On Sunday June 3, 2007, Rendani Nethengwe, ranger at Houtboschrand, reported a fish kill in the Olifants River to Dr Andrew Deacon. He tells the story of the investigation: According to Rendani, the Olifants field rangers observed dead sharptooth catfish or "barbel", Clarias gariepinus, in the river in the area of Shiswayni.
On Monday June 4, Rendani and I visited the area to investigate the situation. I was a bit sceptical about the fact that only sharptooth catfish were apparently dying, since these fish are a very hardy specie and will usually be the last to succumb during adverse circumstances.
We went down to the "Vyeboom" area, between the Mvubu confluence and Shiswayini, (S 24 01 59; E 31 33 56) and found a substantial number of dead catfish - 20 fish in a 500-metre strip, washed up in quiet backwaters and stuck in riffles. It seems that the fish had been dead for at least three days as they were bloated and rotting.
We then went another five kilometres downstream to where the dead fish were first sighted at Shiswayini (S 24 01 43; E 31 36 56). There we found many more dead catfish, 60 fish in 500 metres. However, 1.5 kilometres downstream of Shiswayini, we found no sign of any dead fish! This situation prevailed downstream and as far as Balule, we found no other dead fish.
From Vyeboom, the path turns away from the river upstream and we stopped our search at this point. However, Jacques Venter, KNP river biotechnician, visited the Mamba gauging station about 45 km upstream and found no affected fish so far upstream. The backpack guides also reported dead catfish (6 - 9 June) from upstream of the Tshutsi Spruit confluence to the Vyeboom site, where they get picked up at the end of their trail. The dead fish were not observed in great numbers, but one or two occasionally, all along the route of about 30km.
Jacques also took water samples that were analysed by Foskor's laboratory. Of all the parameters, only nitrate exceeded the Thresholds of Potential Concern (TPC's) and the overall quality was in fact good for this time of year. The first impression was that it might be some kind of disease that influenced the stretch of river that only affected the catfish (also Rendani's comment).
I could not see any other fish influenced, and Dr Piet Kotze supplied me with his fish survey data (Phalaborwa Barrage silt study) for Mamba and Balule collected on 31 May 2007. This indicated that everything was quite normal up- and downstream of the affected river reach at the stage the catfish die-offs started. At Mamba (upstream) 10 unaffected fish species were collected, including six catfish, and at Balule (downstream) 12 species were recorded, but no catfish.
Catfish are not always easy to collect with electro-fishing since they prefer the deeper areas that can only be sampled with nets. Unfortunately at the stage of my visit to the Olifants River, we were unable to collect any dead fish that were fresh enough to do an autopsy on. Thus I left Rendani with our conclusion that it may be some kind of virus that affected the catfish in that reach of the river.
On my way back to Skukuza, I crossed the Sabie River at the low-level bridge near Skukuza, and to my astonishment, I saw several dead catfish in a pool downstream of the bridge. This site is about 150 km south of the Olifants River site and in a completely different catchment.
The next day the Skukuza rangers were on the lookout for more dead fish, but none were found, only the ones in the pool at the bridge. I visited the pool again and realized that it was not isolated, but connected to the mainstream.
That night the Tinga concession rangers reported dead catfish at a seasonal pool in the Muhlambamadube River, a further 20 km to the south. The picture changed now completely, and the factor that is influencing the fish seems to be something that spans the whole southern part of the Park. The only factor that could be influencing fish so far in such a short time span could be aspects of the climate.
I remembered the sudden temperature drop thirteen days ago (fish kills in the Crocodile River). During that event on 22 May, dead tiger fish (Hydrocynus vittatus) were observed in the Crocodile River. During the night the air temperature dropped from a 22.2°C high (previous afternoon), to a very low -0.6°C.
Earlier in the week (normal winter temperatures) the water only dropped to 12.2°C during the night (normal winter temperatures). It became clear that the sudden temperature drop in the upper catchment rapidly cooled the river water down.
As the colder water from upstream flowed through the Lowveld, the sudden drop in temperature gave the temperature-sensitive tiger fish a cold shock which killed some of them. The water mass in a river has a capacity for buffering or resisting temperature change and the greater the water mass the less will be the change.
