Elevated atmospheric carbon dioxide (eCO 2 ) has been shown to have a variety of ecosystem-level effects in terrestrial systems, but few studies have examined how eCO 2 might affect aquatic habitats. This limits broad generalizations about the effects of a changing climate across biomes. To broaden this generalization, we used free air CO 2 enrichment to compare effects of eCO 2 (i.e., double ambient $ 720 ppm) relative to ambient CO 2 (aCO 2 $ 360 ppm) on several ecosystem properties and functions in large, outdoor, experimental mesocosms that mimicked shallow sand-bottom prairie streams. In general, we showed that eCO 2 had strong bottom-up effects on stream autotrophs, which moved through the food web and indirectly affected consumer trophic levels. These general effects were likely mediated by differential CO 2 limitation between the eCO 2 and aCO 2 treatments. For example, we found that eCO 2 decreased water-column pH and increased dissolved CO 2 in the mesocosms, reducing CO 2 -limitation at times of intense primary production (PP). At these times, PP of benthic algae was about two times greater in the eCO 2 treatment than aCO 2 treatment. Elevated PP enhanced the rate of carbon assimilation relative to nutrient uptake, which reduced algae quality in the eCO 2 treatment. We predicted that reduced algae quality would negatively affect benthic invertebrates. However, density, biomass and average individual size of benthic invertebrates increased in the eCO 2 treatment relative to aCO 2 treatment. This suggested that total PP was a more important regulator of secondary production than food quality in our experiment. This study broadens generalizations about ecosystem-level effects of a changing climate by providing some of the first evidence that the global increase in atmospheric CO 2 might affect autotrophs and consumers in small stream ecosystems throughout the southern Great Plains and Gulf Coastal slope of North America.

The direct and indirect regulation of primary productivity has been well established in autotrophic‐based ecosystems; however, less is known about the processes affecting decomposers in detrital‐based ecosystems. Because, small headwater, woodland streams are a dominate feature in most ecosystems and are tightly linked to terrestrial detritus, understanding decomposer‐mediated functions in these systems is critical for understanding carbon processes across the landscape. In this light, we conducted a microcosm and mesocosm experiment to test the direct and indirect food web effects on decomposers in small stream ecosystems. The results from the microcosm experiment supported an existing literature, demonstrating that nutrients directly stimulate decomposers and that microbivores directly reduce decomposers. Based on well‐founded food web theory in autotrophic systems, we predicted that fishes from different trophic‐functional guilds would indirectly stimulate decomposers by enhancing dissolved nutrients and by reducing microbivore densities. Our mesocosm experiment partially supported these predictions. Specifically, we found that fishes that consumed mostly terrestrial foods increased decomposers from the bottom–up by enhancing allochthonous nutrient loading into the stream ecosystems. Contrary to our predictions, however, predatory fishes that consume microbivores did not increase decomposers from the top–down. Rather, in streams with the predatory fish species, microbivores increased (rather than decreased) on leaf litter. This may have resulted from an experimental artifact associated with refuge provided by leaf packs. In conclusion, our data demonstrate that decomposers are regulated by similar direct and indirect processes important in autotrophic‐based ecosystems. This provides further evidence that food web processes can regulate leaf decomposition and flux of detrital carbon through ecosystems.

The independent effects of in-stream structure (ISS) and fish foraging on stream properties have been well documented, but few studies have explored the interactive effects of ISS and fishes on streams. Herein, we tested the independent and interactive effects of ISS and a generalist fish (Blacktail shiner, Cyprinella venusta) on suspended organic matter (SOM), benthic algae, invertebrate density and fish growth using experimental mesocosms. We found that Blacktail shiner foraging affected all of the ecosystem properties; however, in some cases, the fish effects differed between mesocosms with and without ISS. Specifically, mesocosms with ISS provided greater surface area for invertebrate colonisation and enhanced food resources for Blacktail shiner. As a result, benthic foraging by Blacktail shiner was reduced in these mesocosms. The reduced benthic foraging in turn enhanced benthic algae and benthic invertebrates via a bottom-up, nutrient excretion pathway. The ISS-dependent effects of fish on these stream properties, however, were only evident at low and intermediate fish densities (1 and 2 fishÁm À2 respectively). This was likely because at the highest fish density (4 fishÁm À2 ) intense fish foraging overrode any mediating effects of ISS. Furthermore, fish growth decreased with fish density because of intraspecific competition, but this negative effect on growth was reduced in mesocosms with ISS because of the increased forage base. However, the positive effect on fish growth was weak and only marginally significant. Our data suggest that fish-mediated effects on streams are context dependent, changing with microhabitat availability (e.g. ISS) and density of the fish population.

No abstract