Nesofregetta albigularis (Finsch)

Comprehensive Description

provided by Smithsonian Contributions to Zoology
Nesofregetta albigularis (Finsch)


PELAGIC DISTRIBUTION.—This species was widely distributed in the Pacific along equatorial latitudes. The vast majority of sightings in both the central and eastern Pacific is confined to areas between 10°N and 10°S (Figures 122 and 123). Concentrations about the Phoenix and Line islands in the central Pacific are easily referable to the breeding populations of those areas. The equally great concentrations recorded in the eastern Pacific west of the Galapagos Islands are of unknown origin, but most probably from the breeding stations in the Marquesas Islands and Christmas Island. It should be pointed out that our data represent a considerable range extension to the known pelagic range.

TAXONOMY.—To my knowledge this species does not vary throughout its considerable range, and no subspecies have been proposed. Except for Oceanodroma tristrami, N. albigularis is approximately the largest storm petrel in the Pacific area (Appendix Table A). Most POBSP specimens are from the Phoenix and Line islands (Christmas Island) populations.

In the Phoenix group, the majority of birds are of the color morph usually described for the species, i.e., dark upperparts and chest band, with white rump, throat and belly. On both Phoenix and McKean islands, however, every conceivable color variation between the normal form and totally dark individuals may be found (Figure 124). The “Samoan Storm Petrel” (“Fregetta moestissima” Salvin), known from the unique type taken offshore the Samoan Islands, is considered by Murphy and Snyder (1952) to be nothing more than an entirely melanistic example of Nesofregetta albigularis. The appreciably larger dimensions (especially wing) than those of any specimens from the Phoenix, Line, Marquesas or New Hebrides islands indicates it may be from a distinct population, possibly breeding somewhere in the Samoan Islands. Bourne (1957) indicates that the unique, all dark specimen of Fregetta fuliginosa (Gmelin) from Tahiti (specimen subsequently lost) was probably closely allied to the Samoan form. On Christmas Island about half the population of Nesofregetta albigularis shows some dark flecking on the white underparts, but completely melanistic birds have not been seen.

BREEDING BIOLOGY.—Published data on the breeding biology of this species is virtually nonexistent and thus I have entered POBSP data in considerable detail in comparison to other species whose breeding biologies are well known. The breeding range of Nesofregetta albigularis extends from the Line Islands (Christmas Island) and Phoenix Islands (Phoenix and McKean islands) in the central Pacific south to and including the Samoan Islands (?), Marquesas Islands, Fiji Islands, and the New Hebrides Islands. The POBSP surveys covered only the more northern parts of its range, i.e., the Line and Phoenix islands. Since our surveys were fairly complete for the Phoenix group, it is unlikely that N. albigularis breeds on islands other than McKean and Phoenix, although we have several sightings of birds flying over Enderbury Island (25 to 27 September, and 9 to 10 October 1966). A specimen was collected on Howland Island on 24 October 1938, but we have no recent sightings.

In the Line Islands, however, islands such as Maiden and Starbuck are quite large and our coverage was insufficient to rule out the possiblity of birds breeding in those areas. On Maiden Island one bird was noted at night on 31 March 1967. Several islets in the lagoon were not thoroughly investigated and might be utilized by this species for nesting. Other possible breeding areas are the Gilbert, Society, Tuamotu and Marshall islands (King, 1967:42).

The choice of breeding sites on the three central Pacific islands investigated was much the same. Areas supporting bunch grass (Lepturus sp.), large clumps of weeds (Boerhaavia sp.) or Sesuvium, in combination with an abundance of loose rocks, were utilized in practically all cases in the Phoenix Islands. Extensive areas of rocky coral slabs without vegetation, as well as areas of continuous short grass or small annuals, were avoided. Old fallen rock walls which were constructed by the early colonists on the Phoenix Islands were often used as nest sites, although standing walls devoid of vegetation were much more prone to be utilized by Bulwer’s Petrel (Bulweria bulwerii). On Christmas Island most nests on the small islets were under clumps of overhanging grass (Lepturus), in pockets in coral rock shelves also usually sheltered by grass, or in depressions under Sesuvium, Regardless of nest site, the actual nests were similar in all cases. The few pieces of nest material (grass, weed stems, feathers, etc.) invariably present were most likely in the nest cavities to begin with; such material is probably manipulated to some extent, but to my knowledge no material is ever carried into the nests by the birds. Incubating birds are usually hidden from view within nest cavities, but an occasional individual will nest in rocks in such a manner that it can be seen. Schreiber (POBSP) noted that tails of incubating birds were often visible in the shallow nest cavities on the islets in Manulu Lagoon, Christmas Island. Under large, extensive grass or weed clumps, the birds usually retreat to the most concealing portion. The nest cavities may be slightly enlarged under clumps of vegetation, but I have never noted any extensive digging performed by this species. Storm petrels often utilize nest cavities dug out by the abundant Audubon’s Shearwater (Puffinus Iherminieri) on McKean and Phoenix islands.

