Comprehensive Description
provided by Smithsonian Contributions to Zoology
Porzana astrictocarpus
HOLOTYPE.—Nearly complete skeleton (Plates 7, 8, 9). Vertebrate paleontological collections of the National Museum of Natural History, Smithsonian Institution (USNM 175893). Collected from the bank of a dry wash on the east side of the lower portion of Prosperous Bay Valley, St. Helena Island, South Atlantic Ocean (approximately 15°56′18″S; 5°38′5″W), on 12 June 1971 by Storrs L. Olson. The specimen consists of a complete skull (braincase crushed) that includes the palate, sclerotic plates, hyoids, lacrimals, pterygoids, and right quadrate; complete mandible; pelvis (lacking pubic bones); both wings (lacking only digits I and III, radiales, and one ulnare); both scapulae; left coracoid; proximal two-thirds of left clavicle; complete right hind limb including all phalanges and ossified tendons; left femur lacking only the head. The bones are dark tan in color and are only slightly, if at all, mineralized. Unlike most of the specimens that were picked up from the surface or partially embedded in the fine silt in the upper layers of the Prosperous Bay deposits, the type was firmly embedded in an eroded embankment under nearly 2 meters of cindery sediment. In the skull and some of the other bones are minute holes or scars, the edges of which are reddish brown in color and appear almost as if burnt. Other specimens from Prosperous Bay bear these same sort of marks which may have been made by rootlets formerly growing around the bones. The type-specimen is remarkable in retaining almost all the important bones but totally lacking the ribs, vertebrae, and sternum. Since most of the toe bones were found in direct association with the skull, the specimen was probably buried doubled backwards on itself and the ribs, vertebrae and sternum later washed away. The left femur, the only element of that leg found, was discovered in two pieces washed out onto the bottom of the gully, a few inches directly below the rest of the skeleton.
PARATYPES.—The paratypical series consists of USNM 175894–175935, 175945–175949, and 175960.
RANGE.—St. Helena Island, South Atlantic Ocean.
STATUS.—Extinct, probably exterminated after the island’s discovery in 1502.
ETYMOLOGY.—Latin, astrictus, drawn together, plus carpus, referring to the peculiar approximation of the metacarpals in this species. The specific name is considered a noun in apposition.
DIAGNOSIS.—A small, flightless Porzana; pelvis, and hindlimb intermediate in size and proportions between Porzana pusilla and “Porzanula” palmeri (I do not consider Porzanula separable from Porzana, see below); wings and pectoral girdle reduced to the same extent as P. palmeri (for comparisons see Table 4 and Plate 11); interorbital bridge with longitudinal furrow; proximal intermetacarpal space constricted; distal femur much expanded.
DESCRIPTION.—The only available skull, that of the type (Plate 7a,b) is slightly larger than Porzana pusilla and some specimens of P. palmeri and has the typical short bill of a Porzana. The nasal bar is wide and flat. In P. palmeri the bill is longer than P. pusilla, with a longer premaxilla and more elongate nostril. The St. Helena bird falls exactly between these two species in these respects and is slightly more robust than either. A longitudinal furrow creases the interorbital bridge of P. astrictocarpus; only a hint of this may be seen in P. palmeri and it is not found elsewhere in Porzana. The postorbital process is blunt, ventrally directed and with a distinct depression above it. The palate is very similar to P. palmeri. Fortunately, the right lacrimal of the type is complete (Plate 7d). This is nearly identical to that of P. palmeri save that the descending process is longer. It is similar also to P. pusilla except that in that species the horizontal process is straighter and does not curve outward and downward posteriorly. These three species differ from Porzana albicollis, P. porzana, P. Carolina, and P. fusca, in which the lacrimal is wider with a short, anteriorly curved descending process.
The mandible (Plate 7a,c) is a little heavier and more expanded in the articular regions, and in medial view is higher than in either P. palmeri or P. pusilla. The symphysis, which is rather long in P. palmeri and short in P. pusilla and other species of Porzana, is intermediate in P. astrictocarpus.
