Long-term monitoring and modelling of pasture regeneration and water quality from a Bothriochloa pertusa site in the Great Barrier Reef catchments

There has been considerable investment in grazing land management in Australia's Great Barrier Reef catchments targeted at improving off-site runoff and water quality. This study used a systems modelling framework to analyse an 18-year monitoring data set from an Indian couch (Bothriochloa pertusa) dominated hillslope flume site in the Burdekin catchment, Australia. The analyses show important functional differences in B. pertusa pastures, relative to the native tussock pastures they replaced, with implications for how they are monitored and managed: (1) rates of landscape recovery were far slower than rates of degradation. In the practical time frames of most land managers, transition to a B. pertusa state is effectively irreversible, so historical management practices will have to adapt; and (2) transformations in pasture composition and structure can lead to misinterpretation of monitored trends in ground cover (a common proxy for land condition). At 50% pasture cover, B. pertusa pastures have only 33% of the biomass of local tussock pastures. Where B. pertusa invasions are occurring, inferences from positive associations between changes in cover and landscape function can break down: increases in ground cover can instead be accompanied by decreases in biomass, productivity, and runoff retention. The paper contributes an initial calibration of the GRASP pasture grazing systems model for a transformed B. pertusa site, capturing the observed relationships between grazing management, pasture biomass, ground cover and water quality. This serves as a starting point from which these landscapes, of growing extent and concern in Great Barrier Reef catchments, can be better represented in monitoring and modelling assessments, and for revised management options to be explored. The calibrated model was able to accurately represent long-term average runoff and sediment yield (both within 2% of observed) but was only able to adequately represent the year-to-year variation in runoff (not sediment).