The development, implementation and evaluation of an integrated aquatic vegetation management plan for the control of submersed aquatic vegetation in South African impoundments

Abstract

South Africa is an arid country with intermittent, unevenly distributed rainfall and experiences amongst the most variable flow regimes in the world. These climatological realities necessitate careful strategic planning by water resource managers to ensure an equitable distribution of water to all users. This situation has led to substantial investments in the damming of rivers for industrial, agricultural and hydro-electric purposes and the construction of complex inter-basin water transfer schemes to augment water from surplus catchments to catchments where the demand exceeds supply.

The Usutu-Vaal Phase II Inter-Basin Water Transfer Scheme, located in the eastern-central area of South Africa, experienced operational problems caused by nuisance submersed aquatic vegetation (SAV) for many years. Aquatic vegetation interfered with the operation of highcapacity water transfer pumps and other infrastructure of the scheme which resulted in downtime with substantial economic implications. Over the past decades the focus in South Africa was mainly on the control of floating declared aquatic weeds, such as water hyacinth (Pontederia crassipes Mart.). The control of SAV was not a high priority and as a result insufficient information on the management of SAV in South Africa is currently available. The only information on larger scale SAV management operations in the country is the control of SAV in two waterbodies with a full surface area (FSA) of approximately 60ha (Germiston Lake and Potchefstroom Dam) with sterile grass carp as biocontrol agent and in addition to this, the herbicidal control of SAV and macroscopic algae in irrigation canals. Prior to this study, no integration of different control methodologies on large scale SAV management was documented in South Africa.

The study area includes the Usutu-Vaal Phase II Inter-Basin Water Transfer Scheme which comprises the Heyshope Dam (FSA=5,023ha), Geelhoutboom balancing dam (FSA=36ha) and two concrete-lined conduits, Geelhoutboom Canal (21.1km length) and the Balmoral Canal (4.6km length). Water is pumped from Heyshope Dam in the Usutu River catchment via the Geelhoutboom Canal to the Geelhoutboom Dam and from there over the escarpment to the Balmoral Canal which spills into the neighbouring Vaal River catchment. As all components of the scheme experienced significant operational problems with indigenous nuisance SAV, mainly Potamogeton schweinfurthii A. Benn., Stuckenia pectinata (L) Börner and Lagarosiphon muscoides Harv., it was necessary to implement an Integrated Aquatic Vegetation Management Plan (IAVMP) to manage problematic aquatic vegetation in order to ensure a sustainable water supply to the Vaal River System.

The IAVMP on the Usutu-Vaal Phase II Inter-Basin Water Transfer Scheme was developed around an herbicidal component, with a diquat-based herbicide, and sterile grass carp (Ctenopharyngodon idella Val.) as the biological control component, in addition to mechanical adaptations to structures to preclude SAV from entering pumping station forebays and to prevent the biocontrol agent from escaping the target areas.

The primary purpose of the study was to develop and implement tools to prevent nuisance SAV from impacting on the water transfer scheme's ability to sustain an uninterrupted water supply from the Usutu River catchment to the neighbouring Vaal River System. Certain management objectives or aims were set to achieve the goals of the IAVMP. The first management objective was to implement a programme of total aquatic vegetation control during all periods in the Geelhoutboom balancing dam as this was the area experiencing the most significant SAV problems. The aim was to eliminate SAV interference with the operation of the high-capacity pumps (5m3/s) at the Geelhoutboom Pumping Station. The second management objective was to implement a programme of total aquatic vegetation control at all operational periods in both the conduits by securing the integrity of the water conveyance system from Heyshope Dam to Geelhoutboom Dam (Geelhoutboom Canal) and from the Geelhoutboom Dam to the Vaal River System (via the Balmoral Canal). The third management objective was to implement a programme of only partial aquatic vegetation control in Heyshope Dam, with the aim to reduce the original SAV infestation in the impoundment from 700ha to 70ha and to decrease contamination from the impoundment to the rest of the water transfer system. The challenge in Heyshope Dam was to see if the partial control of SAV was achievable with sterile grass carp. The final objective of the study was to develop an IAVMP that can be extrapolated and sitespecifically adapted to impoundments experiencing similar problems with nuisance SAV with maximum effectiveness on the target plants and minimum interruption of the receiving environment (non-target fauna and flora).

Several hundreds of hectares of nuisance aquatic vegetation were treated with a diquat-based herbicide and 36,000 sterile grass carp individuals were stocked in the Heyshope Dam and 1,400 individuals in Geelhoutboom Dam. This translates to 51 units per infested area (hectares) in the Heyshope Dam and 38 fish per ha in Geelhoutboom Dam. Mechanical adaptations to infrastructure at the Geelhoutboom Canal outlet into the balancing dam and the forebay grids and overflow were also implemented to accommodate the biocontrol agent and to prevent aquatic vegetation entering the forebays of water transfer pumps.