The coracoid (Plate 8b,c) is, of course, greatly reduced compared to flying forms such as P. pusilla, and is longer and more slender than in P. palmeri. In P. palmeri the procoracoid process is not as broad, and the fenestra is placed higher than in P. astrictocarpus. The head is narrow, in line with the shaft and not deflected internally in P. astrictocarpus. The sterno-coracoidal process is relatively high and blunt; sternal facet with a well-developed flange on internal side. Sterno-coracoidal impression not particularly deep but with a prominent thickened ridge on its antero-dorsal margin for the attachment of the sterno-coracoid ligament.
Of the furcula there is only part of the left clavicle (Plate 8b). This is a slender rod with a knob at the coracoidal extremity.
The scapula (Plate 8a) is smaller even than in P. palmeri, which it resembles in having the acromion rounded and blunt whereas this process is longer and pointed in other species of Porzana. Porzana palmeri differs from P. astrictocarpus in having the glenoid facet larger and more expanded. In P. astrictocarpus no vestige remains of the tubercle for the attachment of the M. expansor secondariorum tendon, although this tubercle is well-developed in volant species of Porzana and a small bump is still to be found in P. palmeri.
Of the sternum there is known only a left posterior lateral process with a small piece of the sternal plate attached. The anteriormost part of the sternal notch is indicated. The lateral process is very short and the notch shallow (Figure 5), much more so even than in P. palmeri. This abbreviated condition obtains in most flightless rails. I would estimate from this scrap, that the whole sternum of P. astrictocarpus was smaller than that of P. palmeri.
The humerus (Plate 8d,e) is identical to that of P. palmeri in size and proportions, although the distal end is slightly wider. In P. astrictocarpus there is a distinct longitudinal scar immediately below the median crest of the internal tuberosity; this is absent in P. palmeri. The olecranal fossa is wider and shallower and the distal condyles reduced in contrast to P. palmeri and P. pusilla. The brachial depression is much deeper than in the last-named species.
The ulna (Plate 8f) is nearly identical in size and details to P. palmeri but the shaft is not as stout, external condyle slightly larger and rounder, and brachialis impression deeper. The radius (Plate 8f) is very near in size to P. palmeri from which it does not otherwise differ.
The carpometacarpus (Plate 8g,h,i) is extremely similar to that of P. palmeri except in one singular detail. The proximal half of metacarpal III is expanded dorsally and from the opposite side, on metacarpal II, an excrescent ridge extends toward the expansion of metacarpal III, thus narrowing the proximal half of the intermetacarpal space to a small slit. In the more robust carpometacarpus from Dry Gut (Plate 8i), one of the projections of metacarpal II has succeeded in fusing with metacarpal III, leaving a small foramen proximally. This condition cannot be a pathological one, as it occurs identically on both sides of the type and in an entirely different specimen from a different locality. I have not seen a like condition in any other rail. On the dorso-medial surface of the left carpometacarpus of the type was a short portion of ossified tendon, doubtless of M. flexor digitorum profundus which is ossified in this area in other rails.
The pelvis (Plate 9a,b) is of interest for its intermediacy. It is neither as long and narrow as in P. pusilla, nor as short and broad as in P. palmeri. Other than in these general proportions, the pelves of these three species do not differ appreciably.
The femur, also intermediate in proportions (Plate 9c,d), is not as long and slender as P. pusilla or as short and stocky as P. palmeri. However, it differs conspicuously from these and any other crakes in the much expanded distal articulation. This appearance is for the most part due to the greatly enlarged fibular condyle which projects farther laterally and posteriorly than in any related species. The trochanter is better developed also. The posterior intermuscular line is more prominent and displaced medially which may be correlated with the apparent lateral rotation of the entire distal end of the femur.
The fibula (Plate 9e) is the size of that of P. palmeri but the head is somewhat heavier, probably in accordance with the larger fibular area of the femur. Distally it is more attenuated than in P. palmeri.
