Chapter XI: Part V: , Birds. U.S. Army, Signal Corps, Washington, D.C. Pages (5)

Age of the bird affects time and energy commitment to foraging because younger birds are often less skilled at capturing food. This has been noted particularly in long-lived seabird species (Orians 1969; Dunn 1972; LeCroy 1972; Buckley and Buckley 1974; Barash et al. 1975). Older juveniles may be excluded from feeding areas by more experienced, territorial adults (Moyle 1966), whereas immatures are not (Drury and Smith 1968; Ingolfsson 1969).

Nutritional and energetic return obtained from food is a very important factor in foraging strategy that has not received the attention it deserves. Table 1 lists the caloric value of various foodstuffs and illustrates how little is known about foods eaten by seabirds. Although caloric content and abundance of food have often been accepted as the most important determinants of foraging strategies (Bookhout 1958; Emlen 1966; West 1967; Bryant 1973), they may frequently be less important than nutritional value and digestibility, also shown in Table 1 (Pulliam 1974).

Since fish seem to be highly digestible, most of the energy contained in them is available to the consumer. There are, unfortunately, no data on the digestibility of marine invertebrates, but those for insects suggest that digestibility, at least of crustaceans with exoskeletons, is somewhat lower than that for fish. A bird would therefore have to eat a larger biomass of invertebrates than of fish to satisfy the same energetic needs (although cost of procurement might not be as high as for fish).

Table 1. _Nutritional value of foods eaten by birds._ After data
in Hunt (1972) and E. H. Dunn (1973).

A: Kcal/g fresh wt.
B: Digestive efficiency
C: Kcal metabolizable energy/g fresh wt.

------------------------------------------------------------------------------
Percent fresh wt.
composed of:
----------------------
Food type A H₂O Fat Protein B C
------------------------------------------------------------------------------
Vegetable 1.2-5.2 59-86 0.4-3 3-5 30-32 0.3-2.3
Tropical fruits 1.2 75 8 1
Various seeds 4.0-7.3 3-13 1-40 10-29 76-80 3.0-5.2
Various insects 1.4-5.2 65-75 1-3 9-18 66-69 0.9-3.5
Whiting (fish) 1.1 81 79 0.9
Various freshwater fishes 1.2 75 5 18 81
Mix of fish eaten by
double-crested cormorants 1.1 74 1 16 82 0.9
on NE coast
Fresh herring and mackerel 1.9 67 13 19
Garbage ("average" mix) 1.5 67 8 19
Crab with eggs 1.0 68 5 1
Euphausid shrimp 0.8 80 2 1
Clam (edible part only) 0.8 80 1 13
------------------------------------------------------------------------------

A bird must satisfy not only energetic needs, but also nutritional requirements. Fish are high in protein (Table 1), but what little is known of marine invertebrates suggests a low proportion of protein in relation to total bulk. Protein is vital to growth of nestlings (Fisher 1972; Lemmetyinen 1972), and 4-8% protein in the diet seems to be required for minimal maintenance of adults (Martin 1968; Fisher 1972). Some seabirds (such as puffins, _Fratercula_) that eat a varied diet raise their young exclusively on fish (Bédard 1969_b_; Nettleship 1972; Sealy 1973_a_).

Other aspects of nutrition, such as vitamins and minerals, are also important to avian health (Brisbin 1965; Fisher 1972). To further complicate matters, nutritional values vary with season, as do birds' requirements for them (Myton and Ficken 1967; Moss 1972). Adults must adjust their time and energy allocation to foraging to optimize not only energetic, but also nutritional return.

Optimal time and energy allocation has been studied in theory (Orians 1971; Schoener 1971; Pulliam 1974; Katz 1974) and some direct observations have been carried out, largely on seedeaters (Bookhout 1958; Myton and Ficken 1967; Royama 1970; Moss 1972; Willson 1971; Willson and Harmeson 1973). The direct studies, in particular, point out the basic importance of studying cost-benefit ratios by interrelating complex factors of food availability, searching patterns, and type, size, and caloric and nutritional value of foods.

Time-energy Use Beyond the Cost of Living

This section concerns variation in time and energy allocated by seabirds to activities above and beyond the cost of living--particularly to migration, molt, and reproduction. Allocation to such items as avoidance of predation and competition is not considered here, because they are not readily analyzed as activities to which time and energy are devoted in a specific part of the annual cycle.

The previous discussion dwelt on energy considerations and could have referred to almost any group of birds. The following treatment centers on time use of northern seabirds. Little is known of energetics beyond the cost of living, although estimates have been made for certain aspects of migration, molt, and reproduction (Nisbet 1963; Hussell 1969; Hart and Berger 1972; Payne 1972; Ricklefs 1974). Essentially nothing is known, however, of the relationship of such costs to the amount of energy available to the bird once basic maintenance costs have been met (productive energy). Because such complete data are not available, the following account dwells largely on variation in timing and total time devoted to activities beyond basic maintenance.

_Migration_

Among northwestern North American seabirds, most coastal feeders, such as gulls, cormorants, and many alcids and petrels, have only a short southward migratory movement, and many others are more or less resident (Dorst 1961; Ashmole 1971). Terns, on the other hand, migrate long distances in a short time to places where small fish are available near shore in the winter. Other long-distance migrants--Sabine's gull _(Xema sabini)_, jaegers (_Stercorarius_ spp.), pelagic phalaropes (_Phalaropus_, _Lobipes_), and kittiwakes (_Rissa_ spp.)--tend to scatter widely over the southern ocean, concentrating near areas of upwelling (Dorst 1961; Ashmole 1971). Groups such as murres (_Uria_ spp.), eiders (_Somateria_ spp.), and grebes (_Podiceps_ spp.) may move considerable distances by swimming (Dorst 1961; Tuck 1960). True migration tends to take place directly before and after reproduction, whereas dispersal or nomadism takes place over a long period of the winter (Fig. 4).

Among species remaining in the northern hemisphere, younger birds frequently disperse greater distances than do breeding adults (Coulson 1961; Kadlec and Drury 1968; Southern 1967), and the degree of dispersal can vary among colonies of the same species (Coulson and Brazendale 1968).

Energy costs of migration must vary according to distances covered and amount of time allocated to migration. Aside from the references to cost of flight mentioned earlier, however, migratory costs have scarcely been studied. Dolnik (1971) has estimated that chaffinches _(Fringilla coelebs)_ expend about as much energy migrating south as they would on thermoregulation if they overwintered on their breeding grounds. Long-distance migration is presumably selected because the birds are able to collect food more efficiently, because the risks of death or injury in migrating are less than in residency, and so on. Interspecific and intraspecific competition may also be involved (Cox 1968). In other words, migratory patterns are selected to optimize survival and reproduction in alternating environments (Cohen 1967; Drury and Nisbet 1972).

