Conservation genetics of Oreochromis mossambicus across South Africa: foundational knowledge for management of this vulnerable species.

Doctoral Degree. University of KwaZulu-Natal, Pietermaritzburg.

South African freshwater habitats face escalating threats from anthropogenic activities and suboptimal management practices, resulting in a worrisome decline in freshwater taxa. This study addressed a critical knowledge gap concerning the impact of these threats on the genetic diversity and population structure of freshwater fish species in South Africa, focusing on Oreochromis mossambicus. Despite its ecological and economic significance, little is known about the genetic diversity and population structure of both farmed and wild O. mossambicus in the region. Hybridisation with the introduced O. niloticus has further exacerbated the species' vulnerability, leading to its classification as Vulnerable on the IUCN Red List. Establishing baseline genetic data is essential for the effective conservation and management of O. mossambicus populations in South Africa. This research pursued five primary objectives: Firstly, this study assessed the genetic diversity and population structure of wild O. mossambicus in major river catchments across KwaZulu-Natal, Mpumalanga, and Limpopo provinces. Using 14 microsatellite loci, I established baseline genetic data and evaluated if the current water management practices maintain the species' existing genetic structure. Significant genetic differentiation was found among populations, with STRUCTURE analyses revealing 15 geographically correlated genetic clusters. These findings emphasised the role of anthropogenic activities, changes in catchment use, and water management strategies in shaping the genetic structure of O. mossambicus, highlighting the need for conservation-oriented management to preserve existing genetic diversity. Secondly, using 14 microsatellite markers, I determined the genetic diversity and origin of four farmed O. mossambicus populations in KwaZulu-Natal and Mpumalanga provinces, comparing them with wild populations from nearby rivers. The results indicated lower genetic diversity in farmed populations compared to surrounding wild populations. In particular, the uMphafa ponds population exhibited distinctive genetic characteristics, underscoring the need for careful monitoring. This chapter highlighted the potential for using farmed populations from Zini Fish Farm and Fresca Fisheries Farm for selective breeding and broodstock supplementation to maintain genetic diversity in aquaculture practices. Thirdly, this study evaluated the presence of genetic material in farmed and wild O. mossambicus as a potential indication of genetic introgression from introduced O. niloticus and O. aureus using 14 microsatellite loci. Genetic structure analyses revealed evidence of the presence of genetic material and potential introgression primarily between O. mossambicus and O. niloticus in several wild populations across Limpopo, Mpumalanga, and KwaZulu-Natal. Additionally, shared genotypic frequencies between O. aureus and farmed O. mossambicus from uMphafa ponds and Fresca Fisheries Farm were detected. This chapter advocated for stringent measures to preserve the genetic integrity of wild O. mossambicus populations, particularly those showing no signs of introgression by introduced species, to ensure sustainable management of this vulnerable species. Fourthly, this study reviewed the status of COI and 12S rRNA reference libraries for native and introduced freshwater fish in South Africa. An analysis of DNA records available on GenBank and the Barcode of Life Database (BOLD) revealed significant gaps in the records for both markers. Specifically, 34 species, six genera, and zero families of native South African freshwater fish lack COI barcode records, while 86 species, 22 genera, and eight families lack 12S rRNA records. In contrast, non-native fish had complete barcode records for both COI and 12S rRNA. Establishing comprehensive reference libraries for both markers is a crucial first step in developing an eDNA protocol for the non-invasive monitoring of freshwater fish in South Africa. This eDNA protocol may help enhance the effectiveness of monitoring and conservation efforts for threatened species like O. mossambicus and facilitate the early detection and monitoring of invasive species such as O. niloticus. Lastly, the study developed and tested the efficacy of environmental DNA (eDNA) metabarcoding as a non-invasive method for detecting O. mossambicus and the introduced O. niloticus and O. aureus in KwaZulu-Natal. A multi-marker system, incorporating fragments of the cytochrome oxidase I (COI), 12S rRNA, and 16S rRNA gene regions, was employed. The eDNA metabarcoding identified 211 fish-related sequences, including a few sequences of O. mossambicus and O. niloticus, from 481,913 raw reads, demonstrating the method's effectiveness in detecting diverse fish species. However, achieving species-level identification remained challenging, likely due to incomplete reference databases. This chapter highlighted the potential use of eDNA for detecting and monitoring both native and introduced Oreochromis species in South Africa. It emphasised the need to refine the protocol further and design specific primers for more accurate identification. Additionally, this chapter underscored the importance of comprehensive reference libraries and ongoing refinement of eDNA metabarcoding protocols to maximize their potential for freshwater fish monitoring. To enhance conservation efforts, the present study recommends routine genetic monitoring, improved water management practices, expansion of barcode reference libraries, and refinement of eDNA protocols with species-specific primers. Implementing these strategies will aid in better management and preservation of O. mossambicus and other freshwater fish, ensuring their longterm viability and ecological health.