COS 235-1 - Predicting berry plant habitat under climate change in Bristol Bay, AK

Abstract: Climate change is altering the suitable habitat distributions of many species in high latitudes. Fleshy fruit-producing plants (hereafter “berry plants”), important in arctic food webs and as subsistence resources for human communities, may be impacted, but their response to a warming climate at a landscape scale has not yet been examined. Here, we aimed to (Q1) predict current berry plant habitat distribution; (Q2) predict how berry plant habitat distribution might change under future climate conditions; and (Q3) identify dominant environmental drivers of berry plant habitat.

The study area consisted of the Bristol Bay and Togiak Natural Resource Conservation Service (NRCS) survey areas in southwestern Alaska. We built species distribution models (SDMs) using the Random Forests (RF) algorithm and presence/absence records collected by the NRCS to identify characteristics and predict the spatial distribution of habitats suitable for five berry plant species: Vaccinium uliginosum L., Empetrum nigrum L., Rubus chamaemorus L., Vaccinium vitis-idaea L., and Viburnum edule (Michx.) Raf.. Then, we used future climate projections (2081-2100; representative concentration pathways (RCPs) 4.5, 6.0, & 8.5) to predict shifts in species’ suitable habitat distributions based on future climate conditions.

The predicted amount and spatial patterns of suitable habitat for the current time period were variable among species, consistent with species’ diverse life history attributes and habitat preferences (Q1). Future climate models predicted both positive and negative changes to suitable habitat probability for all species across the landscape; future binary classification maps predicted net declines in suitable habitat area for all species and climate scenarios tested (Q2). Models identified elevation, soil characteristics, and January and July temperatures as important drivers of suitable habitat distributions (Q3).

Our predictions contribute to further understanding the response of important berry plant species to climate change at a landscape scale. Shifting and retracting distributions may alter where communities have access to harvesting areas, suggesting that access to these resources may become restricted in the future. Our maps may help to inform climate adaptation planning as communities anticipate shifting access to harvesting locations.