Home > Resources > The 1999 Cape Town Harbour oil spill: one year after the event

The 1999 Cape Town Harbour oil spill: one year after the event


An oil spill in Cape Town Harbour in May 1998 resulted in the oiling of over 500 African Penguins Spheniscus demersus . They were taken to the SANCCOB Seabird Rescue Centre in Table View, cleaned, treated and eventually released. Of 547 penguins released, 266 (48.6%) had been resighted alive at breeding colonies, of which 12 had attempted breeding, during the first year following the oil spill. The fastest bird to return to a breeding colony did so just 2.5 hours after its release. A comparison is made with similar stages of the "Apollo Sea" oil spill of 1994.


On 24 May 1998, a build up of oil pressure in a pipeline below Cape Town Harbour went undetected and led to the pipe fracturing, causing approximately 500 tons of oil to leak into an escape tunnel. Before the fault was discovered, 150 tons of oil had found its way into the harbour and five tons into Table Bay. With large numbers of seabirds foraging in Table Bay, and often in the harbour itself, it was inevitable that some would become oiled. The worst affected were African Penguins Spheniscus demersus and cormorants, mostly Bank Phalacracorax neglectus and Cape P. capensis cormorants. The Southern African National Foundation for Conservation of Coastal Birds (SANCCOB) received 31 cormorants and over 500 African Penguins for cleaning and treatment. Most of the cormorants were caught from around the harbour while the majority of the penguins were captured from Robben Island, along with 71 from Dassen Island and small numbers from Boulders, Jutten Island and Dyer Island. This report details events which occurred during the operation and concentrates on one species, the African Penguin.

Underhill et al. (1999) distinguished six phases in the rescue process between oiling and restoration to breeding productivity. This report is structured to examine these phases in relation to the Cape Town Harbour spill.

Phases 1 & 2. Death of Birds immediately after Oiling, either at Sea or on the Coastline.

This oil spill occurred in an area with a relatively high human population where oiled penguins, dead or alive, would have a good chance of being found. The two islands where most oiled penguins were captured, Robben and Dassen, are both manned and regularly patrolled. As far as is known, no penguins were found washed up dead due to oiling, suggesting that all the oiled birds came ashore alive. Fortunately, all seemed to survive the initial oiling and were captured alive. No recoveries of dead, oiled penguins with flipper bands were made during the "oiling period" of 24-31 May (South African Bird Ringing Unit, unpublished data), nor in the following two weeks.

The behaviour of the penguins after this oil spill was, therefore, in accordance with the hypothesis of Underhill et al. (1999), that "even quite small amounts of oil on their plumage induces them to make a landfall, either on a breeding island or on the nearest mainland." Most of the oiled birds went to the nearest breeding colony, Robben Island, even though this may not have been their "own" breeding island.

Phase 3. Capture, transport of Birds to SANCCOB and Initial Stabilisation

The effect of the spill on African Penguins was not realised until 27 May, when Mr. Mario Leshoro, Robben Island Museum's environmental officer, reported at least 25 penguins oiled on the beaches of Robben Island. A rescue mission, involving staff and volunteers from SANCCOB and staff of the Avian Demography Unit, Marine and Coastal Management and Cape Nature Conservation, was mounted from 28 May. It soon became apparent that many more penguins than first thought had been involved, with nearly 250 being captured on the first day. The birds were caught on the shore, rescuers attempting to move along the low water mark (or sometimes in the water) to surround the loafing penguin groups from both the seaward and landward sides. Penguins were then caught by hand and taken in crates to a central point, where initial stabilisation took place. This involved rehydration of birds with an electrolyte solution and administration of a charcoal solution which helps bind ingested oil, preventing its absorption into the blood stream. The birds were then transferred to purpose-made cardboard boxes for transport, with a maximum of three penguins per box. All the oiled penguins seemed to be on the shore or in the coastal bushes. Searches further inland found no oiled penguins nor any sign of abandoned chicks or eggs, due to the adults being oiled.