Although water temperature patterns follow air temperatures the magnitude of change in the water is not as great, and the change in deeper pools is less than in shallower areas. This brings us to the current morphological situation in the Olifants. Due to excessive erosion in the Olifants River catchment, huge amounts of sediment are deposited annually in the river.
Slower flows in the lower Olifants due to reduction in slope result in sediment deposition, especially in the KNP reach of the river. This process produced a river with extensive very shallow, sandy stretches, in contrast with the historical condition of an overall deeper pool-riffle-channel river.
In the meantime Jacques also reported a pollution incident in the Tshutsi Spruit, a tributary of the Olifants River that enters the main stream ± 30 km upstream of the main fish kill area. They discovered a very serious sewerage point source near Phalaborwa where a pipe was discharging raw sewerage directly into the stream.
I then started to read up in the literature about catfish, sewage and temperature. It is well known that the sharptooth catfish is particularly hardy and can survive adverse environmental conditions without apparent signs of stress. Stress is often an essential requirement for disease in fish.
It is stated that in large bodies of water, environmental conditions are normally stable. If these conditions change, the fish can move away to areas where the stressors are reduced. It is also known that stressful situations may be sufficient to trigger off a disease episode in the population (Hecht et al, 1988).
While catfish can tolerate changes of conditions within the normal range, they become weakened and lose resistance to disease when the changes (stressors) are sudden and approach or exceed the limits of the normal range. The following are examples of stressors which are believed to have adverse influences in Clarias:
The sudden drop in temperature, the shallow stretches of a silted-up river, winter low-flows and the pollution due to inflowing sewage into the Olifants River, before and during the incident, certainly added to the potential stressors.
During the presence of the cold front on 22 May, twelve days earlier, the water must have dropped considerably in temperature, especially in the shallow reaches. Clarias spp appear to thrive best at about 23 to 33°C (30°C optimum), although it can tolerate temperatures that vary from 8-35°C. Temperatures below 10°C may be lethal.
If the morphology of the river is considered, the area where the fish died is perceived as a shallower reach with a sandy bottom, contrasting with the lower bedrock dominated sections and the gorge. Winter is also the low-flow period for the Lowveld rivers and due to the current over-abstraction of water, the river was low during this period.
Catfish in areas with deeper pools should have been able to escape the water by moving to the warmer layers in the pools. However, fish trapped in prevailing shallower stretches of the river were affected by the sudden temperature drop, and it is these fish that were most probably affected. Fish become vulnerable to disease when their natural defence mechanisms for resisting the infection are weakened or lost.
Pathogens thus take hold and rapidly multiply in or on the host, producing more pathogens to further infect the same fish or to infect other fish nearby. Pathogens such as Aeromonas spp. are generally present in water, hence, setting the stage for infectious diseases and resultant mortalities to occur particularly when the fish are stressed and may readily succumb to such diseases.
The sewage spill (with numerous pathogens) in the Tshutsi River was a clear suspect from the beginning, but then Jacques told me that this seasonal river did not flow all the way to the Olifants River at the time of the incident, this factor was dismissed.
The fact that the catfish has a skin with minute scales and do not possess robust scales as most other fish do, has been mentioned by a fish expert as a possible reason why only these robust fish species succumbed. It appears thus that not only one cause was responsible for the bizarre die-off of hundreds of catfish, but a cocktail of stressors affecting the natural resistance of the catfish, allowing a specific pathogen to take advantage of the situation.
In this instance over a wide area (Olifants and Sabie Rivers), but many more fish were influenced in a larger reach of the river in the Olifants River. One can write this fish kill off as a natural occurrence and the way large populations of catfish are controlled by nature.
On the other hand, the unnaturally shallow Olifants River (due to erosion and excessive silting) represents a changed system. The Sabie River still has ample deep bedrock pools as fish refuges. Further upstream from the Park incidents of massive fish kills and dying crocodiles in the Olifants River at Loskop Dam (acid drainage) add to the fears of increasing pressure on the Olifants River. Are these all signs of more things to come?
Dr Andrew Deacon
Skukuza, June 2007