Artificial waterway design affects fish assemblages in urban estuaries

Fishes were collected from residential canal estates leading directly off natural estuaries and artificial lakes with control tidal exchange in south-east Queensland, Australia, to test the model that the difference in artificial waterway design affects fish assemblages. A total of 17 779 fish representing 52 species was caught, including 23 species of economic importance (45% of the catch). Total fish abundance and species richness differed little among sites within or between the two artificial habitats (canals and lakes) in either of the seasons sampled (winter and spring). Multivariate analysis showed, however, that assemblages differed among sites within the same artificial habitat. The differences were best explained by the distance sites were from open water, while salinity, water temperature and dissolved oxygen explained little of the variability. Most species were found in canals and lakes, but there were enough differences in composition between the habitats to detect a difference in both seasons (significant ANOSIM tests). Salinity was lower in lakes because of the tidal restrictions, and while this was only weakly correlated with differences in fish assemblages, it had the most explanatory power of any environmental variable. New recruits arrived later in lakes than canals, perhaps because of the barriers to tidal flow. A survey the following year showed that differences among individual lakes were consistent through time, offering insights into the influence of different tidal barriers on fish assemblages. The design change from canals to lakes has a minor influence on fish assemblages and alters the timing of recruitment. # 2007 The Authors