The tibia (Plate 9e,f) presents intermediate proportions. It is shorter and stockier than P. pusilla but not as short and heavy as P. palmeri. The proximal end is broader than either. Associated with the tibia of the type were six ossified tendons. Three of these were ascertained to be the tendons of M. peroneus brevis, M. tibialis anticus, and M. extensor digitorum longus. On the tibial shank the following tendons may also be ossified in the Rallidae: M. flexor perforans et perforatus digiti III, M. flexor perforatus digiti II, M. flexor hallucis longus, and M. flexor digitorum longus. Any of these may have been ossified in P. astrictocarpus.
The tarsus (Plate 9g,h,i) is stout but not as short as in P. palmeri or as long and slender as in P. pusilla. The internal trochlea extends out farther medially and posteriorly in P. astrictocarpus and in P. palmeri than in P. pusilla, making the distal end wider. The medial side of the hypotarsus is strongly excavated in P. astrictocarpus, even more so in P. palmeri, and less so in P. pusilla. Of the seven ossified tendons associated with the type tarsus, five, from their location before removal, are known to represent the tendons of M. extensor digitorum longus (strongly ossified and triradiate distally), M. flexor hallucis longus, M. flexor digitorum longus, M. flexor perforatus digiti II, and M. flexor perforatus digiti III. The other two are probably of M. flexor perforans et perforatus digiti III, and M. flexor perforatus digiti IV.
Extant Rails
THE RESIDENT Gallinula.—The presence of the Common Gallinule (Gallinula chloropus) on St. Helena was recorded as early as 1715 by Beeckman (in Gosse, 1938:140). Melliss (1870) describes a bird, undoubtedly of this species, that was captured in Lemon Valley. Kinnear (1943) reported a specimen from Fisher’s Valley and referred it to the African subspecies G. c. meridionalis. Benson (1950) mentions Kinnear’s record but did not see the species himself. A single bird and a pair with young were seen by Haydock (1954:69) in Fisher’s Valley and he mentions others seen in James Valley. The Fisher’s Valley stock he says, “have, according to reports, only been there some eight years,” but I am inclined to doubt this seriously. Under the name Gallinula angulata (a lapsus for G. chloropus) Loveridge (1964:430) relates that this species “is now known to nest in Sandy Bay and elsewhere,” and informed me (in litt.) that the birds have been a constant feature of Fisher’s Valley since his arrival on the island in 1957 and that he has seen a nest. Basilewsky (1970) noted the species at Bradley’s Ruins (lower Fisher’s Valley) and less frequently in James Valley.
I found this species abundant throughout the length of Fisher’s Valley, the marshes and sodden yam patches of which I never traversed without seeing and hearing at least a dozen. I also heard them calling from the bottom of Lemon Valley. As a nesting bird it is well known to the islanders who call it “waterduck” or “waterhen” and I was told of its being common in Sandy Bay Valley and of an individual seen at Lemon Tree Gut. The species surely occurs throughout the island wherever there is suitable habitat. During my stay on the island I heard persistent and repeated rumors from a number of sources that another, similar species of bird with fully webbed feet also occurs on the island. Although skeptical I offered a substantial reward for a specimen, but I had no takers. I am quite satisfied that such a bird is imaginary. The legend may have arisen as a result of seeing the Gallinula swimming.
Four gallinules from Fisher’s Valley that I collected, in their small size and dark coloration, agree well with the African race G. c. meridionalis. There has been no decrease in flying ability, as the birds often took to flight readily and flew strongly. Considering the present populations of humans, dogs, and cats on the island, any trend towards flightlessness would no doubt be of great disadvantage. Gallinula chloropus on St. Helena is definitely a well-established, self-introduced, breeding bird.
ERRANT SPECIES OF Porphyrula.—Two species of purple gallinule have occurred on St. Helena as vagrants. The first record of the genus from the island is an immature specimen of the African species Porphyrula alleni, collected July 1938 and reported by Kinnear (1943). I have seen this specimen and can confirm that it was correctly identified. Loveridge (1964) mentions six occurrences of immature Porphyrula on St. Helena between 1959 and 1964. Two of these were identified from photographs by R. E. Moreau as P. alleni. Another individual (taken June 1964 in Jamestown), also identified as P. alleni, was preserved as a skeleton and is now in the collections of the Museum of Comparative Zoology (MCZ 5046). On measuring this specimen, however, i find it is too large for P. alleni (e.g., tarsus 60.2 mm) and it must therefore be referred to P. martinica. This is further corroborated by my collection of an immature individual of P. martinica on 10 June 1971 in Fisher’s Valley (Olson, 1972).
