Josi, D, Taborsky, M and Frommen, JG (2018) Habitat preferences depend on substrate quality in a cooperative breeder. Evolutionary Ecology Research, 19 (5). pp. 517-527. ISSN 1827-7888
|
Published Version
Available under License In Copyright. Download (250kB) | Preview |
Abstract
© 2018 Dario Josi. Background: The evolution of complex social organization is mediated by diverse environmental constraints, including predation risk and the availability and distribution of food resources, mating partners, and breeding habitats. The cooperatively breeding cichlid Neolamprologus pulcher inhabits highly distinct habitats ranging from sheer rock faces to gastropod shells, rubble and sandy bottoms with dispersed stones. Physical habitat characteristics influence predator abundance and consequently the social system and reproductive performance of this species. Under natural conditions, habitat preferences should facilitate optimization of territorial position within a colony. Question: When given the choice, does N. pulcher have a preference for environments differing in structural complexity and the presence of sand? Method: We created breeding groups consisting of a dominant pair and two subordinates. We manipulated structural complexity (low vs. high stone cover) and sandy environments (present vs. absent). We measured habitat preference using a four-factorial design with binary choice options. Predictions: We predicted that groups prefer to settle in a highly structured environment that provides many places to hide from potential predators. We further predicted a preference for a sandy bottom, especially in environments with low structural complexity, because sand allows shelters to be dug out. Results and conclusion: Neolamprologus pulcher favoured more complex over less complex habitats, independently of the presence of sand. When fish faced low structural complexity in both experimental compartments, the presence of sand became a critical factor. Choosing appropriate habitats may help to reduce predation risk.
Impact and Reach
Statistics
Additional statistics for this dataset are available via IRStats2.