MODELING SPATIAL AND TEMPORAL PATTERNS OF SOIL ORGANIC CARBON IN TWO MONTANE LANDSCAPES: THE NORTHERN HARDWOODS, VERMONT AND THE TABONUCO FOREST, PUERTO RICO

Kristofer Dee Johnson, "Modeling spatial and temporal patterns of soil organic carbon in two montane landscapes: The northern hardwoods, Vermont and the tabonuco forest, Puerto Rico" (January 1, 2008). Dissertations available from ProQuest. Paper AAI3328590.

Abstract: 
Forest soils contribute to a significant portion of the world’s carbon flux due to both natural and anthropogenic changes. In terms of human management of carbon pools, forest soil organic matter (SOM) is important because it potentially stores carbon more permanently than living vegetation. Yet, this potential is poorly understood or managed for because of the difficulty in measuring changes in SOM pools over time and space. Modeling combined with intensive field sampling can help overcome these limitations because it extracts from empirically observed relationships to account for the components of SOM formation (topography, time, parent material, organisms and climate[fns2]). This study utilizes intensive field data, statistical models and process-based ecosystem models to investigate the spatial distribution and dynamics of soil organic carbon dynamics in two contrasting ecosystems – the northern hardwood forest in the Green Mountains, VT and the tabonuco forest in the Luquillo Experimental Forest, PR. In both forests landscape position emerged as the dominate factor in explaining SOM distribution. In Vermont, additional variation was explained by aspect and slope and in Puerto Rico additional variation was explained by landscape factors interrelated to soil drainage. Process-based modeling proved to be a useful management and experimental tool in cases were empirical approaches were impractical for both forests. In Vermont, three ecosystem models demonstrated a substantial reduction of soil organic carbon and harvestable biomass due to the removal of woody carbon by logging after 240 years of rotations. In Puerto Rico, the Century model showed that changes in litter quality and quantity were not likely responsible in explaining landscape level SOM differences. Overall, well drained soils located in colder climates stored the highest SOM whereas poorly drained and highly disturbed soils in steep humid climates stored the lowest SOM. This research demonstrates that although SOM amounts are highly variable over many spatial and temporal scales, intuitive relationships are borne out with modeling tools and by careful investigation of the five soil forming factors. Results also raise questions about how these ecosystems and their SOM pools may change in response to changing climate conditions of the future.