The Island Cock of Tristan da Cunha

Albert J. Beintema

The flightless Moorhen of Tristan da Cunha, or Island Cock, Gallinula nesiotis lived on the main island of the Tristan da Cunha Group, and is assumed to have become extinct towards the end of the previous century. An indistinguishable form is surviving on Gough Island, 400 km SE of Tristan. In 1973 Moorhens were discovered on Tristan, and presently they are increasing, and spreading over the island. It is not known whether these Moorhens are survivors of the original Island Cocks, ore descendants from a stock of Gough Island Moorhens, released on Tristan in 1956. The current opinion is that the latter is the case. As such, the Moorhens are considered aliens, and therefore they are not protected under the Tristan da Cunha Conservation Ordinance. The Moorhens pose a management problem, as according to the islanders they prey upon the eggs of the Yellow-nosed Albatross Diomedea chlororhynchos, which has the majority of its world population breeding on Tristan.

Island Cocks were studied on Tristan in February 1993 Tristan by Swales c.s., and in October 1993 by Beintema.

History of the Island Cock

The Island Cock was first described in 1861 as Gallinula nesiotis. Twelve years later, in 1873, the Challenger Expedition failed to find it on Tristan, where the islanders reported it to have become rare. There are no records since. Only two specimens remain, both in the British Museum.

In 1892 a moorhen was described from Gough Island as Porphyriornis comeri, and the Island Cock of Tristan was renamed Porphyrior- nis nesiotis. Generic separation from the genus Gallinula was based on flightlessness. Modern taxonomists consider the moorhens as members of the genus Gallinula again. Moorhens are still common on Gough Island, and specimens have repeatedly been brought to European musea and zoos.

The two specimens from Tristan are indistinguishable from specimens from Gough, the alleged differences falling within the range of individual variation in series of skins from Gough. In this century, doubt was cast upon the former existence of the Tristan Moorhen, as it was thought to be unlikely that a species should evolve into flightlessness on two islands without diverging. When looking at the dates (Tristan Moorhen 1861-1873, Gough Moorhen 1892-present) and the number of specimens available (Tristan 2, Gough unlimited), Eber (1961) concluded that the Tristan specimens had most probably been mislabeled, and in reality emanated from Gough. Thus, the Tristan Moorhen would never have existed. This belief gained currency.

In reality, Island Cocks have been seen (and often caught and eaten) on Tristan long before Sclater described them, with written reports from visitors in 1790, 1811, 1816, 1824, 1835, 1842 and 1856. Specimens have been collected in 1816, 1842, 1856, 1861, and 1872. These have all perished, except the 1861 ones, which Sclater used for his description (Beintema 1972).

In 1973 Island Cocks were reported by islanders from Longwood, an isolated and seldom visited area on the east side of the island. The population was estimated at 170-225 pairs, with densities up to 30 pairs/km2. Four specimens were collected (Richardson 1984). The skins are now in the British Museum. The moorhens are presently increasing and spreading. It is not known whether these moorhens are descendants of surviving Tristan Island Cocks, or derived from Gough Island Moorhens. Between 1950 and 1960, moorhens from Gough have repeatedly been released on Tristan.

Releases of Gough Moorhens on Tristan

None of the possible releases of Gough Moorhens on Tristan has been documented. In 1956, on the way home from the Gough Island Scientific Survey, Michael Swales released moorhens at Pigbite, about 2 km east of the Settlement. This has never been published, and only became known after the moorhens on Tristan were found in 1973. According to Richardson (1984), after correspondence with Swales, seven moorhens were released. But later reports state that there were six pairs.

Basil Lavarello, the factory manager of the fishing company, has kept Gough Moorhens in captivity. It is said that at least twice there have been escapes. The moorhens wandered about in the Settlement some time, and eventually disappeared. This must have taken place in the first half of the 1950's.

Allegedly one of the captains of the fishing company illegally traded in animals from Gough, selling them in Cape Town. He is said to have collected live seals, penguins, and moorhens. When threatened to be checked by the Administrator while fishing off Tristan, this captain is said to have dumped his collections overboard off Sandy Point. It is not known how often this could have taken place, and it is not known whether ever moorhens were involved. Moorhens probably should be able to reach the shore alive. This episode has also taken place in the first half of the 1950's.

Island Cocks a nuisance?

Under the Tristan da Cunha Conservation Ordinance species are protected by naming them, not by excluding others. Thus by default all newcomers are unprotected. This is the case with the Island Cocks (and ironically enough also with newly discovered nesting seabird species, as the Sooty Shearwa- ter Puffinus griseus). Nobody in the island supports the hypothesis that original Island Cocks may have survived extinction. All are convinced that Michael Swales introduced them from Gough Island. Many are even convin- ced that no Island Cocks existed on Tristan in the previous century either.