On Christmas Island the nesting areas are primarily limited to five tiny islets in Manulu Lagoon. These are separated by expanses of water from the main island. A few birds nest on the islets of Motu Tabu, Motu Upua, and on at least one islet in Isles Lagoon. Utilization of these retreats is probably a direct result of feral cat predation on the main island.

From banding and recapture data Schreiber (POBSP) estimated the total population using Christmas Island in 1967 at 350 to 450 birds, although this estimate is probably low. Table 29 summarizes the stage of nesting and nest contents recorded in Manulu Lagoon by POBSP and others.

From the above data it can be inferred that the peak egg period on Christmas Island is from September through November, but obviously some portion of the population is nesting at all times of year. This seasonal spread of nesting apparently allows a greater number of birds to utilize the available nesting area than would be possible if nesting were limited to a specific time of the year for all individuals. One adult, with an egg, banded on 25 August 1967, on Islet #2 in Manulu Lagoon was found incubating an egg on Islet #1 on 28 November 1967. This meager evidence suggests that the species will renest within the same year, and, furthermore, a specific nest site is apparently not adhered to. Another adult in an empty nest cavity on 29 August 1967 was incubating an egg there on 4 November 1967; yet another incubating an egg on 17 August 1967 was still incubating on 4 November 1967. The egg in the latter case was probably infertile, but these examples give an idea of the extensive periods that the birds spend on the breeding islands.

Our data for the Phoenix Islands are less complete because of the extremely short surveys made, but the available evidence shows that on McKean and Phoenix, as on Christmas Island, some portion of the population is nesting during every month of the year. Tables 30 and 31 present POBSP data for McKean and Phoenix islands.

Variation and inconsistencies in the data are due partly to different observers and to the extremely inadequate surveys in most cases. It is nonetheless evident that considerable variation is to be found on the same island for a given month during different years (Table 30, February 1964 and 1965). Despite these differences, considering the data as a whole, there appears to be a breeding peak on McKean from April through at least July or August, with a definite low in October and probably later. For Phoenix Island we do not have data from monthly surveys in successive years, but certainly the high nesting populations recorded from October through December do not correspond to the breeding situation on McKean. This lack of conformity to a set seasonal or annual cycle has been noted for most species on the equatorial islands. Our surveys indicate that the populations on one island appear to operate independently from those on other islands, regardless of the apparent ecological similarities of the islands or their geographic proximity.

On McKean Island there appears to be a definite separation in the timing of nesting of Nesofregetta albigularis and Bulweria bulwerii. This was quite apparent during two fairly extensive surveys in July and October 1968. During July, N. albigularis was definitely undergoing a peak in nesting with an estimated 1000 adults on the island. Bulwer’s Petrel at that time was for all practical purposes absent from the island (one adult calling from an empty nest cavity). During the extensive October survey only about 30 adult White-throated Storm Petrels were present on the island with nine nests (3 with young, 1 with egg, and 5 occupied, but as yet without eggs). Bulwer’s Petrel at this time was approaching a peak nesting situation with a very conservative estimate of 1000 birds on the island. The majority of these were calling from empty nest cavities or from nests with fresh eggs; an estimated five percent had small downy young. Both species tend to utilize the same nest sites to a high degree, and this complementary seasonal nesting arrangement apparently affords both populations minimum competitive activity during their respective breeding periods. Food availability in the nearby ocean may also be a factor in creating alternating nesting seasons for these similar species.

On the breeding islands, White-throated Storm Petrels are usually absent during most of the daylight hours (except for those attending to incubation or brooding duties). On several occasions, however, I have seen adults come to nests in full daylight to feed young. This type of behavior is never performed by any of the small hydrobatids in the eastern Pacific, since any individual showing itself during the day would be immediately set upon and devoured by Western Gulls. The rapidity with which the gulls attack storm petrels appears to be the sole reason why storm petrels never appear overland in daylight hours. In the central Pacific, however, Nesofregetta appears to be without predators. The frigatebirds (Fregata ariel and F. minor) were never noted harassing storm petrels, but seemed to restrict their parasitism to larger birds such as boobies, shearwaters, and terns.

Adult storm petrels normally come in over the breeding islands from the ocean from about one-half to three-quarters of an hour before dusk. This seems to be the period of greatest activity, during which many individuals may be noted flying back and forth over potential or actual nest sites. Birds with active nests seem to circle repeatedly over the general area of the nest site for some time before landing, as do individuals obviously searching for nest sites. When the birds land, they usually do so ungracefully, falling softly and then awkwardly crawling over the terrain toward the nest site or into the nest cavity.

