Dominant cold desert plants do not partition warm season precipitation by event size.

We conducted experiments to examine the quantitative relationships between rainfall event size and rainwater uptake and use by four common native plant species of the Colorado Plateau, including two perennial grasses, Hilaria jamesii (C(4)) and Oryzopsis hymenoides (C(3)), and two shrubs, Ceratoides lanata (C(3)), and Gutierrezia sarothrae (C(3)). Specifically, we tested the hypothesis that grasses use small rainfall events more efficiently than shrubs and lose this advantage when events are large. Rainfall events between 2 and 20 mm were simulated in spring and summer by applying pulses of deuterium-labeled irrigation water. Afterwards, pulse water fractions in stems and the rates of leaf gas exchange were monitored for 9 days. Cumulative pulse water uptake over this interval (estimated by integrating the product of pulse fraction in stem water and daytime transpiration rate over time) was approximately linearly related to the amount of pulse water added to the ground in all four species. Across species, consistently more pulse water was taken up in summer than in spring. Relative to their leaf areas, the two grass species took up more pulse water than the two shrub species, across all event sizes and in both seasons, thus refuting the initial hypothesis. In spring, pulse water uptake did not significantly increase photosynthetic rates and in summer, pulse water uptake had similar, but relatively small effects on the photosynthetic rates of the three C(3) plants, and a larger effect on the C(4) plant H. jamesii. Based on these data, we introduce an alternative hypothesis for the responses of plant functional types to rainfall events of different sizes, building on cost-benefit considerations for active physiological responses to sudden, unpredictable changes in water availability.