Long‐Term Monitoring of Macroinvertebrate Communities Over 2,300 km of the Murray River Reveals Ecological Signs of Salinity Mitigation Against a Backdrop of Climate Variability

Abstract

We investigated the ecological effects of salinity mitigation strategies in the Murray‐Darling Basin (MDB) using macroinvertebrate data collected over 2,300 km of the Murray River between 1980 and 2012. The MDB covers 1 × 106km2 and includes both temperate and semiarid climate zones. It was extensively developed to support irrigated agriculture in the early to mid‐1900s, and the secondary salinization that followed has become a major concern. During 1975–1985 daily salinity levels, measured as electrical conductivity above the Murray River off‐take points for South Australia's major urban water supplies, were above 800 μS/cm for 40% of the time, necessitating mitigation strategies that have reduced the average salinity by about 150 μS/cm since monitoring began. The MDB has also experienced several major floods and droughts during this time, and surface temperatures in the MDB have increased by 0.8 °C since 1910, mostly in the last 50 years. We hypothesized that (1) taxa richness would increase in response to floods; (2) community structure would shift toward tolerant, opportunistic taxa in response to warming; and (3) geographical ranges of species would change in response to shifting stream isotherms and reducing salinity. Our hypotheses were supported, although increases in water temperature appeared to be due principally to the 1997–2009 Millennium drought. Importantly, against a backdrop of significant climate variability, we believe we have distinguished a change in community structure along a salinity gradient and that changes over the 33 years can in part be attributed to mitigation strategies.

Aquatic Macroinvertebrate Community Changes Downstream of the Hydropower Generating Dams in Myanmar-Potential Negative Impacts From Increased Power Generation

“…These changes have altered the river's hydrological regime to one of drought years (61 years per 100 years compared with 5 years per 100 years under natural conditions) (Boulton and Brock, 1999). Recent studies by Paul et al (2018) and Le et al (2020) have shown that the macroinvertebrate fauna are impacted by the changes in hydrology and take over three decades to recover after a flood event (Le et al, 2020).…”

Quantitative Flow‐Ecology Relationships Using Distributed Lag Nonlinear Models: Large Floods in the Murray River Could Have Delayed Effects on Aquatic Macroinvertebrates Lasting More Than Three Decades

“…First, GAMs were fitted for macroinvertebrate taxa richness and total abundance as a function of the physicochemical variables (i.e., EC, pH, WTEMP, TURB, and VEL) and time (coded as TIME, which was denoted by integers indicating the temporal order of the biannual samples) to reproduce the temporal pattern in the macroinvertebrate metrics after controlling for other variables reported by Paul et al (2018).…”

“…In a recent study by Paul et al (2018), it was suggested that the effects of the 1993 flood in the Murray River, Australia, might have persisted for decades. Working with a 33‐year (1980–2012) data set from the River Murray Water Quality and Biological Monitoring Program on macroinvertebrates and hydrological, physicochemical, and climate variables collected at six sites spanning over 2,300 km of the Murray River, these authors noticed a residual temporal pattern in the macroinvertebrate community following the 1993 historic flood at all sites after controlling for the effects of other variables, including contemporaneous hydrological and physicochemical conditions.…”

“…Our aim is to quantify the relationships between flow and macroinvertebrates using the same data set as used by Paul et al (2018), with a hydrological data set extended back to 1961. The period 1961–2012 included major floods, in 1974, 1993, and 2010/2011, as well as several droughts, including the Millennium drought from 1996–2010 (Murray‐Darling Basin Authority [MDBA], 2017a), though at no time did the river fragment into a series of disconnected pools.…”

“…We hypothesized that large floods would have delayed effects on macroinvertebrates that persist for decades. Building upon the work of Paul et al (2018), we used distributed lag nonlinear models (DLNMs; Gasparrini et al, 2010) to quantify the delayed effects of flow at different magnitudes on macroinvertebrate abundance and richness, where the term “delayed effects” throughout this article was employed in the sense of “lagged relationships” to conform to the terminology used in DLNMs. DLNMs enable a parsimonious description of the delayed and possibly nonlinear effects of predictor variables, overcoming the limitations of simpler distributed lag models (DLMs) that assume linear relationships, and which can require estimation of numerous parameters for highly collinear lagged variables.…”