Trophic strategies of garfish, Arrhamphus sclerolepis, in natural coastal wetlands and artificial urban waterways

We used carbon stable isotope and stomach content analyses to test whether snub-nosed garfish, Arrhamphus sclerolepis (Hemiramphidae), in the extensive artificial urban waterways of southeast Queensland, Australia, rely on autotrophic sources different to those in natural wetlands. Carbon isotope values of A. sclerolepis were similar to those in previous investigations, with enriched values in natural habitat (mean = 13.9&, SE=0.6) and depleted values ( 19.1&, 0.1) in artificial habitat. A. sclerolepis in natural habitat consumed large amounts of seagrass during the day and night, and at night also ingested small quantities of crustacean prey. In artificial habitat, A. sclerolepis consumed macroalgae during the night and switched to invertebrates (terrestrial ants) in the day. Values of d N in all the fish were 3–8& more enriched than sources. Mathematical modelling of feasible source mixtures showed that in natural habitat the bulk of the dietary carbon is obtained from seagrass, but the nitrogen is obtained from animal prey. In artificial habitat, carbon is obtained from a mixture of macroalgae and animals. We could not determine the nitrogen sources in artificial habitat of A. sclerolepis since, even after accounting for trophic fractionation of d N, the values were outside the range of potential sources. If the types of animals ingested vary over time, perhaps one or more types of animal important in the provision of nitrogen was not sampled during the study. This study demonstrates that not only does A. sclerolepis occur in both artificial and natural habitats, but it uses the same strategy of bulk herbivory with the inclusion of smaller amounts of animal prey. This understanding of how ecological processes support fisheries production in artificial habitat improves the overall understanding of the effects of urbanisation on coastal food webs.