Table 2. _Wing molt in alcids._ After Stresemann and Stresemann
(1966).

---------------------------------------------------------------------
Rapidity of wing molt Timing of start of molt Species
and (indented) flight
capability in molt

Slow Retained During care of young Cassin's auklet,
parakeet auklet
_(Cyclorhynchus
psittacula)_,
whiskered auklet
_(Aethia pygmaea)_

Retained After young become Least auklet
independent _(A. pusilla)_,
crested auklet _(A.
cristatella)_

Rapid Poor After arrival in Marbled murrelet,
winter quarters Kittlitz's murrelet
_(Brachyramphus
brevirostris)_

Almost synchronous None After end of breeding Xantus' murrelet
_(Endomychura
hypoleuca)_

Synchronous None As soon as young go to Guillemots (_Cepphus_
sea spp.), murres,
razorbill _(Alca
torda)_, dovekie
_(Alle alle)_

None In winter, after body Puffins
molt
---------------------------------------------------------------------

Although one may suspect that location of winter food supply is the main environmental factor affecting migratory patterns, there is little direct evidence on the reasons for, or the benefits accruing from, the different patterns seen in seabirds. Study of cost-benefit ratios of foraging in different stages of migration might help clarify the question.

_Molt_

Patterns of molt vary widely among seabirds. The commonest pattern is for a prenuptial body molt to occur in spring, and for an extended wing molt to begin after the breeding season and continue well into the winter (Fig. 4). In short-distance migrants, molt may overlap slightly with the end of breeding and can last up to 6 months, as in most gulls, terns, alcids, nonmigratory jaegers, and cormorants (Stresemann and Stresemann 1966).

Long-distance migrants frequently delay molt until in the winter quarters (lesser black-backed gull, _Larus fuscus_; Sabine's gull; jaegers; arctic tern, _Sterna paradisaea_; and marbled murrelet, _Brachyramphus marmoratus_) and molt there may occur rapidly (3.5 months in the arctic tern). Certain other long-distance migrants begin molt before leaving the breeding grounds (herring gull; skua, _Catharacta skua_; Leach's petrel, _Oceanodroma leucorhoa_; and fulmar), although molt may be interrupted during migration, as in _Larus argentatus heuglini_ (Stresemann and Stresemann 1966). Duration, timing, and rapidity of molt are particularly varied among the alcids (Table 2).

A few unusual molt patterns are found in northern seabirds. The ivory gull _(Pagophila eburnea)_ has its major annual wing and body molt immediately before it breeds. In several other species such as the glaucous gull _(Larus hyperboreus)_ and Cassin's auklet _(Ptychoramphus aleuticus)_ the molt almost completely overlaps the reproductive cycle (Johnston 1961; Payne 1965). Potts (1971) documented a molt pattern in shags _(Phalacrocorax aristotelis)_ which is more typical of tropical seabirds. Several cycles of wing molt take place simultaneously, each lasting more than a year, and molt ceases in winter. By the time breeding age is reached, each flight feather is replaced once a year.

Within these broad categories of molt pattern there are sometimes variations according to age, sex, and even subspecies (Stresemann and Stresemann 1966). Male common eiders _(Somateria mollissima)_ molt directly after mating, when their reproductive role is completed, whereas females molt only after they have taken their young to sea. Nonbreeders and failed breeders frequently begin molt while other adults are still raising young and not molting--e.g., many alcids, gulls, storm-petrels, and fulmars (Stresemann and Stresemann 1966; Ingolfsson 1970; Harris 1971; Harris 1974; Sealy 1975_b_). In ivory gulls, which molt just before reproduction, and in Sabine's gulls, which complete molt just before breeding, nonbreeders may extend wing feather growth into the breeding season (Stresemann and Stresemann 1966).

There is little information on the energetic cost of molt, although there are indications of at least some expense. Belopol'skii (1961) showed that nonmolting seabird species tended to gain weight after reproduction, whereas those that immediately started molt tended to lose weight. Among other birds, however, it is common for individuals to gain weight just before, and even during molt (Payne 1972). The BMR is known to rise in molting birds (Blackmore 1969; Lustick 1970; Payne 1972), from as little as 5% to as much as 34% above nonmolting levels. In one study, about 35-40% of the increased BMR represented extra thermoregulatory costs incurred by lessening of insulation and increase in heat loss from well-vascularized new feathers; the rest of the increase represented the energetic cost of growing feathers (Gavrilov 1974). The fact that molt rarely overlaps with breeding suggests that the energetic cost, even if slight, may be incompatible with the already high costs of reproduction (Payne 1972). Cassin's auklets, which do molt while breeding, may cease molt while feeding large young (Payne 1965), and certain species interrupt molt during migration (Stresemann and Stresemann 1966). Doubtless a rapid simultaneous molt is more costly than a long gradual one.

Rapid molt appears to occur at a time in the annual cycle when food resources are abundant (spring or late summer), whereas extended molt generally occurs over winter (e.g., Bédard 1969_a_). If one speculates that energy availability is the main determinant of molt patterns, one can also speculate on the cause behind some of the more unusual patterns. Possibly birds in which molt and breeding overlap either have extraordinary available energy at that time or else face shortages in other periods. For example, ivory gulls, which breed in the high Arctic, molt when food resources have become abundant in the low Arctic but before the high Arctic breeding grounds have thawed sufficiently for reoccupation.

Speed of molt may also reflect availability of energy resources or of nutrients needed for feather growth (Payne 1972), but must also be influenced by the need for full flight capabilities to obtain food. The eider duck and many alcids that shed wing feathers almost simultaneously do not need their wings for flight after the young have left the breeding colony. Hydrodynamic considerations suggest that their fishing capabilities may even be improved (Storer 1960). This is not true for the smaller species--e.g., _Aethia_ molts only one feather at a time and retains full flight capabilities (Table 2). Climate may also influence simultaneity of molt if heat loss in rapid molt is particularly severe.

_Reproduction_

Time use of seabirds is best known for the reproductive period, when the birds are on relatively accessible breeding grounds, the weather is most suitable for observation, and academic researchers are freed from their jobs. Even so, the details of timing are known for only a few of the species and localities on the northwest North American coast (e.g., Drent and Guiguet 1961; Drent et al. 1964; Cody 1973; Sealy 1973_a_, 1975_b_, 1975_c_). The following discussion emphasizes the multitude of environmental factors known to influence timing and total length of time devoted to various aspects of the reproductive cycle.

