No abstract

Summary 1. Stream managers need to understand relationships between multiple stressors and ecological responses. We examined responses of benthic invertebrates and algae along two land‐use‐related stressor gradients of concern in running waters. Our correlative study of the consequences of augmented deposited fine sediment and nutrient concentrations was conducted in a regional set of streams and rivers monitored by a water management authority in New Zealand and incorporated a wide range of catchment geologies and stream orders. 2. We used multiple linear regression analysis and an information‐theoretic approach to select the best predictive models for our biological response variables by testing multiple competing hypotheses that include nonlinear subsidy‐stress relationships and interactive effects between the two stressors. 3. Patterns consistent with a subsidy‐stress response to increasing dissolved inorganic nitrogen concentration were found for the relative abundances of the common invertebrate genera Pycnocentrodes and Deleatidium and for the relative abundance of total individuals in the EPT orders (Ephemeroptera, Plecoptera, Trichoptera). 4. Fine sediment seemed the more pervasive stressor, apparently counteracting and overwhelming any initial subsidy effect of increased nutrients, and accounting for more of the variance in biological response variables. Relationships with high nutrient concentrations were weaker and modelled with less certainty, probably reflecting the indirect modes of action of nutrients compared to those underlying sediment effects. Nevertheless, in several cases, the models indicated that nutrients interacted synergistically with fine sediment, lending further weight to the conclusion that managers need to address both stressors to achieve the best outcomes.

1. Large-scale comparative studies of ecological responses to anthropological stressors in rivers require measures that are consistent across a range of spatial scales. The biological trait profile of communities offers an alternative approach to traditional measures of macroinvertebrate taxonomic identity and is less constrained by biogeographic influences. 2. We compared the capacities of taxonomic composition and biological trait composition to discriminate the effects of land use (measured as percentage of the catchment in pasture) across a large geographic zone (the whole of New Zealand) in 30 sub-catchments grouped into five ecoregions throughout the North and South Islands of New Zealand. In addition, we investigated trait consistency (i.e. whether similar traits had similar trait responses to land use at local (catchment) and broad scales). 3. The analysis of taxonomic composition showed that community structure was indeed influenced by land-use intensity, but that relationships differed among ecoregions. In contrast, traditional assessment metrics (Macroinvertebrate Community Index, richness in Ephemeroptera, Plecoptera and Trichoptera taxa) and trait composition were uninfluenced by region. Trait responses were consistent at the broad and catchment scales, with similar traits responding to pastoral land use at both scales. 4. We used general linear modelling to investigate individual trait responses to land-use intensity, catchment area and region, focussing on 15 trait categories known to be influenced by land-use intensity at the catchment scale. Several trait categories varied with land-use intensity and demonstrated consistency at both catchment and broad scales. Of these, the representation of shorter generation time, asexual reproduction and hermaphroditism, ability to lay eggs beneath the water surface, egg protection and respiration types tolerant of oxygen depletion generally increased in assemblages exposed to more intense pastoral land use. At the same time, the representation of short life duration of adults, prevalence of laying eggs at the water surface, sexual reproduction and low body flexibility decreased in assemblages exposed to land-use intensification.

Summary1. Organic farming practices can benefit a wide range of terrestrial biota in comparison to conventional farming but we do not know whether this benefit extends to streams flowing through farmed landscapes. 2. We measured the impacts of organic, integrated management and conventional farming on the taxonomic and trait composition of macroinvertebrate communities and on stream ecosystem functioning (algal accrual on tiles and leaf breakdown). Our study design included five replicates of each farming practice, arranged in blocks, in agricultural land dedicated to the farming of sheep and beef cattle on pasture in southern New Zealand. In each farm stream, we studied both upstream and downstream reaches within the farm's boundaries. 3. The different farming practices were reflected in contrasting stream physicochemistry (total dissolved nitrogen, fine sediment on the bed and glyphosate concentrations in the bed sediment) and in differences in both the taxonomic organization and trait representation of stream invertebrate communities. Conventional farm streams showed the strongest negative responses, whilst the condition of organic and integrated farm streams was similar. 4. Invertebrate trait measures proved as effective as taxonomic measures in their response to agricultural intensity, whereas ecosystem function measures were least sensitive. 5. There were no overall physicochemical differences between upstream and downstream sites within the farms and few notable longitudinal patterns in ecological response variables. 6. Synthesis and applications. Conventional farming had the strongest adverse consequences for stream condition in our study. In contrast, an integrated management system (aimed at reducing pesticide use, increasing beneficial pest predators and encouraging environmentally responsible soil, water and energy management) proved at least as effective as organic farming and, together with organic farming, can be considered better practice.

Input of terrestrial coarse particulate organic matter (CPOM: leaves and small wood, > > > > >1 mm) adds food resources and habitat to streams. Restoration projects often aim to increase CPOM through riparian planting, but this relies on inputs being retained within the restored reach, and quantitative knowledge of factors influencing retention processes is limited. We investigated retention characteristics under base-flow conditions in 65 stream reaches in pasture, native forests and pine plantations (with and without recent logging) in five contrasting New Zealand regions, using three CPOM analogues, namely 30 cm dowels, soaked ginkgo leaves and waterproof paper triangles. These differed significantly in geometric mean retention distance (S p ), with greatest retention (smallest S p ) of dowels > > > > > ginkgos > > > > > triangles. Retention efficiency was greater in riffles and glides than in pools. Objects trapping CPOM differed between land uses and regions in predictable ways. For example, in relatively unshaded pasture streams, instream and encroaching vegetation riparian was more important, and wood less important, than in forest. Catchment-area-and width-specific S p values were more informative for comparisons amongst sites than deposition velocities (v dep ). Catchmentarea-specific S p values varied 10-fold amongst regions and threefold amongst land uses.Stepwise multiple regression models explained 71-74% of variance in S p for the different CPOM analogues and showed that flow (+ + + + + influence; i.e., S p increases with flow) was the most important factor, followed by submerged wood (− − − − −), with periphyton biomass (− − − − −) also significant in all models. A comparison with results from Spain showed similar relationships between S p of plastic CPOM analogues and discharge under base-flow conditions. Restoring riparian vegetation is expected to ultimately enhance instream CPOM by increasing input and retention by wood. Our findings indicate that this should prioritize small streams if the aim is to optimize the restoration of CPOM influence on instream processes.