Interacting effects of vegetation and hydrogeomorphic complexity on nitrate in agricultural waterways

Background/Question/Methods

Intensive cultivation in the California Central Valley has resulted in the loss of ~95% of riparian habitat. In addition, small streams have been replaced by denuded agricultural waterways to convey irrigation drainage high in nitrate and other pollutants. A variety of strategies are being used to increase riparian vegetation along agricultural waterways, but many of the ecological functions of these novel habitat features are unknown. An extensive study was conducted in summer 2014 to investigate which riparian ecosystem features are associated with variable nitrate levels in waterways draining irrigated cropland. Turbidity levels, which are associated with numerous other water quality concerns in this system, were also examined. Eighty sites were selected across two agricultural water districts to represent a wide range of riparian vegetation, hydrogeomorphic features, and agricultural contexts. Sites were visited three times to correspond to the early, mid, and late irrigation season. At each visit, two water samples were taken 300m apart and analyzed for turbidity, pH, temperature, and nitrate. To characterized hydrogeomorphic complexity, the frequency of features associated with transient storage and increased retention time, such as debris jams and side pools, were quantified. Vegetation cover and structure on banks, foreshores, and in channels was also measured. 

Results/Conclusions

Using a generalized linear model, we tested for the interaction of upstream [nitrate-N] with vegetation and hydrogeomorphic variables to predict downstream [nitrate-N], the dependent variable. A similar model was used to test for effects on turbidity. As expected, upstream [nitrate-N] strongly predicted downstream [nitrate-N] and no features directly predicted downstream [nitrate-N]. The frequency of hydrogeomorphic features, however, interacted with upstream [nitrate-N] to predict downstream [nitrate-N], indicating an effect of hydrogeomorphic complexity on nitrate from up to downstream. Vegetation factors did not interact with upstream [nitrate-N], but woody vegetation cover was positively correlated with the frequency of hydrogeomorphic features. This suggests that woody vegetation may indirectly influence nitrate levels by increasing hydrogeomorphic complexity. In the turbidity model, aquatic vegetation cover interacted with upstream turbidity to predict downstream turbidity. No other factors or interactions predicted downstream turbidity. These findings support the use of woody riparian vegetation to increase nitrate retention in agricultural waterways, via an indirect association with hydrogeomorphic complexity. However, turbidity levels were most affected by aquatic vegetation cover, which was inversely related to woody vegetation cover. Whether riparian or aquatic vegetation can be recommended for water quality enhancement may depend on the focal water quality concern.