Timing of the Season

Each species of seabird returns to the colony site when weather conditions have ameliorated sufficiently to meet its particular needs. For example, the early arrivals to islands in the Barents Sea are murres, kittiwakes, and herring gulls, which need only small cracks in the sea ice to meet their feeding requirements (Belopol'skii 1961). Eiders in North America also return early, when a few ice leads have formed (Schamel 1974). Common puffins _(Fratercula arctica)_ and mew gulls _(Larus canus)_ are somewhat later arrivals, and terns and a few parasitic jaegers _(Stercorarius parasiticus)_ are the latecomers to Barents Sea colonies (Belopol'skii 1961).

The timing of the season (as illustrated in Fig. 5) varies widely among localities, and because of local weather patterns and ocean currents, this variation can be unrelated to latitude (Belopol'skii 1961). Examples of such variation are also known in North America: for instance, Leach's storm-petrels in Alaska lay eggs 2 to 3 weeks later than do those in California (Harris 1974); however, the details of timing are largely unknown for many species in this region. Progression of thaw, which also varies from year to year, causes variation in the timing of the breeding season (Belopol'skii 1961; Evans and McNicholl 1972). Fig. 6 shows the diversity in start of the breeding season for different species on the same island in the Barents Sea as well as variation in time devoted to various components of the reproductive cycle.

Prenesting Activities

Some species are apparently able to delay maturity of sexual organs until environmental conditions are suitable for nesting--e.g., burrow and crevice nesters in the Barents Sea do not become sexually mature until snowmelt (Belopol'skii 1961). Many others, however, reach sexual maturity soon after arrival on the breeding grounds, and a few (such as jaegers and kittiwakes) mature in migration or on the wintering grounds (Belopol'skii 1961). Northern phalaropes _(Lobipes lobatus)_ sometimes lay eggs as early as 1 week after arrival (Hilden and Vuolanto 1972). This factor, in combination with timing of arrival, affects the amount of time spent in prenesting activities (Fig. 6). Most species gain weight during this period (Belopol'skii 1961), and the time required for each species to reach full breeding condition must also depend on feeding conditions and the state of the bird on its arrival at the nesting site. These factors help explain why early arriving species are not necessarily early nesters (Fig. 6).

Aside from nest building, most prenesting activity consists of courtship and territorial behavior. These activities have been well described for representative seabird species, but because assessments of time and energy devoted to them have been almost completely neglected, they are not discussed further here. For examples, see accounts in Gross (1935) for Leach's storm-petrel; Tinbergen (1935), Bengtson (1968), Höhn (1971), and Howe (1975) for phalaropes; Storer (1952) for common murre, _Uria aalge_, and black guillemot, _Cepphus grylle_; Tschanz (1959) for common murre; Brown et al. (1967) for Sabine's gull; Tinbergen (1960) for herring gull; McKinney (1961) for eiders; Snow (1963) for shag; Thoresen (1964) for Cassin's auklet; Vermeer (1963) and James-Veitch and Booth (1974) for glaucous-winged gull _(Larus glaucescens)_; and Andersson (1973) for jaegers.

Nest Building

Although many northern seabirds have essentially no nest, they may spend considerable time working or displaying at the site (Belopol'skii 1961). Black-legged kittiwakes _(Rissa tridactyla)_ have substantial nests, but they are built in a comparatively short time (about a week) soon after the birds arrive (Fig. 6). Shags also have substantial nests, but they are not completed until about 1 or 2 weeks before the first egg is laid (Snow 1963). Herring gulls build smaller nests, 5 to 10 days before laying, although in the Far North they and glaucous gulls may not start building the nest until the first egg is laid (Belopol'skii 1961). The eider always begins preparing the nest when the first egg is laid (Belopol'skii 1961; Schamel 1974), and terns and skuas, which build no nests, choose their sites at that time. Murres, which frequently lay their eggs directly on snow, choose a site somewhat earlier and spend considerable time protecting it (Belopol'skii 1961). Burrows may be dug within a period as short as 3 days for Leach's storm-petrel (Gross 1935) to one as long as several weeks in Cassin's auklets (Manuwal 1974_a_). Overall, the prelaying period is longer for burrow nesters than for those using crevices (Sealy 1973_a_).

The amount of time and energy spent by the male and female in nest building differs among species. In Leach's storm-petrel, the male digs the burrow (Gross 1935), whereas in eiders, the nest is built entirely by the female. In most seabird species, the sexes share in nest construction, but roles may still be separated. For example, in shags the male collects the nest material and the female builds the nest (Snow 1963).

Egg Laying

Timing of egg laying is influenced not only by weather (Erskine 1972; Sealy 1975_c_), but also by numerous biotic factors. Smith (1966) showed that where glaucous gulls, herring gulls, and Thayer's gulls _(Larus thayeri)_ breed in mixed colonies, the peak of sexual activity and egg laying in Thayer's gull is about midway between the peaks for the other two species (Fig. 7). In nearby colonies where herring gulls are absent, however, the peak of sexual activity in Thayer's gulls is delayed about a week, and activity continues for a significantly longer period (Fig. 7).

Annual variations in food supply also will affect the start of the egg-laying season. Belopol'skii (1961) cited an example from the Barents Sea in 1940 when a series of storms made it difficult for certain seabirds to find food. Murres and kittiwakes, which were able to catch fish, started reproductive activities on schedule. Gull breeding was delayed, however, and egg laying began in force only after fishing boats arrived and started discarding offal. Onset of egg laying in great cormorants _(Phalacrocorax carbo)_ is correlated to April air temperatures (Erskine 1972), and this may also be related to variations in spring increase of food availability. In certain birds the breeding season has been shown to start particularly early when food supplies are unusually abundant (Högstedt 1974; Källender 1974), but this has not yet been demonstrated in seabirds.

Lastly, age and sex of seabirds are known to affect the timing of egg laying (e.g., Coulson and White 1960; Lack 1966); older, more experienced birds tend to lay earlier than do younger ones. In shags, males tend to breed progressively earlier as they increase in age, but females do not (Snow 1963).

Contrary to the situation in passerines, seabirds tend to lay their clutches with relatively large time intervals between eggs. Eggs may be laid every 2nd or 3rd day in alcids, larids, sternids, stercorariids, and phalacrocoracids (Lack 1968), but every day in phalaropes (Howe 1975). Inasmuch as clutch size in northern seabirds varies from one to five or six, the length of the laying period varies widely among species.

