A major research focus in population and community ecology is to establish a mechanistic understanding of plant interactions and demographic responses. The first step towards this mechanistic approach relies on understanding the differences in stress caused by different environmental conditions. Leaf-level photosynthetic rate (A) within and among plant populations provides important insight into population and community processes, but is difficult to acquire with sufficient replication under field conditions. Instead, chlorophyll fluorescence (F-v/F-m) and predawn water potential (Psi(pd)) are often used in arid and semi-arid ecosystems. 2. F-v/F-m reflects the photoactivation status of photosystem II (PSII), whereas Psi(pd) indicates water availability in the rhizosphere. Here we compare these indices with A in two perennial C-4 grasses (native Heteropogon contortus and invasive Eragrostis lehmanniana) and in seedlings of the C-3 shrub Prosopis velutina growing on highly contrasting sandy loam and loamy clay soils in experimental plots. Measurements were made the day prior to and up to 7 days following a 39-mm rainfall pulse after 2 months of drought. 3. A was more sensitive across a broad range of environmental conditions, whereas F-v/F-m and Psi(pd) only responded to periods of protracted drought. The use of these measures was further complicated because their values varied daily and we observed different time-lags in their response to precipitation pulses. 4. We suggest sampling schemes and a priori measurements to capture the value that is representative for the question of interest, and that match the pulsed biological activity in these ecosystems. Finally, we suggest the use of these measures in combination with measurements providing integration over longer time periods, such as delta C-13, delta O-18 and N concentration in bulk leaf tissue.