[advances in marine biology] advances in marine biology volume 20 volume 20 || competition between...

Competition between Fisheries and Seabird Communities

R. W. Fmess

Zoology Department, Glasgow University, Glasgow, Scotland

I. Introduction . . . . . . . . . . . . . . . . XI. Estimating Food Consumption by Seabird Populations . . . .

A. Field observations . . . . . . . . . . . . B. Bioenergetics equations . . . . . . . . . . C. Input parameters, model sensitivity and output accuracy . .

111. Changes in Marine Ecosystems and Seabird Populations A. British Columbia . . . . . . . . . . B. California current . . . . . . . . . . C. South Africa . . . . . . . . . . . . D. Peru current . . . . . . . . . . . . E. The Southern Ocean . . . . . . . . . . F. North Sea . . . . . . . . . . . .

. . A. Evidence from studies of community strhcture . . B. Evidence from single species studies . . . . . .

V. Acknowledgements . . . . I . . . . . . . VI. References . . .. . . . . .. .. ..

IV. Influences of Food on Seabird Population Ecology

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

..

. . . .

. . . .

. . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . . . . .. . . . . . . I .

. . . .

. . . .

. . . .

225 228 228 229 234 240 240 243 241 259 269 219 292 292 295 291 298

I. Introduction

Seabirds comprise only 3 % of the known avian species, but research into their biology and behaviour has exceeded that on almost all other groups of birds. As a result we know a great deal about their distribution, abundance, breeding biology, behaviour and population dynamics. Seabird studies have been prominent in the development of general theories of animal population

225

226 R; W. FURNESS

regulation. Lack (1954) argued that animals maximize their production of young and their populations are regulated by density-dependent mortality, largely due to starvation outside the breeding season. Three of 15 chapters presenting evidence to support this argument (Lack, 1966) were based on population studies of seabirds. Wynne-Edwards (1962) also made extensive use of seabird studies in order to argue that animals regulate their own density below the potential upper limit set by food. He emphasized colonial nesting, deferred maturity, low clutch size, single brooding and long incubation and nestling periods as characteristic features of seabird biology, as adaptations to avoid overpopulation. Ashmole (1963) argued that pelagic seabird numbers are most likely to be regulated in a density-dependent way by food shortage. He concluded that this is unlikely to occur when the population is dispersed over the oceans, but that competition for food close to breeding colonies would result in birds having to range farther to feed as population increased, eventually resulting in population stability.

Many British seabird populations have increased in numbers during this century and some of these increases have been both rapid and prolonged (Cramp et al., 1974). Such large population changes may suggest that density- dependent control of population is unimportant (Andrewartha and Birch, 1954) or may be taken as evidence that food supplies have improved. Many British seabird populations were exploited during the eighteenth and nineteenth centuries. Coulson (1963) and Potts (1969) have argued that increases of these populations are due to relaxation of such exploitation, and that food supplies were and still are super-abundant. In contrast, Fisher (1952) argued that the dramatic increase of the fulmar Fulmarus glacialis (L.) in Britain and Ireland was caused by the rich new food supply made available by offal from whaling and whitefish trawlers. In order to discriminate between these two lines of argument we need to have a detailed knowledge of the diets of seabirds, the quantity of food they consume in relation to the amount available, and the influence of food shortage on the various aspects of population dynamics.

There has been a tendency to neglect seabird diets and feeding ecology, particularly outside the breeding season, and few biologists have succeeded in placing seabirds in context with the other components of marine ecosystems. One reason for this has been the tendency for seabird biologists to study single species in detail rather than tackle the enormous task of investigating the biological relationships of an entire seabird community. Only a handful of detailed investigations into the feeding ecology of seabird communities have been completed. These include the extensive studies of diets, rates of food consumption and ecological interactions between seabirds of the Barents Sea (Belopolskii, 1961), a similar study at Cape Thompson, Alaska (Swartz, 1966), a study of the comparative feeding ecology of seabirds