Energetic costs of egg laying depend on the actual caloric content of the egg and the speed with which the ova are developed (Ricklefs 1974). The energy in the egg is contained mainly in the yolk, and yolk size depends largely on the developmental pattern shown by the young after hatching (Table 3). Precocial chicks are hatched at a relatively advanced stage, are covered with down, and have open eyes, can maintain reasonably homeothermic body temperature, and leave the nest site to feed themselves after a few hours or days. At hatching, semiprecocial chicks appear similar to precocial chicks, although they are slightly less well developed (Ricklefs 1974; Dunn 1975_a_). In contrast to precocial chicks, they remain at the nest site for some time, are fed by their parents, and tend to grow rather rapidly (Ricklefs 1968). Altricial nestlings hatch at a much less advanced stage of development. They are naked, blind, helpless, essentially poikilothermic, and depend completely on their parents for food and shelter. They usually remain at the nest until full grown. Semialtricial chicks show somewhat intermediate characteristics (Nice 1962).

Table 3. _Amount of yolk in eggs of different types of birds._
After Ricklefs (1974).
------------------------------------------------
Developmental Percent yolk
type (by weight)
------------------------------------------------
Precocial 30-60
Semiprecocial 25-30
Altricial 15-25
+----------------------------------------------+

The amount of yolk (and therefore energy) in an egg is positively correlated to the degree of development at hatching (Table 3). The same is true for egg size: altricial and semialtricial birds have smaller eggs relative to adult body weight than do semiprecocial and precocial birds (Fig. 8). Clutch size, however, is unrelated to energy content of the eggs. For example, shags (which are altricial) and eiders (precocial) have among the largest clutches of northern seabirds (four to six eggs).

The energetic cost of egg laying depends not only on caloric content of the egg and clutch size, but also on speed of development. Ricklefs (1974) has shown that the energetic cost per day of egg laying can be calculated from the energy content of the yolk and white, clutch size, the amount of follicular growth per day, and the laying interval between eggs. The energy content of a single egg (expressed as percentage of BMR) has been estimated as follows: 45 (altricial passerines), 103 (semialtricial raptors), 126 (precocial galliformes), 180 (precocial ducks), 226 (precocial shorebirds), and 320 (semiprecocial gulls and terns). Gulls and terns thus have very costly eggs, as well as a moderately high clutch size (three). However, the development time for a single ovum in the herring gull is unusually long--9 to 10 days (King 1973). Ricklefs (1974), who calculated the energetic cost per day (expressed as percentage BMR), estimated the cost of a clutch in gulls and terns (120% BMR per day) to be similar to that for various groups of precocial birds (about 125-180% BMR per day). Unfortunately, the data required for calculation of the average energetic cost of a clutch are not available for other northern seabirds.

For no species have all the additional factors influencing the energetic cost of a clutch been taken into account. For example, eggs in a clutch may vary in size (and caloric content) according to sequence in the clutch (Preston and Preston 1953; Snow 1960; Coulson 1963; Coulson et al. 1969). Age has a definite effect on laying energetics, as older birds lay larger eggs (Coulson and White 1958; Snow 1960; Coulson 1963; Coulson et al. 1969) and lay larger clutches (Coulson and White 1960; Snow 1960). They also lay, on average, earlier in the season (Coulson and White 1958; Snow 1960; Coulson et al. 1969), and eggs laid late in the season (whether by young birds or older ones in re-nesting attempts) tend to be smaller and contain less energy. In addition, egg quality can vary with food supply: Snow (1960) found eggs to have more yolk in years when food was abundant than in years when food was scarce.

Egg-laying costs are, of course, borne entirely by the female, although males may contribute some time and energy toward egg laying through courtship feeding (Ashmole 1971; Henderson 1972; Nisbet 1973). Courtship feeding takes place in most lariforms but not in eiders, phalaropes, or cormorants.

The time and energy expended on egg laying can be profoundly influenced by the degree of nest destruction, since females usually lay a replacement clutch if the loss of the first does not occur too late in the season. Factors causing egg destruction are numerous, but among the most important in the north is predation. As is shown in Table 4, the degree of egg predation in common murres is correlated to degree of exposure of the nest--so even such an unlikely sounding factor as physical characteristics of the nest site can affect the average time and energy expended on egg laying by a given species or population. Genuine second clutches are occasionally laid by phalaropes (Hilden and Vuolanto 1972) and Cassin's auklets (Manuwal 1974_a_).

Table 4. _Predation on nests of common murres according to degree
of exposure._ From Belopol'skii (1961).

--------------------------------------------
Nests destroyed by
Nest exposure predators (%)
-------------------------------------------
Completely hidden 3.2
Partly exposed 5.8
Largely exposed 13.6
Completely exposed 18.2
--------------------------------------------

In short, time and energy devoted to egg laying depend not only on the species, but also on a multitude of other biotic and abiotic factors, such as age, sex, degree of nest destruction, weather, other species present, and feeding conditions.

Incubation

The total time devoted to incubation does not depend directly on developmental type or egg size but differs markedly among families (Lack 1968). Since incubation period seems to be closely linked to fledging period, factors affecting growth rate (discussed later) apparently affect incubation period as well.

Each species has a different incubation rhythm. In birds in which the sexes share in incubation, the sexes exchange places at intervals that differ widely among different birds: several hours in lariforms and some alcids (Drent 1965; Lack 1968; Preston 1968; Drent 1970); about 4 h in shags (Snow 1963); up to 11 h in the ivory gull (Bateson and Plowright 1959); up to 24 h in certain other alcids (Manuwal 1974_a_; Sealy 1975_a_); 33 h (on the average) in common puffins (Myrberget 1962); 72 h in ancient murrelets, _Synthliboramphus antiquus_ (Sealy 1975_a_); and 96 h in Leach's storm-petrel (Gross 1935). Degree of attentiveness once a bird is on the nest also varies. Petrels may leave the egg for several days (Gross 1935), whereas herring gulls cover their eggs 98% of the time (Drent 1970).

The sexes share in incubation in most seabirds (Snow 1960; Drent 1965, 1970; Bédard 1969_a_), although females frequently take on the greater role (Belopol'skii 1961). Only male phalaropes incubate the eggs, and only female eiders. Eider hens do not feed during the entire incubation period (25 days) and leave the nest only for short periods of about 10 min (Belopol'skii 1961; Schamel 1974).

Several methods exist for calculating the amount of heat input necessary for normal development of a clutch of eggs (Ricklefs 1974). There is controversy, however, as to whether an adult can provide this warmth from excess body heat lost during the course of normal metabolism or whether the adult must raise its metabolic level to produce extra heat (Kendeigh 1973; King 1973; Ricklefs 1974). Several studies of incubating birds suggest that, in at least some situations, adults need not raise metabolic levels, but in others (large clutch, severe weather), they probably do (Ricklefs 1974). Drent (1972) estimated that herring gulls raise metabolic levels to a significant degree during incubation.

