the shore in some places, high densities of B. choanomphala continued out to distances of 80-150 m from the shore at depths of 1-6 m. Thus, the type of bottom rather than depth (at least up to a depth of 7 m) appears to be the main factor determining the maximum distance from the shore at which high densities of B. choanomphala are found. Since the shoreline is undergoing rapid alteration as a result of drifting papyrus islands, shifting agri- culture, and various fishing and housing practices, and since this is leading to changes in the bottom, the distribution of B. choanomphala is likely to change considerably in the future. It would be of interest to repeat the dredging operations in Mwanza at intervals in order to monitor the changing patterns influencing the bionomics of B. choanomphala. BIOLOGY OF BIOMPHALARIA IN LAKE VICTORIA AT MWANZA Table 2. Representative data on the distribution and population densities of B. choanomphala and B. sudanica and their rates of infection with S. mansoni at several sites Lake Victoria at Mwanza a B. choanomphala B. choanomphala DtToa PecnaeDate Total Percentage DeNo. per haul no. collected c infected No. per haul b no. collected c infected site 5 site 38 7 Nov. 1969 20 59 0.0 18 Nov. 1969 31 134 0.0 2 Dec. 1969 33 100 0.0 24 Nov. 1969 51 207 -d 23 Jan. 1970 27 80 -d 14 Jan. 1970 32 162 -d 21 Feb. 1970 36 118 0.0 19 Feb. 1970 123 422 _d 17 April 1970 42 83 0.0 14 April 1970 279 627 0.0 30 June 1970 14 41 0.0 16 May 1970 217 425 0.0 1 Oct. 1970 17 52 0.0 8 June 1970 141 842 0.0 18 Nov. 1970 15 149 2.7 13 July 1970 162 325 0.0 17 Dec. 1970 5 37 0.0 17 Aug. 1970 151 453 0.0 4 Sept. 1970 52 155 0.0 site 49 26 Oct. 1970 110 438 0.0 18 June 1970 269 269 0.7 3 Nov. 1970 70 139 0.7 15 Aug. 1970 156 780 0.0 1 Dec. 1970 30 301 0.0 26 Sept. 1970 50 99 0.0 2 Jan. 1971 35 248 0.0 13 Dec. 1970 23 206 0.0 8 Feb. 1971 45 268 0.5 13 Jan. 1971 85 170 1.1 14 March 1971 124 371 0.3 9 Feb. 1971 38 130 0.8 22 April 1971 60 300 0.0 13 March 1971 27 162 0.0 26 April 1971 57 170 0.6 20 April 1971 89 267 0.4 29 July 1971 89 444 0.0 site 46 Nyegezi site 22 Dec. 1969 39 116 -d 16 Oct. 1969 24 95 0.0 5 Feb. 1970 80 326 -d 22 April 1970 96 289 0.0 23 March 1970 205 387 0.0 16 June 1970 106 351 0.3 30 April 1970 276 612 0.0 22 July 1970 187 374 0.5 12 May 1970 169 339 0.0 24 Aug. 1970 263 1 049 0.2 6 June 1970 173 520 0.0 14 Oct. 1970 111 277 0.0 7 July 1970 24 48 0.0 8 Nov. 1970 96 383 0.5 4 Aug. 1970 34 135 0.0 2 Dec. 1970 21 170 1.1 12 Sept.1970 49 98 0.0 31 Dec. 1970 18 126 0.0 16 Oct. 1970 15 74 1.4 30 Jan. 1971 12 108 0.9 25 Nov. 1970 21 148 0.0 1 July 1971 133 530 0.2 4 Jan. 1971 19 132 0.0 6 Aug. 1971 96 382 0.5 a Complete data for all 70 sites are available at the East African Institute of Medical Research, Mwanza. b Dredge dropped at a distance of 40 m from the shore. The number of snails collected per 40-m haul is the average number per haul c Total number of snails collected at 20, 40, and 60 m from the shore. d Not recorded. 335 M. MAGENDANTZ Table 2 (continued) B. choanomphala B. choanomphala NoDerhula ToaePrenaeDate Total PercentageDate No. per haul ano. collected b infected per haul ano. collected15inN.fn.ecllctecbtnfete site 46 Nyegezi site 9 Feb. 1971 13 54 0.0 5 Sept. 1971 436 486 0.2 22 March 1971 17 149 0.7 14 April 1971 60 240 0.4 30 July 1971 77 540 0.2 B. sudanica B. sudanica Date Date Total no. collected b Percentage infected Total no. collected b Percentage infected site 9 site 38 28 April 1970 88 0.0 17 Dec. 1969 53 _c 10 May 1970 100 0.0 19 Feb. 1970 407 0.0 24 June 1970 312 0.0 25 April 1970 164 1.8 2 Sept. 1970 47 2.1 25 May 1970 139 0.7 12 Oct. 1970 160 0.0 3 July 1970 65 0.0 3 Nov. 1970 253 0.0 13 Aug. 1970 33 0.0 4 Dec. 1970 151 0.0 3 Sept. 1970 86 0.0 7 Jan. 1971 201 0.0 17 Oct. 1970 57 1.8 2 April 1971 193 1.0 26 Nov. 1970 67 0.0 12 Sept. 1971 51 0.0 2 Dec. 1970 66 0.0 2Jan. 1971 89 0.0 3 Feb. 1 971 37 0.0 a Dredge dropped at a distance of 40 m from the shore. The number of snails collected per 40-m haul is the average number per haul. b Total number of snails collected at 20, 40, and 60 m from the shore. c Not recorded. Seasonal variations in population density and age structure Seasonal fluctuations in the population densities of B. choanomphala were recorded during the period October 1969 