The islanders do not like Island Cocks. They say that the birds damage the populations of Yellow-nosed Albatrosses, by eating their eggs. Allegedly, they operate in pairs, where one bird attracts the attention of the incubating albatross, while the other sneaks in from behind, to punch the egg and empty it. So far, nobody outside the island has believed this story. It is generally believed that Island Cocks frequent albatross nests to scavenge for spilt food, a habit already described by Richardson (1984) when moorhens were first (re)discovered. Also, they may easily take deserted eggs.

Egg desertion in albatrosses occurs when the incubating partner has to wait too long for a returning mate. In some years this happens more frequently than in others, which is most likely linked with sea (and food) conditions in a wide ocean area, as the years with high egg loss on Tristan and Gough Island coincide (J. Cooper pers. comm.). On Gough, a deserted egg will be removed by skuas within minutes. On Tristan, skua populations have been much reduced, and deserted eggs could remain available for other scavengers, like moorhens, for a considerable time.

The islanders are very insistent on their story of moorhens taking eggs. They also say that albatross populations in the eastern part of the island have already decreased, but unfortunately, none of this can be substantiated by counts. There are many eyewitness accounts. Many men claim to have seen moorhens taking albatross eggs. When tested, most of these cases do not hold. There are cases where moorhens have been seen around the nest, which were later found empty. Predation was then easily implied, but it could also have been scavenging after desertion.

Also, punched, empty eggs have been found, whith a hole that could only be made by a moorhen bill. Again, the attack of the egg could have been preceded by desertion, instead of vice versa. Most eyewitnesses say to have seen it happen only once or twice, and usually the predation had not been actually checked, or had been deduced from signs that could have been interpreted differently.

A curious account comes from an older islander, who claims to have been more often on the mountain than anyone else. He says that Island Cocks most certainly take albatross eggs, but he also thinks that nobody ever actually has ever seen it. There remains one account which can not easily be interpreted in a different way. This islander saw the moorhen peck the egg, while the albatross was sitting on it. He checked and found the egg holed. He took the egg with him, to show it to the Administrator. But he first had to shear a sheep, and hid the egg under a bush, where he unfortunately later forgot it.

In conclusion, we have to treat the story with some care, and we may not exclude the possibility that moorhens do prey on eggs from time to time, even if most of their visits to nests are aimed at scavenging. How unlikely it seems, it may be feasible for a moorhen to damage an egg under a sitting bird. Albatrosses show no form of defense towards intruders like man or dog, apart from a little bill snapping, and often not even that. Albatrosses show no interest at all in moorhens walking around their nests. This lack of interest reminds of Chinstrap Penguins Pygoscelis antarctica ignoring egg stealing Sheathbills Chionis alba in the Antarctic, while at the same time they are very aggressive towards skuas. Such behaviour suggests that egg stealing by otherwise scavenging Sheathbills is a recent development. This could also be the case with the moorhens.

There is no indication that moorhens on Gough Island attack eggs of albatrosses or other seabirds.

Present status of the Island Cock

As on Gough Island (Watkins & Furness 1985) the moorhens on Tristan are strictly limited to the fernbush zone. The fernbush covers a belt of 2.5-3 km wide all around most of the island, on the cliffs and the Base. Above the Settlement Plateau this belt has been considerably reduced by grazing and cutting (map 2). The fernbush thus covers a total of approximately 50 km2, or almost half the island. This is much more than on Gough, where the fernbush covers 10-12 out of c 60 km2, i.e. less than 20%. The fernbush extends from near sea level to c 1000 m altitude, marking the range in altitude of the occurrence of Island Cocks. In most places the lower 300 m of the slopes offer too little cover for the birds, and then they are only found on the upper half of the cliffs and on the adjacent Base. Only near Sandy Point adequate cover approaches sea level, often in the form of fern- covered boulders at the bottom of gulches. Here the birds can be seen near sea level.

The moorhens usually stay well in cover, and only rarely venture into the open. Therefore, they are more easily heard than seen. They may be seen on patches of grassland, but always with dense bushes very near. If the observer sits still, the birds may approach and show themselves at close range. Favoured spots are sheltered bottoms of gulches, and other broken terrain, with a great variety of fern vegetations, and dense stands of Island Trees. On the Base they prefer the same places as the Yellow-nosed Albatrosses do for nesting, i.e. the lower half of the fern-covered part of the Base. In the windswept upper part, with more monotoneous stands of tree ferns, they are less common. Moorhens are also numerous in planted forest with undergrowth of ferns and brambles, as is the case at Sandy Point. The distribution of the moorhens is not associated with open water, and they have never been seen swimming.

