AUTHOR=Agboola Olalekan A. , Downs Colleen T. , O’Brien Gordon TITLE=A Multivariate Approach to the Selection and Validation of Reference Conditions in KwaZulu-Natal Rivers, South Africa JOURNAL=Frontiers in Environmental Science VOLUME=8 YEAR=2020 URL=https://www.frontiersin.org/journals/environmental-science/articles/10.3389/fenvs.2020.584923 DOI=10.3389/fenvs.2020.584923 ISSN=2296-665X ABSTRACT=

The use of reference conditions is essential to the monitoring and management of aquatic ecosystems. We examined existing and potential reference sites through historical data, maps, and field data collected from river sites in KwaZulu-Natal (KZN), South Africa. In our study, we applied nine criteria that best reflect the characteristics of South African rivers on 24 a priori selected reference sites. These nine criteria comprised of catchment conditions (flow modification and natural landscape) and site-specific attributes (water quality, human disturbance, river channel, water abstraction, riparian vegetation, riparian zone modification, and instream habitat quality). The a priori selected reference sites were subjected to validation using multivariate methods, such as analysis of similarities (ANOSIM), similarity percentages (SIMPER), and non-parametric multidimensional scaling (MDS) based on the macroinvertebrate fauna by applying a SASS5 threshold considered to be an indicator of undisturbed sites in South African rivers. We identified differences in the macroinvertebrate assemblages of the reference conditions for each river group based on their ecoregions, geomorphology and seasonal variations. Ecoregions and river geomorphology were better in the grouping of sites with similar reference conditions than the seasons. Our findings indicated that all of the selected sites could be considered as valid reference sites; however, caution should be taken in applying this method to lowland rivers because of their noticeable seasonal variability and habitat instability which tend to alter their reference states. We recommend that a type-specific reference condition be developed for lowland rivers. Also, statistical validation of reference conditions should be a continuous process in river biomonitoring.