to September 1971 at four sites in Mwanza North and South bays and in Nyegezi Bay. In Mwanza South and Nyegezi bays from March to August 1970, snail densities were 3-6 times those found from November 1969 to January 1970 and from September 1970 to February 1971 (Table 2). In 1971 the population densities began increasing at one site in January, while at two other sites and at Nyegezi the population increase took place in March and April. An upsurge in population density took place at one site in the North Bay in April 1970, and a decline occurred in late October and November (Table 2). In Nyegezi, on the other hand, B. choa- nomphala densities were highest from August to October, had fallen 14-fold by December, and by March 1971 had again increased to one-third of the density recorded in August 1970 (Table 2). Repeated collections at four sites showed that the age structure of B. choanomphala populations changed during the year. The most marked changes are seen when the September-November and the January-July collections from Mwanza South are compared: in the former, 50-80% of the snails were 6-8 mm in diameter, but in the latter, 50-70% were immature or young mature snails less than 5.5 mm in diameter. Marked changes in age structure also occurred in the Nyegezi population, but they did 336 BIOLOGY OF BIOMPH4LARIA IN LAKE VICTORIA AT MWANZA not coincide in time with those in the Mwanza Bay population. During October 1970, over 50% of the snails in the Nyegezi population were juveniles, whereas only 20% in the Mwanza populations were immature. Variations in shell growth in populations of B. choa- nomphala In Mwanza Bay and nearby bays large and small forms of B. choanomphala, which are possibly eco- phenotypes, have been found. The larger form has been collected in Mwanza North Bay and the smaller form in Nyegezi, Mwanza South, and Butimba bays. In Mwanza North Bay, particularly at sites near the outflow of the Mirongo stream, 10-15% of the snails in dredge hauls are 8.0-9.5 mm in diameter, whereas in Pasiansi, Mwanza South, and Nyegezi bays the number of snails exceeding 8.0 mm in diameter collected at any time of the year seldom exceeds 5 %. Both the larger and smaller forms are found in fairly shallow water (0.5-3.0 m), but the larger Mwanza North Bay form is found on sand/mud bottoms inshore from silt-laden substrata. The greater shell growth of the Mwanza North Bay form possibly reflects a richer supply of organic and min- eral nutrients in the North Bay. The Mwanza North Bay form corresponds to that described by Mandahl- Barth (1957) as the " Mwanza form Egg production The egg masses of B. choanomphala, like those of all planorbid snails, are oval in shape, and are commonly found on mollusc shells and decaying vegetation. They can be distinguished from the egg masses of Bulinus truncatus trigonus, which is also abundant at some sites on the lake bottom, because the latter have thicker egg capsules and a more pronounced terminal tail. The egg masses of these two species can also be readily identified by examin- ing the hatchling snails. Fluctuations in B. choanomphala densities during the year may be the result of variations in egg pro- duction (natality) and/or variations in survival rates (mortality) of snail embryos or snails in various size classes. It was observed that the number of egg masses in dredge hauls varied considerably during the year. At Mwanza South Bay large numbers (100-300 egg masses per 5 40-m dredge hauls) were collected between February and July 1970, while between October and December 1970 egg masses were found only with difficulty (50 egg masses in 5 40-m hauls). The mortality rates of embryos collected through- out the year ranged from 22% to 59%, but higher rates did not coincide with lower snail densities. Predation by ciliate protozoa and annelid worms was an important cause of mortality. S. mansoni infection rates in B. choanomphala in the lake Infected B. choanomphala were found in Mwanza South Bay at 10 sites, in Mwanza North Bay at 4 sites (including the Karumo ferry landing), and at 1 site in Nyegezi Bay. Pollution of the water with human faeces was evident at all these sites, particularly at a site where there is only one public water tap on a 3-km stretch of Mwanza South Bay shore, and where