“Frequency versus amount” – Understanding the impact of rainfall extremes on Australian grasslands

Background/Question/Methods

Given the strong link between plant productivity and rainfall inputs (Harper et al., 2005, Walter et al., 2010), climate model predictions of decreasing frequency and amount of rainfall are likely to affect the composition and functioning of terrestrial ecosystems across the globe. Australia naturally experiences a high degree of inter-annual climate variability. Predicted decreases in the frequency of rainfall events and seasonal reductions of up to 50% in rainfall amount will increase both the severity and duration of drought stress, and push Australian grasslands into uncharted climate territory. In 2013 we established a rainfall manipulation experiment at a grassland site in Western Sydney with two aims: 1) quantifying the impacts of rainfall extremes on the plant community and key ecosystem processes, and 2) establishing the relative importance of changes in the frequency versus the amount of rainfall for ecosystem function. Rainfall shelters are used to exclude ambient precipitation, with water reapplied to achieve the following treatments: 1) ambient, 2) 50% reduction, 3) 50% increase, 4) reduced frequency (RF, ambient rain amount reapplied in full, on one occasion every 3 weeks), and 5) summer drought (12 week rainfall removal).

Results/Conclusions

There was clear evidence of shifts in plant community composition in the first year of this study. Plots receiving rainfall addition (+50%) had increased abundance of the native C3 grass, Microlaena stipoides, while those receiving reduced rainfall (-50% and summer drought) shifted towards dominance by the C4 grasses Cynodon dactylon and Eragrostis curvula. Reduced rainfall frequency was associated with increased abundance of Cymbopogon refractus and E. curvula, both C4 grasses. Aboveground productivity was significantly affected by rainfall treatment, with the highest biomass in +50% and RF treatments, and the lowest in plots experiencing summer-long drought. Belowground productivity was also related to rainfall, with the lowest root biomass in RF plots. Complementary glasshouse experiments suggest that reduction in rainfall frequency drives a shift towards deeper rooting strategies, while reduced rainfall amount is associated with a shallower rooting profile. Overall, early results suggest that changes in rainfall amount and frequency are both driving major shifts in plant community composition, but that the size of rainfall events is a more important determinant of aboveground productivity than rainfall frequency in these plant communities.