Thesis (Ph.D. (Zoology))--North-West University, Potchefstroom Campus, 2005

Ph.D. / The current study evaluated a bio-monitoring technique developed in the USA by Adams, Brown and Goede, 1993. This project was sponsored by the Department of Water Affairs and Forestry (DWAF), to enable testing of the Health Assessment Index (HAI) under South African conditions. Testing took place in the Olifants River system, one of the most polluted river systems . in South Africa. Initially two river points were tested using Oreochromis mossambicus (Robinson, 1996), Clarias gariepinus (Marx, 1996) and Labeo rosae (Luus-Powell, 1997). The current study re-tested the HAI at the same two sample sites, namely Mamba and Balule in the Kruger National Park, using 0. mossambicus and C. gariepinus respectively. Two additional sites were tested in the upper catchment area, namely Loskop Dam and Bronkhorstspruit Dam. The current study further enabled the comparison of HAI results collected during drought and flood conditions. Results obtained after deployment of the HAI were corroborated using chemical analysis of water, sediment and biota. Water and sediment analysis was carried out by the Institute for Water Quality Studies using standard techniques. Bio-accumulation of aluminium, copper, iron, lead, manganese, nickel, strontium and zinc was assessed in the gills, liver, skin and muscle tissue of sample fish using standard Atomic Absorption Spectrometry techniques. Modifications made to the original HAI involved the inclusion of variable ranking in the assessment of fish parasites, with endo- and ectoparasites evaluated separately. Testing of this parasite hypothesis lead to the development of a Parasite Index component to the HAI. Assessment of water, sediment and fish tissue determined that the Olifants River system is indeed exposed to macro and heavy metal pollutants, which negatively affect aquatic health. Constituents posing the greatest threat are chlorides, fluorides, phosphates, total dissolved solids, copper and iron concentrations. Testing the HAI and parasite hypothesis using C. gariepinus, provided the most meaningful results. During testing of the parasite hypothesis both endo- and ectoparasite numbers conformed to the suggested idea that higher endoparasite numbers will occur at highly impacted areas, whereby ectoparasite numbers will be low. This was particularly evident in the lower catchment area, whereby comparisons between drought and flood conditions were carried out. Subsequent decreases in water quality directly after the flood were noted using water and sediment analysis. This observation reflects the results gathered using the HAT and during testing of the parasite hypothesis at all four sample sites. During statistical analysis of the HAI, using logistic regression analysis, parasite numbers, more specifically endoparasite numbers, were the most indicative of fish health. Environmental stressors (flood conditions) result in immunological responses observed in fish, and are reflected statistically using the HAI as changes in WBC %. It is suggested that endoparasites and WBC % provide the best overall assessment of fish condition. These variables should thus not be eliminated, in order to streamline the HAI evaluation procedures. Testing of this bio-monitoring technique under South African conditions provided meaningful results. This indicates that the HAI can be used to assess water quality, with existing water monitoring programmes further benefiting from its incorporation.