some 2 350 people are dependent on the lake for their entire water needs. S. mansoni infection rates in B. choanomphala ranged from 0.2% to 3.3 %. These data on seasonal infection rates, though insufficient, suggest a tendency towards higher infection rates when seasonal den- sities are at their lowest and when snail populations consist mainly of mature individuals. For example, infection rates in Mwanza South Bay between October 1970 and January 1971 were generally 1.5-3 times those during March and April 1971. Further studies over a period of several years should be made, uniformly large samples of snails from several sites being screened to determine whether seasonal variations in infection rates occur. All infected snails were collected from depths of 0.5-3.0 m in the various bays. In Pasiansi Bay infected snails were found on one occasion only, but they occurred in most of the monthly collections from Nyegezi Bay. Biomphalaria sudanica: distribution along the lake shore B. sudanica snails were found in numerous habi- tats above the lake level, particularly in grassy seep- ages and irrigation furrows. Seasonal variations in population density and structure Thousands of young snails migrated from shallow grassy seepages to shallower upshore seepages at the time of heavy flooding in April and May 1970. In June, those seepages farthest (about 60 m) from the shoreline, where grass was sparse, quickly dried out and the numerous exposed snails died within a month. On the other hand, about 20 m from the shore, where there were taller clumps of grass pro- viding shade, 30% of the snails survived for 3 months until September, after which time no live snails were 337 M. MAGENDANTZ recovered. Still closer to the shoreline (about 5-10 m), behind a zone of Typha sp. in small pockets of water that periodically filled and dried up, a few surviving B. sudanica snails were found until March 1971. No snails were found during the rains of April and May 1971; in the previous year there was a migration of thousands of snails, and infected snails were collected on several occasions. At a site in Mwanza South Bay young snails (.<6.0 mm in diameter) constituted about half the B. sudanica population during the long rainy season from February to May 1970; from July to Sep- tember 30-40% were juveniles, and between October and December 1970, when the rainfall was half the average figure, less than 20% were young snails. In a markedly different site, a group of irrigation furrows in Mwanza North Bay, young snails were rare during the heavy flooding in April. In this habitat, the main period of reproductive activity appeared to be delayed until the early part of the dry season. Moreover, high population densities were maintained throughout the dry season, as the water level only slowly receded and the irrigation furrows never completely dried out. Such high den- sities during the dry season were also described by Berrie (1964) for a B. sudanica population in a road- side ditch or a papyrus swamp near Kampala. Rates of S. mansoni infection in B. sudanica B. sudanica snails infected with S. mansoni were collected at 4 sites in the North Bay and at 6 in the South Bay. Much pollution with human faeces was observed at these sites, which were all close to open beaches. Infection rates in the two bays ranged from 0.5% to 7.7%, averaging 1.6 %. Lateral- spined S. mansoni ova were recovered from mice exposed to mammalian schistosome cercariae pooled from naturally infected B. sudanica. No S. rodhaini ova, as described by Schwetz (1951), were recovered from the mice. No infected B. sudanica were found along the South Bay in irrigation furrows in an area where there is neither access to the lake nor signs of obvious faecal pollution in the dense growth of papyrus and sedges. DISCUSSION The marked variations in population density of B. choanomphala snails in different sites in Lake Victoria at Mwanza are probably influenced by a number of environmental factors, the main one being, perhaps, the nature of the lake bottom. The highest snail densities occurred on mixed sand/mud substrata that had a rich algal growth and contained a moderate amount of organic matter. On the other hand, the reduction of B. choanomphala observed on predominantly muddy substrata may be related