The Manx shearwater had been closely studied (Lockley, 1942) on Skokholm, where they breed in the burrows in the loamy earth above sandstone. On Skomer they find the sandy deposits of the plateau easy to excavate; although most of the work has already been done for them by the rabbits they like to enlarge and spring-clean the nesting chamber during the early days of spring. They first come to land on mild dark nights in February, and by mid-March the majority of the tens of thousands of old breeding stock have staked a claim to the familiar nesting burrows, easily ousting the rabbits which may have used the holes as winter refuge.The hooked beak of the shearwater is a weapon which can draw blood from human flesh and a rabbit must find it too formidable to argue with.
As in so many colonial nesting species this early return to the nesting ground is undoubtedly associated with the shortage of desirable nestsites. As the date of egg-laying approaches, one of the pair (and sometimes both) remains in constant possession of this coveted territory. (This is also the case in the gannet, guillemot and storm- petrel especially, and to a lesser degree in the crowded colonies of many kinds of gulls and terns and razorbills.) Possession is nine-tenths of the law, it is said; and many seabirds become so firmly attached to a nest site that they appear to possess a powerful moral indignation which in turn gives strength to the attack which they will make upon the casual visitor of the same species attempting to usurp the site in the temporary absence of the owners. This we suspect to be the case with the Manx shearwater; the fact that stray (unringed) birds found visiting (a common habit early in the year) the home of a marked pair rarely succeed in dispossessing them subsequently is an indication of a strong territory- defending ability in this bird; as of course the survival of the species requires.
The nest-site appears to be the meeting and mating place of the pair; it is unlikely that many seabirds recognise each other at sea. Although Richdale believes that royal albatrosses pair for life and spend the off-season at sea together, we have, at present, no evidence of this being the case in any British seabird. But we may well believe that, in the majority of instances, the same birds pair each spring. Male and female have grown used to the same burrow or nest-site, and make straight for it at the commencement of the breeding season, and so meet there once more to resume their former relationship in the annual business of mating, incubation and the rearing of the young.
Sight recognition of a mate at sea or in flight may or may not be possible-we do not know; but recognition by sound or call-note exists and must be particularly important in the nocturnal burrow-nesting birds. Shearwaters and petrels are most noisy on dark nights when the incoming bird must find it harder to locate its nesting hole among the thousands of other holes in a large breeding warren or heap of rocks. The unearthly screams and cries of the in- flighting birds are answered by the cries of the birds which have been in the burrows all day, and must help to localise the flight of the arriving bird and guide it home. On bright moonlit nights shearwaters are much more silent, and very much fewer come to the island then; this on Skomer is believed to be due partly to fear of the predatory gulls which actively hunt for shearwaters on moonlit nights. But Glauert (1946) has shown that in Western Australia the little shearwater (Puffinus assimilis) visits Eclipse Island in the same numbers both on dark nights (when it calls loudly) and on moonlit nights (when it is almost silent), and there it has no predatory enemies. Large powerful members of the petrel tribe with no natural enemies tend to be diurnal when on land, e.g. the albatrosses and the great shearwater, while the fulmar is active throughout the day.
On Skomer members of the survey party found a good deal of healthy amusement in going out on dark nights 'shearwatering'. This activity took place round about midnight as soon as the cries of the incoming shearwaters had begun to fill the air about the old farm- house. Members worked in pairs, one catching the birds, the other clipping the rings on. We encouraged this occupation, on the principle that the more we ringed the greater the number of recoveries. The shearwater was temporarily dazzled by the beam of a torch and could be quite easily and gently picked up, and, if care was taken to hold the bird with one hand close behind the neck and over the closed wings, bites from the hooked beak were avoided. This island night's entertainment was a delightful experience: above was the vast black clouded dome filled with a bedlam of weird screams; beneath was the bracken (filled with the fairy lights of the glow-worms) and the heather, and the silent, still or suddenly scurrying forms of the shearwaters; now and then came a sigh from the wind in the rocks or the distant sea, and much laughter, punctuated with agonised curses when a beak gripped a tender finger, from the 'shearwatering' party. In the middle of the summer the really dark nights fit for 'shearwatering' were all too few for the stream of eager volunteer watchers which each week arrived at (and departed from) Skomer. One thousand six hundred and sixty-seven shearwaters were ringed during this summer.
