Scientific Foundations for an IUCN Red List of Ecosystems
Example using stream flowthrough data as percent of mean unregulated flows (aqua line joining filled circles) for the Murray River adapted from , see Appendix S2.8. There is uncertainty in both the rate of decline in flowthrough (two alternative regression lines) and the level of flowthrough at which the water-dependent ecosystem would collapse (shaded area). The threshold of collapse is the level of stream flowthrough that would result in widespread tree death and replacement of forest vegetation (most likely by shrubland). This was estimated to occur when mean flowthrough (as estimated by long-term regression) falls to 0–10% of unregulated flow levels (shown as a bounded estimate c1–c2, dashed lines), as widespread tree dieback began to occur when flowthrough was zero in several year of the past decade (see Appendix S2.8 for process model and justification). Based on a best-fit Gaussian regression model of the flowthrough data (dark blue line), the mean flowthrough fell from 71% in 1960 (dotted line a1) to 50% in 2009 (dotted line b1). A beta regression model (red line) gave an improved fit to the data and indicates a decline in mean flowthrough from 63% in 1960 (a2) to 31% in 2009 (b2). A standardised estimate of the relative severity of hydrological degradation over the past 50 years = 100×(b-a)/(c-a). The minimum plausible estimate = 100×(b1–a1)/(c1–a2) = 100×(71–50)/(71–0) = 30% and the maximum plausible estimate = 100×(b2–a2)/(c2–a1) = 100×(63–31)/(63–10) = 60%. Based on uncertainty in the flowthrough regression models and collapse threshold, a bounded estimate of hydrological degradation in this ecosystem is therefore 30–60% over the past 50 years.