Abstract
Streams and rivers are globally threatened ecosystems because of increasing levels of exploitation, habitat degradation and other anthropogenic pressures. In the Lake Victoria Basin (LVB) in East Africa, these threats are mostly caused by unsustainable land use; however, the monitoring of ecological integrity of river systems has been hampered by a lack of locally developed indices. This study assessed the health of four rivers (Nzoia, Nyando, Sondu–Miriu and Mara) on the Kenyan side of the LVB using physicochemical water quality parameters and a fish-based index of biotic integrity (IBI). Fish tolerance ranking was derived from principal component analysis of water quality parameters, and the concept of niche breadth (NB). The relationship between fish species and water quality parameters was examined with canonical correspondence analysis, whereas community metrics and stressors were evaluated through Pearson network correlation analysis. Fish species richness, trophic structures, taxonomic composition and species tolerance were used to generate the metrics for fish-based IBI. NB showed that most of the fish species were moderately tolerant to poor water. Moderately tolerant and intolerant fish species were negatively correlated with a high level of organic loading in the Mara River. Fish-based IBI scores for the rivers ranged from 26 to 34, with Sondu–Miriu scoring the lowest. Our results show that the cumulative effect of stressors can adequately rank fish species tolerance according to the disturbance gradients and further develop regional metrics to assess river health. Despite the fact that fish communities are declining, continual management and enforcement of environmental regulations are important, with conservation and management of headwaters and low-order streams being essential while they are still species rich.
Highlights
River catchments are some of the most vulnerable ecosystems through being increasingly exposed to multiple anthropogenic stressors, including habitat degradation, flow alteration, increased water demand, urbanization and agricultural intensification (Dudgeon et al, 2006; Mamun and An, 2020)
Riverine fish species richness and composition in the Lake Victoria Basin (LVB) have declined in the past decades in response to increasing complexity and multiple stressors in the catchments of many rivers
Our results demonstrate that the cumulative effect of stressors can adequately rank fish species tolerance to disturbance gradients and help to further develop regional metrics to assess and monitor river health
Summary
River catchments are some of the most vulnerable ecosystems through being increasingly exposed to multiple anthropogenic stressors, including habitat degradation, flow alteration, increased water demand, urbanization and agricultural intensification (Dudgeon et al, 2006; Mamun and An, 2020). There are many aquatic organisms that can be included in the evaluation of river health (Herman and Nejadhashemi, 2015), regional indices have widely focused on macroinvertebrates (Dickens and Graham, 2002; Thirion, 2007; Masese et al, 2013, 2020c) that are rarely identified to species level and have some inaccuracy due to ecological and physiographical diversity between regions (Hering et al, 2010) Despite this constraint, these approaches are applied to studies within the Afrotropical region; the identification keys and indices are typically developed elsewhere. The Fish Response Assessment Index that was developed more than a decade ago (Kleynhans, 2007) has not been widely adopted in the Afrotropical region, yet fish communities have significantly declined
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
