RESUMEN
In the semi-arid grasslands of the southwest United States, annual precipitation is divided between warm-season (July-September) convective precipitation and cool-season (December-March) frontal storms. While evidence suggests shifts in precipitation seasonal distribution, there is a poor understanding of the ecosystem carbon flux responses to cool-season precipitation and the potential legacy effects on subsequent warm-season carbon fluxes. Results from a two-year experiment with three cool-season precipitation treatments (dry, received 5th percentile cool-season total precipitation; normal, 50th; wet, 95th) and constant warm-season precipitation illustrate the direct and legacy effects on carbon fluxes, but in opposing ways. In wet cool-season plots, gross primary productivity (GPP) and ecosystem respiration (ER) were 103% and 127% higher than in normal cool-season plots. In dry cool-season plots, GPP and ER were 47% and 85% lower compared to normal cool-season plots. Unexpectedly, we found a positive legacy effect of the dry cool-season treatment on warm-season carbon flux, resulting in a significant increase in both GPP and ER in the subsequent warm season, compared to normal cool-season plots. Our results reveal positive legacy effects of cool-season drought on warm-season carbon fluxes and highlight the importance of the relatively under-studied cool-growing season and its direct/indirect impact on the ecosystem carbon budget.
RESUMEN
Shrub encroachment is a common ecological state transition in global drylands and has myriad adverse effects on grasslands and the services they provide. This physiognomic shift is often ascribed to changes in climate (e.g. precipitation) and disturbance regimes (e.g. grazing and fire), but this remains debated. Aeolian processes are known to impact resource distribution in drylands, but their potential role in grassland-to-shrubland state changes has received little attention. We quantified the effects of 'sandblasting' (abrasive damage by wind-blown soil) on the ecophysiology of dryland grass vs shrub functional types using a portable wind tunnel to test the hypothesis that grasses would be more susceptible to sandblasting than shrubs and, thus, reinforce transitions to shrub dominance in wind-erodible grasslands when climate- or disturbance-induced reductions in ground cover occur. Grasses and shrubs responded differently to sandblasting, wherein water-use efficiency declined substantially in grasses, but only slightly in shrubs, owing to grasses having greater increases in day/nighttime leaf conductance and transpiration. The differential ecophysiological response to sandblasting exhibited by grass and shrub functional types could consequently alter the vegetation dynamics in dryland grasslands in favour of the xerophytic shrubs. Sandblasting could thus be an overlooked driver of shrub encroachment in wind-erodible grasslands.