This doctoral dissertation research project will analyze the consequences of woody plant encroachment on ecosystem carbon stocks and albedo changes at regional scale in the Western Pampas grasslands, Argentina. Land cover change dynamics alter ecosystem and climate functioning primarily through changes in matter cycles and energy fluxes. Woody plant encroachment in semiarid rangelands is a global process of land cover change driven primarily by changes in land use practices. The widespread increase in woody plant cover alters ecosystem distribution of carbon and other nutrients and influences the radiative balance due to the different reflective properties of grasses, soil and woody plants. Despite the fact that rangelands comprise almost 45% of the global land surface, our understanding of the process of encroachment is limited, in most cases, to fine resolution studies of the northern hemisphere. In this study a space-for-time substitution approach will be used to analyze changes in C stocks in the woody plant, grass, litter and soil organic carbon pools and changes in albedo along an encroachment gradient. Field measurements along this gradient will be scaled up with the use of sub-pixel woody plant cover estimates derived from moderate resolution satellite imaginary. To understand the temporal component of the process of encroachment, woody plant cover change dynamics will be analyzed between 1880, 1960 and 2009. Historical land surveys (1880) will be digitized and sampling and statistical procedures will be used to develop estimates of woody plant cover. Aerial photographs (1960) and high resolution panchromatic satellite images (2009) will be processed using spatial wavelets and texture analysis, and moderate resolution images (2010) using multi-date spectral mixture models, in order to assess the changes in woody plant cover over time.

This project will be among the first to investigate spatio-temporal dynamics of woody plant encroachment and its consequences on C stock and albedo at a regional scale in the southern hemisphere. Results will be critical for more complete accounting of C and the radiative forcing associated with woody encroachment in semiarid rangelands across the globe. As a Doctoral Dissertation Research Improvement award, this award also will provide support to enable a promising student to establish a strong independent research career.

Project Report

Forest and savanna ecosystems provide multiple benefits to society in the form of food, fiber, soil protection, and climate control just to mention a few examples. Therefore, it is extremely important to understand woody plant cover dynamics to better manage and conserve the ecosystems and the benefits they provide. In this project we studied long term woody plant cover dynamics in semiarid savannas of central Argentina. This area is an active agriculture frontier where both gains and losses in woody plant cover have occurred in the last 120 years. We used historical maps from the 1880s to reconstruct vegetation cover in the area before the massive influx of European settlers. We then assessed changes in woody plant cover by comparison with current land cover maps. We found that approximately 14% of the study area had experienced increases in woody plant cover in the last 120 years, while 6% had been converted from savannas to cropland and 8% to pastures. Describing changes in vegetation cover was interesting, but in order to understand the processes driving the change we decided to evaluate the effect of different factors. We found that increases in woody cover had occurred in areas closer to woodlands present in the 1880s, at higher elevation, steeper slope, with poorer soils and closer to Native American trails. This last finding was surprising and could indicate the presence of legacy effects. The main woody plant in this region (Prosopis caldenia) is dispersed by cattle, and it seems like Native Americans movement of cattle from the east to the west of the region, favored the expansion of woody plants. Conversion from savannas to cropland occurred primarily in areas with higher annual rainfall, richer soils and closer to cities. The opposite was found for conversion to pastures, which were located in areas with less rainfall and poorer soils. Land cover changes affect the composition, structure and function of ecosystems. In order to understand how these changes had affected carbon storage in the region, we designed and implemented an extensive field campaign to measure carbon stocks in soils, woody plants, and grasses. We found that on average closed woodlands contain almost 4 times more C than grasslands (16.4 kgCm-2 vs 4.5 kgCm-2). Given that measuring carbon stocks is time consuming and expensive, we developed simple equations to predict ecosystem carbon stocks from easily measured variables (woody plant cover and soil texture). This equations could be used by ranchers or government officials to obtain fast and reliable estimates of ecosystem carbon stocks. We finally used the results from the historical land cover change analysis and the extensive fieldwork to estimate the net regional change in C stocks for the last 120 years. The increase in woody plant cover increased the storage of carbon in the Caldenal savannas by ~96 TgC. However, the conversion to cropland and pasture generated a loss of ~167 TgC. This means that the region lost ~10% of the carbon the ecosystem had in the 1880s. Moreover, these losses could be even larger in the future if a new phase of agriculture expansion occurs in this energy crops, as some analysts predict will happen. Results from this study are important for the scientific community because they provide accurate and detailed measurements of land cover change and carbon stocks which are greatly needed to improve global models of land cover change. Moreover, the information is useful for local governments to implement more informed land use policies. For example, a market of payments for ecosystems could be more easily developed now that we have better estimates of carbon stocks in a range of land cover classes.

National Science Foundation (NSF)
Division of Behavioral and Cognitive Sciences (BCS)
Standard Grant (Standard)
Application #
Program Officer
Thomas J. Baerwald
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Duke University
United States
Zip Code