In spite of the lack of quantitative data, one can surmise that the cost of incubation varies among seabirds. Precocial and semiprecocial birds tend to have a larger clutch weight relative to body weight than do altricial birds (Fig. 8; Lack 1968), and therefore require greater heat input to the eggs. These costs may be reduced by heavily insulating the nest (e.g., eiders), or by nesting in burrows, which have much more moderate and even climates than do external nests (Richardson 1961; Manuwal 1974_a_). Other semiprecocial species, however, such as the murre, may sometimes lay eggs directly on snow or ice (Belopol'skii 1961)--presumably at increased incubation costs. Lastly, certain species incubate eggs with their feet (e.g., cormorants), rather than develop featherless brood patches. There are no measurements of comparative heat flow from feet versus brood patches.

Raising Nestlings

The length of the nestling period (hatching until departure from the nest) varies greatly among northern seabirds (Fig. 6). Nestling period depends on the stage of growth at which the young leave the nest and the rate at which they attain that stage. Growth rate in turn depends largely on body size and developmental type.

The stage of growth attained when birds leave the nest varies considerably (Fig. 9). Precocial eiders leave the nest within a day of hatching, whereas altricial shags remain until completely grown. The young of semiprecocial species, on the other hand, leave the nest at all stages between these extremes. Larids normally remain at the nest until 75-90% grown, but certain alcids leave much sooner--well before the young can fly.

Growth rate depends both on body size and developmental type (Fig. 10). The length of stay at the nest for precocial young is unaffected by growth rate (which is typically very slow), since they leave soon after hatching. The nestling period of semiprecocial and altricial seabirds is, however, affected by the rate at which the young grow to the nest-leaving stage. This depends mainly on body size (Fig. 10) and to a certain degree on developmental type, as some semiprecocial species grow rather slowly. Certain seabirds with clutches of one egg grow particularly slowly (petrels, some alcids, sulids). Several other alcids with single-egg clutches, however, grow at rates normal for semiprecocial chicks (Fig. 10). Very slow growth may be related to food stress (Lack 1968; Ricklefs 1968) or to reduction of reproductive effort in the adults (discussed later). Contrary to Cody (1973), slow growth in alcids does not correlate to the distance adults must commute for food. (Cody tried to directly compare growth in birds of different sizes.) Chicks in nocturnal species, however, tend to have slow growth rates (Sealy 1973_b_).

Daily time budgets of adults raising nestlings also vary widely, depending on the amount of brooding required, food requirements of the young, and foraging costs (which differ in the breeding season from those at other times of the year).

Nestlings have imperfect control of body temperature at hatching (Fig. 11) and develop this capacity only gradually. Altricial birds are hatched at a particularly undeveloped stage; e.g., double-crested cormorants attain reasonable control of body temperature in moderate ambient temperatures only after about 14 days (Fig. 11; Table 5). Semiprecocial seabirds, which are more fully developed physically at hatching, attain control of body temperature much sooner, in a matter of several days, and precocial eiders can thermoregulate within a few hours after hatching (Table 5).

Until the age of temperature control, nestlings must be brooded almost constantly, and occasional brooding takes place for some time afterward, especially in severe weather, in all species studied (Tinbergen 1960; Belopol'skii 1961; Weaver 1970; Dunn 1976_a_, 1976_b_). Thermoregulatory capabilities in cold weather are better in ducklings of species nesting at high latitudes than at lower ones (Koskimies and Lahti 1964), and the same may be true of gull species (Dawson et al. 1972). The cooling mechanisms of double-crested cormorants are better than in the more northerly distributed pelagic cormorant, _Phalacrocorax pelagicus_ (Lasiewski and Snyder 1969). Thus, variation in cost of thermoregulation due to different environments may be reduced through adaptation.

Food requirements of the chick depend on growth rate, amount of fat deposition, cost of thermoregulation, degree of activity and other factors (E. H. Dunn 1973). Estimated energy budgets for nestling double-crested cormorants and herring gulls in the same year and locality (Fig. 12) indicate that these factors vary according to developmental type, and comparison with budgets for nonseabird species suggests wide variation within developmental types according to the particular adaptations of each species to its own environment (E. H. Dunn 1973).

Thus, the energy demands of nestlings are not easy to predict. Brood size differences multiply variation in food demand on adults (except in precocial birds whose young feed themselves). Energy demands are labile, however, particularly in requirements for activity and growth, and adults can frequently raise young successfully without providing optimum amounts of food (Spaans 1971; Kadlec et al. 1969; LeCroy and Collins 1972; Lemmetyinen 1972; Cody 1973; E. H. Dunn and I. L. Brisbin, manuscript in preparation). Studies of double-crested cormorants by Dunn (1975_b_) and pigeon guillemots _(Cepphus columba)_ by Koelink (1972) have suggested that each adult providing optimum amounts of food to a normal-sized brood would have to approximately double the amount of food gathered each day over the amount gathered by nonbreeders. This relation does not imply, however, that the time and energy allocation of the adults would be the same for the two species.

Cost-benefit ratios of food gathering in the nestling period differ from those at other times. Besides facing increased food demands, costs of delivery to the nest, and changes in food availability, the parents' choice of foods is constrained by the need to forage within reasonable commuting distance of the nest and perhaps by concentrated competition with conspecifics and other seabird species. In addition, small nestlings are frequently unable to eat foods normally eaten by adults (Drent 1965; personal observation). In the face of these constraints, adults often shift food preferences while raising nestlings (Belopol'skii 1961). For example, female mew gulls in the Barents Sea forage in the tidal zone, eating more small invertebrates than at other times of the year, while males continue to forage at sea and consume larger quantities of fish (Fig. 13).