Richardson (1984) estimated the density of Island Cocks in the Longwood area at 20-30 pairs/km2, giving a total of 175-255 pairs. On Gough Island, he found the density to be much lower, c 12-20 pairs/km2. Thus he arrived for Gough, with 25 km2 of suitable habitat, at a total population of 300-500 pairs. Watkins & Furness (1985) found much higher densities on Gough: up to 230 pairs/km2. They classified only 10-12 km2 as suitable, so their total estimate for Gough is 2000-3000 pairs. They do not suppose an increase during the ten years since Richardson visited Gough, but apparently suppose that Richardson's estimates of densities are very conservative.

If the Island Cocks occupy the entire fernbush zone on Tristan in densities of 20-30 pairs/km2, the island can harbour 1000-1500 pairs. If Richardson's estimates of densities are as conservative as they have been thought to be on Gough, this figure may be tenfold.

In February 1993 Swales divided the island into sectors, according to lines radiating from the Peak. He counted 181 birds calling in 83 degrees, which would give a total of 785 birds for the full circle. However, only 200 degrees of the circle have as yet been colonised. Thus, Swales arrived at 436 calling birds. Assuming that only males were calling, he interpreted these as pairs. However, in the Common Moorhen Gallinula chloropus both sexes give advertisement calls, so it is very likely that this is also the case in the Island Cock. On the other hand, not all birds present necessarily call during a visit. Thus, there may be between 200 and 400 pairs, which seems very low as compared to Richardson's estimate in 1973.

Swales probably did not visit all available habitat within his sectors. In 30 degrees encompassing Big Green Hill and The Ponds, he counted 56 birds calling. In the same area I counted 22 birds along one route from Big Green Hill to The Ponds, and 26 along another route, both routes together taking up c 6 km. I estimate a maximum recording distance of 200 m, which would yield a density of 40 birds per sqkm, or between 20 and 40 pairs. This agrees with Richardson's density estimate. Along the cliff edge, and on the steep slopes below, densities are higher. I counted 12 birds along 1 km of the edge, which may give a density of 50-60 per sqkm.

At Sandy Point the density is higher. Swales found 77 birds in 13 degrees, which is 3.2 times more per degree that between Big Green Hill and The Ponds. Of these 77 birds, 62 were found below the Base, in an area totalling c 1 sqkm. I counted 48 birds during a 4 km walk and climb, including part of the Base, which comes to the same density of c 60 per sqkm. On the Base above Sandy Point Swales counted 15 birds, but again, he probably did not penetrate far into the available habitat. Near Stony Hill Swales estimated the density (in birds per degree) as being of the same order of magnitude as between Big Green Hill and The Ponds.

Island Cocks have not yet colonised the area between a point somewhere between Hottentot Gulch and Burntwood, and Cave Gulch near Stony Hill. The colonised part takes up c 200 degrees out of the full circle, but due to the uneven distribution of the habitat around the mountain, this amounts to c 80% of the available habitat, or 40 sqkm. With an average density of 30-50 pairs per sqkm (in the centre this will be higher, near the egdes lower), the present population can be very roughly estimated at 1200 - 2000 pairs. Once the circle is completed, and the birds everywhere reach densities like near Sandy Point, the population may grow to c 3000 pairs.

In 1973 the birds only inhabited a small area of less than 8 sqkm. By 1983 the birds had reached First Lagoon Gulch, just west of The Ponds (Swales pers. comm.). In 1987 the first birds were recorded west of Hottentot Gulch (Tristan Times). In 1988 one stray bird was heard near Burntwood, but not again until 1993. On 23 October 1993 one bird was again heard above Burntwood. This place had been visited several times during the preceding weeks, without hearing them. No birds were heard during walks between Burntwood and and Stony Hill.

It has taken the population 20 years to expand their range by 10 km both ways, i.e. 500 m/year. If the birds proceed at the same pace, they will close the circle in about 10 years from now.

Breeding biology of the Island Cock

Moorhens on Gough have a protracted breeding season, but probably do not nest all year round, with the earliest small chicks being reported on October 12th, and the latest on March 29th (Watkins & Furness 1985).

On Tristan, Richardson (1984) found a male in breeding condition in December, and juveniles in February and April. In February-April 1993 Swales c.s. saw families with large juveniles. Individual moorhens seen could be either black (adult) or brown (immature).

On October 2nd and 3rd, above Sandy Point, I heard moorhens giving a soft call suggestive of parents leading chicks, but the birds could not be seen. This call was heard at several locations, suggesting that families with chicks were common. On October 21st, near Big Green Hill, the same call was heard, and although again the chicks could not be seen, I could get close enough in the dense vegetation to hear them peep, and record them on tape. All moorhens seen in October were adults; brown birds were not observed. No nests were found.

These observations suggest a similar breeding season as on Gough, with egg laying starting in September.

