/ L.' JOURNAL THE EAST AFRICA NATURAL HISTORY SOCIETY AND NATIONAL MUSEUM 19 May 1983 No. 179 PREDATION BY MONGOOSES, RODENTS AND SNAILS ON SITALA JENYNSI (PER.), ACHATINA FULICA BOWDICH AND OTHER LAND SNAILS IN COASTAL TANZANIA. P.F. Kasigwa, A.J. Mrema * and J.A. Allen + Department of Zoology, P.O. Box 35064, University of Dar es Salaam, Tanzania SUMMARY The broken shells of Sitala jenynsi, Achatina fulica and other land snails were collected from two sites near Dar es Salaam, Tanzania. Observations in both field and laboratory suggested that most of the shell damage was the result of predation by a snail iEdentulina obesa), the Banded Mongoose (Mungos mungo), rodents and (perhaps) birds. E. obesa was shown to eat some snail species readily, yet neglect others. Details of its predatory behaviour were obtained and there is some evidence that it prefers smaller specimens of A. fulica. Field observations of mongooses revealed that they have a number of different ways of breaking snail shells and these methods dif- fer from those used by captive rodents. Birds of an undetected species may use yet another method. Some idea of the relative numbers of S. jenynsi eaten by these four groups of predators was obtained by first clearing the empty shells from an area of ground and then removing the sheik that accumulated over each of four successive two or four-monthly intervals. The relative population sizes of the living snails over this period were also estimated. Forty to one hundred percent of the snails eaten by shell-damaging predators was ascribed to E. obesa and there are in- dications that the proportion of the population that was destroyed differed between the four sam- ples. INTRODUCTION The dead shells of tropical African land snails have received some attention from collectors and taxonomists, but the ecology of the living animals is still poorly known. We have been studying the land snails inhabiting the scrubland around Dar es Salaam, Tanzania, and in this paper provide information on the identity, behaviour and relative importance of some of their predators. For practical reasons we have concentrated on the commoner molluscs and on those predators that damage the shell in some way. METHODS Field Sites The snail populations studied were in and around the grounds of the main campus of the University of Dar es Salaam and at Wazo Hill, respectively 10 and 17 kilometres north-west of the city of Dar es Salaam. The two areas have a similar altitude (University: 60 - 1 10m above sea level, Wazo: 80-110 m above sea level). The vegetation in both places consists of thick scrub (shrubs, small trees and climbers of various species) except for areas where herbs and grasses predominate after the land * Present address: Science Research Council, Dar es Salaam, Tanzania. Page 2 No. 179 had been cleared for agriculture at some time in the past (Kasigwa, 1975). One important geological difference is that whereas the University site is on the edge of a clay-bound sand plateau, Wazo Hill is topped by a raised Pleistocene coral reef (Temple, 1970). The presence of the limestone at Wazo Hill is reflected not only by the pH of the soil (Wazo: 7.0, University: 6.0), but also by the greater mean weight of the shells of the Wazo snails (Kasigwa, 1975). The annual climate at both sites is dominated by the occurrence of two rainy seasons (roughly, March to May and October to December), the second one being the less predictable. Bargman (1970) gives more information on the climate of the region. Prey We concentrated our attention on Sitala jenynsi (Pfr.) (Pulmonata: Urocyclidae) (Fig. 1). This was the most accessible snail at the two sites and occurred at quite high densities (approx.3m"^ ; Kasigwa, 1975) and in discrete populations. During the day, and whatever the season, the snails remain on the leaves and branches of the vegetation (particularly shrubs and trees). S. jenynsi is an ‘annual’ species; the snails are usually born at the onset of the first rainy season and rarely live beyond 14 months (Kasigwa, 1975). 10mm a Fig. 1. (a) Sitala Jenynsi, The next most studied species was Achatina fulica (Bowdich) (Pulmonata: Achatinidae), the Giant African Snail of Mead (1961) (Fig. 1). This species aestivates below ground in response to prolonged dry conditions and is therefore most obvious during the rainy seasons. Even when ac- tive, in our study areas it tends to remain close to the ground. Eight other species featured less prominently in our work, mainly because they were uncommon and/or very seasonal. These were: Edentulina obesa (Gibbons) and Gulella laevigata (Dohrn) (Streptaxidae); Edouardia alycaeoides (Verdcourt), E. tumida (Taylor), Rachis punctata (Anthon), Rachistia mozambicensis (Pfeiffer) and Rachadina braunsi (von Martens) (Enidae); and Tropidophora letourneuxi (Bgt.) (Pomatiasidae). No. 179 Page 3 Identification of predators Broken and unbroken empty shells were collected from the ground at the two sites in 1971 and 1972. These were