The capture operation on Robben Island continued for a week. During the first three days, more than 450 oiled penguins were captured. They were taken by lorry to the airfield from where Court Helicopters transferred the birds to their helicopter terminal at the Waterfront. From here, the penguins were transported by road, courtesy of the South African Air Force, to SANCCOB's seabird rescue centre at Table View. On subsequent days, the smaller numbers of penguins captured were transported to the Waterfront by Robben Island Museum's ferry service and then by road to SANCCOB. By now, word had come from Dassen Island that a number of oiled penguins had been found on the island's beaches. They were captured in similar fashion to those at Robben Island by Anton Wolfaardt and his staff, of Cape Nature Conservation, and one member of staff from Marine and Coastal Management. The birds were also stabilised and boxed before being transported to Cape Town Harbour, by a Marine and Coastal Management patrol boat, and then taken to SANCCOB by road. Four penguins from Jutten Island were transferred to the mainland by boat, by South African National Parks, and then taken to SANCCOB by road. Most birds found at other localities were collected by SANCCOB's own transport and brought to the rescue centre by road. The success of this stage of the operation was highlighted by the fact that all oiled birds were apparently captured and none died in transit to SANCCOB's rescue station.

Phase 4. Treatment at SANCCOB

Birds arriving at SANCCOB were initially stabilised, if this had not been done immediately after capture, with rehydrate solution and charcoal. All incoming birds were also treated with Teramycin, to prevent eye infection, and given an iron injection to counter potential anaemia brought on by ingestion of oil. Each bird was given a plastic tag with an identity number and a card on which its place of capture, medical history and other details could be recorded. The long and labour intensive job of washing, rinsing and feeding over 500 African Penguins then began. SANCCOB staff and many volunteers, some visiting from overseas, worked long, hard hours for the next few weeks, sometimes by candlelight when the somewhat temperamental power supply dictated, to ensure that as many birds as possible could be returned to the wild. No birds were released until they were completely free of oil and their natural waterproofing had returned. Other stringent conditions which had to be met were weight (no birds under 2.5 kg were released), blood tests to check that counts of both red and white blood cells were normal, and blood smears to ensure that birds were free from any sign of infection. All the penguins were fitted with metal flipper-tags prior to release, by staff of the Avian Demography Unit, SANCCOB and several other experienced and registered bird ringers. This would allow their progress in the wild to be monitored.

During the cleaning and treatment of birds at SANCCOB, only 16 penguins failed to survive the ordeal.

Phase 5. Post-Release Survival

A total of 547 African Penguins was released following the oil spill. They were released from Melkbosstrand north of Table View and some 10 km from Robben Island. Immediately before release, penguins were marked with picric acid, a yellow/orange dye, on the neck (or belly in the case of immature birds) and birds from later releases were also marked with Porcimark, a pink dye, on the belly. This made them instantly recognisable as victims of the Cape Town Harbour oil spill, even if the flipper-band number could not be read. The first 40 birds were released on 10 June 1998 with the final 25 returning to the wild on 29 July 1998. The immediate response of the first 159 birds to be released was monitored by PAW on Robben Island and by Anton Wolfaardt at Dassen Island.

The first birds to return to Robben Island did so by the morning of the 11 June, the day following their release. By midday, at least 12 of the 40 birds released on 10 June were back on the island's landing beaches. By June 20, 159 birds had been released over four separate release dates. Of the 142 birds in this total which had been rescued from Robben Island, between 46 and 61 (32%-43%) had reappeared at the island by 20 June. The fastest return took place on 19 June, when one bird returned within 2.5 hours of release, and a further three were back on the island within three hours of release. Only one bird had returned to Dassen Island during this period, a juvenile which had been rescued from Robben Island. The first of the birds rescued from Dassen Island to return did so on 21 June, two days after its release.

By the end of the first year following the oil spill, a total of 266 African Penguins (48.6% of those released) had been resighted alive at penguin breeding colonies. Only two had been found dead, one shortly after release on Dassen Island and the other, about six months after release, on Robben Island. Neither bird had been previously seen alive after release. Given that adult African Penguins have an annual mortality rate of 10-12% and immature birds about 50%, we would expect about 70 of the released birds to die of natural causes between June 1998 and the end of May 1999, irrespective of them being oiled. The probability of a dead, banded bird being found and reported to the South African Bird Ringing Unit (SAFRING) is about 2.3% of those that die, so we would expect 1.8 penguins to be reported dead in the first year. Two recoveries of dead penguins are therefore in keeping with what would be expected from a total of 547 unoiled, birds and, therefore, provide no cause for concern.

Of the 266 birds resighted, 130 were first reported on Robben Island, 131 on Dassen Island, 3 on Jutten Island, one on Vondeling Island and one at Boulders. Of the 71 birds found oiled at Dassen Island in the current incident, 34 (48% of those released) had been seen back in the wild, all at Dassen Island. Of those found oiled at other localities, principally Robben Island, 232 (49% of those released) had been seen alive (and two found dead) at breeding colonies. Of 230 oiled birds that were captured on Robben Island, 129 were first seen at Robben Island, 96 first seen at Dassen Island, 3 at Jutten Island, one at Vondeling Island and one at Boulders. One bird found oiled at Jutten Island was resighted at Robben Island and a bird found oiled at Boulders was resighted at Dassen Island.

