Experimental increases in detritus boost abundances of small‐bodied fish in a sand‐affected stream

Abstract

Abstract Restoration projects often rely on the assumption that a local intervention will restore diminished populations, without fully understanding the constraints that limit the target species in the first place. In rivers, one common restoration technique is to place large structures, such as wood and boulders, on the bed, with the assumption that fish will subsequently arrive and use them. Nonetheless, providing large habitat structure may not overcome demographic or resource constraints on fish populations, and thus may not aid recovery. We aimed to test if resource constraints (food and cover) are limiting local densities of fishes in a degraded stream by experimentally alleviating these constraints. If the abundance of one or more species is constrained by resource availability, then local numbers of these species should increase following an increase in resources. To test our prediction, we increased the availability of food and microhabitat complexity (cover) at sites in Hughes Creek, a degraded stream in south‐east Australia that has extensive accumulations of sand and limited in‐stream structure. At treatment sites, we hammered pairs of wooden stakes (25 cm apart) into the stream bed so that the ends of stakes protruded just above the water surface at moderate flows. Stakes effectively trapped passing sticks and leaves, which increased local detrital densities and, subsequently, invertebrate densities, hence providing food and cover for fish. Over the course of a year, we compared the changes in fish abundances at treatment sites to unmanipulated control sites. Fish responded quickly to enhanced retention of detritus, with assemblage differences observed between treatment and control sites. We caught more river blackfish (Gadopsis marmoratus), southern pygmy perch (Nannoperca australis), Macquarie perch (Macquaria australasica), and mountain galaxias (Galaxias olidus) at treatment sites on some occasions, indicating that these species may be subject to resource constraints in this stream. The magnitude of observed positive fish responses was influenced by the life‐stage of individuals and local stream conditions. Importantly, treatment effects varied through time and were no longer observed after a large flood affected the study stream. Our results show that resource constraints limit local species abundances and demonstrate a novel method of overcoming these constraints in a small, degraded stream. This is a necessary first step but future work is needed to examine whether increases in abundance are due to the provided resources increasing growth rates, survivorship, or reproduction. This work also highlights the importance of understanding species’ life histories, the broader landscape setting, and the disturbance regime when undertaking site scale restoration.