2020 ESA Annual Meeting (August 3 - 6)

COS 68 Abstract - Accounting for hostplant availability and quality to predict the distribution of the Western monarch population under climate change

Alma Carvajal Acosta, Ecology and Evolutionary Biology, University of California, Irvine, Irvine, CA, Kailen A. Mooney, Ecology and Evolutionary Biology, University of California, Ivine, Irvine, CA and Anurag A. Agrawal, Ecology and Evolutionary Biology, Cornell University, Ithaca, NY
Background/Question/Methods

Species distributions are recognized to be driven by both abiotic factors and biotic factors such as the availability of critical resources, although less is known of the importance of variation in the critical resource quality. Disentangling the relative importance of these factors – abiotic environment, presence and quality of critical resources – will be critical to predicting species response to climate change. We used SDMs to address these questions for the western populations of Monarch Butterflies, a species that obligately feeds upon plants in the genus Asclepias and for which hostplant quality varies among species in this region by an order of magnitude. We modeled the distribution of 24 Asclepias species to develop and compare three monarch distribution models with increasing levels of ecological complexity: (1) a null model using only environmental factors, (2) a model using environmental factors and Asclepias spp. distribution, (3) and a model using environmental factors and Asclepias spp. distribution weighted by hostplant quality assessed through bioassays of larval performance. Model comparison and performance was determined based on AUC and AICc values. These models were used to project and estimate changes in distribution of the western monarch.

Results/Conclusions

Our Asclepias models predicted that approximately 75% of the Asclepias spp. in this study will expand their ranges and/or shift their distribution poleward to various degrees. Performance analysis of monarch models revealed that the worst-performing model was the environment-only model (AICc=168.52, AUC=0.799). Including hostplant distribution produced the best model (AICc=0.0 and AUC=0.803) while accounting for hostplant quality did not improved model performance (AUC=0.80, AICc=123.50). The variable that contributed most to all three models was the minimum temperature of the coldest month, contributing 26-43% to the models. Host plant distribution was the second most important variable for models including Asclepias distribution and quality, contributing 22.5 and 25.5 %, respectively. Although the contemporary monarch distribution estimated by the three models were strikingly similar, range estimates differed drastically when projecting onto a future climate change scenario. All three models predicted monarch range expansions, but models including hostplants almost doubled the estimated range of the environmental-only model.Hostplant models predicted a poleward range expansion for the western monarch, whereas the environment-only model predicted no poleward expansion. These models suggest that current and future monarch range is co-limited by cold temperatures and hostplant availability, but that hostplant quality is of lesser importance.