Light, nutrient availability, and flow are strong factors controlling the elemental composition and biomass of epilithon in temperate stream ecosystems. However, comparatively little is known about these relationships in tropical streams. We investigated how gradients of light and nutrient availability, seasonality, and habitat influenced epilithon biomass, chlorophyll a, and nutrient ratios in montane streams of Trinidad, West Indies. We sampled 4 focal tributaries of a single river, 2 of which had canopies experimentally thinned, every other month over a 2-y period to observe temporal dynamics and light effects on epilithon. We also sampled 18 sites across Trinidad's Northern Range Mountains once each in a wet and dry season to examine the effects of naturally occurring differences in light and dissolved nutrient availability on epilithic characteristics. We found greater chlorophyll a concentrations in habitats with greater light availability, but the effect of light on epilithon stoichiometry differed between the site-survey and focal-tributary data. In general, epilithic C:nutrient ratios decreased with increasing dissolved nutrient concentrations, but relationships between nutrient availability and biomass probably were obscured by naturally high dissolved N and P concentrations in many of the streams. Season and habitat type had profound effects on epilithon variables. Biomass and % C generally decreased in riffles and under wetseason conditions. These results suggest multiple controls for the quantity and quality of stream epilithon and have important implications for in-stream consumers.

Decades of ecological study have demonstrated the importance of top‐down and bottom‐up controls on food webs, yet few studies within this context have quantified the magnitude of energy and material fluxes at the whole‐ecosystem scale. We examined top‐down and bottom‐up effects on food web fluxes using a field experiment that manipulated the presence of a consumer, the Trinidadian guppy Poecilia reticulata, and the production of basal resources by thinning the riparian forest canopy to increase incident light. To gauge the effects of these reach‐scale manipulations on food web fluxes, we used a nitrogen (15N) stable isotope tracer to compare basal resource treatments (thinned canopy vs. control) and consumer treatments (guppy introduction vs. control). The thinned canopy stream had higher primary production than the natural canopy control, leading to increased N fluxes to invertebrates that feed on benthic biofilms (grazers), fine benthic organic matter (collector‐gatherers), and organic particles suspended in the water column (filter feeders). Stream reaches with guppies also had higher primary productivity and higher N fluxes to grazers and filter feeders. In contrast, N fluxes to collector‐gatherers were reduced in guppy introduction reaches relative to upstream controls. N fluxes to leaf‐shredding invertebrates, predatory invertebrates, and the other fish species present (Hart's killifish, Anablepsoides hartii) did not differ across light or guppy treatments, suggesting that effects on detritus‐based linkages and upper trophic levels were not as strong. Effect sizes of guppy and canopy treatments on N flux rates were similar for most taxa, though guppy effects were the strongest for filter feeding invertebrates while canopy effects were the strongest for collector‐gatherer invertebrates. Combined, these results extend previous knowledge about top‐down and bottom‐up controls on ecosystems by providing experimental, reach‐scale evidence that both pathways can act simultaneously and have equally strong influence on nutrient fluxes from inorganic pools through primary consumers.

The influence of specific stressors, such as nutrient enrichment and physical habitat degradation, on biotic integrity requires further attention in Midwestern streams. We sampled 53 streams throughout Illinois and examined relationships between macroinvertebrate community structure and numerous physicochemical parameters. Streams were clustered into four major groups based on taxa dissimilarity. Habitat quality and dissolved nutrients were responsible for separating the major groups in a nonmetric multidimensional scaling ordination. Furthermore, the alignment of environmental factors in the ordination suggested there was a habitat quality-nutrient concentration gradient such that streams with high-quality habitats usually had low concentrations of nutrients. Discrimination by community measures further validated the major stream groups and indicated that forested streams had generally higher biological integrity than agricultural streams, which in turn had greater integrity than urban streams. Our results demonstrate that physical habitat degradation and nutrient pollution are important and often confounded determinants of biotic integrity in Illinois streams. In addition, we suggest that management of Midwestern streams could benefit from further implementation of multivariate data exploration and stream classification techniques.

An understanding of the distribution patterns of organisms and the underlying factors is a fundamental goal of ecology. One commonly applied approach to visualize these is the analysis of occupancy-frequency patterns. We used data sets describing stream insect distributions from different regions of North America to analyze occupancy-frequency patterns and assess the effects of spatial scale, sampling intensity, and taxonomic resolution on these patterns. Distributions were dominated by satellite taxa (those occurring in or=90% of sites) determined the overall modality of occupancy-frequency patterns. The proportions of satellite taxa increased with spatial scale and showed positive relationships with sampling intensity (r2=0.74-0.96). Furthermore, analyses of data sets from New York (USA) showed that generic-level assessments underestimated the satellite class and occasionally shifted occupancy-frequency distributions from unimodal to bimodal. Our results indicate that, regardless of species- or generic-level taxonomy, stream insect communities are characterized by satellite species and that the proportion of satellite species increases with spatial scale and sampling intensity. Thus, niche-based models of occupancy-frequency patterns better characterize stream insect communities than metapopulation models such as the core-satellite species hypothesis.