first sorted according to the pattern of shell breakage (if any) and we then attempted to identify the predator responsible for each category of damage. Four groups emerged. Later these were associated (with differing degrees of confidence) with four classes of predator: the streptaxid snail Edentulina obesa (also one of the prey species), the Banded Mongoose Mungos mungo (Gmelin), rodents (probably of at least tw*© species) and, possibly, birds of unidentified species. Some broken shells could not be ascribed to any of the four categories with any degree of certainty — this applied to all the broken T. letourneuxi and to about 10 per cent of the S. jenynsi. We were unable to discover the cause of death of unbroken shells, but suspect senescence for the older (age >1 1 months) S. jenynsi. Two of the agents (£’. obesa and M. mungo) were identified through direct observation in the study areas. The rodents were implicated from the nature of the damage to snails when offered to captive rats, while the evidence for bird predation is circumstantial and rests mainly on the presence of beak-shaped marks on some of the shells. Behaviour of predators E. obesa Information on the behaviour of E. obesa was obtained from observing the snail in both field and laboratory. In the laboratory twelve adult snails were housed, usually individually, in transparent plastic sandwich boxes (volume approx.O.OOlm^ lined with tissue paper. The paper was kept moist and was replaced frequently. Some experiments were designed simply to test the acceptability of different snail species to E. obesa. A single individual of a potential prey species was introduced and the box was then checked periodically for evidence of predation. Snails that were still alive after 48 hours were defined as ‘safe’. No more than one snail per day was offered to each E. obesa and the sequence in which different prey species were offered was randomized. It was quite clear that some species were eaten avidly while others were neglected. This sort of design was also used to obtain details of the sequence of events when E.- obesa successfully attacks a snail, except that observations were more thorough, and were continuous during the early stages of predation. In a series of experiments designed to test whether E. obesa selects different size classes of A. fulica, two E. obesa were housed in separate glass tanks (150mm x 150mm x 1 30mm). Six ‘small’ (shell height approximately 15mm, range 13 - 17mm) and six ‘large’ (shell height approx. 25mm, range 22 - 27mm) A. fulica were added to each tank. The containers were checked frequently and each eaten snail was replaced by another of the same size-class to keep the 1:1 ratio constant. Banded Mongoose All observations on M. mungo were made in the field. Binoculars were usually used and the observer hid behind vegetation for cover. Rodents In order to test whether rodents prey on S. Jenynsi. live snails of this species were left overnight on three different occasions in eight cages each containing a single rodent which had been caught on the University site. The rodents were three Praomys (Mastomys) natalensis (A. Smith), three Mus musculus L., one Acomys sp. and one Praomys (Praomys) sp. Each animal was offered six snails at a time and was also provided with water and commercial rat pellets. The snails were examined the following day for evidence of predation. Birds (?) No data on the behaviour of presumed avian predators were obtained. Relative importance of predators of S. jenynsi An attempt was made to assess the relative importance of the four groups of predators in the mortality of S. jenynsi. In mid- April 1973 a small (10m x 12m) area of ground in one particular sampling site (‘Hall 6 locality’) at the University was cleared of all dead S. jenynsi shells. This plot was then checked in November 1973, March 1974 and May 1974. On each occasion all the dead shells were collected and the number destroyed by each predatory agent was noted. At the same time we also recorded the total number of live S. jenynsi seen on a set route walked through the Hall 6 locality. The shrubs and trees on this route were thoroughly searched and the numbers of 5. Jenynsi counted. This method of ‘counting heads’ is quite efficient, and accounts for over 60 per cent of the population in the area sampled (Kasigwa, 1975). It provided a measure of relative population size in the Hall 6 locality over the 10 month period. Page 4 No. 179 RESULTS AND OBSERVATIONS Predation by E. obesa Casual observations suggest that E. obesa occurs at a density appreciably less than one per 1 00 m*. When active (in the rainy seasons) it appears to confine itself to shrubs and trees, existing on a diet of other arboreal molluscs. Of these, S. jenynsi is usually the most widespread and abundant and is likely to form the major dietary component. Attacks by E. obesa on S. jenynsi have been witnessed on at least 20 occasions in the field, and in plastic boxes in the laboratory on many