Time lags, multiple sources, mixing and more: the complications of conducting long-term research on rivers

Rivers provide a wide range of services important to human society, from purely aesthetic (top) to utilitarian (above) (Picture: Mitzi du Plessis) SAEON technician Thabo Mohlala has been sampling water quality at 10 sites in the lower Olifants River catchment since July 2009 SAEON has begun to monitor the abundance of river-dependent birds on a stretch of the Selati River (Picture: Mitzi du Plessis)

The biodiversity of freshwater ecosystems (rivers, wetlands and estuaries) is declining faster than any other type of ecosystem.

In South Africa, contamination of rivers with man-made pollutants has occurred for decades, with extensive and increasing agricultural, industrial, mining and residential development within the catchments of all our major rivers. The resulting impacts on aquatic biodiversity have degraded a range of related ecosystem services, from aesthetically based services (such as recreational fishing and swimming) to more critical services involving the provision of clean water (hundreds of thousands of rural South Africans still rely on river water as a source of drinking water and for washing clothes and themselves).

Coverage of water pollution issues in the national media has increased markedly over the past few years - some experts claim that there is a “water crisis” in South Africa - and there is hardly a need for scientists to inform the public about the problem. The causes of declining water quality are also well known, and often widely publicised. Legislation to prevent water pollution is well established. So why then the need for scientific research, of the type SAEON is involved in, on the topic?

There are a number of reasons why intensive, long-term environmental research is required to improve the management of freshwater resources in South Africa.

Establishing how pollution affects aquatic ecosystems

Firstly, while the direct effects of increasing water pollution for human well-being are reasonably well understood, it is not clear how this pollution affects aquatic ecosystems, and how those effects then translate into indirect negative effects for society. Long-term research is needed to address this question, as the impact of pollution can take many years, or longer, to manifest, and occurs via very complex ecological processes.

Secondly, the regulatory side of water management could also benefit from such research. When companies or individuals pollute a river to the extent that they violate the law, the Department of Water Affairs must punish them to discourage further transgressions. Such punishment typically involves a large fine, the magnitude of which is determined by the extent of the damage caused by the pollution - the so-called “polluter must pay” principle.

In cases where water is needed for human consumption within the stretch of a river that has been polluted, the financial cost of the damage is relatively easy to determine. But how does one calculate the financial costs of damage to ecosystems, and related degradation of ecosystem services? Financial values of ecosystems services can be estimated, but the real challenge lies in clearly demonstrating a link between pollution, changes in aquatic ecology and degradation of related ecosystem services.

Finally, what about cases where pollution is insidious, occurring in small amounts over long periods, and with significant impacts, to nature or people, only occurring many years or even decades later? The on-going discharge of untreated or partially treated sewerage is a case in point. How does a regulatory body punish transgressors in these cases? Often they are unable to do so as the occurrence of small levels of pollution is very difficult to detect, and the long-term effects even harder to determine. When there are multiple sources of this type of pollution, from multiple transgressors within the same stretch of a river, then prosecution is even more difficult. Research is needed to find methods to better detect such pollution, and tease out the sources, and consequences, of multiple types and sources of pollution occurring in close proximity.

Much still needs to be learnt about the ways in which aquatic ecosystems accumulate, process and transport such pollution, so that its broader impact can be more accurately determined. This would also enable regulatory bodies to determine more appropriate fines for transgressors and defend these with confidence in a courtroom.

Measuring water quality

The Lower Olifants River monitoring project conducted by the SAEON Ndlovu Node, is an attempt at providing some of the research results needed. SAEON researchers use the abundance of various types of macro-invertebrates (insects and other bugs that live in the water) as the primary measure of water quality, following the well-established SASS methodology (a standardised system used throughout the world and customised for South Africa following rigorous testing in the 1990s and 2000s).

The advantage of this method is that it doubles as an indicator of the biodiversity of the aquatic ecosystem. Thabo Mohlala, a SAEON technician based in Phalaborwa, has been sampling SASS, and other measures of water quality, at 10 sites in the lower Olifants River catchment since July 2009.

Degradation of this section of the Olifants River is of concern to the Kruger National Park, where most of the pollutants released into the river upstream eventually arrive. Furthermore, South Africa has an obligation to Mozambique to ensure a flow of clean water in the Olifants where it leaves Kruger. Monitoring the biodiversity of the river upstream is critical to meeting this international obligation, not only because the state of biodiversity provides a convenient measure of water quality, but also because the many hundreds of species living in the river, from microscopic diatoms to algae, insects, fish and aquatic plants, play an important role in trapping and removing contaminants from upstream before the river flows across the border.

Recent data from the project reveal a disturbing decline over the past two years in the quality of water where the Olifants enters the Kruger National Park (Figure 1). Surprisingly no decline is evident for the years preceding the onset of crocodile deaths further downstream in the Olifants in 2008 (see article in this edition of SAEON eNews). Is this because water quality was not measured frequently enough to detect major pollution spills that were responsible? Perhaps, but it is more likely that the SASS monitoring method used is not sensitive enough to detect the type of low-level but continuous pollution that is most likely responsible for the crocodile deaths.

