Effects of salinity and flow interactions on macroinvertebrate traits in temporary streams

Increasing salinity in freshwater ecosystems is globally widespread, especially, in arid and semi-arid regions, and can co-occur with flow intermittency, particularly in temporary streams. Both these stressors are known to affect macroinvertebrate traits individually, but their interactive effects have not been previously considered. There are inconsistencies reported in the literature regarding the response of particular traits to flow or salinity, and accordingly, we hypothesized that interactive effects between these two stressors may underlie inconsistencies in the literature. We used multivariate and univariate approaches to investigate the effects of salinity and flow interactions on macroinvertebrate traits using 13 years of data sampled across multiple sites in South Australia, the driest state in the driest inhabited continent in the world. Ovoviviparity, multivoltinism, aerial respiration and strong fliers were favoured as salinity increased, while medium-high physiological sensitivity to salinity and respiration via gills decreased. During low flows, holometaboly, univoltinism, high rheophily, cool eurythermality, streamlined body shape and gill respiration decreased, while aerial respiration and fliers and high crawling rate increased. Interestingly, traits with inconsistent responses (e.g. burrowing, tegument respiration and collector-gathering traits) in the literature were associated with interactions between flow and salinity in our study. These traits showed a similar interaction, by being least abundant in streams with high salinity and low flows, and low salinity and high flows. The interactions seem to be driven by the differential response of different taxa with the same trait category being abundant in different parts of the interaction plot. Our findings suggest that, in addition to differences in methodological and analytical approaches, interactions may also underlie inconsistencies in trait responses to flow and salinity. Finally, to foster the operative use of traits to resolve the effects of multiple stressors on ecosystems, there is the need for a better mechanistic understanding of how specific stressors (e.g. flow and salinity) act as trait filters, potentially through the use of experiments, to ensure that each of the stressors is strong enough to produce clear trait responses.