Going beyond richness and diversity to capture the consequences of land use on stream fish in Ohio

Background/Question/Methods

Stream habitats experience significant degradation in watersheds that are dominated by agricultural land use. Channelization, sedimentation and nutrient input degrade the physical structure of streams. Stream algal communities undergo composition changes when streams experience nutrient loading; so while productivity may increase, food availability for grazers may not. Beyond the typical scopes of studies that consider changes in richness or diversity only, we asked, how do fish communities vary along a gradient of agricultural land use? How well does land cover composition predict biotic integrity? To this end, we analyzed the effects of land use on fish communities as represented by several biotic metrics, including the abundance  of specific trophic groups and a regionally developed Index of Biotic Integrity (IBI). We used stream habitat and fish community data from the Ohio Environmental Protection Agency’s Fish Database, a long-term monitoring project with more than 2,700 sites located throughout the state. With landcover data from the USGS Land Cover Database (2011), we calculated land use profiles for sites throughout all four major ecoregions in the state.

Results/Conclusions

In the two eastern ecoregions, land use is most heterogeneous with the lowest agricultural land use. In those areas the stream habitat quality, land use composition in the watershed, and ammonia concentration were the strongest predictors of fish community integrity; species richness and Shannon diversity had no relationship with any physical or chemical factors. In the western ecoregions, which primarily consist of row crop farming and urban development, percent agricultural land use did not have a direct relationship with fish community integrity. However, communities appeared to be more severely impacted in those areas and showed negative relationships with other factors thought to be influenced by agricultural land use, indicating that the relationship between land use patterns and biotic communities are complex. In particular, the herbivorous fish community decreases in both abundance and as a proportion of the overall community as agricultural land use increases, indicating that the detrimental effects of such land use outweighs any potential increase in food availability for that trophic group. While we observed some of the expected relationships between water chemistry, stream habitat quality, and agricultural land use, our results imply that land use data cannot be used to determine stream health without local assessment.