Vegetation change can affect the magnitude and direction of global climate change via its effect on carbon cycling among plants, the soil and the atmosphere. The invasion of non-native plants is a major cause of land cover change, of biodiversity loss, and of other changes in ecosystem structure and function. In California, annual grasses from Mediterranean Europe have nearly displaced native perennial grasses across the coastal hillsides and terraces of the state. Our study examines the impact of this invasion on carbon cycling and storage at two sites in northern coastal California. The results suggest that annual grass invasion has caused an average drop in soil carbon storage of 40 Mg/ha in the top half meter of soil, although additional mechanisms may also contribute to soil carbon losses. We attribute the reduction in soil carbon storage to low rates of net primary production in non-native annuals relative to perennial grasses, a shift in rooting depth and water use to primarily shallow sources, and soil respiratory losses in non-native grass soils that exceed production rates. These results indicate that even seemingly subtle land cover changes can significantly impact ecosystem functions in general, and carbon storage in particular.

We sampled isolated trees and tree clusters from a blue oak, Quercus douglasii, savanna to determine the spatial heterogeneity of fine root biomass and soil carbon across the landscape as a function of tree size and configuration. We aimed to understand how fine root structure enables sustained ecosystem metabolism through a summer of limited moisture and high heat and facilitates resource acquisition during the short period of high resource supply. An additional goal was to provide a basis for upscaling root biomass and soil carbon to the landscape scale. We sampled trees of different size and tree clusters via a stratified sampling scheme that accounted for spatial heterogeneity in root biomass and soil carbon with lateral distance from the tree bole, or cluster centre, and soil depth. We upscaled these estimates using site-specific information from a lidar survey. We found that fine roots and soil carbon are spatially heterogeneous in their landscape distribution and greatly increase with tree size. We also found that Q.douglasii possesses a dimorphic fine root architecture, uniquely suited to the region's climatic constraints and exhibits morphological plasticity among trees of different size and physical setting.