The origin of Island Cocks on Tristan

There are two hypotheses for the explanation of the present day occurren- ce of Island Cocks on Tristan:
  1. Island Cocks have been introduced from Gough Island.
  2. Island Cocks never became extinct on Tristan.
Ad 1. The main argument supporting this hypothesis is the fact that no Island Cocks have been observed between the turn of the century and 1973, that there have been some releases in the 1950's, and that there are no differences to be seen between Island Cocks from Gough and Tristan.

Ad 2. If Island Cocks could have escaped attention for almost 20 years, between 1956 and 1973, they might as well have escaped attention for four times 20 years. The heart of the area of rediscovery lies in very difficult terrain, where people hardly ever penetrate. Islanders attending their sheep get all over the mountain, but they get around in the heath- and grass zone, above the fern bush zone. From above, sheep only penetrate the upper parts of the fern bush zone, where dense stands of Island Trees are absent. From below, people get all around the island along the shores, where they penetrate into the gulches, up to the edge of the Base, but access to the cliffs and the Base itself is extremely limited in the eastern part of the island. From below, sheep grazing has cleared part of the fern bush, but this is mainly limited to the low-lying plateaus and the lower half of the slopes. Above Sandy Point, for instance, there is a sharp demarcation line, above which sheep do not penetrate into the bush. So, although the islanders state that they get everywhere, there may be a fairly wide belt of densely wooded fern bush, which is never visited, roughly from a little east of The Ponds, to the hills above Stony Beach. Some maps show a path through this area, connecting The Ponds with Sandy Point, but the islanders say this path does not exist, and the crossing is not possible, which indicates that one does not go there.

All releases have taken place on the northern side of the island. Swales released his moorhens at Pigbite, below Big Green Hill. The question is how likely it is that such birds would climb all the way up, and cross the Base all the way to Longwood, a distance of 6 km, before getting settled, while there is plenty excellent Island Cock habitat halfway the slope to Big Green Hill already. Once established, Island Cocks appear to be very sedentary, expanding their range by no more than 500 m/year.

Under either hypothesis there is the question why Island Cocks can survive and expand on Tristan today, where they could not in the previous century. Firstly, there is the possibility that by heavy selection and adaptation the birds changed their behaviour towards rats, for instance by running away from them when they see or hear them, or by sleeping in trees. Such a development would be more likely in a remnant population after many generations under predation pressure, than in a flock of newcomers from Gough. Secondly, crucial changes may have taken place in the environment. Rat populations may have stabilized at much lower levels that during their first population explosion, due to the impoverished fauna of small seabirds.

Feral dogs and cats are no longer to be found on the island. Thus the predation pressure, from rats, cats and dogs combined, may be much lower than hunderd years ago. In addition, the food situation may have improved, since nesting seabirds on Tristan are being protected since the 1950's. Populations of nesting albatrosses have increased maybe tenfold (Richardson 1984), and also the ground-nesting larger seabirds may have increased, where especially the winter breeding large Pterodroma species are of interest as a resource for scavengers. These are mostly found on steep grassy edges of the cliffs just below the Base.

Finally, the invertebrate fauna may have increased significantly. There are many alien species, and especially the larger ones amongst them, like centipedes and isopods, may be of importance as food for moorhens. For example, the common alien isopod Porcellio scaber is very abundant nowadays, occurring all over the island in litter and in the stems of the tree ferns.

References

Beintema, A.J. 1972. The history of the Island Hen (Gallinula nesiotis), the extinct flightless gallinule of Tristan da Cunha. Bull. Br. Orn. Club 92: 106-113.

Eber, G. 1961. Vergleichende Untersuchungen am flugf„higen Teichhun Gallinula chl. chloropus und an der flugunf„higen Inselralle Gallinula nesiotis. Bonn. Zool. Beitr. 12(3/4):247-315.

Richardson, M.E. 1984. Aspects of the ornithology of the Tristan da Cunha Group and Gough Island, 1972-1974. Cormorant 12(2):123-201.

Wace, N.M. & Holdgate M.W. 1976. Man and Nature in the Tristan da Cunha Islands. IUCN Monogr. 6.

Watkins, B.P. & J. Cooper, 1983. Scientific research at Gough Island, 1869- 1982: a bibliography. S. Afr. J. Antarct. Res. 13: 54-58.

Watkins, B.P., J.Cooper & I.P. Newton,1984. Research into the natural sciences at the Tristan da Cunha Islands, 1719-1983: a bibliography. S. Afr. J. Antarct. Res. 14.

Watkins, B.P. & R.W. Furness, 1985. Population status, breeding and conservation of the Gough Moorhen. Ostrich 57: 32-36.



Copyright A.J. Beintema, February 1997