Inferring landscape-scale land-use impacts on rivers using data from mesocosm experiments and artificial neural...

Area of Science:

• Ecology
• Environmental Science
• Freshwater Biology

Background:

• Land-use changes, such as livestock grazing, significantly impact river ecosystems.
• Correlations between variables like nutrients and sediments complicate identifying specific drivers of river condition changes.
• Standard metrics may not always reflect ecological impacts, necessitating advanced analytical approaches.

Purpose of the Study:

• To investigate the impacts of catchment-scale domestic livestock grazing on river macroinvertebrate communities.
• To differentiate the roles of nutrients and fine sediments as mediators of grazing impacts.
• To utilize artificial neural networks (ANNs) to analyze landscape-scale ecological data.

Main Methods:

• Conducted a gradient survey of river macroinvertebrate communities across varying levels of upstream catchment grazing.
• Employed a stream mesocosm experiment to independently quantify the effects of nutrients and fine sediments on macroinvertebrates.
• Trained artificial neural networks (ANNs) using experimental data to predict nutrient and fine sediment impacts at survey sites based on community composition.

Main Results:

• A correlative approach showed a strong relationship between macroinvertebrate community structure and the proportion of catchment area under grazing.
• ANNs indicated that fine sediments, not nutrients, were significantly related to the extent of catchment grazing.
• Macroinvertebrate communities at grazed sites resembled those exposed to high fine sediment levels in experimental settings.

Conclusions:

• Fine sediment is a critical mediator of land-use impacts on river macroinvertebrate communities.
• Artificial neural networks (ANNs) are effective tools for identifying subtle ecological effects and disentangling correlated variables.
• Experimental data can inform and explain patterns observed at larger landscape scales.