to the presence of large amounts of organic matter and organic acids. Analyses of bottom mud from Lake Victoria have shown that its organic content shows little tendency to decompose (Fish, 1955). Generally, in the temperate zone, molluscs are abundant in eutrophic lakes with hard water and a high calcium content, but they tend to be scarce in acid dystrophic lakes. Similarly, some tropical lakes (including Lake Victoria) that are rather alka- line support relatively large populations of molluscs, at least locally. Lakes that are surrounded by sphagnum bogs or papyrus swamps (e.g., Lake Nabugabo near the north-western shore of Lake Victoria) have low pH, oxygen, and salinity levels, and contain very few molluscs (Beadle & Lind, 1960). Similar conditions probably prevail in the muddy substrata of Lake Victoria, particularly offshore from extensive areas ofemergent grasses and papyrus. Although the actual factors that determine the population density of molluscs in such conditions have not been carefully examined, laboratory studies have shown that B. sudanica fails to develop under conditions of low salinity and oxygen levels (Beadle & Beadle, 1969). Probably, very muddy substrata are hazardous mainly to the embryos and hatchlings. The decline of B. choanomphala in increasingly muddy substrata is not due to the depth of water since these substrata often occur even in shallow water up to 1.0 m deep, compared with the 3-12 m depth at which B. choanomphala was found by Webbe (1962). However, the maximum depth of water in which B. choanomphala can survive may be deter- mined by oxygen depletion in the bottom layers resulting from thermal stratification in bays (Talling, 1957), since this species has not been observed to rise to the surface for air. Other lake-dwelling pul- monates, including two species of Bulinus, have been collected from even greater depths (up to 104 m) in Lake Malawi (Wright et al., 1967). Although B. choanomphala is associated with par- ticular substrata, no association with emergent vege- tation was observed. The absence of these snails from such vegetation contrasts with the marked attractiveness of certain aquatic plants for some other lake-dwelling snail hosts of schistosomes, such as Bulinus truncatus rohlfsi in the new Volta Lake in Ghana (Paperna, 1969) and Bulinus (Physopsis) 338 BIOLOGY OF BIOMPHALARIA IN LAKE VICTORIA AT MWANZA africanus and Biomphalaria pfeifferi in Lake Kariba in Zambia (Hira, 1969). The marked seasonal fluctuations in density and age structure of B. choanomphala, and variations between the populations of different bays, could not be accounted for by sampling error of the dredge method alone. Even if the dredge collected only 50% of the snails on the bottom in a transect, differences in snail densities of 10-13-fold at different sites and times of the year could readily be detected. It is suggested that the pronounced seasonal fluctuations in snail densities may depend in part on environ- mental factors in the lake that may periodically stimulate increased reproductive activity and enhance survival rates. The most obvious cyclic factor is the annual rise and fall in the levels of water averaging 0.5 m (Temple, 1964). In the present study it was seen that peak densities of B. choanom- phala did not coincide in time in the different popu- lations. Whereas low densities occurred between September 1970 and March 1971 in Mwanza South Bay, high densities were found during the same period in Pasiansi Bay. This suggests that seasonal variations in density are not in fact linked with changes in the level of the lake. On the other hand, fluctuations might be related to more local pheno- mena, such as the upwellings reported by Fish (1954) and Kitaka (1969), which allow for the mix- ing of bottom and surface waters and influence the fertility of coastal regions. Thus, it would be of interest to study the annual levels of dissolved nutrients and algal growth in shallow bays in rela- tion to fluctuations in the snail population. Studies on fluctuations in the populations of the ciliates and annelids that are predatory on B. choanomphala, might reveal the possibility of biological control. Other more localized phenomena, such as the inter- mittent inflow of streams carrying nutrient-rich water and