SPATIO-TEMPORAL PATTERNS OF MACROINVERTEBRATE FUNCTIONAL FEEDING GROUPS IN THE SOSIANI-KIPKAREN RIVER, KENYA

Abstract

Different land-use practices in river basins affect the hydrological characteristics, water quality and alter the complex biotic and abiotic processes that govern the functioning macroinvertebrate communities. Studies utilizing macroinvertebrate functional feeding groups to assess the ecological condition of streams in Kenya are scarce. This study set out to investigate changes in total abundance, taxon richness and biomass of macroinvertebrate functional feeding groups (FFGs) in response to land-use change and assess the suitability of abundance- versus biomass-based metrics as surrogates of ecosystems attributes and ecological integrity in the Sosiani-Kipkaren River in western Kenya. A total of 21 sites were sampled during the wet season (July-August 2018) and 14 of the same sites during the dry season (February-March 2019) along a land-use gradient. Four land-use categories; Forest (n = 5), Mixed (n = 6), Agriculture (n = 6) and Urban (n = 4) were sampled. Macroinvertebrates were collected seasonally, identified, assigned to functional feeding groups and used to derive the five metrics utilized as surrogates of ecosystem attributes. Water and habitat quality variables were also measured seasonally and their data used to correlate with the macroinvertebrates assemblages. There were significant (p < 0.05) spatial variation in habitat quality, organic matter standing stocks, electrical conductivity, temperature, sodium, potassium and nutrient concentrations across land-uses, with forested sites recording lowest values. Forest land-use sites had good habitat quality (QHEI Score of 56) while the rest were marginal. Macroinvertebrates total abundance was significantly higher (R-statistic = 0.30, p < 0.007) during the wet season (35,827 individuals) than dry season (7,652 individuals). Responses in macroinvertebrates differed among functional feeding groups, with biomass-based metrics responding more strongly to land-use change while richness-based metrics being the least predictive, indicating replacement of taxa within functional feeding groups across land-uses. Higher shredder abundance, biomass and richness were recorded in forested streams and lowest in urban streams during both seasons. Collector-gatherers dominated agricultural streams during both the wet and dry seasons, while predators dominated urban streams. Scrapers responded positively to increased nutrient levels and open canopy in mixed and agricultural streams. Abundance-based metrics were better predictors of ecosystem attributes, and displayed greater response to changes in stream size than biomass-based metrics. There was incongruence between abundance- and biomass-based indicators for Production/Respiration and coarse particulate organic matter to the fine one. Catchment land-use did not influence metric performance, suggesting that reach scale influences played a predominant role in structuring communities and determining ecosystem functioning. Even though there is need for more studies to refine the metrics used and establish thresholds for the various attributes, this study established that there is indeed potency of functional feeding groups approach to be used as a means of assessing both the ecological integrity and functioning of Afrotropical streams.