| In an Australian semi-arid grassland, recurring droughts repeatedly reset the abiotic and biotic components of the ecosystem. This temporal variability, coupled with spatial heterogeneity, allows rainfall, geology, and grazing to interact in a complex web of non-linear, indirect and feedback effects, which control primary production and regulate nitrogen cycling. I investigated the feedback effects of grazing on nitrogen cycling and primary production by adding nutrients and excluding grazers in a series of replicated factorial experiments.; Grazing by kangaroos, wombats and rabbits was significant, reducing above ground biomass (AGB) by approximately 50–80%. Without grazing, nitrogen additions increased the AGB and the nitrogen concentration of standing vegetation and litter. Grazers preferred nitrogen-amended plots, and the standing biomass the grazed plots was similar, regardless of nutrient amendments. Grazers reduced the mass of litter by 17–92%, and increased the nitrogen concentration of the litter 20–30%, diverting nitrogen from litter into more rapidly-mineralized urine and fecal pellets, reducing immobilization of soil nitrogen in decomposing litter, which probably accelerated nitrogen cycling. Grazing increased the plant uptake of nitrogen by stimulating regrowth and increased the dominance of unpalatable, mostly introduced species, which produced litter higher in nitrogen, and probably accelerated litter decomposition and nitrogen cycling. The usual positive correlation between palatability and decomposability of vegetation was reversed because many of the unpalatable species were defended by secondary compounds (alkaloids), which have not been reported to slow decomposition.; Drought-pulses and spatial heterogeneity organize the trophic dynamics and nitrogen biogeochemisty in this semi-arid ecosystem. The biota modify or redirect the effects of the abiotic forces and create or exploit temporarily or locally concentrated resources. I suggest that management efforts be similarly opportunistic: targeted in location, flexible and focused in timing, and prepared to exploit the drought-pulse as a subsidy. Such “opportunistic management” might assist in the development of economic and social subsystems better matched to the spatial patterns and temporal dynamics of their larger ecological support systems. |
