Summary Plant productivity in deserts may be more directly responsive to soil water availability than to precipitation. However, measurement of soil moisture alone may not be enough to elucidate plant responses to precipitation pulses, as edaphic factors may influence productivity when soil moisture is adequate. The first objective of the study was to determine the responses of the aboveground annual net primary productivity (ANPP) of three perennial species (from different functional groups) in a Chihuahuan Desert grassland to variation in natural precipitation (annual and seasonal) and a 25% increase in seasonal precipitation (supplemental watering in summer and winter). Secondly, ANPP responses to other key environmental and soil parameters were explored during dry, average, and wet years over a 5‐yr period. ANPP predictors for each species were dynamic. High ANPP in Dasylirion leiophyllum was positively associated with higher soil NH4‐N and frequent larger precipitation events, while that in Bouteloua curtipendula was positively correlated with frequent small summer precipitation events with short inter‐pulse periods and supplemental winter water. Opuntia phaeacantha was responsive to small precipitation events with short inter‐pulse periods. Although several studies have shown ANPP increases with increases in precipitation and soil moisture in desert systems, this was not observed here as a universal predictor of ANPP, particularly in dry years.

Global climate models predict that in the next century precipitation in desert regions of the USA will increase, which is anticipated to affect biosphere/atmosphere exchanges of both CO(2) and H(2)O. In a sotol grassland ecosystem in the Chihuahuan Desert at Big Bend National Park, we measured the response of leaf-level fluxes of CO(2) and H(2)O 1 day before and up to 7 days after three supplemental precipitation pulses in the summer (June, July, and August 2004). In addition, the responses of leaf, soil, and ecosystem fluxes of CO(2) and H(2)O to these precipitation pulses were also evaluated in September, 1 month after the final seasonal supplemental watering event. We found that plant carbon fixation responded positively to supplemental precipitation throughout the summer. Both shrubs and grasses in watered plots had increased rates of photosynthesis following pulses in June and July. In September, only grasses in watered plots had higher rates of photosynthesis than plants in the control plots. Soil respiration decreased in supplementally watered plots at the end of the summer. Due to these increased rates of photosynthesis in grasses and decreased rates of daytime soil respiration, watered ecosystems were a sink for carbon in September, assimilating on average 31 mmol CO(2) m(-2) s(-1) ground area day(-1). As a result of a 25% increase in summer precipitation, watered plots fixed eightfold more CO(2) during a 24-h period than control plots. In June and July, there were greater rates of transpiration for both grasses and shrubs in the watered plots. In September, similar rates of transpiration and soil water evaporation led to no observed treatment differences in ecosystem evapotranspiration, even though grasses transpired significantly more than shrubs. In summary, greater amounts of summer precipitation may lead to short-term increased carbon uptake by this sotol grassland ecosystem.

Arid and semi-arid environments are dynamic ecosystems with highly variable precipitation, resulting in diverse plant communities. Changes in the timing and magnitude of precipitation due to global climate change may further alter plant community composition in desert regions. In this study, we assessed changes in species richness and plant density at the community, functional group, and species level in response to variation in the magnitude of natural seasonal precipitation and 25% increases in seasonal precipitation [e.g., supplemental watering in summer, winter, or summer and winter (SW)] over a 5-year period in a sotol grassland in the Chihuahuan Desert. Community species richness was higher with increasing winter precipitation while community plant density increased with greater amounts of winter and summer precipitation, suggesting winter precipitation was important for species recruitment and summer precipitation promoted growth of existing species. Herb and grass density increased with increasing winter and summer precipitation, but only grass density showed a significant response to supplemental watering treatments (SW treatment plots had higher grass density). Shrubs and succulents did not exhibit changes in richness or density in response to natural or supplemental precipitation. In this 5-year study, changes in community species richness and density were driven by responses of herb and grass species that favored more frequent small precipitation events, shorter inter-pulse duration, and higher soil moisture. However, due to the long life spans of the shrub and succulent species within this community, 5 years may be insufficient to accurately evaluate their response to variable timing and magnitude of precipitation in this mid-elevation grassland.