As at Ascension, it is now certain that both species of Porphyrula occur at St. Helena. The records from both islands have nearly all been of subadult birds that appeared between May and September. Loveridge was under the impression that P. alleni might be breeding on St. Helena, but there is no evidence for this and I feel quite certain that no Porphyrula breed on any of the South Atlantic islands (see below).
Haydock (1954) discusses a bird that he saw on St. Helena that he thought to be Porzana porzana. Although possible, in the absence of a specimen the record is best considered questionable.
Environment and Lives of the Rails
St. Helena provided a much more equable environment for terrestrial birds than Ascension. It is larger (122 km2; 47 mi2) and displaced farther eastward and thus closer to West Africa than is Ascension. The oldest rocks from the island have been dated at 14.3±1.0 million years (Baker, 1970) so it is also considerably older than Ascension.
A rugged island, entirely volcanic in origin, St. Helena juts up abruptly out of the sea and is characterized by steep cliff faces and deeply dissected valleys. Access to the sea is limited to a few valley mouths. Melliss (1875:221) quotes the noted botanist Hooker as saying that the island “when discovered … was entirely covered with forests, the trees drooping over the tremendous precipices that overhang the sea.” This must be to a certain extent an exaggeration of the facts. Today, much of St. Helena is barren and arid. The areas below 500 meters (1500 ft) receive little rainfall (Jamestown, near sea level, averages a little over 200 mm per year) and there is no reason to assume that they were much wetter in the immediate past. That there are a number of endemic invertebrates and one endemic species of bird that are adapted to these open, arid areas argues convincingly against the entire island having been covered with forest in the past. However, the areas above 500 meters, the “green belt,” receive nearly four times as much precipitation and were once well vegetated with many endemic plants (Decelle, 1970). There are abundant clear streams of fresh water originating in this higher portion of the island.
After the discovery of St. Helena in 1502, man systematically set about his usual destruction. The forests were decimated by goats and by cutting for timber and fuel for lime burning. Hogs, rats, cats, and mice set upon the native avifauna, extirpating about 15 species. Today only two native landbirds (counting Gallinula) and six seabirds breed on the island and offlying rocks and the native vegetation is reduced to a small area on the central ridge, which is fast being encroached upon by introduced species of plants. With the destruction of the native vegetation, a number of species of endemic land snails also became extinct.
Unlike Atlantisia elpenor, there is no written record of either A. podarces or Porzana astrictocarpus. We can only guess at what their niches might have been. In contrast to the Ascension rail, the St. Helena rails shared their environment with other landbirds, for on the island there were at least a plover, a dove, and a large species of hoopoe present in addition to the rails (Olson, in prep. (b)).
St. Helena supports a much larger and richer fauna of terrestrial invertebrates than Ascension. This fauna was extensively collected by members of the Museé Royal de L’Afrique Central, Tervuren, Belgium, the first of several volumes of their results having already appeared (e.g., Basilewsky, 1970). In addition, the island once was a major breeding station for thousands of seabirds—almost all of which have now been extirpated. There were probably at least six species of shearwaters and petrels, two frigatebirds, boobies, and Sooty Terns breeding in large colonies on the mainland (Ashmole, 1963b; Olson, in prep. (b)). As on Ascension, these immense seabird colonies would have provided a veritable cornucopia for rails, and at least Atlantisia podarces, by virtue of its large size, would have been well equipped to turn this source to its advantage. The Porzana, like its counterpart on Laysan, also may well have taken sustenance from the seabird colonies. Another prime source of food would have been the diverse fauna of pulmonate snails formerly found on the island. Wollaston (1878) lists about 20 species of land snails in 5 genera, most of which are known only from subfossil remains. Snails are a favored food item of many continental rails and seem to be especially preferred by members of Porzana. They very likely were taken by P. astrictocarpus. Some of the St. Helena snails were far too large to be utilized by P. astrictocarpus but may well have been preyed upon by A. podarces. The great disparity in size of the two rails would have precluded much overlap in their respective food items.