more. There is no evidence that S. Jenynsi perceives its aggressor when beyond a range of about 4cm; that is, virtually until just before tactile contact. The predator often appears to ‘strike’ in a fast movement from a distance of 1-2 cm, with its front end fully extended. The response of the prey, if it is not already retracted, is to withdraw rapidly into the shell. In the field this often results in S. Jenynsi rolling from its position, usually into the undergrowth below. It is tempting to suggest that this behaviour is adaptive, for should E. obesa make firm contact with its prey predation nearly always follows. Observations on captive snails have revealed the sequence of events in the feeding behaviour of E. obesa. After a successful ‘strike’ the front end of the body is inserted into the aperture of the shell and grips the body of the prey, which in response attempts to withdraw into the upper whorls. As the predator extends into the distal regions of the prey, the two shells come to touch one another, aperture facing aperture. E. obesa then retracts its tail end, and the bodies of both snails become scarcely visible. This position may be maintained for up to 12 hours, with the predator occasionally moving its shell from side to side (presumably as it delves deeper into the prey, or as the mouth alters its grip). During this operation copious amounts of mucus are produced (by the predator?), securely anchoring the shell of the prey to the substratum. Lastly, the predator relaxes its hold, withdraws its head end into its own shell, and then thrusts the posterior part of its body into the aperture Qf the prey. In this position it falls into a ‘dormancy’ lasting several hours before it finally abandons the prey. A shell which has just been attacked by E. obesa is identifiable not only by the dried mucus attaching it to its resting site but also by the pale brown E. obesa faecal matter on or near it. Predation by E. obesa is often so thorough that none of the soft tissues of the prey remain. In addition, the interior of the shell itself becomes corroded, particularly the columella and the in- ternal surfaces of the whorls (Fig. 2). Commonly all that is left is a translucent ‘ghost’, consisting mainly of the outer proteinaceous periostracum. Shells attacked by E. obesa are readily identified among recently dead shells on the ground. Other streptaxids, such as E. affinis Boetger are also known to attack the calcified layers of the shell (Williams, 1951). Fig. 2. Predation by the snail Edentulina obesa on two specimens ((a) and (b)) of S. Jenynsi. Stippling in- dicates areas of shell corrosion. (Drawn from photographs by Kasigwa, 1975.) No. 179 Page 5 E. obesa will prey on other snail species when available. In the field we have observed it feeding on A. fulica (small juveniles only), Edouardia spp., Rachis punctata, Rachistia mozam- bicensis and Rachadina braunsi. The last four of these, like S. jenynsi, have a pronounced ten- dency to climb and are therefore likely to be encountered by E. obesa. The only climbing snail we have never seen being eaten is the prosobranch Tropidophora letourneuxi. Laboratory ob- servations on two E. obesa showed that they would actually attack T. letourneuxi when out of its shell, but would desist after it had retracted and closed its operture with its operculum. Williams (1951) makes a similar observation for another prosobranch, Maizania magilensis (Craven) when attacked by E. affinis. Other snails which live close to the ground, such as Gulella laevigata and Pseudopeas sp. were never seen to fall prey to E. obesa in the field. In part this may be due to the difficulties of us fin- ding snails in the dense vegetation at ground level. However, we have tested the acceptability of two of these species to captive E. obesa and one of them, Gulella laevigata, appears to be resistant to predation. This species, like T. letourneuxi, attracted attention when mobile, but after it retrac- ted it was soon neglected. Undoubtedly this was partly due to the small size of the aperture of this small snail (shell 8mm x 4mm), but the teeth projecting from the sides towards the centre of the aperture may also have made it difficult for the predator to gain entry. Further evidence for this is that Gulella usugarica Crosse (a larger species with a larger aperture) and G. alleni Verdcourt (similar in size to G. laevigata), both from the West Usambara mountains in Tanzania and both ‘toothed’, were also safe from predation. We also presented single Achatina fulica snails to E. obesa, and these were readily accepted when they were below a certain size. Not surprisingly, large snails (shells over 100mm x 50mm) were disregarded. (This vacant niche for such a molluscan predator is perhaps filled elsewhere in East Africa by Edentulina affinis a snail about twice the size of its congener (Williams, 1951).) However, young snails (shell height < 5mm) were readily accepted. There is also evidence from choice experiments