Influences of land-use change and edges on soil respiration in a managed forest landscape, WI, USA

A critical issue in landscape ecology is to understand the effects of disturbances on landscape processes. While there is increasing interest in examining the ecological consequences associated with changes in landscape structure and composition, little is known about how soil respiration across landscapes might be affected within this context. We examined field-season soil respiration rates (SRR) and soil carbon fluxes in a managed landscape in northern Wisconsin related to land-use change, area-of-edge influence (between non-forested bare ground (NFBG) and other forest cover types), and spatial variation. We estimated SRR for all six cover types classified from Landsat images of 1972 and 2001 across the entire landscape using published SRR-T-s (soil temperature at 5-cm depth) models developed based on data collected in the study area. Between the 2 years, the landscape-level mean SRR increased by 0.5% (mainly as a result of a 33% increase in the proportion of young forests proportion), varying from -2.6 to 4% among 16 sub-landscapes, suggesting spatial variation in land-cover composition can affect landscape-level SRR significantly. If the influence of edges was considered, landscape mean SRR further increased by 0.03% in 1972 and 0.16% in 2001 as a result of an increase in NFBG from 3.1% in 1972 to 9.2% in 2001. Both land-use change and edges influence landscape-level SRR, but at different magnitudes. The former caused +/- 4% variation in landscape-level SRR while the latter, associated with increase in NFBG, introduced an additional 0.01-0.26% increase in landscape-level SRR. Smaller than expected edge influences on landscape-level SRR were partly due to the offset of T-s changes in the forested and NFBG sides of the edges. We present an empirical equation to predict edge influences on net changes in landscape-level SRR using net changes in two representative, class-level (NFBG) indices that are easily calculated from an existing spatial pattern analysis program. (c) 2005 Elsevier B.V. All rights reserved.