PhD / An important component of fish health is an optimally functioning reproductive system. The Luvuvhu River Catchment in the Limpopo Province, South Africa, is a tropical, high-risk malaria area where 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane (DDT), an endocrine disrupting chemical (EDC), has been used annually since 1945 as a malaria vector control. DDT is known to affect testes morphology and motility of fish sperm. As such, testicular histology and sperm motility (kinematic) parameters were studied as reproductive indicators of the reproductive capacity for two wild, indigenous fish species (Oreochromis mossambicus and Clarias gariepinus) from the currently DDT sprayed area. Three field studies were carried out over two years (2007 – 2008), including two high flow (HF) periods and one low flow (LF) period [HF 1 (March 07), LF (October 07), HF 2 (February 08)]. Both species were sampled from three sites on the Luvuvhu River for testicular histology and computer assisted sperm analysis (CASA), during all three field studies. The sites included a reference site outside the DDT sprayed area, Albasini Dam (AD), and two exposed sites within the DDT sprayed area, Xikundu Weir (XW) and Nandoni Dam (ND). CASA, based on open-source software, was used for the first time in South Africa to assess sperm kinematic parameters of indigenous fish species in field conditions. These included percent motile sperm (% MOT), curvilinear velocity (VCL μm s-1), velocity of an average path (VAP μm s-1), straight line velocity (VSL μm s-1), linearity (LIN %), progression (PROG μm), and average efficiency (AVE. EFF.). Water and sediment samples were collected during all field studies from the three sites for metal and EDC analysis. Controlled laboratory studies were also carried out on the sperm of both species, externally sourced from aquaculture farms equipped to breed and raise fish in toxicant free water. The laboratory studies involved in vitro exposure of spermatozoa to two different, but environmentally relevant, concentrations of both DDT (DDT 1: 0.27 μg L-1; DDT 2: 0.5 μg L-1) and 1,1-dihloro-2,2-bis(p-chlorophenyl)ethylene (DDE) (DDE 1: 0.11 μg L-1; DDE 2: 1.0 μg L-1) with the aim to provide data to support the possible outcomes found in the field studies using CASA. Furthermore, peroxidation of sperm lipids was assayed by production of malondialdehyde (MDA) after in vitro exposure of spermatozoa to DDT and DDE. DDT and its metabolites were found in varying concentrations in the water from all three sites (0.1 μg L-1 – 1.2 μg L-1). Levels of dieldrin (3.5 μg L-1) and lindane (9.4 μg L-1) residues were also found at XW in HF 2. The histological results revealed alterations to testis tissue of both species at all three sites. The testes were assessed through the identification of alterations and an organ index was calculated: Testes Index (IT). The index is indicative of the histological response in the respective tissue type. O. mossambicus at XW had the highest mean IT value during LF (7.45 ± 5.73) and for all field studies combined (5.47 ± 4.63), primarily due to the occurrence of testicular oocytes (intersex), where the frequency of prevalence was 72.73% and 58.82% respectively. These results were statistically higher than the laboratory control (C) group. The CASA results showed statistical differences primarily for O. mossambicus, where motility parameters were lower at XW when compared to AD. Laboratory exposures found a decrease in sperm motility (% MOT) between the control (C) group and the DDT 1, DDE 1 and DDE 2 exposed groups for C. gariepinus. No significant differences were seen for lipid peroxidation (MDA). On the other hand, no significant differences were seen in CASA parameters between the control and exposed laboratory groups for O. mossambicus, but there was an increase in MDA production from the control to the DDT 1 exposure group.

Freshwater fish species have been introduced into freshwater systems around the world, primarily for aquaculture, ornamental fish trade and sport fishing. Their introduction into estuarine systems is uncommon however, instances do occur and their establishment success and impacts on these estuarine systems is not well documented. The extent of invasion by four freshwater fishes in a RAMSAR listed estuarine system, the Wilderness Lakes, Western Cape was investigated. This thesis determined the relative abundance and distribution of alien fishes in relation to the native fish biota, their establishment success in the system, what factors inhibited their establishment and whether the introduction of alien fishes negatively impacted the native fish biota. The distribution and abundance of fishes were assessed primarily using fyke nets, seine nets and gill nets in each of the lakes, interconnecting channels and the Touw Estuary. The fish fauna was made up of euryhaline marine species comprising 46%, native estuarine species comprising 18%, catadromous species comprising 7% and freshwater alien species comprising 29% of the total biomass sampled. A total of 26 species were sampled in the system, three of which were considered alien; Oreochromis mossambicus, Gambusia affinis and Cyprinus carpio, and Micropterus salmoides were not sampled but confirmed in the system. Establishment success was determined by evidence of: a sustainable breeding population, a wide distribution, abundant in the sampling area, and all size classes of fish sampled. Gambusia affinis and Oreochromis mossambicus have been recorded in the system for a minimum of 13 and 26 years respectively. They were widely distributed and highly abundant and are established in the system. Micropterus salmoides was first recorded in the system in 1985 but abundances have remained low with fish appearing to be limited to Island Lake and Langvlei. Reproduction appeared to be limited by higher salinity and these factors indicated that this was a casual species which relies on repeated introductions for population maintenance. Cyprinus carpio spawned successfully in 2010 and was widely distributed but abundances were still low with a total of 15 fish being sampled throughout the system, and was thus in the establishing phase. As these are considered freshwater alien species, the physico-chemical parameters in the estuarine environment inhibiting the establishment success of the alien fishes were investigated. Gambusia affinis and O. mossambicus were not limited by the physical environment, and while O. mossambicus cannot tolerate temperatures below 11⁰C, temperatures only dropped below its tolerance for a total of two days between February 2010 and February 2011. Cyprinus carpio and M. salmoides were restricted by salinity in Rondevlei and Langvlei but could tolerate salinity in Island Lake and the Touw Estuary during closed mouth phases. While adults appeared to tolerate the salinity in the system, egg and larval development could be affected thus reducing the viability of the population. The abundance of alien fishes did not negatively impact the abundance of native fish species. The interactions between the native and alien fish biota tended towards biotic acceptance where, as alien abundance increased so did native fish abundance. The impacts of the alien fishes on the native fish biota were assessed by comparing the fish community from a study completed in 1985 to the findings of this study. From these two studies there were no apparent negative impacts on the native fish biota and the fish community composition would most likely be structured by estuarine mouth opening events.