that E. obesa prefers the smaller of two juvenile size-classes of A. fulica. The total numbers of the two size-classes eaten by a pair of snails are given in Table 1. Each snail took a statistically significantly higher proportion of the smaller prey. We suspect, therefore, that selec- tion varies with the size (and therefore age?) of the prey, but the full details of the response will only come from further experiments. Predation by two E. obesa when presented TABLE 1 with equal numbers of two sizes of A. fulica E. obesa A E. obesa B 15 mm No. eaten 11 21 A. fulica 25 mm No. eaten 3 6 The values are based on a 1:1 expectation ^ ^ (1) p 4.57 <0.05 8.33 <0.01 We have found, to our cost, that E. obesa will eat its own kind if more than one are kept in a container in the absence of alternative food. We have never observed such cannibalism in nature, but it has been recorded for two other streptaxids, E. affinis and Gonaxis kibureziensis E.A. Smith in coastal Kenya (Williams, 1951). Page 6 No. 179 Predation by the Banded Mongoose Mungos mungo uses at least four different methods of breaking snails, perhaps depending on the size and strength of the shell. 1 . In the case of juvenile S. jenynsi ( < 4mm diameter), it places the entire animal in its teeth and swallows the soft parts and most of the shell. 2. Adult S. jenynsi have a considerably thicker shell and are first held in the forepaws in a ver- tical position, apex upwards. Then, with its lower incisors against the lower side of the body whorl and its upper incisors on the upper side of one of the spire whorls, the mongoose bites off a portion of the shell and pulls out the body with its teeth, leaving the empty shell characteristically agape on one side (Fig. 3). A similar technique is used on E. obesa but the position of the bite is more variable. Of eight specimens eaten by mongooses, three were bitten at a perpendicular or oblique angle to the long axis of the shell; the remainder had a portion bitten from one side. a Fig. 3. Two methods of predation by the Banded Mongoose, Mungos mungo, on S. Jenynsi: (a) by biting a piece off the side of the shell (removed area indicated by stippling), (b) by breaking successive pieces from the aperture and along the lower whorls, ((a) Drawn from photograph by Kasigwa (1975); (b) drawn from specimen.) 3. Sometimes M. mungo deals with 5. Jenynsi by biting sections off the shell, starting at the aper- ture and working along the whorls until the withdrawn body is accessible to the teeth. Shells so attacked have the lower portion of the columella characteristically exposed (Fig. 3). A. fulica with shells below about 50mm in height are also broken in this way (Fig. 4). This method has been recorded for small mammals preying on helicid snails in Europe (Kerney and Cameron, 1979). Fig. 4. Two methods of predation by the Banded Mongoose, M. mungo, on A. fulica: (a) by biting successive pieces from the aperture and along the lower whorls, (b) by breaking the shell (stippled area) byj throwing it against a rock. (Drawn from photographs by Kasigwa, 1975.) No. 179 Page 7 4. Larger, more resistant /I . /wZ/ca are broken against rocks or tree-trunks. For this operation the mongoose assumes the spectacular postures described by Hinton and Dunn (1967) and Eisner and Davis (1967). The animal holds the prey in both forepaws and stands bipedally on its hind legs with its back to the ‘target object’. It then swings its forelegs back-and-forth beneath itself for a number of times (normally 2-4) before throwing the snail under its body; at this point the hindlegs leap into the air, above the snail eti route to the target. This operation is repeated until the soft parts become accessible, usually because of a crack in the middle of the whorls (Fig. 4). Williams (1951) reports that a Black -legged Mongoose in Zanzibar, Bdeogale tenuis (Thomas and Wroughton), also feeds on A. fulica but makes no mention of the methods described above. Instead he describes a fifth behaviour: ‘The animal holds the mollusk between its paws and strikes it rapidly against some hard object until the shell is broken’. Mead (1961) provided indirect evidence for similar behaviour by the mongoose Herpestes edwardsi Geoffroy when preying on A. fulica in Ceylon (Sri Lanka). Predation by Rodents Empty S. jenynsi shells lacking the top 3 or 4 whorls were commonly found on the ground in the study area. Rodents were suspected as candidates for such predation and this hypothesis was tested by experiments with eight captive animals: of these, only the single Acomys and the three P.