The rock pools on the river bank of the Comoé National Park (West Africa) provide a very diverse and unpredictable environment for anuran larval development. Because rock pools differ considerably in biotic and abiotic parameters, it should be adaptive for reproducing anurans to choose the most suitable oviposition sites. During the beginning of each rainy season (March to May), from 1991 to 1995, we investigated the allocation of breeding sites by Hoplobatrachus occipitalis, counted the number of eggs, and measured several biotic and abiotic factors that might influence the choice of a spawning site. The probability of predation by conspecific cannibalistic tadpoles and the water-holding capacity (WHC) of pools were the best predictor of number of eggs laid. We experimentally investigated the influence of these two parameters on egg-laying and showed that adults potentially can assess the presence, density, and size of tadpoles in pools by chemical cues. Likewise, manipulation of the water-holding capacity caused a rapid change in egg-laying behavior. To assess the risk of desiccation, frogs have to visit familiar pools repeatedly to monitor the decrease in volume over time and thus gain information about the pools' water-holding capacity.

No abstract

Most anurans of the species‐rich community of the Comoé National Park (Ivory Coast) use ephemeral savannah ponds to spawn. Owing to the great risk of desiccation and the large number of predators, the mortality for tadpoles is very high in these ponds. Therefore, colonization of other spawning habitats might be highly advantageous. Such spawning sites are presented by the Comoé river, which is characterized by frequent and unpredictable changes of the water level. Only Bufo maculatus (Anura: Bufonidae) and one other bufonid species breed in small and shallow inlets and puddles created by the rising and falling of the predator‐rich Comoé river. We observed that predatory fish advanced to the spawning sites of the toads, when the water level rose. If the toad larvae were attacked by these predators, they formed dense aggregations of up to several thousand individuals. These aggregations were maintained for longer periods on one place at the river’s edge where they are inaccessible to the larger predators. In field experiments we ascertained that this aggregation behaviour in tadpoles was caused by a combination of two stimuli: first, a chemical cue from injured tadpoles and second, a mechanical stimulus caused by rapid movements of aquatic predators. Initial trials indicated that tadpoles aggregating at the bank were, however, disadvantaged compared with free‐swimming larvae in having slower growth and an increased risk of desiccation. This is presumably why aggregations broke up within 24 h after the predators had left these puddles, as the water level fell. At this point these tadpoles either spread out or formed loose swarms near the bottom of the puddles. This swarming behaviour differed considerably from that induced by aquatic predators.

No abstract

The nematode Rhabdias bufonis is an anuran lung parasite with a heterogonic life cycle, characterized by alternation of gonochoristic (free-living) and hermaphroditic (parasitic at late larval and adult stages) generations. Adult parasites containing unhatched embryos were isolated from infected amphibians, but all other stages were cultured under controlled laboratory conditions with bacteria on agar plates. Development was studied with particular reference to a comparison of the two generations. No deviation from the strict heterogonic life cycle could be induced by alterations in the conditions of culture. The early embryonic lineages in the two generations are similar but not identical. Early development in both generations takes approximately twice as long as that in Caenorhabditis elegans, and the sequence of divisions is more variable. Nuclei counted in late embryos were similar in number (mean numbers: free-living generation 502; parasitic generation 570). Their number increases only moderately in the free-living generation, reaching a mean of 873 for non-gonadal tissue in adults, similar to that (816) in C. elegans. In contrast, parasitic adults contain in total approximately 4700 somatic and of these 2600 intestinal nuclei -about five times and one hundred times, respectively, the numbers present in free-living adults. We conclude that the dramatic differences in body size and cell numbers between the two generations can be attributed mainly to additional gut tissue-specific cell divisions during the late larval or adult stages of development of the parasitic generation.