Abstract1. Network analysis is increasingly widespread in ecology, with frequent questions asking which nodes (typically species) interact with one another and how strong are the interactions. Null models are a way of addressing these questions, helping to distinguish patterns driven by neutral mechanisms or sampling effects (e.g. relative abundance of different taxa, sampling completeness) from deterministic biological mechanisms (e.g. resource selection and avoidance), but few "off the shelf" tools are available.2. We present econullnetr, an r package combining null modelling and plotting functions for networks, with data-export tools to facilitate its use alongside existing network analysis packages. It models resource choices made by individual consumers, enabling it to capture individual-level heterogeneity and generalising to a wider range of data types and scenarios than models applied directly to network matrices.The outputs can be analysed from the level of individual links to whole networks.3. We describe the main functions and provide two short examples, along with the results of a benchmarking exercise to provide guidance about the statistical power and error rates. Our hope is that econullnetr provides a basis for more widespread use of null modelling to assist ecological network interpretation. K E Y W O R D Sfood webs, plant-pollinator networks, prey choice, resource selection

Summary Although agriculture is amongst the world's most widespread land uses, studies of its effects on stream ecosystems are often limited in spatial extent. National monitoring data could extend spatial coverage and increase statistical power, but present analytical challenges where covarying environmental variables confound relationships of interest.Propensity modelling is used widely outside ecology to control for confounding variables in observational data. Here, monitoring data from over 3000 English and Welsh river reaches are used to assess the effects of intensive agricultural land cover (arable and pastoral) on stream habitat, water chemistry and invertebrates, using propensity scores to control for potential confounding factors (e.g. climate, geology). Propensity scoring effectively reduced the collinearity between land cover and potential confounding variables, reducing the potential for covariate bias in estimated treatment–response relationships compared to conventional multiple regression.Macroinvertebrate richness was significantly greater at sites with a higher proportion of improved pasture in their catchment or riparian zone, with these effects probably mediated by increased algal production from mild nutrient enrichment. In contrast, macroinvertebrate richness did not change with arable land cover, although sensitive species representation was lower under higher proportions of arable land cover, probably due to greatly elevated nutrient concentrations. Synthesis and applications. Propensity modelling has great potential to address questions about pressures on ecosystems and organisms at the large spatial extents relevant to land‐use policy, where experimental approaches are not feasible and broad environmental changes often covary. Applied to the effects of agricultural land cover on stream systems, this approach identified reduced nutrient loading from arable farms as a priority for land management. On this specific issue, our data and analysis support the use of riparian or catchment‐scale measures to reduce nutrient delivery to sensitive water bodies.

Land‐use change can alter trophic interactions with wide‐ranging functional consequences, yet the consequences for aquatic food webs have been little studied. In part, this may reflect the challenges of resolving the diets of aquatic organisms using classical gut contents analysis, especially for soft‐bodied prey. We used next‐generation sequencing to resolve prey use in nearly 400 individuals of two predatory invertebrates (the Caddisfly, Rhyacophila dorsalis, and the Stonefly Dinocras cephalotes) in streams draining land with increasingly intensive livestock farming. Rhyacophila dorsalis occurred in all streams, whereas D. cephalotes was restricted to low intensities, allowing us to test whether: (i) apparent sensitivity to agriculture in the latter species reflects a more specialized diet and (ii) diet in R. dorsalis varied between sites with and without D. cephalotes. DNA was extracted from dissected gut contents, amplified without blocking probes and sequenced using Ion Torrent technology. Both predators were generalists, consuming 30 prey taxa with a preference for taxa that were abundant in all streams or that increased with intensification. Where both predators were present, their diets were nearly identical, and R. dorsalis's diet was virtually unchanged in the absence of D. cephalotes. The loss of D. cephalotes from more intensive sites was probably due to physicochemical stressors, such as sedimentation, rather than to dietary specialization, although wider biotic factors (e.g., competition with other predatory taxa) could not be excluded. This study provides a uniquely detailed description of predator diets along a land‐use intensity gradient, offering new insights into how anthropogenic stressors affect stream communities.