The work had its reward in several interesting recoveries, and confirmed for Skomer shearwaters the migration, already discovered as part of the routine flight of the Skokholm shearwaters, south to the Bay of Biscay (in April and May many birds were recorded from near Bilbao and Biarritz) on a feeding expedition (Recovery of Marked Birds, Brit. Birds, December 1947 and 1948). Almost more interesting than these was the recovery on Skokholm on May 27th, 1947, of a shearwater ringed on Skomer as an adult on July 6th, 1946 (this bird we suspect may in 1946 have been a 'sweethearting' non- breeder); and of a second ringed as a young bird on Skomer on September 14th, 1946, recovered on Skokholm on June 3rd, 1947. We may appropriately record here too the recovery on Lundy Island (where only a small colony exists) seven days later of a shearwater ringed on Skokholm on July 11th, I947; and on April 30th, I948, another was recovered on Lundy which had been ringed on Skokholm on June 23rd, 1946; thirdly, there has been an interchange the other way: a shearwater ringed on Lundy, July 18th, 1947, was recovered on April 11th, I948, on Skokholm. These valuable records show that there is an interchange of individuals between the three Bristol Channel breeding grounds of the shearwater. They are valuable because they prove that the species is not so insular as might have been supposed - a fact which reduces the possibilities for subspeciation; and also it is significant that some of these birds were ringed in the latter part of the season and were probably birds which had not yet bred or established homes in the area; suggesting again that it is the young birds which wander most and found new colonies, the old birds being content with their well-established sites.
This wandering and exploration is characteristic of many birds in their first year. Numbers of migratory birds, obviously immature, appear on the islands in late May, June and July, wandering almost as if lost, while adults of the same species are well advanced in breeding operations in their normal environment. It is to be observed in the wanderings of young and adult gannets which we have ringed on Grassholm; the young birds go as far south as Senegal in their first winter, but the majority of the adults stay in home waters.
What happens to the storm-petrel at sea ringing has so far not recorded; of about one thousand individuals which have been ringed on Skokholm since the Bird Observatory was opened there in 1933 none has been recovered except where ringed (i.e. in nesting-holes). On Skomer storm-petrels were scarce- there were only three small colonies (twelve pairs at the North Landing, six pairs at the Basin, and four pairs at the Waybench) and a few isolated pairs in the hedgewalls-in all we discovered about thirty pairs. (In 1939 about six hundred pairs bred on Skokholm, where they arrive at the end of April.) On Skomer they were first heard calling on May 11th It is not easy to prove why they are so scarce on Skomer but it has been suggested that they (or their small chick, which is very feeble at birth) are subject to persecution from the numerous small mammals of the island which are known to eat flesh: the Skomer vole, the woodmouse and the shrews. As already mentioned, the little owl, as on Skokholm, persistently preyed on the storm-petrel on Skomer. Leo Drury describes in the log-book how the petrels breeding at the North Landing June 15th, 1946) ceased their courtship flight and purring song when a little owl wailed in the distance and later perched upon the wall in which the petrels were nesting. They appeared terrified and did not recommence their activities until half an hour had elapsed after the owl had flown away. It is obvious that this owl could be the cause of the extirpation of the small colony of storm-petrels on Skomer.
We may reflect here on the strange fact that so many of the seabirds, which reproduce slowly at the rate of one egg per pair per year, are heavily preyed upon by species which reproduce at a much greater rate (gulls lay two to three eggs and owls three to five eggs), yet the preyed- upon remain far more numerous than the predators. For the annual mortality rate of the former must be, contrary to appearances on Skomer, in proportion to the number of eggs laid, assuming a reasonably static population, which is the case. We shall refer to this later.
The slaughter of the helpless shearwaters and puffins by the gulls is a prominent everyday feature on Skomer, and they are also killed by buzzards, peregrine falcons and ravens. Disaster to the egg or young chick in the crowded burrows is not infrequent either. Later, at fledging time, the young birds, deserted by the adults, are killed by the predators. Even when these young birds reach the sea they are sometimes caught in severe equinoctial gales and blown far inland or battered to death on a lee shore. More rarely death is due to striking wire fences on the breeding islands; we have seen shearwaters and puffins caught by the wing upon barbed wire. On September 28th,1946, one fledgeling shearwater was found in a fork of ragwort, hanged by the neck - an unusual accident.