As with the "Apollo Sea" post-release situation some four years previously, there was no evidence of any mass-mortality following the release of cleaned penguins. The number of birds seen alive at breeding colonies a year after the "Apollo Sea" disaster was 1673 (41%). By the end of July 1998, four years after this incident, the total of birds seen alive was 70% of those released. There is no reason to suggest that the future results from the Cape Town Harbour incident will not be equally impressive, assuming that regular monitoring can be continued.

Phase 6. Breeding Attempts

Of the total number of birds resighted, 12 (4.5%) had been seen with eggs or chicks subsequent to their release. One of these breeding attempts was actually interrupted by the oil spill itself. The bird concerned (banded as S24611) was from Robben Island and had also been oiled in the "Apollo Sea" oil spill of 1994. Two days after its release, it was seen with its mate and two feathered chicks back on Robben Island. The mate had continued rearing the chicks alone during the four week absence of S24611. A similar instance during the "Apollo Sea" spill was documented by Underhill et al. (1999). None of the other eleven breeding attempts were closely monitored. Five had produced chicks while the other six birds were seen incubating eggs. Only one of the breeding attempts was recorded from Dassen Island, the others relating to birds at Robben Island.

The proportion of birds attempting to breed was lower than that at a similar stage following the "Apollo Sea" oil spill. After the latter incident, 556 (33.2% of birds resighted) had attempted breeding by the end of the year following their release. The most likely explanation for the lower proportion of birds observed breeding after this spill in comparison with the "Apollo Sea" incident, is the difference in observation intensity. Most of the birds affected in the "Apollo Sea" incident were from Dassen Island, a colony which has been continually monitored since July 1994. In contrast, Robben Island, which probably contributed most of the birds oiled in the Cape Town Harbour incident, has less regular monitoring, with thorough searches for banded birds being made on a quarterly basis. It should also be noted that the peak of the breeding season for 1999 could be yet to come and the proportion of birds recorded breeding should therefore increase.

The age structure of the birds oiled in the Harbour Spill incident is unknown. The number of birds oiled in juvenile plumage was 57 (10%) in the current incident as compared to 340 (8%) in the "Apollo Sea" incident. A large number of birds banded prior to oiling (215) gave some insight into the age structure of birds involved in the "Apollo Sea" oil spill. Of 88 banded birds of known age, 31 (35%) were likely to be too young to breed, leaving 65% which could have been potential breeders (Underhill et al. 1995). In the current incident, only four of the 12 birds banded prior to oiling were of known age but all 12 would have been of breeding age at the time of oiling (Table 1). There is no biological evidence to suggest that there should be any difference in the number of breeding attempts between the two incidents based on what little is known of the age structure of birds oiled.

Origin of Oiled Birds

Twelve of the penguins oiled in this incident bore flipper-bands, providing a little information on where the penguins could be from (Table 1). All 12 birds were found oiled on Robben Island. Eight of the 12 had been banded at Robben Island, four of which were known to be part of the breeding population. The other four penguins had already been involved in previous oiling incidents, three of which related to the "Apollo Sea" spill of 1994. One of these three was known to be a breeding bird from Robben Island, another had last been seen on Dassen Island in 1996. The fourth bird had been found oiled at Boulders in 1997 and released on Robben Island in January 1998. Thus the Cape Town Harbour spill resulted in its second oiling within six months.

Movements of Birds after release

Following their initial sightings in the wild, 36 penguins were subsequently observed at a different locality. The majority of movements involved birds going from Robben Island to Dassen Island. Nineteen birds made this journey, one subsequently returning to Robben Island. None of these 19 birds had been found oiled at Dassen Island. It is not known what the true origin of any of these birds is. Five of these 19 birds were oiled in juvenile plumage and could be wandering birds from other colonies. Thirteen birds first seen at Dassen Island were later seen at Robben Island, three of them later returning to Dassen Island. Of these 13, only one had been found oiled at Dassen Island and it was one of the three to return there. Two of the birds moulted at Robben Island, one later returning to Dassen Island. One of these 13 birds was oiled in juvenile plumage.

The remaining movements all involved birds first seen after release at Dassen Island, one going to Boulders, one to Vondeling Island (30 km north of Dassen Island) and two to Jutten Island in Saldanha Bay (40 km north of Dassen Island). Only one of these four birds was found oiled at Dassen Island. One of the birds seen at Jutten Island was moulting and was later seen back at Dassen Island.

