Primary Production and Abiotic Controls in Forests, Grasslands, and Desert Ecosystems in the United States

This paper, a synthesis based on data generated by the International Biological Program, deals with the relationships among biotic factors at the ecosystem level. Emphasis is placed on aboveground net primary production (ANPP), a major component of energy that drives ecosystem processes, and on potential evapotranspiration (PET), the abiotic variable most often used to explain variation in ANPP. The question addressed is: can ANPP be related to combinations of biotic and abiotic factors such that the relationships are independent of ecosystem type, whether it be forest, grassland, or desert? ANPP as a function of peak foliar standing crop (FSC) was best explained by models which showed a reduction in ANPP/FSC as FSC increased. Thus, deserts had a higher ANPP per unit of FSC than did other systems. As expected, photosynthetic efficiency (PE) was highest for forests, °100 times greater than for deserts. However, when PE was evaluated per unit of foliage, the differences in PE of ecosystems were much less. In fact, a hot—desert site had the highest PE/FSC. In terms of a theoretical maximum, the PE of forests was only 6—25% of the maximum value. Systems with nearly steady—state aboveground standing crop (ASC) showed an exponential decrease with decreased water availability (potential evapotranspiration minus precipitation). For these same systems, the ratio of ANPP to ASC increased with decreased water availability, suggesting that water—stressed systems need more energy from ANPP to drive internal processes. A model predicting ANPP of desert—shortgrass steppes was structured in terms of FSC, water availability, and temperature. The predictive power was found to be very high, and the model was successfully validated in two of three cases with an independent data set. A model predicting ANPP of forests was structured in terms of FSC, radiation, ASC, and temperature. The deviation of the observed ANPP relative to that calculated was 17%. Deviations from predicted values were highest for deciduous stands with high ANPP and low FSC. Most relationships exhibited good correlations between ANPP and the various independent variables including both biotic, abiotic, and combinations of the two. However, in many instances the data tended to be grouped by ecosystem type, suggesting that variation in ANPP can be reduced if ecosystem type is an added independent variable. It was surprising to find that, with the limits of our data, differences in ANPP at the ecosystem level are not glaring, especially considering that soil factors were not included in our analyses. When considering the broad range of genotypes in each ecosystem, and the much broader genotypic range representing all ecosystems, the control that native ecosystems have over abiotic factors in producing ANPP is evident but not large.