Better desiccated than eaten by fish: Distribution of anurans among habitats with different risks to offspring

Abstract

Abstract Freshwater species with complex life cycles face a trade‐off between the risks of offspring mortality due to desiccation in temporary habitats and due to predators common in long‐duration habitats, especially fish. In real‐world conditions, intermediate‐gradient areas that are highly suitable for some ecological specialists are often limited. I examined the relative significance of drivers of the permanence–predation gradient in habitat selection by pond‐breeding anurans. Anuran oviposition was investigated at the community level in fish pond landscapes with only three types of habitat, permanent fish‐containing ponds and highly desiccation‐prone pools, either fishless or recently emptied of fish and with fish odour (fish cue). Altogether, 65 ponds/pools, interspersed in four clusters, were examined for the presence of egg masses. Bufo bufo was the only species preferring permanent ponds. Egg masses of Rana sp. (Rana arvalis/Rana temporaria) occurred in all fishless pools, and less frequently in the ponds. Bufotes viridis and long larval period species Pelobates fuscus and Hyla orientalis oviposited almost exclusively in fishless pools. Fish‐cue pools were avoided by all species. Anuran richness was higher in fishless pools than in permanent ponds. Species distribution between the two habitats was nested, with both common and rare species occurring in fishless pools. The results indicate the primacy of offspring predation risk over hydroperiod constraints in oviposition decision‐making by fish‐intolerant anurans. The absence of some species from ponds and pools with fish or fish cues shows that non‐consumptive interactions may better explain the scarcity of anuran larvae in waters dominated by fish than actual consumption. The strict avoidance of fish habitats despite their proximity to fishless patches indicates fine‐scale assessment of predation risk. However, rigid habitat selection against fish predation on offspring may prove maladaptive if habitats with high desiccation risk are the main alternative and are indiscriminately preferred. Maintenance of fishless wetlands with a hydroperiod sufficient to allow completion of metamorphosis should be a conservation priority for anuran diversity in areas where fishless habitats are limited.