Patchiness reduces temporal variability in aboveground biomass: Landscape-level diversity—stability in rangeland
Ecological theory predicts that diversity decreases variability in ecosystem function. We predict that functional landscape diversity—as in, a spatially-heterogeneous landscape mosaic created by patches that differ in time-since-disturbance—will decrease variability in aboveground plant biomass. We test this prediction with several years of aboveground plant biomass data from seven experimental tallgrass prairie landscapes continuously grazed with moderately-stocked cattle. Each landscape was divided into 1, 2, 3, 4, 6, or 8 patches (defined by prescribed fire units), which created spatial heterogeneity through the timing of spatially-discrete fires and the resulting mosaic of plant succession stages following variable time-since-disturbance. Using a random-effects regression variance partitioning method, we determined the amount of variation attributable to spatial and temporal components for each pasture. We also tested for the portfolio effect and a relationship between temporal variability and synchrony in aboveground plant biomass as mechanisms behind the diversity-stability hypothesis.
We show that increased functional landscape diversity driven by spatially-patchy fire reduces temporal variability in aboveground biomass. We found a negative relationship between functional landscape diversity and temporal variability. The pattern is associated with both statistical evidence for the portfolio effect and a positive relationship between temporal variability and synchrony in aboveground biomass across plant functional groups. As disturbance from fire and grazing interact to create a shifting mosaic of spatially-heterogeneous patches within a landscape—an ecological disturbance known as pyric-herbivory—temporal variability in aboveground biomass can be dampened. However, although the relationship between temporal variability and spatial heterogeneity is linear, the relationship does not proceed ordinally with number of patches within a landscape—polynomial fits against number of patches suggests temporal variability and spatial heterogeneity are lowest and greatest, respectively, around 4-6 patches per landscape and diminish thereafter. These results suggest that managing rangeland with spatially-heterogeneous disturbance regimes maximizes the full portfolio of functions provided by rangeland vegetation, including wildlife habitat, fuel, and forage. A shifting mosaic of spatially-heterogeneous patches—i.e., patches with contrasting vegetation structure within a landscape that shift through space and time—has potential to buffer ecosystem function against global environmental change.