As already stated, the true origin of most of the birds involved in this oil spill was not known. There is, however, circumstantial evidence to suggest that some of the penguins found oiled at Robben Island actually came from Dassen Island, while those found oiled at Dassen were predominantly Dassen Island birds. This could have two implications:

  1. African Penguins at Dassen Island are foraging fairly widely over fairly large distances. This is supported by evidence from resightings of other birds and by satellite transmitter readings from a breeding adult (B.M. Dyer pers. comm.).
  2. Oiled birds frequently make landfall at the nearest "safe" (i.e. predator free) piece of land, rather than attempting to return to their "home" colony.


The procedure during phases 1 to 4 of the Cape Town Harbour spill were considered to be satisfactory. All oiled birds appeared to be captured, there were no deaths of penguins during transportation and stabilisation and very few penguins died during the treatment phase. The signs are that the rehabilitation of birds to the wild (phase 5) has also been successful. There is little evidence as yet of restoration of birds to the breeding population (phase 6), but this will hopefully improve as more birds return to breed. It is possible to compare the phases following this spill with those following the sinking of the "Apollo Sea" near Dassen Island in 1994 (Table 2). It should be noted that the number of birds affected during the current spill is only about 5% of the 10 000 oiled after the "Apollo Sea" sank, and was a far more "manageable" number for the rescuers and SANCCOB to handle.

Mortality of penguins in phase 1 was probably much the same in both instances, i.e. very few, if any, birds died at sea. While there are no accurate records of how many penguins died before they were discovered after the "Apollo Sea" went down, none of the penguins oiled in the Cape Town Harbour spill are known to have died before they could be captured. The biggest difference lies in phase 3 of the operation, the number of penguins dying during transportation and stabilisation. Phase 3 accounted for the death of 29% of all penguins captured in the former incident (Williams 1994), whereas mortality was zero at this stage of the Cape Town Harbour spill. This is due to the fact that far fewer birds had to be transported and that co-ordination of the transportation phase involved fewer people and was much improved from that of the 1994 incident. The number of birds which died within 24 hours of arriving at SANCCOB in the "Apollo Sea" incident, amounted to 10% of all birds captured (Williams 1994). This may have resulted, in part, from the length of time taken to get oiled birds from the islands to SANCCOB, such that many may have already been very weak on arrival. In 1994, most of the oiled birds were collected from Dassen Island, which is further from Cape Town and logistically more difficult to access and transport birds from than Robben Island, where most of the Cape Town Harbour spill birds were collected. Penguins dying in phase 4 (treatment at the rehabilitation station) was also much reduced in the latter incident when compared to the "Apollo Sea". Again, sheer volume of birds involved in 1994 must have made it very difficult to treat all birds quickly and efficiently enough to prevent a proportion from dying. However, the greatly reduced mortality in 1998 is probably also due to improvements to the co-ordination and stabilisation procedures of birds treated at SANCCOB. Phase 5 of the operation seems equally successful in both instances. The proportion of birds seen alive after a year following release is remarkably similar and the mortality of released birds very low in both cases. Breeding performance of birds released from the current spill has not been investigated preventing a comparison of phase 6 from being made. The large difference in the number of birds which had attempted breeding during that first year in each incident, could well be due to differences in monitoring levels as previously described. A more meaningful comparison may be possible after a second year has passed.


British Petroleum and the International Fund for Animal Welfare are gratefully acknowledged for their support of the cleaning and monitoring of oiled seabirds. PAW acknowledges a Jagger Scholarship of the University of Cape Town. Staff of the Avian Demography Unit, Marine and Coastal Management, Cape Nature Conservation, Court Helicopters, Robben Island Museum, South African National Parks, South African Air Force and SANCCOB played an important part in the rescue operation. Anton Wolfaardt, of Cape Nature Conservation, kindly made all data from Dassen Island available for this analysis. The Dept. of Chemistry at the University of Cape Town provided picric acid dye. Dr R.J.M. Crawford, Prof. L.G. Underhill and Mr. A.C. Wolfaardt commented on earlier drafts of this report. A great debt of thanks is owed to the staff and many volunteers at SANCCOB, who gave so much of their time, energy and patience to ensure that the rescue operation was a resounding success.