(M.) natalensis accepted the snails, which they did on all three nights. They apparently treated the snails in the same way, removing the tops of the spires and leaving an empty shell similar to those found in the field. We were unable to observe the predators in action but deduce that the rodent holds the snail in its teeth (and forepaws?) with the apex within its mouth, bites through both sides of whorls 3 and/or 4, swallows the apical portion (including the shell) and then pulls out the soft parts. The lower whorls are therefore left clean but with a characteristically ragged edge where the snail has been cracked by the teeth (Fig. 5). Similar evidence for rodent predation has not been obtained for other snail species. It is interesting to note that the captive rodents did not use the method of breaking successive pieces off the aperture, as European rodents apparently sometimes do when feeding on helicids (Kerney and Cameron, 1979) and as we have recorded oc- casionally for mongooses. European rodents have also been recorded as using the same method as their African relatives (Bang and Dahlstrom, 1974; JAA, personal observations). a b Fig. 5. Predation by rodents on two specimens ((a) and (b)) of S. jenynsi. Stippling indicates external areas of the shell that have been removed. (Drawn from photographs by Kasigwa, 1975.) Predation by birds (?) The marks on some dead and living S. jenynsi snails suggested to us that they might have been attacked by birds, but we never witnessed such an event. Such specimens are similar to those eaten by rodents in that the tip has been nipped off, but the cut is cleaner and closer to the apex (i.e. at about the 3rd whorl). Snails which have recently been killed in this way sometimes have Page 8 No. 179 the bulk of the body remaining in the lower whorls. Some live S. jenynsi and E. obesa have been found with two V-shaped (beak?) indentations on the tip of the shell in the position expected from the appearance of the de-tipped shells. These marks could result when the snails slip out of the bird’s grip during predation. The bird which is our prime suspect is the White-browed Coucal, Centropus superciliosus Hemprich and Ehrenberg, a species which frequents the study areas, feeds occasionally on slugs (Moreau, 1935) and although not reported as eating snails, is generally omnivorous (Mackworth-Praed and Grant, 1957). It also has a beak corresponding in size to the marks found on some shells. Mead (1961) observed the related species C. chlororhynchus Blyth pecking holes in the shells of live A. fulica in Ceylon. Relative importance of predators of S. jenynsi Table 2 gives the numbers eaten by each class of predator for the four sampling occasions in 1973/74, along with the number of live S. Jenynsi observed in the whole of the Hall 6 locality. The data all refer to the same cohort, which was born in March and April, 1973. TABLE 2 Recorded predation on the 1973174 generation o/S. jenynsi at a site in the Hall 6 locality, University of Dar es Salaam. Total num- Month of Age ber sample (months) live snails counted in Hall 6 Numbers eaten in small site by: E.obesa M.mungo Rodents Birds? All No. of empty but unpred- ated shells July 1973 November 4 1099 17 1973 8 651 23 March 1974 12 461 42 May 1974 14 93 5 - - - \ 11 14 5 — 42 \ 9 10 7 9 68 74 4 3 - 12 41 By May 1974 the numbers in the 13114 generation had declined to below 10 per cent of their July 1973 value. Throughout this period E. obesa was clearly the most prominent of the four groups of predators, taking 40 to 100 per cent of the total recorded predation. It is clear that E. obesa did not feed in direct proportion to the density of S. Jenynsi: the numbers taken on the four occasions differ statistically significantly from an expectation based on the relative population sizes (goodness-of-fit X2,3) = 49.92, p < 0.001). Disproportionately very few snails were taken in July 1973 and disproportionately many more taken in March 1974. A number of factors probably contributed to this effect. The July snails were small and we might have been more likely to overlook them when the ground was searched. They were probably more liable to be destroyed totally by E. obesa (and the same applies to the other predators, none of which was recorded for July). If the results for July are ignored there is still appreciable overpredation for March and un- derpredation for November ( 14.49, p. < 0.001). The March data cover the start of the main rainy season, when there is a corresponding increase in the activity of E. obesa. The snail breeds at this time and if each individual feeds more and if there are more snails feeding (because of births) there will be increased predation. The data for predation by M. mungo and rodents during the last three sampling intervals do not differ statisticallv significantly from an expectation based on the availability of S. Jenynsi ( X’2,2, = 1.70 and X'^ (2-) = 4.39, respectively) on thej three occasions when data were collected. No. 179 Page 9 DISCUSSION Not surprisingly, the different species of predator have different ways of getting to the soft parts of snails. E. obesa differs from the other three groups in that the shell damage is caused after the snail has entered the aperture and eaten most of the body. Furthermore, the damage occurs from the inside, to the extent that the presence of the undigested periostracurti is often the only