Influence of land change on stream fish communities resampled after 15 years

Background/Question/Methods

The Alameda Creek Watershed is the largest watershed draining directly into San Francisco Bay and supports a diverse fish assemblage. This watershed has experienced considerable human population growth in recent years, leading us to ask how land change has influenced the fish community. In 2009, we resampled 70 sites across the watershed that were initially sampled between 1992-1994. We hypothesized that we would detect a significant change in land cover and associated fish assemblages across the watershed, but that mid-elevation sites would shift towards more mixed native-nonnative associations following recent urban expansion. Separate analyses were performed for six sub-watersheds within the larger Alameda Creek Watershed.  We used supervised classification with Landsat imagery to classify land cover into four classes (forest, grassland, urban, agriculture). We compared changes in species abundance and watershed/riparian land cover across time using Wilcoxon paired sampled tests. Additionally, we used nonmetric multidimensional scaling and a permutational analysis of variance to quantify shifts in community structure after 15 years.

Results/Conclusions

Preliminary results show significant changes in land use (forest/grassland to urban/agriculture) in mid-elevation regions, while high elevation-headwater and low elevation-densely urban regions have remained roughly unchanged. Twenty-six species were captured in both time periods and twenty-four of these were common to the two time periods. There were no basin-wide extirpations of native fishes, however there was an overall decrease in mean native species diversity and relative abundance (Wilcoxon, p<0.05). The proportion of Sacramento sucker (C. occidentalis) increased in half of the regions, perhaps indicating a shift towards a system dominated by tolerant native fishes. The change in land cover within a 3km watershed buffer had the strongest relationship with the change in fish community structure (R²= 0.54), suggesting that more localized scales of land use were most important in terms of explaining fish shifts across the 15 years. These results emphasize the need to consider watershed-scale factors in management efforts to conserve native fishes of the region.