Dryland Floodplain Ecosystems: influence of flow pattern on fish production

"2002".

Project Number: Native Fish Recruitment and Flood Pulse Water Quality Monitoring on Lindsay Island - M/BUS/83.

MDFRC item.

44 pages.

Management decisions to improve Australia’s native freshwater fish populations are based largely on the Flood Pulse Concept (Junk et al. 1989). Floods are commonly thought to both stimulate spawning in native fish and provide sufficient food for the recruitment of larvae. More recently, however, this notion has been challenged (Humphries et al. 1999, 2002). Many native species have been shown to spawn irrespective of flow events, and the decline of native species has been related to a lack of slow flowing and productive habitats suitable for recruitment of larvae. This study examines the use of different but interconnected flow habitats by adult, juvenile and larval fishes on Lindsay Island. Thirteen fish species were collected as adults or juveniles from three different flow habitats (fast creek, shallow pond, weir pool). Large bodied fish, including Murray cod, golden perch and carp, were most abundant in the fast creek habitat, followed by the shallow ponded habitat and the weir pool, respectively. Smaller bodied fish, including gudgeon, hardyhead and Australian smelt, were most abundant in the shallow creek habitat, followed by the weir pool and fast creek habitats, respectively. With the exception of golden and silver perch, all species collected as adults/juveniles were also collected as larvae. The species diversity of larval catch was similar in all three habitats, though many species were collected only in either the fast creek or the slow flowing (weir pool and shallow pond) habitats. Only three species, flathead gudgeon, Murray rainbowfish and Australian smelt, were collected as larvae from all three flow habitats. Larval abundances in the three flow habitats varied temporally and were highly dependent on the timing of individual species spawning. Overall, the weir pool yielded the greatest larval abundance, followed by the shallow pond and the fast creek. This catch pattern supports the ‘low flow recruitment hypothesis’ of Humphries et al. (1999). Comparison of the timing of the appearance of larvae in Lindsay Island with that in the Broken and Campaspe rivers (Humphries et al. 2002) indicates that in both upland and lowland systems, spawning of many native species occurs at similar times of the year independent of temporal changes in flow. This refutes the commonly held notion that a flood pulse is necessary to stimulate spawning and recruitment of native fishes. A new model, the temporal trophic cascade, is suggested as a more appropriate model for the management of native fish populations.