Vegetation Responses to Climate Change: Do Biotic Interactions Matter in Predicting Species Distributions in Northern Alpine Tundra?

A rapid shift in Arctic and alpine plant community distribution, composition and structure is likely as species respond to climate change. In northern alpine tundra ecosystems, like those found in the southwest Yukon, Canada, these movements may have significant impacts on biodiversity, habitat quality and structure as suitable areas diminish in size at higher elevations. We used bioclimatic envelope models to generate current day vegetation-climate relationships. We further developed these bioclimate envelopes in a spatially explicit spatio-temporal landscape model that could account for competitive interactions and processes.

We used Nonparametric Multiplicative Regression Modeling to determine the relationship between climate variables and the distribution of metrics for alpine tundra vegetation (cover and height) in Kluane, Yukon. The climate model was produced using the North American Regional Reanalysis downscaled from 32 km grids to 30-100 m to spatially match the ecosystem models. For precipitation, we relied on a linear model of orographic precipitation modified to track air mass and dynamically calculate nucleation and fallout timescales. Our temperature downscale calculated free-air and inversion lapse rates to adjust mid-tropospheric temperature to ground-level. Other variables in our model included solar radiation, snow-free days and topographic measures.

We integrated multiple vegetation species/functional group layers, competition and transition rules into the spatial-temporal model to project regional distributions of vegetation cover and height, with and without interspecific interactions, under various climate chance scenarios.

We demonstrate that changes in vegetation cover and height at our site are driven by differing climate forces and that the nonlinear relationship between the two metrics can become increasingly uncoupled over projected time under climate change. We show that by refining bioclimate envelope models to include both vegetation cover and height along with dynamic interactions within a spatio-temporal approach, we may be better able to predict, at the regional level, changes to vegetation composition and structure under various climate change scenarios.