Agricultural disturbance affects taxonomic and functional diversity of Afrotropical macroinvertebrate composition in a South African river system

Developing species-level biomonitoring tools to monitor riverine systems threatened by anthropogenic pollution, including local agricultural activities in the Afrotropical region, remain a critical challenge. Here we explored the utility of taxonomic-based (diversity, richness, and composition) as well as functional-based (functional diversity) indices to examine the effects of agricultural disturbance on macroinvertebrate communities in the Kat River, Eastern Cape Province of South Africa. We collected physicochemical parameters and macroinvertebrates from eight sites delineated into four land-use categories (highly impacted, HIC; impacted category, IC; moderately impacted, MIC and least impacted, LIC) using agricultural land cover. We recorded 70 macroinvertebrate taxa belonging to 49 families and 48 genera in the Kat River. Redundancy analysis (RDA) and Pearson correlation analysis revealed that species of Lymnaeidae, Belostomatidae, Planorbidae and Libellulidae families and class Oligochaeta were tolerant to agricultural disturbance, as they were dominant in the highly impacted sites and were significantly associated with high salinity, temperature, total dissolved solids (TDS), flow velocity and nutrients. Conversely, species of Baetidae, Caenidae and Potamonautidae were negatively associated with the highly impacted sites and high salinity, temperature, and nutrients. On the other hand, taxonomic indices showed more sensitivity to indicators of agricultural pollution than functional indices, with taxon richness, Shannon index, Simpson's index and Margalef's index declining significantly in the highly disturbed sites (p < 0.05). They were negatively associated with high electrical conductivity, large river width, and high nitrite and nitrate concentrations; hence they were identified as indicator metrics sensitive to agricultural pollution. Overall, our study revealed that agricultural disturbance could differentially affect the structure and function of macroinvertebrates, and indicator taxonomic and functional indices were identified for long-term monitoring of rivers that drain agricultural landscapes.