As with other hydrobatids and procellariiforms in general, Nesofregetta lays only a single egg. Occasionally a bird is found incubating two eggs, but one is invariably an infertile or abandoned egg from an earlier nesting. Relative to other storm petrels, the eggs are large for the size of the bird. A series of 20 eggs from Phoenix, McKean, and Christmas islands show the following measurements: length (mean with standard error) 37.95 ± 0.41 mm (SD=1.85); breadth 27.45 ± 0.18 mm (SD=0.82). The eggs have a creamy-white background color, with reddish and lavender dots which usually form a prominent wreath about the large end, but are scattered less finely over the entire surface.

The few observations available on marked nests show that both birds are present in the nest cavity for some time (days?) before the egg is laid. Copulation probably takes place at the nest site. During these times and at least during early incubation, adults utter a soft guttural “grrrrr-grrrrr-grrrrr” call from the nest cavity. Both sexes incubate, but the rate of changeover, as well as the length of the incubation period, is unknown. After the chicks are a few days old the parents only come to the nest to feed them. The young hatch blind, but are profusely covered with medium-gray down which is quite short on the underparts and long on the dorsum. A chick estimated to be five days old had down to 20 mm long on the back. I doubt if the young can use vision in the nest until they are in the large, fully-feathered stage, since the entire orbital region is completely covered by shaggy down. When the chicks are approximately two-thirds grown, they have a definite bi-colored aspect. The dark and light colors correspond roughly to the black and white areas of the adults. This bicolored effect is brought about by the growth of new feathers beneath the down. As the feather coat progresses, more and more down is shed until only patches remain about the head, dorsum and flanks (Figure 125). It is not known how soon the young leave the island after attaining their juvenal plumage, but apparently it is rather soon, as birds with traces of down still clinging to the feathers were noted attempting to fly at night; they were at least many yards away from any possible nest site (see comparative data on Oceanodroma leucorhoa socorroensis young).

The White-throated Storm Petrel exhibits a peculiar flight at sea which has not been noted in any other species of hydrobatid. On many occasions these birds were seen kicking off waves with considerable force, using their extremely long legs and spatulate feet. The moment their feet hit the water surface, the birds instantly change direction and sharply increase their speed, attesting to the force applied to the water’s surface. The usual procedure was for the individual to “kick off” a wave, glide for approximately 20 to 30 seconds and again “kick off.” I have on occasion noted a bird parallel a ship for a mile or more in this manner. Both Gallagher (1960) and Schreiber and Ashmole (1970) have noted birds performing this behavior in the quiet lagoon on Christmas Island. Wilson’s Storm Petrels often dabble their feet in the water, but this appears to be more of a balancing maneuver than an actual thrust motion as in Nesofregetta. This mode of travel might be a means of expending less energy in equatorial regions where wind is often at a minimum. At any rate, the extremely long tarsi and spatulate feet of Nesofregetta make possible its unique locomotory behavior.


The hydrobatids form a rather homogeneous group widely distributed over the Pacific regions. In many cases their breeding distribution and pelagic movements appear to be random, but undoubtedly reflect patterns that have been dictated by seasonal abundance and distribution of food items, interspecific competition, and the need to occupy the limited breeding areas during favorable periods. These patterns have undoubtedly undergone considerable fluctuation over time; the present view of the family shows a few exceedingly abundant forms and many uncommon to rare species. In the Pacific region, individuals of the combined subspecies of Oceanodroma leucorhoa by far outnumber all other species combined.

The northern breeding forms of O. leucorhoa show south-north migration patterns and practically inundate the central and eastern equatorial Pacific during the nonbreeding season (September-March). Wilson’s Storm Petrel, a strictly southern hemisphere breeder, shows the inverse north-south movements in the western Pacific and along the South American coast during migration and nonbreeding season from the Antarctic. Another southern hemisphere breeder (White-faced Storm Petrel) demonstrates an east-west pattern, moving east from its Australian breeding grounds to the South American coast, thence north along the Peru Current and west through the Galapagos area with a subsequent return to the Australian region. Possibly Wilson’s performs more of a circular route in its north-south movements. A number of forms such as Oceanodroma leucorhoa chapmani, O. l. socorroensis, and O. homochroa apparently do not move any great distance from the breeding islands during the nonbreeding season. For many Pacific hydrobatids the nonbreeding distribution is imprecisely known; for others the stated breeding grounds are merely conjecture.

