We explored the notion that small canals could be good experimental proxies of streams by documenting physiochemical parameters and macroinvertebrate community development in an aridland irrigation canal. Further, we tested the production-compensation hypothesis between benthic invertebrates and invertebrates in the water column (drift). If the hypothesis held, invertebrates in the drift would be low until the benthic carrying capacity was reached; then organisms in the drift would increase as individuals avoided overcrowded conditions in the substrate. In a small, naturalized freshwater canal, we sampled macroinvertebrates in the substrate and those in the drift once every 2 weeks over 170 days (May–October). We placed macroinvertebrates into functional feeding groups (FFG) and examined these groups along with total density and taxa richness. We found no density-dependent relationship either in FFG or total density between the benthos and invertebrates in the drift. Our negative results might indicate that the stream benthos did not reach carrying capacity (partially affected by adult emergence in the autumn), the invertebrate dynamics in the canal did not adequately represent those occurring in a natural stream, or the duration of our study was too short. However, the invertebrate community in the canal did follow community buildup patterns for small streams reported in the literature, and it also resembled the community in a nearby natural stream. We suggest that naturalized canals could be used as “mesocosms,” mimicking small natural streams, in which highly manipulative experiments could be conducted. In addition, the effects of temporary and permanent canals across the arid western landscape have been understudied and represent a new area of ecological research.