Seabirds which nest in grassy situations may be attacked by grass-ticks (Ixodes, spp.); puffins sometimes have their faces quite darkened with the swollen bodies of these blood-sucking parasites, which when full-fed drop to the ground and then moult or (if mature females) lay over 2,000 eggs in the earth or grass roots. These eggs hatch in about two months and the young larval ticks seek a warm-blooded vertebrate host. They seem to be able to withstand salt water during the visits of the seabird host to the sea, but obviously those that are attached to the bird when it has moved out into the ocean on its winter migration must perish as soon as they are engorged and relax their grip. Although ticks are known to carry the germs of certain diseases (as of redwater in cattle), so far they have not been proved to act as host in the few known diseases of seabirds.
The only disease noticed in the crowded seabirds' colonies of Skomer was an epizootic affecting the shearwater. David Surrey Dane in 1947 investigated the epidemiology of this disease which was previously unknown, although some of the symptoms were described by Lockley (1942) in shearwaters on Skokholm before the war. Surrey Dane (1948) records that the disease was first discovered on August 31st, 1946, when two fledgeling shearwaters (deserted by the adults, as is usual in this species) were found in North Valley, Skomer, with blisters on the webs of the feet; many more were later found in the same area, some with blistered feet, some with conjunctivitis, and some with both conditions. Many of the affected young shearwaters died, some becoming blind before death. In 1947 Surrey Dane found that there was no conjunctivitis but only blistered feet in the victims, whose movements he studied at night by means of leg-rings. Transmission of the disease seemed to be by contact of the webbed feet with soil and stones, already contaminated from burst blisters, during the nightly scramble of the independent young bird to high points on hedge-walls and banks where, for perhaps an hour or so, it exercises its wings before returning to its burrow or (after several nights of such wing-beating) taking to the sea. The incubation period of the disease appeared to be about eight days or less, but the manner in which it arises is obscure; it is possible that other hosts (a trombiculid mite is mentioned) exist besides the shearwater. It is also possible that the same disease may attack other seabirds. In August 1947 three juvenile herring-gulls with blistered feet died on Skokholm, after a few days of sickness.
The life histories of the Manx shearwater and the storm- petrel have been described elsewhere. These birds have unusually long incubation and fledging periods. In the Manx shearwater these average 51 and 72 days respectively; and in the storm-petrel 38 and 61 days. This means that they are the last to leave the island at the end of the summer. There were still numbers of young shearwaters in the burrows when the island was evacuated by its human inhabitants in early October; although the last adult shearwaters had abandoned their chicks by the end of September. Storm-petrels would then still be feeding their young - such as had survived the little owls (and one petrel at least was seen on October 1st).
The fulmar petrels had been seen daily at sea from March 25th. They were first watched on the cliffs on April 7th and rapidly increased during that month. In May and June courtship was in full swing. A typical entry in the records may be quoted:
But no egg-laying took place anywhere on these and other favoured ledges; and from their peak numbers of 33 on May 7th they gradually dwindled to ones and twos in mid-July. In August one bird was seen flying over the sea on five occasions but it did not settle on the ledges. One was last seen on September 2nd and 4th, flying in North Haven. In 1947 the fulmar appeared in the same numbers but again failed to breed.
This non-breeding in apparently adult birds is a fact which the field ornithologist constantly stumbles upon. It is plain in the fulmars which came to Skomer, haunting the northern cliff ledges until after midsummer and then disappearing without breeding. Darling in Bird Flocks and the Breeding Cycle (1938) has shown that there is a numerical 'threshold' for certain species below which it does not thrive and may not survive; this threshold, which is however more variable than Darling suggests, must be reached too before a species can successfully colonise a new territory. On the island of Lambay, for example, half a dozen fulmars were able to start breeding in 1933 and the same number bred at New Quay, Cardiganshire, in 1947. On Skomer in 1946, however, although thirty-three individuals were seen in one day on ledges, no eggs were laid.