factor preventing local collapse. It is not certain whether all Edentulina-aita.ckQd snails suffer in this way — a certain fraction could be killed without noticeable shell damage. The adaptive significance of shell corrosion is clear: by first dissolving the prey’s shell and then absorbing the mineral salts E. obesa obtains a ready supply of ions for its own shell. Mongooses and rodents were sometimes recorded eating pieces of shell, but this behaviour may be due more to accident than design. The strategy adopted by a given predator when attempting to eat a snail will presumably depend, at least in part, on the size, strength and shape of the shell. These properties (which are not necessarily independent of one another) will clearly vary from species to species. It is for this reason, perhaps, that M. mungo employs a number of methods to crack A.fulica shells that it does not use on the much smaller S. jenynsi. There is also considerable variation in shell properties within species and this is manifest in two ways. First, within populations there is variation that will depend partly on the different ages of the individuals and partly on the different external (environmental) and internal (genetic) factors that act on them. Thus populations of A. fulica show considerable variation in shell size in com- mon with populations elsewhere (Mead, 1961). Our choice experiments suggest that E. obesa prey selectively on these snails, tending to take the smaller of the two size-classes presented. Life table analysis of A. fulica populations in Hawaii also provides evidence that other streptaxids iGonaxis quadrilaterilis (Preston), G. kibweziensis (E.A. Smith) and Euglandina rosea (Ferussac)) preferen- tially attack smaller snails (Nishida and Napompeth, 1975). It seems likely that this behaviour is more related to some characteristic of the soft parts of the prey rather than to the size or strength of the «hell per se. The properties of the shell also vary between populations of a given species. For example, the shells of Wazo Hill S. jenynsi tend to be larger, of a different shape, and heavier than those of their University counterparts. Although too few data are available to make an ac- curate comparison of the behaviour of predators at the two sites, we suspect that the choice of method used by M. mungo is related to the strength (= thickness?) of the shell. Compared with the other predators, E. obesa appears consistently to take more S. Jenynsi throughout the year, at least in the area that was sampled. It is probable that E. obesa depends en- tirely on moUuscan prey and, at least in our study areas, on S. Jenynsi in particular. Both the Ban- ded Mongoose, M. mungo, and one of the rodents, P. (M.) natalensis, are renowned as op- portunistic omnivores (Delany and Happold, 1979) and the importance of snails in their diet is unknown. Page 10 No. 179 ACKNOWLEDGEMENTS We are grateful for the facilities provided by the Department of Zoology, University of Dar es Salaam and for financial assistance in the form of two In-Country Scholarships to PFK from the Federal Republic of Germany. Dr B. Verdcourt helped in identifying the snails. We thank Simon Lawrence, Jon Cooper, Keith Anderson and Liz Hogg for helpful discussion and/or criticism of the manuscript. Helen Linford and Sue Coxson did most of the typing. REFERENCES BANG, P. & DAHLSTRON, P. (1974). Animal Tracks and Signs. Collins, London. BARGMAN, D.J. (1970). The climate of Dar es Salaam. Tanzania Notes and Records 7L 55 - 64. DELANY, M.J. & HAPPOLD, D.C.D. (1979). Ecology of African mammals. Longman, London. EISNER, T. & DAVIS, J.A. (1967). Mongoose throwing and smashing millipedes. Science 155: 577 - 579. HINTON, H.E. & DUNN, A.M.S. (1967). Mongooses: their natural history and behaviour. Oliver and Boyd, Edinburgh. KASIGWA, P.F. (1975). Studies on the ecological genetics of the land snail Sitala jenynsi (Pfr.). M.Sc. thesis. University of Dar es Salaam. KERNEY, M.P. & CAMERON, R.A.D. (1979). A Field Guide to the Land Snails of Britain and North-West Europe. Collins, London. MACKWORTH-PRAED, C.W. & GRANT, C.H.B. (1957). African Handbook of Birds. Series 2. Birds of Eastern Africa Vol. 1. Longmans, London. MEAD, A.R. (1961). The Giant African Snail: a problem in economic malacology. University of Chicago Press. MOREAU, R.E. ( 935). A synecological study of Usambara, Tanganyika Territory, with particular reference to birds. J. Ecol. 23:1-43. NISHIDA, T. & NAPOMPETH, B. (1975). Effect of age-specific predation on age distribution and survival of the Giant African Snail, Achatina fulica. Proc. Hawaii, ent. Soc. 22: 1 19-123. TEMPLE, P.H. (1970). Aspects of the geomorphology of the Dar es Salaam area. Tanzania Notes and Records 71: 21-54. WILLIAMS, F.X. (1951). Life-history studies of east African Achatina snails. Bull. Mus. Comp. zool. Harv. 105: 295-317. (Received 20 September 1982) Joint Editors: M.E.J. Gore, Shereen Karmali No. 179 Page 11 Page 12 SMITHSONIAN INSTITUTION LIBRARIES