Biodiversity and ecological structures of an African subtropical river and associated floodplain pans

Abstract

Although only recently added to the many existing threats that freshwater ecosystems already face, climate change is one of the most prominent to inland waters in the 21st century. Through extended periods of drought and extreme flooding events, climate change is predicted to severely affect freshwater ecosystems, particularly those that are already exploited by humans. Unfortunately, it is still not well understood exactly how climate change will affect these ecosystems, particularly in semi-arid countries such as South Africa. There is also still a conflict of interest in terms of the needs of aquatic ecosystems and the needs of socio-economic development, particularly in a developing country such as South Africa where growth of industry is necessary for economic growth. Increased development however, also leads to additional pollution of the environment and loss of biodiversity and ecological integrity. Thus, making the conservation and proper management of freshwater ecosystems more difficult. Although progress has been made in South Africa, through the National Water Act of 1998, the protection and restoration of riverine habitats is still somewhat limited due to a continued poor understanding of these ecosystems and their natural functions. Through construction of dams, irrigation schemes and mismanagement, particularly of flood releases, large floodplain ecosystems in Africa are under threat and the lower Phongolo River and associated floodplain is no exception. Located in northern KwaZulu-Natal, the Phongolo floodplain is South Africa’s largest inland floodplain ecosystem and of great economic and socio-ecologic importance. Unfortunately, this ecosystem has been regulated by the Pongolapoort Dam since the 1970’s and the upper reaches of the river utilized for anthropogenic activities (mostly irrigation) since at least the 1950’s. The river and associated floodplain stretches approximately 80 km downstream of the dam, through the Ndumo Game Reserve (NGR) which is the only protected section of the floodplain, and into Mozambique. The floodplain and NGR are considered a biodiversity hotspot due to the large number of bird, frog and fish species (among other fauna) that are found in this region, many of which are highly dependent on the floodplain and the services it provides. Despite all this, flood releases from the dam are mismanaged, focusing mostly on providing water for socio-economic demands and not the ecological needs of the floodplain. The Phongolo River ecosystem was quite extensively studied in the 1970’s and 1980’s after which little research took place until an extensive Water Research Commission research project took place between 2012 and 2014. Unfortunately, a supra-seasonal drought event occurred in South Africa shortly after this project was completed, and rainfall was severely reduced across the country, including the Phongolo region. The last flood release from the Pongolapoort Dam occurred in December 2014 after which only a baseflow has been released in accordance with the “compensation flows” agreement with Mozambique. No documentation is presently available on the possible effects low water flow and drought may have on this ecosystem. In addition, concerns have previously been raised regarding the flooding regime and anthropogenic activities surrounding the area. Thus, the overarching goal of this study was to determine the present environmental state of the lower Phongolo River and associated floodplain ecosystem and assess the possible effects drought may have on such an ecosystem. In order to achieve this aim, several surveys were conducted between 2016 and 2017 along various types of aquatic habitats found on the floodplain during which time both abiotic and biotic factors were assessed. For the abiotic assessment, two types of data analysis took place. First, historic water quality data (from the 1970’s to present) were obtained from three South African Department of Water and Sanitation (DWS) monitoring sites (gauging stations) within the Phongolo River catchment in order to assess long-term water quality of the Phongolo River, both upstream and downstream of the Pongolapoort Dam and determine potential changes that have occurred since the 1970’s. Secondly, additional water samples were collected from three sites on the Phongolo River during the present study (two sites outside NGR and one within) in order to assess the present water quality of the lower Phongolo River during the supra-seasonal drought. With the use of multivariate statistical analyses as well as the Physico-chemical Driver Assessment Index (PAI) various in situ and chemical water variables, including salinity and organic nutrients, were assessed. The key findings of the long-term analysis were that water quality of the upper Phongolo River is affected by numerous anthropogenic activities before it enters the Pongolapoort Dam and has degraded since the 1970’s. The primary cause of this is increased organic nutrients (NO3/NO2 and PO43-) and salinity (conductivity and K) from the large irrigation schemes found in the upper reaches of the river that have grown over the past 40 years. It was further found that the Pongolapoort Dam appears to be trapping nutrient rich sediments which has led to nutrient depleted water entering the lower Phongolo River. This, however, increases as the river flows downstream into NGR through naturally saline groundwater seeping into the river as well as nutrient rich soils and fertilizers. The supra-seasonal drought did not appear to have any significant effect on water quality as the PAI remained similar to that of before the drought. For the biotic assessment various aspects were analysed. The first of these was an in situ field mesocosm experiment that aimed to determine whether length of hydrological connectivity between the Phongolo River and a floodplain wetland would affect aquatic invertebrate communities. Nine mesocosms were set up within NGR next to a floodplain pool selected as the model for the assessment. These mesocosms were separated into three groups for various experimental treatments and water quality and aquatic invertebrate data collected over a five-week period. To prevent physical reduction of aquatic invertebrates these biota were identified and counted as live specimens and placed back into each mesocosm on the day of sampling. The key findings from this assessment were that hydrological connectivity does not affect aquatic invertebrate communities and this was related to the adaptations these biota possess that allow them to resist changes in water quality and water levels. Water quality also changed in a similar manner between the three groups suggesting that a floodplain wetland would need to be flushed with a great deal of water in a short period of time in order to prevent decreasing water quality. The next biotic assessment was to determine current aquatic invertebrate community structure and aquatic food webs and fish consumer diets, with the use of stable isotope analysis (SIA), in three ecologically important aquatic habitats found in the lower Phongolo region. The first habitat was an ecologically important and unique temporary wetland located within NGR that