Forecasting climate-driven extinction risk and distributional change of the Karner blue butterfly to guide managed relocation

Published: 7 July 2022| Version 1 | DOI: 10.17632/46pxcbzty8.1
Yuid Li


The Karner blue butterfly (Kbb) (Plebejus samuelis) is a federally listed endangered species whose decline is exacerbated by climate change. Evaluating how climate shapes population dynamics and the distribution of the Kbb may help develop adaptation plans. We applied mixed endo-exogenous models to examine the extinction risk and the likely distributional shift of some existent Kbb populations under several climatic change scenarios (elevated temperature and increased precipitation variance) by 2049. The predictions showed that local management of Kbb populations in Central Wisconsin and at Indiana Dunes National Park might not have been sufficient to mitigate adverse effects of climate change, and Kbbs were projected to reach higher abundance in the northern Midwest. These results suggest that some threatened populations might benefit from managed relocation and that it would be possible to reintroduce the Kbb back to Minnesota. To identify target sites for relocation, we overlapped model projections with the Minnesota soil classification map. We found a large expanse of sandy and loamy sandy soils geographically coincident with high climatic suitability for the Kbb. If those areas were planted with the host wild lupine (Lupinus perennis) and other native nectar plants, they have the potential to be developed as climate refugia of this imperiled butterfly. However, habitat management in currently occupied sites can be a more effective conservation strategy for populations with lower extinction risks. The uploaded data include 1. future climate envelops of the upper Midwest region till 2049 we extracted from the four RCPs of GFDL in terms of mean, max, and min temperature and total precipitation as well as three additional scenarios (+0.5/1%, +1.0/2%, +1.5/3%), which are the inputs of our endo-exogenous models for prediction; 2. all the pseudo-extinction risk outputs for each population and ecoregion, as well as four-way ANOVA test results of pairwise comparison between treatments (i.e., four climate scenarios, two habitat areas, four time windows, and two models); 3. all the projected maps of suitable habitats in the upper Midwest for the populations in each of the five ecoregions; 4. Some demo codes for extracting raw climate date and making predictions.


Steps to reproduce

Due to confidentiality, the original long-term population count data should be requested directly from the owners (see Table A.2 for more info). PRISM historical climate records and GFDL future climate pathways could be downloaded from their respective websites below (they are too large and cannot be compressed). After accessing these data, read them in the corresponding demo R codes. The reproductivity could be checked with the outputs we provided.