llillllllilll 3 9088 01313 7500 No. 179 NOTICE TO CONTRIBUTORS Papers on natural history subjects will be considered on the understanding that these are being offered exclusively to the Society and have not been submitted, at the same time, to any other journal. They should be addressed to the Editors, Journal of the East Africa Natural History Society and National Museum, P.O. Box 44486, Nairobi, and be accompanied by a covering letter from the author responsible for correspondence and decisions regarding the typescript. Two copies of the contribution should be forwarded to the Society; a further copy should be kept by the author for reference. The typescript should be on A4 size paper. Use one side only, double-spaced with margins of at least 5 cm at the top and left-hand side of the pages to allow for editorial instruction to printers. The pages must be numbered and all measurements are to be in metric units. The title page should contain the title of the article followed by the name of the author(s) and institution(s) where the research was carried out. A present address, if applicable, should be included as a footnote. Abstract to be on the second page and contain not more than 150 words. State the purpose of the study, procedures, main findings and conclusions. This should be intelligible in itself without reference to the paper. Nomenclature In cases where there are well known and accepted English names, e.g. for birds and larger mammals, these should be used throughout the text, accompanied by the scientific name on first mention. Authorities should be given beside the generic and specific names, when quoted for the first time, for all groups, with the exception of ornithological papers not dealing with taxonomic discussion. Text This is usually, but not necessarily, divided into parts with the headings Introduction, Methods, Results and Discussion, these being broken up with sub-headings where necessary. Main heads should be centred on the page and typed in capital letters. Subheads should be ranged to the left margin and typed in capitals. Any further headings should start with a capital letter followed by small letters on the left margin. No headings should be underlined. Underlining is an instruction to the printer to use italic type and should be reserved for all generic and species names, foreign words not adopted into English and for descriptions of bird or animal vocalisation. Acknowledgements As acknowledgements may imply endorsement of the data and conclusions contained in the authors work, it is advisable to obtain permission before quoting persons who have made contributions to the study. \ References Care should be taken to see that all references quoted in the text are given in full, in this section. To avoid confusion, references to books and periodicals should not be abbreviated and neither should they be underlined. For example; (in text) Cave & Macdonald (1955) or (Cave & Macdonald 1955) (in reference list) CAVE, F.O. & MACDONALD, J.D. 1955. Birds of the Sudan, Oliver & Boyd, London. Tables Each table must be typed, double-spaced on a separate sheet and numbered. Leave good spaces between columns and omit internal vertical and horizontal rules. Cite tables in the text in consecutive order. Illustrations Submit two sets of figures which should be well drawn on bristol board or heavy tracing paper with a photographed or xeroxed copy. Lettering must be of first quality (e.g. lettraset) and consideration must be given to size when reduced. Typewritten or freehand lettering will not be accepted unless put onto an overlay which can be followed by the printers for typesetting. Each figure should have a label pasted on its back giving the number of the figure, caption details and accreditation. Cite each figure in the text in consecutive order. A separate caption sheet should be added to the typescript. Photographs are acceptable if they have good definition and contrast and should be produced on glossy paper. Proofs The author will be sent one set of galley proofs for correction of typographical errors. Authors may be charged for any changes from the original typescript made by him at this stage, and which result in additional printing expense. Plates or page proofs will not be sent. Published material The author is entitled to 25 copies free of charge. Further copies can be supplied at cost price, provided the order is placed when the typescript is submitted. If there is more than one author, 30 copies will be sent to the corresponding author for distribution. The society is supported entirely by membership subscriptions and may not, for financial reasons, be able to accept all submissions of value. Wherever possible an author should seek a grant, from his institution or other source, to help with costs of publication. Published by The East African Natural History Society, Box 44486, Nairobi, Kenya. Phototypesetting by Kul Graphics Ltd., P.O. Box 18095 and printed by Manhattan Displays, P.O. Box 30434, Nairobi.