Table 5. _Age of thermoregulatory control in various species of
northern seabirds._

Species Age when Source
moderate
temperature
control is
attained
(days)

Common eider 0.1-0.3[41] V. V. Rolnik, in Belopol'skii
(1961)

Herring gull 1.5-2 V. V. Rolnik, in Belopol'skii
(1961)

2-3 E. H. Dunn (1976_b_)

Leach's storm-petrel [2] Ricklefs (1974)

Mew gull 2-3 V. V. Rolnik, in Belopol'skii
(1961)

Lesser black-backed gull 2-3 E. K. Barth (in Farner and
Serventy 1959)

Greater black-backed gull 2-3 E. K. Barth (in Farner and
Serventy 1959)

Pigeon guillemot 2-4 Drent (1965)

Common tern 3 LeCroy and Collins (1972)

Roseate tern _(Sterna 3 LeCroy and Collins (1972)
dougallii)_

Common murre 3 V. V. Rolnik and Yu. M.
Kaftonowski (in Sealy 1973_b_)

Razorbill _(Alca torda)_ 3 V. V. Rolnik and Yu. M.
Kaftonowski (in Sealy 1973_b_)

Black guillemot 3-4 V. V. Rolnik, in Belopol'skii
(1961)

Tufted puffin 3.5[42] Cody (1973)

Northern phalarope 4-5[43] Hilden and Vuolanto (1972)

Cassin's auklet 5-6 Manuwal (1974_a_)

Horned puffin _(Fratercula 2-6 Sealy (1973_a_)
corniculata)_

Common puffin 6-7 V. V. Rolnik and Yu. M.
Kaftonowski (in Sealy 1973_b_)

Black-legged kittiwake 6-7 V. V. Rolnik, in Belopol'skii
(1961)

Double-crested cormorant 14 Dunn (1976_a_)

Shag 12-15 V. V. Rolnik, in Belopol'skii
(1961)

Commuting distances vary tremendously among species (Fig. 14), but the number of feeding trips to the nest per day does not correlate with foraging distance (Cody 1973; Sealy 1973_a_, 1973_b_). There is not, therefore, a simple relationship between time and energy expenditures of the adults and foraging distances. Nocturnality, on the other hand, correlates with reduced feeding rates (usually one visit to the nest each night). Seabirds feeding far from the colony tend to show adaptations for bringing larger amounts of food per visit, such as carrying more than one fish at a time, as in tufted puffins, _Lunda cirrhata_, and rhinoceros auklets, _Cerorhinca monocerata_, vs. guillemots and murres (Richardson 1961; Cody 1973; Sealy 1973_a_, 1973_b_); developing a sublingual storage pouch, as in Cassin's auklets (Speich and Manuwal 1974); or concentration of food into stomach oil, as in petrels and albatrosses (Ashmole 1971). Commuting costs are largely eliminated when the young leave the nest, but only in the alcids does nest leaving occur long before attainment of full growth. Early nest leaving may allow adults and young to disperse to better feeding areas than are exploitable from the colony site (Sealy 1973_b_) and probably involves a major change in optimal food size and type as well (Lind 1965).

Patterning of adult time budgets may differ between geographical regions. For example, rhinoceros auklets are nocturnal in the far north (where the summer night is particularly short), crepuscular in the Olympic Peninsula, and mainly diurnal in the Farallon Islands (Manuwal 1974_a_).

Food demands of nestlings have a great influence on the time and energy allocation of breeding over nonbreeding seabirds. Because food is particularly abundant in the reproductive season, however, one cannot ascertain whether the vulnerability of breeding birds to time or energy crises is far different from that at other times of the year.

Post-fledging Care

Little is known about the amount of care provided by adults to young after they are fully grown. At least some species, such as gannets and procellariiformes (Ashmole 1971), are known to desert their young, whereas others are known to feed their young, at least occasionally, for some weeks or months after they can fly--e.g., terns and gulls, many alcids, and shags (Snow 1963; Vermeer 1963; Drury and Smith 1968; Ashmole and Tovar S. 1968; Potts 1968; Ashmole 1971; LeCroy 1972).

Annual Time and Energy Budgets

The discussion of time and energy allocation during reproduction was complex and detailed because so much more is known about the influences altering budgeting during this period than during other times of the year. It is likely that influences on molt and migration will prove to be equally complicated, once more is learned about them.

If all data on time and energy allocation for a single species were known, it would be possible to make up detailed budgets for birds of different age, sex, and experience throughout the year. However, such detailed data have not been collected for any species. An annual time budget for male and female yellow-billed magpie, _Pica nuttalli_ (Verbeek 1972), points out the great amount of difference between the sexes (Fig. 15). A time and energy budget for the reproductive season only (Fig. 16) shows large differences between two closely related species, as well as between sexes; it also indicates the wide difference between the budgeting of energy as opposed to budgeting of time. All other time-energy budgets to date are for nonseabird species and for only a portion of the annual cycle (Verbeek 1964; Verner 1965; Schartz and Zimmerman 1971; Stiles 1971; Wolf and Hainsworth 1971; Smith 1973; Utter and LeFebvre 1973).

Time-energy budget analysis can be useful in determining the leeway a bird has in surviving unusual stress at different times of the year. For example, a study by Feare et al. (1974) showed that rooks _(Corvus frugilegus)_ in the dry part of the summer spent 90% of 15 h of daylight to collect 150 kcal of food energy. In winter, foraging in snow, the same birds were able to collect 240 kcal of food in only 30% of a 10-h day. This suggests that rooks would be far more vulnerable to unexpected periods of stress in late summer than in winter. Such information would clearly be useful in making management decisions.

A more precise measure of vulnerability, although much more difficult to determine, is that of productive energy--the amount of caloric intake left over after the birds' cost of living (metabolic functions and procurement and processing of food) have been accounted for. Costs are highest when temperatures are extremely hot or cold or when food is most difficult to obtain. Productive energy is highest in summer (Kendeigh 1972), and that is presumably why reproduction normally takes place then. It is unknown whether birds are more vulnerable to time and energy shortages in the harder nonbreeding season or in the breeding season after the extra demands of reproduction have been accounted for. Vulnerability may also differ between sexes and among age groups.

Time-energy studies, although useful in comparing ecology, determining vulnerability, and cataloging location of birds, do have limitations. Careful studies are time-consuming and are not the best approach to determining key factors influencing population increase or decrease. Even when different kinds of data are being sought, however, it is worthwhile keeping the time-energy framework in mind as a "big picture" into which other facts can be fitted and their significance considered.

Life History Strategies

The study of life history strategies is largely theoretical, and in the following discussion I do not comment on current theoretical arguments. On the other hand, life history strategies can be regarded as time and energy allocation on a grand scale, and it therefore seems appropriate to look briefly at their implications for seabird management.

Annual reproduction evidently has a negative effect on resources remaining for other functions, and may reduce the chances for an organism to reproduce again in a later season (Cody 1966, 1971; Williams 1966; Gadgil and Bossert 1970; Gadgil and Solbrig 1972; Hussell 1972; Trivers 1972; Calow 1973). If the chances of survival to another breeding season are small, the selective advantage lies with the bird putting the most effort into early reproduction, in spite of its negative effects on survival, because future chances of reproduction are small. If chances of survival are good, however, it may be more advantageous to reduce annual reproductive effort and allocate resources to other functions.

