Variation of soil respiration at three spatial scales: Components within measurements, intra-site variation and patterns on the landscape

Soil respiration is an important component of terrestrial carbon cycling and can be influenced by many factors that vary spatially. This research aims to determine the extent and causes of spatial variation of soil respiration, and to quantify the importance of scale on measuring and modeling soil respiration within and among common forests of Northern Wisconsin. The potential sources of variation were examined at three scales: [1] variation among the litter, root, and bulk soil respiration components within individual 0.1 m measurement collars. [2] variation between individual soil respiration measurements within a site (<1 m to 10 m), and [3] variation on the landscape caused by topographic influence (100 m to 1000 m). Soil respiration was measured over a two-year period at 12 plots that included four forest types. Root exclusion collars were installed at a subset of the sites, and periodic removal of the litter layer allowed litter and bulk soil contributions to be estimated by subtraction. Soil respiration was also measured at fixed locations in six northern hardwood sites and two aspen sites to examine the stability of variation between individual measurements. These study sites were added to an existing data set where soil respiration was measured in a random, rotating, systematic clustering which allowed the examination of spatial variability from scales of <1 m to 100+ m. The combined data set for this area was also used to examine the influence of topography on soil respiration at scales of over 1000 m by using a temperature and moisture driven soil respiration model and a 4 km(2) digital elevation model (DEM) to model soil moisture. Results indicate that, although variation of soil respiration and soil moisture is greatest at scales of 100 m or more, variation from locations I m or less can be large (standard deviation during summer period of 1.58 and 1.28 mu mol CO(2) m(-2) s(-1), respectively). At the smallest of scales, the individual contributions of the bulk soil, the roots, and the litter mat changed greatly throughout the season and between forest types, although the data were highly variable within any given site. For scales of 1-10 m, variation between individual measurements could be explained by positive relationships between forest floor mass, root mass, carbon and nitrogen pools, or root nitrogen concentration. Lastly, topography strongly influenced soil moisture and soil properties, and created spatial patterns of soil respiration which changed greatly during a drought event. Integrating soil fluxes over a 4 km2 region using an elevation dependent soil respiration model resulted in a drought induced reduction of peak summer flux rates by 37.5%, versus a 31.3% when only plot level data was used. The trends at these important scales may help explain some inter-annual and spatial variability of the net ecosystem exchange of carbon. (C) 2009 Elsevier Ltd. All rights reserved.