Two morphological adaptations of A. podarces seem to be directly correlated with the nature of the island’s terrain. St. Helena is possessed of many valleys that are exceptionally deep and precipitous. A. podarces, being a large and heavy flightless bird, would likely have had difficulty getting out of such steep-sided, rocky ravines. But as we have seen, compared to other large flightless rails from islands with more even terrain, A. podarces has proportionately better developed wings (Table 3) and much longer claws on the toes (Plate 6a). These I believe may have been adaptations for climbing and fluttering up the valley walls. In the National Zoological Park in Washington, I have seen the African rail, Limnocorax flavirostra (ordinarily a very aquatic species), scale a nearly vertical rock lace, accomplishing this feat with much fluttering of its wings, although never actually flying. Hagen (1952:197) observed that Atlantisia rogersi used its wings “as brakes when jumping down small declivities,” and one in captivity climbed tussock stems “fluttering with the wings to aid the climbing.” Atlantisia podarces probably used much the same technique. Porzana astrictocarpus, weighing approximately as much as A. podarces, would not have been in as much need of pronounced climbing adaptations.
The extinct rails of St. Helena would have had but very few natural enemies. Frigatebirds may have presented a possible, but not too likely, hazard to chicks of either species or even adults of the Porzana. On the other hand, A. podarces, being of such large size, very conceivably may have consumed both the young and eggs of P. astrictocarpus on occasion. Although there is no direct proof of the cause of the extinction of the two endemic rails of St. Helena, there can be little doubt that they survived up until man’s arrival whereupon they succumbed to predation by man and his domestic animals and to their respective destruction of the habitat.
The establishment of Gallinula chloropus is an obviously recent event which probably did not occur until after man’s coming. First of all, the nature of the marshy vegetation in the valleys it now occupies is the result of changes brought about by man and livestock. For instance, E. L. Jackson (1905:26) informs us that “shortly after the establishment of the colony [ca 1660] a species of yam had been introduced from Madagascar. Planted in the valleys it throve wonderfully well, as it requires a constant soak of water …” Today this yam is one of the dominant plants of the marshy valley bottoms and its growths are the constant resort of gallinules. Secondly, had the gallinules attempted to colonize the island prior to the extinction of A. podarces, they most likely could not have competed successfully against such a well-established and adapted species. The recency of the arrival of Gallinula chloropus is testified to by the fact that it does not sensibly differ from the African mainland form and by its total absence from the fossil record.
RAILS OF THE TRISTAN DA CUNHA GROUP AND GOUGH ISLAND
In the remote Tristan group, consisting of the three associated islands of Tristan da Cunha, Inaccessible, and Nightingale, and on Gough Island, removed some 400 km SSE of Tristan, two vagrant and three endemic forms of rails are known. Although I have no new information concerning these birds, a short synopsis of them is included for the purposes of later discussion.
Easily the most interesting of the Tristan rails is the minute, flightless, Atlantisia rogersi, confined entirely to Inaccessible Island. A rail said to inhabit Inaccessible was described to members of the Challenger expedition in 1873 but they did not encounter it (Moseley, 1879). Nor was the Shackleton-Rowett expedition able to secure specimens of it in 1922 (Lowe, 1928:101). Not until 1923 did specimens collected by Rev. H. M. C. Rogers, then resident chaplain on Tristan, arrive at the British Museum where they were examined by Percy Lowe and briefly described as a new genus and species, Atlantisia rogersi, in honor of Rev. Rogers (Lowe, 1923). Lowe later (1928) gave a rather comprehensive treatise of the species, including anatomical notes and a color plate. The dark brown and black plumage of A. rogersi is fluffy and decomposed, appearing almost hairlike. The adults are variably barred with white or buffy on the flanks, thighs, wing coverts, scapulars, secondaries and even in the distal parts of the primaries, while the juveniles are entirely dark (Hagen, 1952). The iris is red in adults and dark brown in juveniles.
