Monday, August 3, 2009 - 2:50 PM

COS 12-5: Ecological consequences of niche evolution in willows (Salix) in southeastern Minnesota

J.A. Savage, University of Minnesota and J.M. Cavender-Bares, University of Minnesota.

Background/Question/Methods Thirteen willow (Salix) species occur in southeastern Minnesota and often co-occur within the same wetlands. This high local diversity is challenging to explain since closely related species are often functionally similar, and density-dependent interactions such as competition and susceptibility to pests and pathogens should limit their co-occurrence. However, if willow species are partitioning resources, or if willow communities are phylogenetically structured so that closely related species rarely co-occur, the impact of these density-dependent processes would be reduced.  To investigate the role of niche partitioning in maintaining local willow diversity, we examined the relationship between species physiology and their distributions across a water availability gradient. We randomly-established 50 plots in three preserves in southeastern Minnesota. In each plot, we measured species abundances and total basal area and measured environmental conditions monthly for two years. In both the field and a greenhouse common garden, we measured functional traits with known associations to drought and waterlogging tolerance. We investigated species functional traits and phylogenetic similarity in relation to their ecological distributions. In particular, we examined patterns of species co-occurrence and niche overlap in relation to their functional and phylogenetic diversity.
Results/Conclusions

We found that willow species demonstrate non-random patterns of co-occurrence in southeastern Minnesota and some species pairs infrequently co-occur in close proximity (based on plot C-scores). Although there is evidence that species segregate across a water availability gradient, there is more variation in species niche overlap (Pianka’s index) than expected by chance. This high variation is indicative of a guild structure where species cluster into groups in different habitats. Additionally, we found a strong correlation between species wood density, stomatal pore index (SPI), seedling growth rate, seed viability, and their distributions across a water availability gradient. These correlations demonstrate that there is a physiological basis for species niche separation, and provide evidence for a trade-off between waterlogging tolerance and growth. Overall, local willow species break down into a set of guilds that exhibit niche partitioning across a water availability gradient. The high level of co-occurrence among species in the same functional and phylogenetic guilds indicates that local willow diversity cannot be exclusively explained by niche partitioning, and that other mechanisms are likely to be important in structuring these plant communities.