Seabirds are generally long-lived, have small clutches, and generally delay first breeding until at least the 2nd year, and usually longer (Table 6). Phalaropes seem to differ from this pattern (Hilden and Vuolanto 1972; Howe 1975). Several ecological factors (not entirely independent) are believed to contribute to the evolution of the long life and low reproductive effort pattern favored by seabirds.

First, if population size is determined largely by density-dependent mortality, individuals may be favored that allocate resources to attaining longer life (and more chances to reproduce) or insuring greater chances of survival of their offspring (Murphy 1968; Hairston et al. 1970). Density-independent mortality, on the other hand, is so unpredictable that there is no advantage in allocating resources toward protection against it (Gadgil and Solbrig 1972).

Two factors closely linked with density-dependence are high levels of competition, and perennial difficulties in obtaining food. In adapting to these difficulties, a bird may be selected which develops more efficient foraging techniques, wider dispersal, or better abilities to defend nesting territory--all of which may reduce resources available for reproduction. As mentioned earlier, marine foods tend to be patchily distributed, and a long learning period seems to be necessary before seabirds become proficient at foraging. In addition, there is evidence that food availability is low, at least in the tropics, and perhaps in the winter in other regions (Ashmole 1971). If nesting places are in short supply, long life may be favored so that the bird can live long enough for a place to become vacant. Several authors feel that competition is a serious factor in the life of seabirds, both for food (Lack 1966; Cody 1973) and for nesting space (Snow 1960; Belopol'skii 1961; Lack 1966; Manuwal 1974_b_). Others, however, disagree, at least for the breeding season (e.g., Pearson 1968).

Table 6. _Life history data for certain northern seabirds._[44]

Annual adult Age at first
survival breeding Clutch
Species (%) (years) size

Fulmar 94 7+ 1
Gannet _(Morus bassanus)_ 94 (4)-5+ 1
Manx shearwater 93-96 (4)-5+ 1
Shag 85 (♂) (2)-3 3-4
80 (♀)
Herring gull 91-96 3.5 (♂) (2)-3
5 (♀)
Black-legged kittiwake 88 4-5 (♂) 3
3-4 (♀)
Arctic tern 89-91 (2)-3+ 2
75
82[45]
Common murre 87 3+? 1
Black guillemot 88+[46] 3?[46] 2[46]
Cassin's auklet 83[47] 3[47] 1

There is some evidence of density-dependent population size control in seabirds, although much of it is circumstantial. For example, there are large nonbreeding populations in such diverse species as shags, herring gulls, and Cassin's auklets, which move into a breeding area when established adults are removed or colonize new breeding areas (J. C. Coulson, personal communication; Kadlec and Drury 1968; Drury and Nisbet 1972; Manuwal 1974_b_). Lack (1966) and Ashmole (1971) presented other arguments for density-dependence. Density-dependent mortality is difficult to demonstrate, at best, and may be obscured by interpopulation movements (Drury and Nisbet 1972).

If long life is a life history option, a low annual reproductive effort could be favored in several ways. First, it may be necessary for insuring long life, if breeding has a serious negative feedback on life expectancy (Calow 1973). Second, if survival of offspring is more unpredictable than that of adults, low annual effort may be selected so that reproductive effort will not be wasted in years when young have poor chances of survival. Unpredictable and variable first-year survival in seabirds has been documented (Potts 1968; Drury and Nisbet 1972). In addition, some seabirds show adaptations that allow high reproductive success in any given year but which do not drain off resources if the season turns out to be poor (e.g., small last eggs in the clutch or asynchronous hatching, both of which lead to elimination of the smallest chicks when conditions are poor [Parsons 1970; E. H. Dunn 1973]).

It should be emphasized that the factors involved in the evolution of life histories are complex and poorly understood, and simple formulas should not be expected to apply to all situations (Wilbur et al. 1974).

In the framework of life-history strategies, small clutch sizes and slow growth rates exhibited by some seabirds can be explained as adaptive reductions in annual reproductive effort, rather than as responses to immediate food shortages. Arguments for this view are presented on theoretical grounds (Dunn 1973) and by the fact that many seabirds are able to raise larger than normal broods in certain situations (Vermeer 1963; Nelson 1964; Harris 1970; Hussell 1972; Ward 1972; Corkhill 1973). In addition, seabirds with particularly slow growth rates all grow at about the same rate, regardless of body size (contrary to the situation in other birds). This suggests that low growth rates do not reflect variations in feeding abilities among species (Ricklefs 1968).

Several conclusions relating to management of seabird populations can be drawn from the above discussion. First, if population size is determined largely by density-dependent factors, the birds are not adapted to precipitous and unexpected declines in population levels. Because there is low annual reproductive effort geared to a world in which there is slow turnover in population, seabirds are not able to rebound quickly from disasters. Provision of excess food should not be expected to improve breeding performance, at least in experienced birds.

Second, because seabirds are able to reproduce in many different seasons and are adapted to a low reproductive effort within a given season, one should expect them to be easily disturbed and to fail to complete the reproductive cycle during any given breeding attempt. A few indications of such failures have already been observed (Erskine 1972; Manuwal 1974_a_; Nettleship 1975).

Again, the tentative nature of this discussion should be emphasized, and conclusions drawn from it may not apply equally to all seabird species.

Conclusions

In this discussion I have tried to emphasize the variety of factors affecting seabird life cycles and the diverse responses among different species to their environment. The main conclusion I stress is that each species (and age group and sex within that species) has a different vulnerability to stress, which may be most severe at different times of the year for each group. To determine these periods of stress, researchers may find a time-energy approach to be useful.

As for northwestern North American seabirds in particular, ignorance is vast. Twelve years ago, Bourne (1963:846) noted the following needs in seabird research (among others): "The investigation of seabird biology has been reduced to a routine, but there is a great need for more study of some other aspects of the life or annual cycle, including events in the period immediately after fledging, and behaviour and survival in the immature period and outside the breeding season. Much more accurate information is needed about breeding distribution and seasons in many parts of the world, about molting seasons and ranges in most parts, and the distribution of birds of different age groups during these periods in practically all areas."

Since the time of Bourne's remarks, a number of excellent studies have provided data on the breeding biology of certain northwestern seabird species. Scientists remain largely ignorant, however, about where birds of different age groups are located throughout the year. Such knowledge is necessary for effective protection and is basic to understanding population dynamics, even if it does not elucidate causes. Studies of timing of annual cycles and movements should be carried out hand in hand with resource analysis--not just finding what birds eat, but discovering where the food is at what times, how hard it is to catch, and what the nutritional return is. Much careful field work must be done before effective management of most of our northwestern seabirds can become a reality.

