Spatial heterogeneity of tundra vegetation response to recent temperature changes

The spatial heterogeneity of recent decadal dynamics in vegetation greenness and biomass in response to changes in summer warmth index (SWI) was investigated along spatial gradients on the Arctic Slope of Alaska. Image spatial analysis was used to examine the spatial pattern of greenness dynamics from 1991 to 2000 as indicated by variations of the maximum normalized difference vegetation index (Peak NDVI) and time-integrated NDVI (TI-NDVI) along latitudinal gradients. Spatial gradients for both the means and temporal variances of the NDVI indices for 0.1 degrees latitude intervals crossing three bioclimate subzones were analyzed along two north-south Arctic transects. NDVI indices were generally highly variable over the decade, with great heterogeneity across the transects. The greatest variance in TI-NDVI was found in low shrub vegetation to the south (68.7-68.8 degrees N) and corresponded to high fractional cover of shrub tundra and moist acidic tundra (MAT), while the greatest variance in Peak-NDVI predominately occurred in areas dominated by wet tundra (WT) and moist nonacidic tundra (MNT). Relatively high NDVI temporal variances were also related to specific transitional areas between dominant vegetation types. The regional temporal variances of NDVI from 1991 to 2000 were largely driven by meso-scale climate dynamics. The spatial heterogeneity of the NDVI variance was mostly explained by the fractional land cover composition, different responses of each vegetation type to climate change, and patterned ground features. Aboveground plant biomass exhibited similar spatial heterogeneity as TI-NDVI; however, spatial patterns are slightly different from NDVI because of their nonlinear relationship.