Intraspecific diversity in arctic charr, salvelinus alpinus, in iceland: II. which environmental factors influence resource polymorphism in lakes?

Background: The mechanisms causing resource polymorphism are not well understood, but likely include frequency-dependent selection. However, other selection mechanisms could also explain the development of resource polymorphism. Comparative analyses of polymorphic and monomorphic systems are uncommon, making it difficult to distinguish the effects of geography, frequency-dependent selection, niche expansion, and species interactions. Detailing ecological conditions associated with the development of resource polymorphism is necessary to discern demographic and environmental processes that may cause it. Goal: Test for environmental correlations with (a) the presence of resource polymorphism and (b) the degree of differentiation in polymorphic systems, to evaluate the hypotheses that the development of resource polymorphism results from (1) frequency dependence, (2) expansion to include a zooplanktivorous niche, or (3) lower survivorship due to predation on intermediate trait values. Trends in prey consumption, as they related to the presence of polymorphism and limnetic lake characteristics, were also analysed. Organism: Arctic charr (Salvelinus alpinus) populations from lakes sampled in 1994-2004 across Iceland. Methods: Random forest and multiple regression models assessing the presence and degree of resource polymorphism using environmental variables reflecting physical, chemical, and biological conditions. Prey consumption was fitted to the presence of polymorphism and brown trout abundance in negative binomial generalized linear models. The proportion of individuals consuming zooplankton within monomorphic versus polymorphic populations was also measured to test the idea that more widespread zooplankton consumption reflects niche expansion. Results: In Iceland, polymorphic populations tended to occur in cooler lakes with few brown trout (Salmo trutta), a trophic competitor, and in lakes where Arctic charr consumed more zooplankton. Lakes with greater limnetic habitat, fewer nutrients, and greater potential to consume zooplankton appeared to promote resource polymorphism, thereby supporting the presence of niche expansion. The overall results supported the frequency dependence hypothesis, as well as the niche expansion hypothesis in most cases. However, morphs thatdiffered in consumption of fish or chironomid pupae rather than zooplankton were also evident, indicating that mechanisms other than niche expansion may also be important. Relative resource availability and its link with the environment need to be accounted for when trying to predict the occurrence of polymorphism.