Abstract:
The hump-shaped relationship that predicts highest species richness (species/unit area) at intermediate levels of productivity was examined for woody plants across the Pacific and Inland Northwest of the U.S. Many studies have examined this relationship at regional scales, but commonly use species range maps and surrogate measures of productivity (e.g. evapotranspiration.) In this study, productivity during the growing season was modeled over the entire study area at a monthly time step with a physiologically-based process model using climate data extrapolated for terrain effects. The productivity information was validated in Oregon by converting it to site index and comparing it with field data; reasonable agreement was found. The productivity-species richness relationship was then examined with two datasets of tree species richness, one of shrub species richness and one dataset of woody plant species richness in the states of Oregon,
Washington, Idaho, and Montana. The relationship between predicted productivity
and species richness was analyzed at multiple grain sizes by spatial aggregation at 1, 100, and 1000 km². After taking into account the effect of area, improvements in
the hump-shaped relationship emerged at progressively larger spatial units of
aggregation for all datasets. In every case, a hump-backed curve was found with different degrees of fit. One tree species dataset was examined in detail in the states of Oregon and Washington at the scale of 100 km² in a comparison among productivity and surrogates measures (i.e. potential and actual evapotranspiration). Modeled productivity best accounted for patterns in tree species richness and the statistical model was validated with independent field data and performed well. In conclusion, where climate data are available to model productivity spatially, the common surrogates of productivity such as potential or actual evapotranspiration are better avoided as productivity substitutes. Scaling up to coarser scales proved valuable in removing error of unknown plot locations and producing a more complete species tally with which to compare productivity.
