Long-term monitoring data are rightly seen as a valuable resource for identifying trends and changes in endpoints. However, condition monitoring programs are of limited use for informing management decisions because they don’t establish causality between drivers and outcomes. Targeted, complementary research is needed to fill these knowledge gaps. This presentation will illustrate this outcome using our experience with 12 years of data collection (so far) under the Commonwealth Government’s Flow-MER program, which monitors the effects of environmental flows on multiple ecosystem outcomes in the Murray-Darling Basin.
As part of our original monitoring plan, we developed a conceptual model linking flows to ecological outcomes for the priority monitoring endpoints, bank vegetation and fish. The model hypothesized that flow alone could not explain variation in these endpoints, but relationships with intermediate factors, such as bank condition, stream metabolism, and macroinvertebrate biomass would also shape the response. However, more than a decade of data collection has struggled to confirm or falsify these linkages. Mostly, we see loose associations of outcomes among the individual monitoring endpoints, but without sufficient evidence to provide confidence in a causal relationship.
These monitoring programs were set up to test causal relationships between river flows and individual environmental outcomes. In this aim, they have been successful for several endpoints. However, targeted research has been necessary to better test links among environmental outcomes in our conceptual model.
Our experience with one of the best-funded long-term monitoring programs in Australia shows that it is difficult to infer causality with monitoring data alone. Instead, the data can point towards possible relationships that must then be assessed using targeted research. This does not diminish the value of long-term data for detecting trends or changes. We merely contend that to understand causality, and therefore inform management, more than long-term monitoring is required.