At sea most storm petrels are either solitary or join with members of their own species. In certain cases two or more species will congregate, often in response to a temporarily plentiful food supply. They normally do not mix with other seabird species, even with their nearest relatives, the procellariids. I believe the presence of storm petrels in mixed feeding flocks of procellariids, terns, boobies, etc., is usually a random occurrence. On many occasions when I have observed storm petrels in the vicinity of mixed feeding flocks, the hydrobatids appeared to lack the intense feeding interest of the other species and usually moved out of the immediate area of intense feeding. It should be stressed that in most cases POBSP observers did not distinguish, in a recorded species association, those members that were actually feeding or traveling together. If members of two or more species were together at the time of initial sighting, they were usually recorded as such. The imprecision of these observations was the result, usually, of the impossibility of detailed notations of individuals in large mixed feeding or traveling flocks where often hundreds or even thousands of birds were concerned. In such cases a gross account of numbers was about the best that could be acquired. Thus the presence of one or more hydrobatids in the immediate vicinity of a flock would be recorded as positive, although, as I have often stated from personal observations, this is usually not the case.

The amount of intraspecific sociability can be calculated in a gross manner quite easily from our data. From Table 32 it is immediately apparent that individuals of some species are far more prone to associate intraspecifically than others. Fork-tailed and Wilson’s Storm Petrels commonly associate with other members of their respective species. On the wintering grounds Leach’s Storm Petrels disperse singly or in very low densities unless a local food source happens to bring large numbers together temporarily. During migration they commonly group and are noted resting on the water in large rafts. Even with our sightings of migratory and feeding flocks included, the low sociability ratio of 1.6 for nearly 10,000 sightings (involving over 14,000 individuals) demonstrates their usual solitary nature. Other species, such as the White-throated Storm Petrel, show this solitary nature to a greater degree and are almost invariably noted singly. For most of the less abundant species, although the available evidence suggests solitary occurrence, we do not have enough observations to demonstrate their overall social behavior at sea.

The solitary nature of most hydrobatid species away from the breeding grounds is perhaps necessary because of the foraging behavior (flight low over the water with a limited area view) to distribute themselves as widely as possible over their range to take utmost advantage of the widely distributed small organisms which make up their diet. Members of a species, or of several species, will gather readily at an unusual, abundant food source such as an oil slick. At such gatherings I have never noted either inter- or intraspecific hostile behavior between individuals, behavior that certainly some (if not all) species are capable of demonstrating as witnessed by intraspecific battles over nest sites on the breeding grounds (Leach’s, Least, Black, and White-throated Storm Petrels).

To my knowledge all members of the Hydrobatidae are strictly surface feeders, that is, they do not submerge the body beneath the surface of the water to obtain food. All Pacific species were observed to glean the surface of the water for floating objects, often with wings held aloft for balance. The foot dabbling behavior with wings held aloft for balance in the Wilson’s Storm Petrel is well known and documented. Leach’s, Harcourt’s, and most especially Least Storm Petrels will hold the wings aloft in a V formation when feeding over an oil slick or other concentrated food source. In this respect they appear as so many butterflies on a wet sand bar. Least Storm Petrels were never noted dabbling their feet the way Wilson’s do. They usually feed while resting on the water with wings up until wanting to move. Then with a few quick wingbeats the bird moves to a new position, usually still holding the wings aloft.

The preceding species accounts provide a basis for understanding the distribution and abundance of Pacific hydrobatids and indicate the vast amount of data yet to be gathered concerning these forms. While it is unlikely that a program devoted to pelagic ornithology in the Pacific of the magnitude of the POBSP will occur again in the near future, important contributions can be made to existing knowledge by participation in cruises devoted to any type of pelagic investigation. The great amount of information on hydrobatids gathered during the EASTROPAC cruises is a good example of this.

While it is understandable that we know little or nothing about the breeding biology of many of the southern hemispere hydrobatids, it is less explicable why information of this sort is lacking for most of the species and subspecies inhabiting the coastal islands of the western United States and Mexico. There is a decided need in these areas for long range studies with banded birds similar to the investigation of the nominate race of Leach’s Storm Petrel on Kent Island, New Brunswick, Canada, which was begun by Gross (1935) and continued by Huntington (1963) and his co-workers.

A fascinating problem awaiting study is the elucidation of the relationships and interactions of winter and summer breeding populations of Leach’s Storm Petrel on Guadalupe Island. A second problem is to account for the clumping of completely dark hydrobatids along the California and Baja California coasts. In the Hydrobatidae, where white patterning is widespread, the lack of this color in so many forms (four species and three subspecies) common to this area can hardly be coincidental. The factor of predation by the various gull species (especially the Western Gull) is probably only part of the reason. The dark rump color predominates in all forms of hydrobatids along the coast of Japan also, possibly for the same reasons.
bibliographic citation
King, Warren B. 1974. "Pelagic studies of seabirds in the central and eastern Pacific Ocean." Smithsonian Contributions to Zoology. 1-277. https://doi.org/10.5479/si.00810282.158