A fairly complete account of the life and habitat of this species has been given by Hagen (1952), and additional comments rendered by Elliott (1957). Uninhabited Inaccessible Island is small (12 km2) and is about 40 km removed from the main island of Tristan and 22 km from Nightingale. It is well vegetated, especially with dense fern and tussock-grass (Spartina arundinacea). In the past it has suffered somewhat at the hand of man and at one time was said to have been overrun with wild pigs and goats as well as being fired by sealers. By 1937, however, there were only seven sheep and one pig on the island, rats and mice apparently not having reached it (Hagen, 1952:210). The rails appear not to have been detrimentally affected. Hagen (1952:197) estimated the population of A. rogersi at around 1200, but Elliott (1957:578) felt that this was much too low and that the figure would lie somewhere between 5000 and 10,000. The rails are found in small groups throughout the island in the dense fern and tussock growth, and communicate by various twittering trills. Three stomachs examined by Hagen contained seeds, berries, and insects. Atlantisia rogersi is not known to feed on the eggs or regurgita of seabirds even though Inaccessible is the breeding station for thousands of petrels and penguins. The rails are preyed upon, one being recovered by Hagen from the stomach of a Skua (Catharacta), but such predation is probably not significant.
The first rallid described from the Tristan group was a gallinule “closely allied in general aspect to our Common Water-hen (G. chloropus), though readily distinguishable on accurate comparison” (P. L. Sclater, 1861:260). P. L. Sclater received specimens of this bird in 1861 “brought from the island of Tristan d’Acunha to Cape Town by a person formerly in the service of Sir George Grey” the then governor of Cape Colony who forwarded the birds (skins, specimens in spirits, and a living example) to London. The species was described as being slightly larger than Gallinula chloropus, with darker plumage, heavier legs and pelvis, reduced wings and sternum, and stouter bill, and was given the name Gallinula nesiotis by Sclater.
A similar bird was later collected on Gough Island, over 400 km from Tristan, by George Comer. This was described as a new species, comeri, by Allen (1892) who erected for it and nesiotis the genus Porphyriornis, based solely on the shape of the bill and the heaviness of the feet. Ripley (1954) quite justifiably sank Porphyriornis into Gallinula and considered comeri merely as a subspecies of nesiotis. Rand (1955), who saw no specimens, followed suit in the latter decision but was ambiguous about whether the species may have been derived from Gallinula, Porphyrula, or Porphyrio (let me emphatically interject here that the latter two genera have no close relationship to the Tristan-Gough birds) and he maintained the genus Porphyriornis. Voous (1961) and especially Eber (1961) have given comprehensive accounts of the Gough bird, comeri, both concluding that the species was clearly derived from Gallinula stock and should be included in that genus. With this there can now be little dispute.
Eber (1961:254), however, felt that it was wholly improbable that two such similar flightless species as G. nesiotis and G. comeri could develop on islands so distant from one another. Because no specimens were taken on Tristan after 1861 and because of the “round about way” (Umweg) that the original specimens reached Sclater, Eber concluded that the original specimens of G. nesiotis actually came from Gough Island and that the species never existed on Tristan (thereby making comeri a junior synonym of nesiotis). But, what Eber overlooked or disregarded were the several early accounts of the island of Tristan that mention a flightless gallinule.
Beintema (1972) has presented a thorough summary of the old accounts of gallinules on Tristan, and a review of the history of extant specimens attributed to G. nesiotis. A report of a gallinule on Tristan earlier than any cited by Beintema may be found in Pike (1873:34) who mentions the visit of a Captain Patten from the sealer Industry who stayed on Tristan from August 1790 to April 1791 and described there “a bird something like a partridge, only it is black, and cannot fly” as being “abundant.” The accounts of Lambert (cf. Holdgate, 1958:19), Carmichael (1818), Earle (1832), Stirling (in Stoddart, 1971), and Gurney (1853) establish beyond doubt that a gallinule was resident on the island of Tristan da Cunha and was common enough to have played a fairly important role in the diet of the islanders.