The IAVMP proved to be successful as the area of SAV infestation was reduced from 700ha in 2001 to 375ha in 2006, and 69ha in May 2008, an area slightly smaller than the set management objective of 70ha, which calculates to 1.5% of the FSA of Heyshope Dam. No pump failures were experienced due to interference by aquatic vegetation over the total study period and unscheduled downtimes could not be attributed to nuisance aquatic vegetation blockages of grids or water transfer pumps. Savings in pump maintenance were found to be significant and Department of Water and Sanitation (DWS) officials were of the opinion that the programme was cost-effective and worth the investment.

As part of the ecological component of the monitoring programme, ecological sensitive areas in Heyshope Dam were identified, demarcated and protected as far as possible from interference by the aquatic vegetation control activities. When the aim of a 70ha reserve of SAV in Heyshope Dam was reached in the winter of 2008, a moratorium was placed on any further stockings of sterile grass carp in the impoundment and herbicidal treatments were restricted to a demarcated area of 11ha near the pump abstraction location at the dam wall area. The SAV survey of May 2010 however, revealed that the SAV surface area was further reduced from 69ha calculated in 2008 to a mere 1ha. It was clear that the anticipated impact of piscivores on the fingerling/juvenile sized grass carp survival at the initiation of the programme was overestimated and the palatability of the hardy, fibrous P. schweinfurthii, as grass carp nourishment, was underestimated.

Although the Heyshope Dam is an unnatural constructed reservoir, the altered habitat created after construction (from a lotic to a lentic ecosystem) has a definite conservation value as the impoundment plays refuge (especially during drought periods) to many waterbird species frequenting the numerous wetlands in the larger catchment. The conservation value of the adapted aquatic ecosystem should be protected as far as possible and the impact of the IAVMP on non-target species should be kept to a minimum without compromising the original intentions for construction of the scheme, i.e., water augmentation to the Vaal River System.

In addition to the SAV management programme, the effect of the IAVMP on the receiving environment (i.e., non-target fauna and flora) was monitored and evaluated in the Heyshope Dam with the aim to generate enough information to develop a protocol for IAVMP's to be applied to other impoundments, with similar nuisance SAV problems. This was done by monitoring the water quality of the tributaries and the two impoundments, evaluating the trophic status of Heyshope Dam, and monitoring the effect of the removal of SAV on macroinvertebrates, fish and waterbird communities.

The results concluded that Heyshope Dam is an oligotrophic, turbid system. It was also clear that grass carp preferred the more palatable softer, finer structured Stuckenia pectinata to the more fibrous, hardy Potamogeton schweinfurthii, as S. pectinata was removed by preference from the impoundment before P. schweinfurthii. Twenty-four macro-invertebrate taxa were sampled during April 2000, in comparison to 19 taxa observed during April 2007. Thus, five taxa associated with SAV as substrate were not recorded after the implementation of the IAVMP. Three indigenous fish species; the Bushveld Smallscale Yellowfish (Labeobarbus polylepis Boul.), Sharptooth Catfish (Clarias gariepinus Burch.), and the Banded Tilapia (Tilapia sparrmanii Smith); and three exotic species, Carp (Cyprinus carpio L.), Largemouth Bass (Micropterus salmoides Lac.) and Mosquitofish (Baird & Girard) were sampled at the selected sites in Heyshope Dam. The only noticeable change in the fish community since the inception of the study was the possible disappearance of the Southern Mouthbrooder (Pseudocrenilabrus philander Weber), which needs to be verified in a follow-up monitoring programme.

The positive effects of the implementation of an IAVMP on the fish community is that Smallscale yellowfish numbers increased significantly from 2000 to 2007, and that no evidence could be found of the biocontrol-agent migrating out of the target area. The sterile grass carp stocked in the Geelhoutboom balancing dam in 2002 (400 units) and 2004 (1,000 units) died out towards the end of 2021 as the original target SAV was returning to the waterbody and no grass carp activity was observed in the balancing dam. It can therefore be concluded that the biocontrol agent has a life expectancy of up to 18 years under favourable local conditions. The introduction of the IAVMP on Heyshope Dam resulted in a sharp decline in the numbers of the two true open water avian herbivores on the impoundment, Fulica cristata and Thalassornis leuconotus. F. cristata aggressively protects its territory and scares away other competitive waterbird species and they are therefore considered to be very strong territorial birds. It is anticipated that the reduction in F. cristata numbers may facilitate waterfowl species diversity.

This study was the first on the implementation and evaluation of a large scale IAVMP on SAV in South Africa. The study highlights the successes and shortcomings in the development of an IAVMP for the Usutu-Vaal Phase II Inter-basin Water Transfer Scheme. It is accepted that, with progressive deteriorating surface water quality, problems with nuisance SAV will increase through South Africa. This study will contribute to the knowledge base on the management of SAV in South Africa as it represents the first large scale programme to manage SAV. This thesis provides baseline information for the development of IAVMP's in South Africa and will provide a guideline in the development, implementation and evaluation of IAVMP's on SAV in the country.