Figure 1. Changes in the water quality of the Olifants River where it enters the Kruger National Park (Mamba) and 60 km upstream (R40). The upper graph shows the “average score per taxon” of the SASS method, which provides an overall index of water quality (lower values indicate greater contamination by pollutants). A large flood in January 2012 had a short-term impact on these values, resulting in very low values in April 2012. While a rapid recovery occurred at the R40 site, values have stayed low at Mamba. At the same time conductivity (a measure of the amount of chemicals dissolved in the water) has been increasing (lower graph). This suggests that on-going pollution somewhere between the two sites is the cause. SASS data prior to September 2009 courtesy of Palabora Copper LTD

The difficulty of detecting pollution events, and determining exact sources of these, is further illustrated by results from the Selati River (Figure 2). The Selati is the only major tributary of the Olifants River, between the R40 and the Kruger National Park (see Figure 1). It receives all water-borne pollution from the mines and industries outside Phalaborwa and is most likely responsible for the recent decline in quality where the Olifants River enters the Kruger National Park. However, SASS results from monitoring sites on the Selati indicate that while water quality did drop at this time, it has been at similarly low levels many times over the past decade.

Figure 2. Long-term SASS data for the Selati River, measured downstream and upstream of the mining-industrial complex at Phalaborwa. A large spill of highly acidic water from a fertiliser plant occurred in late December 2013 and early January 2014, resulting in a fish kill. However, SASS values suggest that pollution has been occurring for many years, although at a lower magnitude and with less obvious impacts. SASS results from the site upstream of the mining-industrial complex indicate considerable pollution inputs from upstream as well. SASS data prior to September 2009 courtesy of Palabora Copper LTD

While large, discrete pollution events - such as that which occurred in the river in December 2013 - have a clear short-term effect, there must have been similar incidents in the past, or alternatively, a series of more regular but less intense events. And while the SASS results provide a nice integrated picture of this pattern, the method cannot reveal exactly which pollutants were responsible, making it very difficult to detect pollution sources upstream.

Figure 3. Dead fish found on the banks of the Selati River, and further downstream along the Olifants River in the Kruger National Park, following the large spill of acidic waste water from a fertiliser plant. Water with high concentrations of phosphoric and sulphuric acid (with pH <2.5) discharged into the river for eight days, resulting in rapid die-off of adult and juvenile fish, of the full range of species present in the river, including (clockwise from top left): Sharptooth Catfish, Red-breasted Tilapia, Three-spotted Barb (left) and Yellowfish (right), and Straightfin Barb.

The large and obvious spill that occurred in December 2013 generated extensive publicity in the national and international media, and prompted a strong response from the Department of Water Affairs. As the impacts on the ecosystem and at least one ecosystem service (recreational fishing) were clear, the prosecution of the offending company should be straightforward.

Yet a multitude of other pollution inputs have gone unnoticed and unpunished in the past, and their impacts on the aquatic ecosystems of the Selati, and of the Olifants River in Kruger downstream, are barely understood. These impacts are extremely difficult, if not impossible, to determine retrospectively. For example, the SASS results from a site on the Selati upstream of Phalaborwa mines and industries are often the same, or even lower, than those downstream (see Figure 2). This indicates that a significant amount of pollution is entering the river upstream, and the faulty sewerage works of Phalaborwa are the obvious source (Figure 4).

Figure 4. Results of sampling for the presence of E. coli bacteria in the tributaries that drain into the Selati River, some carrying the ouput from sewerage works and related pump-stations. Numbers indicate counts of colony-forming units of E. coli (per millilitre of water), and range from values expected of natural rivers that are unpolluted (green boxes), to values which indicate an input of sewerage, to that extent that the water poses a risk contracting gastrointestinal disease if drunk (orange boxes) or even if just used for swimming or washing (red boxes).

This makes research on the impacts of the Selati-borne pollutants even more difficult to determine, as the variety of pollutants is substantially broader than those from the industrial-mining complex alone. It also leads to questions regarding who is to blame when detrimental effects are detected downstream. Are the current low values in the Olifants (at Mamba) a result of pollution from the Phalaborwa industrial complex, or the sewerage inputs upstream? Or a combination of both? And how far downstream do these various pollutants affect the aquatic ecosystems in Kruger - is it just near the Kruger boundary, or do their long-term impacts extend all the way to Mozambique?

Exploring new ways to measure water quality

The complicated results emerging from this long-term project, and the long list of questions they generate, highlight the need for a fresh look at water quality monitoring and the effect of pollution on aquatic ecosystems. New long-term observation methods are being explored: SAEON began monitoring the abundance of river-dependent birds on a stretch of the Selati River; chemical analysis of sediments in the river is currently being conducted in collaboration with the University of Limpopo; and SAEON researchers are investigating the use of a hyperspectral camera fitted to a radio-controlled aircraft that can patrol up and down the river to identify sources of pollution inputs. Finding new cost-effective methods of sampling key chemicals that can indicate the occurrence of specific forms of pollution is another clear research challenge.

Ultimately, cost-effective methods must be found that will allow for rapid detection of short-duration, point-source pollution, as well as longer duration, diffuse pollution (and at a large number of sites that is currently being achieved). Furthermore, the long-term impacts of these on the biodiversity and functioning of aquatic ecosystems, and the related ecosystem services, remains a field of research with wide horizons.