In his study of the gallinule specimens allegedly from Tristan, Beintema (1972:112) concluded that only the skin of the type of G. nesiotis and two incomplete skeletons could be considered as definitely having been taken on Tristan rather than on Gough. He could find no external differences between G. nesiotis and G. comeri but reported that the smaller sterna and pelves of the two skeletons of G. nesiotis indicated a bird of “much less heavy build” than G. comeri. He felt that the two forms “obviously belong to the same species, Gallinula nesiotis Sclater (1861)” (Beintema, 1972:107).
What is known of the disappearance of G. nesiotis, late in the 19th century, is documented by Beintema (1972) and is presumed to have resulted from the combined depredations of cats, rats, man, and dogs. On the other hand, G. comeri on uninhabited Gough Island, still exists and in good numbers, despite a large population of introduced mice. Holdgate (1958) has given an account of this species in its native haunts. It is an inhabitant of the forest and tree-fern zones of the island but not the open mountain areas. Elliott (1957:579) also found the birds “very common in the boggy areas bordering the Glen” and observed them feeding on grass heads. Like Atlantisia rogersi, the Gough Island gallinule probably finds its only natural enemy in the skua.
The peregrine Porphyrula martinica once again is known as a regular vagrant to Tristan da Cunha. Rand (1955) has summarized numerous records of this bird, all apparently from the main island of Tristan. Voous (1962:110) adds another record, describing the species as a “regular visitor.” As at Ascension and St. Helena, the birds are mainly immatures and occur during the northern spring and summer. The species is of such regular occurrence on Tristan that the natives even have a vernacular name (“guttersnake”) for it (Hagen, 1952:201). There is no evidence of its breeding on the island, however.
W. L. Sclater (1911:94) discusses a specimen of what he describes as a quite young coot “very closely allied to Fulica armillata and F. ardesiaca from South America, though it presented some differences from those species,” as having been obtained by a “Mr. Keytil” [sic] and forwarded to him by the director of the South African Museum. It is this same specimen that is referred to by Winterbottom (1958) as Fulica armillata and said to have been collected on Tristan and received from Keytel by the South African Museum in 1910. Winterbottom (in litt., to G. E. Watson, 21 Feb. 1972) reports that this specimen is currently under study by Beintema and “that it is an immature of one of the South American coots (definitely not F. americana …, though armillata is one of the possibilities).”
ORIGIN AND EVOLUTION OF THE SOUTH ATLANTIC RAILS
… imprison’d in the viewless winds,
And blown with restless violence round about
The pendent world.
Shakespeare
Measure for Measure
Act III, Sc. 1
The various facts previously outlined may be coordinated to arrive at a coherent explanation of the origins and relationships of the different groups of South Atlantic Rallidae. Each group contributes something different to this explanation and a fairly logical progression of ideas obtains. The distribution and status of the South Atlantic rails may be summarized as follows:
Ascension Island
Atlantisia elpenor (extinct endemic)
Gallinula chloropus (vagrant)
Porphyrula martinica (vagrant)
Porphyrula alleni (vagrant)
St. Helena Island
Atlantisia podarces (extinct endemic)
Porzana astrictocarpus (extinct endemic)
Gallinula chloropus (abundant breeding resident, recent colonizer)
Porphyrula martinica (vagrant)
Porphyrula alleni (vagrant)
Tristan da Cunha group
Atlantisia rogersi (resident endemic, Inaccessible Island)
Gallinula nesiotis (recently extinct endemic, Tristan)
Porphyrula martinica (regular vagrant, Tristan)
Fulica sp. (rare vagrant, Tristan)
Gough Island
Gallinula comeri (resident endemic)
- bibliographic citation
- Olson, Storrs L. 1973. "Evolution of the rails of the South Atlantic islands (Aves: Rallidae)." Smithsonian Contributions to Zoology. 1-53. https://doi.org/10.5479/si.00810282.152
