1. This study assessed the ecological effects of substituting pasture for forest through changes to benthic macroinvertebrate assemblages. 2. Rapid river assessment protocol (RRAP) scores were determined for eight river stretches, four stretches from each one of the two matrix category, where benthic macroinvertebrates were also sampled. 3. All forest fragment stations were sorted into the natural category, but one, which was categorized as altered; all pasture stations were sorted into the impacted category; there were statistically significant differences in terms of RRAP scores and macroinvertebrate assemblage structures between the two matrix types; 18 taxonomic units contributed the most to this difference; Concerning their water-quality tolerance, seven were considered intolerant, eight tolerant, and one resistant to pollution while two could not be placed into any pollution resistance category; 39.44% of all specimens collected was herbivores, 8.06% was detritivores, 7.92% was predators, and 44.57% of them could not be categorized into a feeding category; forest fragment river stretches had greater taxon richness and relative abundances of pollution-intolerant taxa and predators than pasture river stretches; a positive relationship was found between RRAP score and intolerant taxa abundance. 4. These results may be a consequence of changes to habitat properties related to the matrix substitution, such as greater fine sediment suspension, greater nutrient concentration, and lower availability of food resources; the stress imposed by these changes mainly selects the more resistant taxonomic groups over the more sensitive ones, such as Ephemeroptera, Plecoptera, and Trichoptera; this indicates a reduction in water quality of river stretches without riparian forests.