COS 26-4 - Niche modeling of North American soil fungi predicts rapid transitions in boreal climates

Monday, August 15, 2022

4:15 PM – 4:30 PM EDT

Location: 514B

Presenting Author(s)

Clara Qin

PhD Candidate
University of California, Santa Cruz, United States

Slides (PDF)

Coauthor(s)

PP

Peter Pellitier

Postdoctoral Scholar
Stanford University
Stanford, CA, United States
MV

Michael Van Nuland

Postdoctoral Scholar
Stanford University, United States
KP

Kabir G. Peay

Associate Professor
Stanford University, United States
KZ

Kai Zhu

Associate Professor
University of California, Santa Cruz, California, United States

Background/Question/Methods

Soil fungi are key moderators of global biogeochemical cycles, but the large-scale responses of fungi to climate change are understudied relative to plants and animals. Furthermore, it remains unclear where fungal community turnover is driven by the magnitude of climate change as opposed to the underlying climate sensitivities of fungi. To describe the rate of change in the composition of soil fungal species pools with respect to climate change, we introduce a novel metric, the Sørensen sensitivity index SS, which is based on the overlay of individual species’ climate niches. We overcome data limitations by joining internal transcribed spacer (ITS) sequences from across two continental-scale sampling networks in North America, clustering them into operational taxonomic units (OTUs) at 97% similarity. Using a class of ecological niche models (ENMs) based on ridge-regularized logistic regression models, we estimate the climate niches of fungal OTUs and SS across a gridded climate space. We construct a permutation-based null model in order to standardize and conduct significance testing of SS. Finally, we map standardized SS onto geographic space for present-day climates and 2070 climates projected under RCP8.5, in order to identify geographic patterns in the climate sensitivity of soil fungal species pools.

Results/Conclusions

ENMs were fitted for the 8,597 most prevalent OTUs in our dataset, with a mean out-of-sample predictive accuracy of 0.76 and a mean true skill statistic of 0.45. Standardized SS displays non-uniform distributions across climate space, peaking in climates that correspond to the boreal forest biome. This rapid transition zone is projected to advance northward under climate change, bringing sweeping changes to the composition of soil fungal communities in Canada and Alaska, and signaling the potential northward migration of the boreal–temperate forest ecotone. Previous research has demonstrated that ectomycorrhizal fungi in the boreal–temperate forest ecotone are vulnerable to ongoing climate change; our findings suggest that this reflects a rapid transition zone for soil fungi in general. By contrast, soil fungal species pools in western montane ecosystems are predicted to be among the most resilient to climate change; however, even these species pools will begin to respond to climate change by 2070 under RCP8.5 projections. We propose that conceptualizing climate response in terms of niches in climate space allows for more generalizable predictions of ecological thresholds across geographic space, and connects community-level metrics such as species turnover with the autoecological intuitions of the Hutchinsonian niche.