  1. UNDERHILL, L.G., THORNTON, M., OATLEY, T.B., WHITTINGTON, P.A., CRAWFORD, R.J.M., DYER, B.M., UPFOLD, L., WILLIAMS, A.J., NEL, D.C., GILDENHUYS, A.G. & BAUMANN, L. 1995. Mortality and survival of Jackass Penguins Spheniscus demersus involved in the Apollo Sea oil spill incident: results from flipper banding. ADU research report 13: 1-22.
  2. UNDERHILL, L.G., BARTLETT, P.A., BAUMANN, L., CRAWFORD, R.J.M., DYER, B.M., GILDENHUYS, A.G., NEL, D.C., OATLEY, T.B., THORNTON, M., UPFOLD, L., WILLIAMS, A.J., WHITTINGTON, P.A. & WOLFAARDT, A.C. 1999. Mortality and survival of African Penguins Spheniscus demersus involved in the Apollo Sea oil spill: an evaluation of rehabilitation efforts. Ibis. 141: 29-37
  3. WILLIAMS, A.J. 1994. Factors to consider in the capture and transport of penguins. In: BARRETT, J., ERASMUS, Z. & WILLIAMS, A.J. (Eds) 1994. Proceedings. Coastal oil spills: effect on penguin communities and rehabilitation procedures: 15-18. Cape Town: Cape Nature Conservation.
  4. WILLIAMS, A.J., CRAWFORD, R.J.M., NEL, D.C. & UNDERHILL, L.G. Submitted. The rescue, rehabilitation and restoration of oiled African Penguins and Cape Gannets in southern Africa.

Table 1. Details of banded birds oiled in the Cape Town Harbour oil spill, 1998

A00094 14/08/1997 Roben Is. Adult   28/05/1998 Roben Is. oiled
A00100 15/08/1997 Roben Is. Adult   25/03/1998 Roben Is. Breeding (with 2 chicks)
          28/05/1998 Roben Is. oiled
A02181 20/01/1998 Roben Is. Adult yes (oiled) 30/05/1998 Roben Is. oiled
S04540 01/07/1991 Roben Is. chick   28/03/1996 Roben Is. Breeding (with 2 chickS)
          28/05/1998 Roben Is. oiled
S09550 29/05/1990 Roben Is. chick   05/08/1997 Roben Is. Breeding (with 2 chicks)
          01/06/1998 Roben Is. oiled
S11987 21/07/1993 Roben Is. chick   24/05/1996 Roben Is. with mate
          02/06/1998 Roben Is. oiled
S14302 24/09/1993 Roben Is. Adult   30/05/1998 Roben Is. oiled
S22700 19/08/1994 Silvers-Troom Adult yes (oiled) 06/07/1995 Roben Is. Breeding (with one chick)
          05/06/1998 Roben Is. oiled
S24611 07/08/1994 Silvers-Troom Adult yes (oiled) 19/12/1996 Roben Is. loafing
          02/06/1998 Roben Is. oiled
S25200 25/08/1994 Silvers-Troom Adult yes (oiled) 07/02/1996 Dassen Is. on shore
          02/06/1998 Roben Is. oiled
T2586 30/06/1993 Roben Is. Adult   11/08/95 Roben Is. Breeding (2 eggs)
          29/05/1998 Roben Is. oiled
T6515 06/11/1991 Roben Is. chick   10/02/1997 Roben Is. on shore
          06/06/1998 at SANCCOB oiled

Table 2. Comparison of phases during the "Apollo Sea" oil spill of 1994 (Williams 1994, Underhill et al. 1999) and the Cape Town Harbour oil spill of 1998.

African Penguins oiled c. 10 0001 c. 563
PHASE 1 - Death at sea Few or none died at sea. Few or none died at sea.
PHASE 2 - Death on coastline Not documented. None appeared to die before capture.
PHASE 3 - Death during transport and stabilisation 1800 penguins died during transportation (19% of those captured), 1000 died within the first 24 hours after arrival at SANCCOB.2 No penguins died during transport or stabilisation.
PHASE 4 - Death during treatment c. 2400 penguins died during treatment (25% of those captured).2 16 penguins died during treatment (3% of those captured).
PHASE 5 - Death after release 32 birds out of 4076 released with flipper bands (0.8%) died within the first year after release. 2 birds out of 547 released (0.4%) died within the first year after release.
PHASE 6 - Reproductive "death" Restored penguins bred successfully with normal breeding productivity when conditions were good. Their productivity was below that of unoiled birds during adverse breeding conditions (Williams et al. submitted). Breeding performance not monitored. Very few birds attempted to breed.

1 Data from Underhill et al. 1999
2 Data from Williams 1994
Phases 1 to 6 describe the situations in which birds can die after being oiled. Phase 6 applies to birds which make the transition back to the wild, but fail to reproduce (and are therefore effectively dead).