References

Andersson, M. 1973. Behaviour of the pomarine skua _Stercorarius
pomarinus_ Temm. with comparative remarks on Stercorariinae.
Ornis Scand. 4:1-16.

Ashmole, N. P. 1971. Seabird ecology and the marine environment.
Pages 223-286 _in_ D. S. Farner and J. R. King, eds. Avian
biology. Vol. I. Academic Press, New York.

Ashmole, N. P., and H. Tovar S. 1968. Prolonged parental care in
royal terns and other birds. Auk 85:90-100.

Barash, D. P., P. Donovan, and R. Myrick. 1975. Clam dropping
behavior of the glaucous-winged gull _(Larus glaucescens)_.
Wilson Bull. 87:60-64.

Bateson, P. P. G., and R. C. Plowright. 1959. The breeding biology
of the ivory gull in Spitsbergen. Br. Birds 52:105-114.

Bédard, J. 1969_a_. Feeding of the least, crested and parakeet
auklets around St. Lawrence Island, Alaska. Can. J. Zool.
47:1025-1050.

Bédard, J. 1969_b_. Adaptive radiation in alcidae. Ibis 111:189-198.

Belopol'skii, L. O. 1961. Ecology of sea colony birds of the
Barents Sea. (Transl. from Russian.) Israel Program for
Scientific Translations, Jerusalem. 346 pp.

Bengtson, S. 1968. Breeding behavior of the grey phalarope in west
Spitzbergen. Var Fågelvärld 27:1-13.

Berger, M., and J. S. Hart. 1974. Physiology and energetics of
flight. Pages 415-477 _in_ D. S. Farner and J. R. King, eds.
Avian biology. Vol. IV. Academic Press, New York.

Berger, M., J. S. Hart, and O. Z. Roy. 1970. Respiration, oxygen
consumption and heart rate in some birds during rest and flight.
Z. Vgl. Physiol. 66:201-214.

Birkhead, T. R. 1974. Movement and mortality rates of British
guillemots. Bird Study 21:241-254.

Blackmore, F. H., Jr. 1969. The effect of the temperature,
photoperiod and molt on the energy requirements of the house
sparrow, _Passer domesticus_. Comp. Biochem. Physiol. 30:433-444.

Bookhout, T. A. 1958. The availability of plant seeds to bobwhite
quail in southern Illinois. Ecology 39:671-681.

Bourne, W. R. P. 1963. A review of oceanic studies of the biology
of seabirds. Proc. Int. Ornithol. Congr. 13:831-854.

Brisbin, I. L., Jr. 1965. A quantitative analysis of ecological
growth efficiency in the herring gull. M.S. Thesis. Univ. of
Georgia, Athens. 28 pp.

Brown, R. G. B., N. G. Blurton-Jones, and D. J. T. Hussell. 1967.
The breeding behaviour of Sabine's gull, _Xema sabini_. Behaviour
28:110-140.

Bryant, D. M. 1973. The factors influencing the selection of food
by the house martin (_Delichon urbica_ (L.)). J. Anim. Ecol.
42:539-564.

Buckley, F. G., and P. A. Buckley. 1974. Comparative feeding
ecology of wintering adult and juvenile royal terns (Aves:
Laridae, Sterninae). Ecology 55:1053-1063.

Calder, W. A. 1974. Consequences of body size for avian energetics.
Pages 86-144 _in_ R. A. Paynter, ed. Avian energetics. Publ.
Nuttall Ornithol. Club 15.

Calder, W. A., and J. R. King. 1974. Thermal and caloric relations
of birds. Pages 259-413 _in_ D. S. Farner and J. R. King, eds.
Avian biology. Vol. IV. Academic Press, New York.

Calow, P. 1973. The relationship between fecundity, phenology, and
longevity: a systems approach. Am. Nat. 107:559-574.

Cody, M. L. 1966. A general theory of clutch size. Evolution
20:174-184.

Cody, M. L. 1971. Ecological aspects of reproduction. Pages 461-512
_in_ D. S. Farner and J. R. King, eds. Avian biology. Vol. I.
Academic Press, New York.

Cody, M. L. 1973. Coexistence, coevolution and convergent evolution
in seabird communities. Ecology 54:31-44.

Cohen, D. 1967. Optimization of seasonal migratory behavior. Am.
Nat. 101:5-17.

Cone, C. D., Jr. 1964. A mathematical analysis of the dynamic
soaring flight of the albatross with ecological interpretations.
Va. Inst. Mar. Sci., Spec. Sci. Rep. 50:1-104.

Corkhill, P. 1973. Food and feeding ecology of puffins. Bird Study
20:207-220.

Coulson, J. C. 1961. Movements and seasonal variation in
mortality of shags and cormorants ringed on the Farne Islands,
Northumberland. Br. Birds 54:225-235.

Coulson, J. C. 1963. Egg size and shape in the kittiwake _(Rissa
tridactyla)_ and their use in estimating age composition in
populations. Proc. Zool. Soc. Lond. 140(2):211-227.

Coulson, J. C., and M. G. Brazendale. 1968. Movements of cormorants
ringed in the British Isles and evidence of colony-specific
dispersal. Br. Birds 61:1-21.

Coulson, J. C., G. R. Potts, and J. Horobin. 1969. Variation in the
eggs of the shag _(Phalacrocorax aristotelis)_. Auk 86:232-245.

Coulson, J. C., and E. White. 1958. The effect of age on the
breeding biology of the kittiwake, _Rissa tridactyla_. Ibis
100:40-51.

Coulson, J. C., and E. White. 1960. The effect of age and density
of breeding birds on the time of breeding of the kittiwake,
_Rissa tridactyla_. Ibis 102:71-80.

Cox, G. W. 1968. The role of competition in the evolution of
migration. Evolution 22:180-192.

Cramp, S. 1972. The breeding of urban wood pigeons. Ibis
114:163-171.

Cullen, J. M. 1957. Plumage, age and mortality in the arctic tern.
Bird Study 4:197-207.

Dawson, W. R., and J. W. Hudson. 1970. Birds. Pages 223-310 _in_ G.
C. Whittow, ed. Comparative physiology of thermoregulation. Vol.
I. Academic Press, New York.

Dawson, W. R., J. W. Hudson, and R. W. Hill. 1972. Temperature
regulation in newly hatched laughing gulls _(Larus atricilla)_.
Condor 74:177-184.