was experiencing desynchronised wet/dry cycles. This site is ecologically important as it is the only temporary pool within the Ndumo region that has ever been found to have populations of the annual spotted killifish Nothobranchius orthonotus. This site was assessed to determine if certain ecological processes were being disrupted by the out of season drying event induced by the supra-seasonal drought and to determine the diet of the spotted killifish with the use of SIA. Aquatic invertebrate communities and water for select chemical analyses were collected over four surveys, twice before and twice after the site had dried out, and data for food web and dietary analysis collected over two surveys, once before and once after the site had dried out. The aquatic invertebrate community analysis indicated that biological processes remained intact and that changes in the community structure followed the natural changes present in temporary wetlands. This included a decrease in biodiversity as the pool dried up and water quality decreased (largely due to an increase in salinity) and increases after the site had received fresh, low conductivity rainwater. Stable isotope analysis indicated that there was nitrogen enrichment in the food web as all collected food web components had increased nitrogen signatures (δ15N) after the site had received water. This indicates towards nutrient inputs from anthropogenic activities occurring outside NGR. It was further found that the spotted killifish, which has always exclusively been considered a predator, also feeds on a large amount of plant-based food sources which has never been documented before and are able to shift their diets to include a greater percentage of low-nutrient food sources when necessary. Although there were no apparent long-term adverse effects at the time of analysis, continued nutrient inputs may eventually have negative implications for this aquatic habitat. In addition, an extended shift to low-nutrient plant food sources by the killifish may be the first signs of effects that drought and climate change may have through desynchronization of consumers with their preferential, high-nutrient food sources. The second habitat that was selected is the ecologically important large saline lake located within NGR. This site is the only naturally saline lake found in the region and was selected as a study site as the water levels were severely reduced during the supra-seasonal drought and water became highly saline as a result of this. In addition, fish for SIA analysis and aquatic invertebrates for diversity analysis were previously sampled at this site in pre-drought conditions (2010 and 2014 respectively) thus providing the present study with data from before the supra-seasonal drought. The site was assessed to determine whether the supra-seasonal drought and accompanying increased salinity had an effect on the aquatic invertebrate community, food web and fish consumer diets present in this site by comparing to the data previously collected. It was found that aquatic invertebrate diversity and abundance was severely affected during the peak of supra-seasonal drought in 2016 with only a few highly tolerant taxa able to survive. Through rainfall and small floods from the Usuthu River at the end of 2016 and beginning of 2017, water became less saline and several fish species and aquatic invertebrate taxa previously found started to return to the lake. Ecosystem resilience was evident at this site since aquatic invertebrate diversity and abundance was able to recover rather quickly after the site received some fresh water and the SIA indicated that fish consumers were able to adapt to the changes in the habitat by switching to generalist feeding behaviour in 2017 compared to a more specialised feeding behaviour evident in 2010. The final habitat was the lower Phongolo River of which three sections were studied, two of these located in the communal area of Ndumo among anthropogenic activities and one just as the river flows into NGR. In addition to many human activities that take place the lower Phongolo River, the supra-seasonal drought that occurred in the region prevented the release of large floods from the Pongolapoort Dam and only a baseflow was released during the study period. These sites were therefore assessed to determine whether the low baseflow as well as the anthropogenic activities had any effect on the ecology of the river. This was accomplished by studying aquatic invertebrate community structure, aquatic food webs and fish consumer diets of the river. To determine if there was any change in the aquatic invertebrate diversity and community structure results from the present study were compared to those of previous studies on the river and floodplain in 2012 and 2014 respectively. It was found that the supra-seasonal drought did indeed affect the aquatic invertebrate community structure through a loss of rheophilous and pollutant sensitive aquatic invertebrate taxa. However, no loss of the total aquatic invertebrate diversity occurred due to a replacement with more tolerant aquatic invertebrate taxa and frequent taxonomic turnover. In addition, the presence and high abundance of pollutant tolerant aquatic invertebrate taxa within NGR indicated that anthropogenic activities in the surrounding region were most likely affecting the Phongolo River, particularly the section located within the NGR since there was a general increase in nitrogen in the aquatic food webs at all sites. Even though there was nitrogen enrichment in the aquatic food webs as well as a shift towards production of inorganic carbon by autochtonous sources, the food webs remained stable throughout the study period. This suggests that the supra-seasonal drought and anthropogenic activities are most likely not affecting the aquatic food webs. The drought as well as anthropogenic activities did however appear to affect the fish consumers as their diets consisted largely of generalist and opportunistic feeding strategies often found in disturbed ecosystems. This study found several ecological changes and effects along the lower Phongolo River and its associated wetlands caused by the supra-seasonal drought with its accompanying effects and numerous anthropogenic activities that occur in the region. However, presently these changes and effects are not severe, and the ecosystem appears to be resilient and adaptable with the ability to rapidly recover from disturbances, such as absence of flood releases, anthropogenic activities and drought. It was further found that the lower Phongolo is a resilient ecosystem able to adapt to and recover from disturbances quite rapidly. However, it is quite likely that even this resilient ecosystem will be severely affected if the drought were to continue over an extended period or anthropogenic inputs were to increase, as has been suggested by future predicted climate change and continuously growing human population of this region. It is therefore still imperative that appropriate management and conservation of the region take place. This study also recommends that there is continued monitoring of the lower Phongolo River and ecosystem in order to confirm that recovery from the drought is indeed taking place and that no long-term effects occurred during the supra-seasonal drought.