“Predicting habitat choice after rapid environmental change”
Philip H. Crowley, Pete C. Trimmer, Orr Spiegel, Sean M. Ehlman, William S. Cuello, and Andrew Sih (May 2019)
Choosing where to settle is a critical decision various animals have to take, whether it is a bird choosing its nest site, or a fish larvae settling on a home coral. While we know that environmental changes are impacting many habitats and can have major fitness implications, it is harder to predict how these changes in the availability and quality of natural habitats will affect the settlement decisions of different species. A small number of species are actually thriving following environmental changes (e.g. pigeons and other urban exploiters), while many others are making poor habitat choices and consequently decline. We have developed an analytical model to predict: 1) which kinds of environmental change have large, negative effects on fitness (e.g., which is worse: lower average quality or frequency of finding patches?); 2) how species’ evolutionary histories affect their susceptibility to environmental change (are those that are used to rare habitats beforehand less affected?); and 3) how much lost fitness can be recovered via re-adjustment after environmental change (and which scenarios are non-fixable?). In our model, animals search for habitable patches in an otherwise inhospitable matrix. They are assumed to settle when they find a patch which exceeds their threshold of necessary quality. We consider decisions and fitness before environmental change, immediately following change (i.e., if the animals continue to use their existing decision thresholds), and after optimal re-adjustment (e.g., via learning or evolution). We find that decreases in survival during searching (per time step), and declines in habitat quality or availability, generally have stronger negative effects than reduced season duration. Animals that are adapted to good conditions remain choosy after conditions decline and thus suffer more from environmental change than those adapted to poor conditions beforehand. Re-adjustment can recover much of the lost fitness in some situations, such as a reduction in average habitat quality, but re-adjustment recovers much less of the lost fitness when environmental change has reduced habitat availability, or has increased death-rates during the search process. Taken together, these findings increase our ability to predict which species will be more vulnerable to environmental changes and to better prioritize conservation efforts.
Decisions made while searching for settlement sites (e.g., nesting, oviposition) often have major fitness implications. Despite numerous case studies, we lack theory to explain why some species are thriving while others are making poor habitat choices after environmental change. We develop a model to predict: 1) which kinds of environmental change have larger, negative effects on fitness; 2) how evolutionary history affects susceptibility to environmental change; and 3) how much lost fitness can be recovered via re-adjustment after environmental change. We model the common scenario where animals search an otherwise inhospitable matrix, encountering habitats of varying quality, and settling when finding a habitat better than a threshold quality level. We consider decisions and fitness before environmental change, immediately following change (assuming that animals continue to use their previously adaptive decision rules), and after optimal re-adjustment (e.g., via learning or evolution). We find that decreases in survival per time step searching, and declines in habitat quality or availability, generally have stronger negative effects than reduced season duration. Animals that were adapted to good conditions remained choosy after conditions declined and thus suffered more from environmental change than those adapted to poor conditions. Re-adjustment recovered much of the fitness lost through a reduction in average habitat quality, but recovered much less following reductions in habitat availability or survival while searching. Our model offers novel predictions for empiricists to test, as well as suggestions for prioritizing alternative mitigation steps.