predation by fishes, might have more limited effects on particular populations. The observations on S. mansoni infection rates in B. choanomphala and B. suidanica are preliminary, and further studies over a period of several more years, large samples of snails being screened monthly at several sites, are needed. However, it can at least be said that the relatively high frequency of infection in B. choanomphala, despite the large dilution factor in the lake, is no doubt partly a result of the high susceptibility of the species to S. mansoni, as shown in laboratory studies. Prentice (1970) reported an infection rate of 38-77% in an Entebbe strain of B. choanomphala exposed to 3 miracidia of a local strain of S. mansoni. A 57% infection rate has been obtained with a Mwanza strain of B. choanomphala exposed to 10 miracidia (Magendantz, unpublished data). Besides variations in the susceptibility of dif- ferent strains, the following environmental factors might also influence natural infection rates in dif- ferent populations of B. choanomphala: (1) snail densities and age structures on the lake bottom; (2) distance offshore and the depth at which snails occur; (3) the degree of faecal pollution and the nature of the shoreline; and (4) the number of people using particular sites and the proximity of houses and latrines to the lake shore. Snail densities do not appear to be the decisive factor in determining infection rates since popula- tion densities in Pasiansi Bay were as high as those in Mwanza South Bay, but few infected snails were collected from Pasiansi Bay. Second, infected snails were found in Mwanza North Port on several occa- sions when snail densities were lower than at any time in Mwanza South Bay. In fact, infection rates tended to be directly proportional to the age of the snails and inversely proportional to their density, the highest infection rates occurring when seasonal snail populations contained the largest proportion of adult snails but were at their lowest density. Infection rates also seem to be related to the dis- tance offshore and the depth at which the snails occur. In Pasiansi Bay, where infected snails were collected on only one occasion, the highest densities were found as far as 150 m offshore at depths of 1-3 m, while in many Mwanza North Bay and South Bay sites where infected snails were collected frequently, the highest densities occurred within 30-40 m offshore at depths of 1-3 m. With regard to the nature of the shoreline in relation to infection rates, there is probably a greater chance that faecal matter will be washed into the lake from a sloping shore with continuous seepage, as in Mwanza South Bay, thus allowing for higher infection rates, than from a raised dry shore, as found in much of Pasiansi Bay. Another factor that may contribute to the lower snail infection rates in Pasiansi Bay is the absence of houses and latrines close to the shore, while at Mwanza South there are many houses within 10-50m from the shore. Nevertheless, despite low snail infec- tion rates, the high prevalence of S. mansoni infec- tions in inhabitants using the lake at Pasiansi (McCullough & Eyakuze, unpublished data) sug- gests that snail infection rates are not necessarily proportional to schistosomiasis prevalence rates. It is possible, however, that if the Pasiansi shore were 339 M. MAGENDANTZ more heavily inhabited and polluted, snail infection rates and schistosomiasis prevalence would be higher than they are. It has been suggested that the discharge of sewage from lake steamers contributes to S. mansoni trans- mission in Mwanza North Port. Infected B. sudanica and B. choanomphala have been found near the Karumo ferry landing, but although the steamers dock nearby, the heavily polluted shore behind the ferry landing is undoubtedly a major source of infec- tion. This landing is considered to be important because of the large numbers of people who pass to and from the ferry and, in the absence of a public latrine and water tap, use the lake. The present study revealed a relatively high fre- quency of infected B. sudanica in Mwanza in com- parison with the results of a recent study on the north shore of Lake Victoria, in which none of 24 500 B. sudanica snails from Kampala, Port Bell, and Entebbe were found to be infected (Prentice et al., 1970). Varying infection rates in B. sudanica, as in B. choanomphala, may be due to both intrinsic and extrinsic (environmental) factors. Different strains of B. sudanica seem to vary considerably in susceptibility. McClelland (1962) obtained a 52% infection rate with a Mwanza strain of B. sudanica exposed to 8 miracidia. Prentice et al. (1970) reported infection rates of 0-16% in B. sudanica from Entebbe and the West Nile exposed to 3 mira- cidia originating from each of these regions, but an infection rate of 41 % with an Entebbe strain of B. sudanica exposed to 10 miracidia of a local strain. Local differences in natural infection rates of B. suda- nica at Mwanza are probably largely the result of external factors; the main factor contributing to high frequencies of infected snails along sections of Mwanza North and South bays appears to be the proximity of snail habitats to open shores where there is much bathing, washing, and fishing, and pollution arising from many houses and latrines. The latter are sometimes so poorly built and badly sited that they drain into the lake. The high fre- quency of infected B. choanomphala and B. sudanica suggests that people who are dependent on the lake in any way are likely to be exposed repeatedly to infection. The fact that on average 49% of the children living along the Mwanza South shore are infected with S. mansoni (McCullough & Magen- dantz, 1972) is evidence of intense and regular transmission. In view of the high prevalence of S. mansoni among the inhabitants living near the lake shore, the heavy faecal pollution of the lake in these places, and the high population densities of both B. choanomphala and B. sudanica, a control programme involving the relocation of housing, improved sanitation, mollusci- ciding, and chemotherapy will be necessary to reduce S. mansoni transmission effectively in Mwanza. Blanket mollusciciding along the entire shoreline (over 13 km), however, would be unwarranted and uneconomic. After the relocation of housing away from the lake shore, land-filling the shoreline, and the provision of more public water taps and latrines, the author would favour selective focal molluscicid- ing of the lake bottom at selected sites where there is unavoidable human contact with the water, such as at fishing boat landings, and in places where there are high densities of snails. Such focal molluscicid- ing would not upset the overall balance of aquatic life in the various bays. Clearance of vegetation along the Mwanza shore is to be recommended where it is accompanied by the complete filling-in of the shore and by prohibition of housing, bathing, and defaecation along the shore. By itself, the clearance of emergent vegetation opens up new areas of contact with the lake, removes the source of large amounts of organic matter that are deterrents to the growth of mollusc populations on the lake bottom, and is followed by the formation of grassy seepages and the digging of irrigation furrows in which B. sudanica thrive. In Mwanza, a coordinated " environmental " plan involving the town's medical, water and engineering authorities as well as local leaders and factory man- agers is urgently needed to safeguard the health of the population. Such a plan should include health edu- cation, improved sanitation and water supplies, vector control, and the reduction of industrial pollution. ACKNOWLEDGEMENTS The author thanks Dr V. M. Eyakuze, Director, East African Institute for Medical Research, for helpful advice and for the provision of laboratory space and facilities; Dr G. Webbe, Reader in Medical Parasitology, London School of Hygiene and Tropical Medicine, for stimulating discussions; Mr I. Justine for assistance in the field; and Mrs F. McCullough for secretarial and other assistance. The study was supported byaWHO Research Grant. 