Canopy scale measurements of CO2 and water vapor exchange along a precipitation gradient in southern Africa


Todd M. Scanlon, e-mail:


Short-term measurements of carbon dioxide, water, and energy fluxes were collected at four locations along a mean annual precipitation gradient in southern Africa during the wet (growing) season with the purpose of determining how the observed vegetation–atmosphere exchange properties are functionally related to the long-term climatic conditions. This research was conducted along the Kalahari Transect (KT), one in the global set of International Geosphere-Biosphere Program transects, which covers a north–south aridity gradient, all on a homogenous sand formation. Eddy covariance instruments were deployed on a permanent tower in Mongu, Zambia (879 mm of rainfall per year), as well as on a portable tower in Maun (460 mm yr−1), Okwa River Crossing (407 mm yr−1), and Tshane (365 mm yr−1), Botswana for several days at each site. The relationships between CO2 flux, Fc, and photosynthetically active radiation were described well by a hyperbolic fit to the data at all locations except for Mongu, the wettest site. Here, there appeared to be an air temperature effect on Fc. While daytime values of Fc routinely approached or exceeded −20 μmol m−2 s−1 at Mongu, the magnitude of Fc remained less than −10 μmol m−2 s−1 when the air temperature was above 27°C. Canopy resistances to water vapor transfer, rc, displayed an overall decline from the wetter sites to the more arid sites, but the differences in rc could be almost exclusively accounted for by the decrease in leaf area index (LAI) from north to south along the KT. Ecosystem water use efficiency (WUE), defined as the ratio of net carbon flux to evapotranspiration, showed a general decrease with increasing vapor pressure deficit, D, for all of the sites. The magnitudes of WUE at a given D, however, were dissimilar for the individual sites and were found to be stratified according to the position of the sites along the long-term aridity gradient. For example, Mongu, which has the wettest climate, has a much lower WUE for like levels of D than Tshane, which historically has the most arid climate. Given the similar inferred stomatal resistances between the sites, the disparate carbon uptake behavior for the grass vs. woody vegetation is the likely cause for the observed differences in WUE along the aridity gradient. The short-term flux measurements provide a framework for evaluating the vegetation's functional adaptation to the long-term climate and provide information that may be useful for predicting the dynamic response of the vegetation to future climate change.