COS 9-10 - Incorporating biotic interactions improves the prediction of mountain ecosystem species abundance and distribution

Monday, August 7, 2017: 4:40 PM
D131, Oregon Convention Center
Joshua S. Lynn1,2, Stephanie N. Kivlin1,2, Melanie R. Kazenel1,2 and Jennifer A. Rudgers1,2, (1)Rocky Mountain Biological Laboratory, Crested Butte, CO, (2)Department of Biology, University of New Mexico, Albuquerque, NM
Background/Question/Methods Understanding the relative importance of biotic and abiotic controls of the distribution and abundance of species can improve forecasts on the ecological consequences of climate change. Here, we constructed statistical ecological niche models that incorporate a large suite of biotic interactions as well as traditional abiotic environmental data to ask: What is the relative importance of biotic interactions versus the abiotic environment in predicting the local abundance and distribution of a species? We assessed the presence/absence and abundance of three focal alpine restricted plant species (Poa alpina, Festuca brachyphylla, and Elymus scribneri) over 6 mountains and 67 sites in the Upper Gunnison Basin, Colorado, USA. Variables relating to geography (e.g. slope), the abiotic environment (e.g. soil depth, plant available nutrients), and biotic interactions (e.g. herbivory, mammal disturbance) were assessed as predictors of focal species abundance at each site. All predictor variables were scaled to a mean of zero and standard deviation of one, allowing for the comparison between predictors as effect sizes, and then assessed for collinearity issues (using VIF analyses) prior to analysis of abundance distributions. We then used these variables to predict focal species presence/absence and abundance using maximum likelihood and AICc model selection statistical methods.

Results/Conclusions The best models of all three species’ elevational abundance distributions included biotic interactions. Poa alpina abundance decreased severely with high levels of herbivory, which had the greatest effect on P. alpina abundance. The next three predictors of P. alpina abundance with the greatest effect size were conspecific vegetation cover, bacterial/fungal damage, and fossorial mammal disturbance, which were all negatively related to abundance. Abundance of F. brachyphylla showed the strongest declines in abundance with herbivory and fossorial mammal disturbance, and presented increases in abundance with soil depth. High rates of fossorial mammal disturbance and soil potassium predicted declines in the abundance of E. scribneri. In all cases, a combination of abiotic and biotic factors were better models for predicting abundance, suggesting that models lacking information on biotic interactions could fail. Furthermore, the species-specific results presented here suggest that climate change may have non-symmetric effects on regionally coexisting species, such that a failure to account for the effects of non-climatic niche dimensions in the movement of a species’ range will be inaccurate. Experiments aimed at understanding how biotic interactions may change with climate are required to establish causality behind model findings.