340 BIOLOGY OF BIOMPHALARIA IN LAKE VICTORIA AT MWANZA 341 RISUMt LA BIOLOGIE DE BIOMPHALARIA CHOANOMPHALA ET DE B. SUDANICA AU REGARD DE LEUR ROLE DANS LA TRANSMISSION DE SCHISTOSOMA MANSONI DANS LE LAC VICTORIA A MWANZA (TANZANIE) Une 6tude de la biologie des mollusques h6tes inter- mediaires de Schistosoma mansoni dans le lac Victoria a e entreprise a Mwanza (Tanzanie) en octobre 1969. On se proposait de determiner la r6partition et les variations saisonnieres de Biomphalaria choanomphala et de B. sudanica en fonction de la nature des sediments du lac et des caracteristiques de la rive et d'analyser les facteurs agissant sur l'intensit6 de la transmission de S. mansoni. Les recherches ont ete effectuees en 70 endroits, pres de Mwanza ou dans les baies voisines, les mollusques etant recueillis a l'aide d'un filet metallique traine sur le fond du lac. On a releve une nette correlation entre d'une part les variations de la r6partition et de la densit6 de B. choanomphala et d'autre part les caracteristiques physiques du fond et de la v6getation de la rive. Pres de la rive, oui le fond etait form6 d'un melange de sable et de vase, la densite des mollusques atteignait 1-20/M2; plus au large, oui la vase predominait, elle etait inferieure I 1/rM2. Les populations de B. choanomphala pr6sentaient des variations saisonnieres de la structure par age et leur densite fluctuait dans des proportions atteignant 10-13 fois. Des amas d'ceufs ont ete recolt6s en abondance et en nombre variable suivant la saison. Deux formes du vecteur, une grande et une petite, correspondant peut- etre a des 6cophenotypes, ont et6 identifiees. B. choanom- phala etait infecte par S. mansoni dans la proportion de 0,2 a 3,3 %; les taux d'infection semblaient plus 6leves quand la population etait formee en majeure partie de mollusques adultes. On a decouvert B. sudanica dans de nombreux habitats situes au-dessus du niveau du lac, principalement dans les zones herbeuses oiu l'eau s'etait infiltree et dans les rigoles d'irrigation. La periode de reproduction maximale de 1'espece s'etendait sur la saison des pluies et le debut de la saison seche. Des populations denses se sont main- tenues dans les rigoles d'irrigation pendant la longue saison seche de 1970-71; dans les zones herbeuses asse- chees, les mollusques n'ont pas survecu plus de 3 mois. On a frequemment trouve des B. choanomphala et B. sudanica emettant des cercaires de S. mansoni dans des endroits fortement pollues par des matieres fecales et le long des portions de rive les plus frequentees. Ces facteurs, de meme que la frequence des approvisionne- ments en eau, la proximite des habitations et la presence d'habitats de mollusqUes a faible distance de la rive semblent en grande partie responsables des variations des taux d'infection suivant les endroits. REFERENCES Baalawy, S. S. (1971) E. Afr. med. J., 48, 385-388 Beadle, L. C. & Beadle, S. F. (1969) J. exp. Biol., 50, 491-499 Beadle, L. C. & Lind, E. M. (1960) Uganda J., 24, 84-98 Berrie, A. D. (1964) Ann. trop. Med. Parasit., 58, 457-466 Cridland, C. C. (1955) J. trop. Med. Hyg., 58, 1-11 Fish, E. A. (1954) In: East African Freshwater Fisheries Research Organization Annual Report for 1953, Jinja, Uganda, pp. 4-5 Fish, E. A. (1955) In: East African Freshwater Fisheries Research Organization Annual Reportfor 1954-55, Jinja, Uganda, pp. 3-4 Hira, P. R. (1969). Nature (Lond.), 224, 670-672 Kitaka, G. E. B. (1969) In: East African Freshwater Fisheries Research Organization Annual Report for 1968, East African Community, Jinja, Uganda, pp. 39-43 Mandahl-Barth, G. (1957) Bull. Wld Hith Org., 16, 1103-1146 McClelland, W. F. (1962) In: East African Institute for Medical Research Annual Report for 1961-62, Mwanza, Tanzania, East African Common Services Organi- zation, pp. 16-17 McClelland, W. F. & Jordan, P. (1962). Ann. trop. Med. Parasit., 56, 396-400 McCullough, F. S. et al. (1972) E. Afr. med. J. (in press) Paperna, I. (1969). Bull. Inst. fr. Afr. noire, A, 31, 497- 499 Prentice, M. A. (1966). In: East African Institute for Medical Research Annual Report for 1964-65, Mwanza, Tanzania, East African Common Services Organi- zation, pp. 8-9 Prentice, M. A. (1970). In: Proceedings of the OAU Symposium on Schistosomiasis, November 1970, Addis Ababa Prentice, M. A. et al. (1970). Ann. trop. Med. Parasit., 64, 339-348 Schwetz, J. (1951). Ann. Parasit. hum. comp., 26, 323-333 Talling, J. F. (1967). Proc. roy. Soc. B, 147, 57-83 Temple, P. H. (1964). Proc. E. Afr. Acad., 2, 50-58 Webbe, G. (1962). In: Wolstenholme, G. E. W. & O'Connor, M., ed., Bilharziasis, London, Churchill, p. 7-22 Wijers, D. J. B. & Munanga, P. N. (1971). E. Afr. med. J., 48, 136-140 Wright, C. A. et al. (1967). J. zool. Res., 151, 199-209