Modeled Seed Accumulation Patterns Explain Spatial Heterogeneity of Shrub Recruitment Within the Taiga-Tundra Ecotone

Abstract

Arctic shrub productivity trends display variability at multiple spatial scales. Fine-scale studies have generally observed the greatest shrub expansion in landscape positions that accumulate water and nutrients. While considerable work has focused on the mediating effect of these resources on growth responses to warming, less is known about the mechanisms constraining recruitment-driven expansion. Given the low seed viability of many Arctic shrubs, spatial patterns of seed dispersal may play an important role in constraining fine-scale variability of shrub recruitment. This variability may also be driven by ground cover suitability, though these relationships are understudied in undisturbed sites. Here, we developed models representing seed accumulation mechanisms around Alnus alnobetula (green alder) patches within the taiga-tundra ecotone of the Northwest Territories and compared these with observations of seed and seedling density. We also investigated relationships between seedling abundance, topographic position, and ground cover. Observed patterns of recruitment were complex, with preferential expansion occurring beneath alder patches only on the steepest slopes. Seed accumulation models representing overland flow, wind, and source distance were important predictors of seedling recruitment. This provides indirect evidence of localized seed limitation around patches, suggesting future recruitment may not respond as expected to changing environmental conditions. Sphagnum cover also predicted recruitment, indicating the importance of seedbed conditions for establishment. We propose that developing models of shrub expansion that include both dispersal and environmental constraints may increase our ability to predict patterns and rates of expansion. Such predictions are necessary to understand future biosphere-atmosphere interactions in a rapidly changing Arctic.