Using high frequency digital repeat photography to quantify the sensitivity of a semi-arid grassland ecosystem to the temporal repackaging of precipitation

Across large portions of the western United States, drought intervals are increasing, often accompanied by larger-magnitude rainfalls. Semi-arid ecosystems are expected to be especially responsive to such temporal repackaging of rainfall because of their high sensitivity to variation in soil moisture. We conducted a field manipulation experiment to evaluate the impacts of summer rainfall repackaging (small/many events to large/few events with a fixed total seasonal amount) on a semi-arid mixed annual/perennial bunchgrass ecosystem. We monitored the sensitivity of plant greenness and productivity to soil moisture under rainfall pulses by combining automated, high-frequency repeat digital images at both plot- and plant functional type- level, whole plot CO2 uptake measurements, and continuous in-situ soil moisture data. We found that plot greenness was closely correlated with gross primary productivity across all rainfall repackaging treatments (R = 0.82). Plot greenness was weakly correlated to soil moisture at the beginning of the growing season but showed a significant positive correlation during peak and late growing season. Notably, we found a significant lag time of similar to 5 days between changes in soil moisture and canopy greenness under large/few events, while there was no lag under small/many events. This time lag was found to be driven by perennial grass greenness and its sensitivity to the relatively deep soil water infiltration observed with large/few rainfall events. Predicting semi-arid grassland responses to soil moisture dynamics under ongoing rainfall repackaging, and subsequent impacts on the regional to global carbon budget, should consider plant communities' phenological characteristics and functional type composition.