COMPETITION BETWEEN FISHERIES AND SEABIRDS 227

breeding on a tropical island (Ashmole and Ashmole, 1967), and a detailed investigation into the feeding biology of seabirds breeding on the Farne Islands, Northumberland (Pearson, 1968). Swartz (1966) attempted to assess the quantity of food consumed by the Cape Thompson seabirds. He ignored non-breeders, chicks and food requirements for egg production or moult, and simply multiplied the numbers of breeding adults by an estimated average daily food intake per bird, and multiplied this by the number of days for which the birds were present at these breeding colonies each year. In this way he estimated a total food consumption of I3 100 tonneslyear by the 421 000 seabirds breeding on the 12 km of cliff coastline at Cape Thomp- son. Although the estimate is crude, it is clear that the seabirds consume a significant quantity of the production of the local marine ecosystem. In recent years studies of the metabolic rates of birds in controlled conditions in captivity have shown that a series of equations relating metabolic costs to body mass can be applied to all species of birds, providing they are sub- divided into passerines and non-passerines. Using such equations and a knowledge of the biology of the particular species of interest it is possible to estimate the energy consumption of their populations in the wild (Kendeigh et al., 1977). Using this approach Wiens and Scott (1975) estimated that the seabirds of coastal Oregon consume 22 % of the fish production of the area, while Furness (1978b) calculated that the seabird populations of Foula, Shetland, consume a quantity equivalent to 29 % of the fish production within a 45 km radius of the colony.

Increasing exploitation of fish stocks throughout the world has led to a focussing of attention on the management of marine ecosystems to maximize their yield to man. Rather than treating each fish population as if it were an isolated stock, it has become clear that we must manage whole ecosystems in order to optimize the yields of different commercially valuable organisms. Commercial fisheries exhibit the effects of competition and predator-prey interactions among species of fish (Andersen and Ursin, 1977) and may reveal management problems involving interactions between several trophic levels (May et al., 1979; Vesin et al., 1981). Seabirds are generally top predators in marine ecosystems, and as such are potential competitors with commercial fisheries. There have been demonstrable changes in ecosystem structure in many seas and oceans as a result of overfishing as well as natural climatic or oceanographic fluctuations. It is inevitable that these will influence seabird populations through alterations in the availability, quantity or quality of their food supplies. In order to make management decisions, or simply to predict the effect of such changes, it is important to know what effects changing fishing practices are likely to have on seabird populations, and conversely, whether seabird predation competes with and significantly reduces catches by commercial fisheries; see review by Blaxter and Hunter (1982).

228 R. W. FURNESS

In this review I shall first discuss the evidence suggesting that seabirds are an important component of many marine ecosystems, in that they consume more than a trivial quantity of the production of lower trophic levels, and examine the changes in seabird abundance which have taken place in many parts of the world, apparently as a response to alterations in lower trophic levels of the marine ecosystems. I will then examine the evidence concerning ways in which seabird population dynamics may be affected by food quality or availability, ‘and assess the likely influence of some current trends in the management of fisheries.

II. Estimating Food Consumption by Seabird Populations

A. Field Obsrrvutions

In many seabird communities one or two species are numerically dominant and are responsible for almost all the food consumption of the community. Field observations can be used to determine how many feeding trips each adult makes per day and a sample of adults returning to the colony with food can be shot to discover how much food is carried in an average meal. The number of adults can then be multiplied by the number of feeding trips and the average weight of a meal to give the daily food consumption of the population, and this can be multiplied by the number of days that the birds spend at the colony each summer to give the annual food consumption of the adults in this area. Chick food requirements can be determined by feeding experiments in captivity, or preferably in the field, or can be estimated by the same method described for adults. This direct field approach is a useful independent check on the estimates obtained from bioenergetics modelling, but is not very accurate. For example, several studies have been carried out at common guillemot Uria aaZge (Pontopp.) and Briinnich’s guillemot Uria lomviu (L.) colonies. At the Seven Islands Reserve, Murman, Kaftanovski (1951) estimated adult food intake at 30 g/day. Also on the Murman coast, Belopolskii (1961) estimated adult food consumption at about 60 g/day and that of chicks at 20 g/day, On Novaya Zemlya, Uspenski (1956) calculated that adults consumed 100gfday and chicks 30 to 45g, while Tuck and Squires (1955) at Akpatok Island, Ungava Bay, estimated that adults ate 220 g/day, and from feeding experiments determined that chicks consumed 13.4 g of food for each gram increase in body mass, or the equivalent of about one-half their body mass in food per day.

Tuck’s and Squires’ value of 220g/day is twice as high as Uspenski’s, and seven times as high as Kaftanovski’s calculation. Belopolskii (1961)

COMPETITION BETWEPN FISHERIES AND SEABIRDS 229

states that Kaftanovski accepts that his value is too low and that 60 g/day is a better estimate, but this is still only one quarter of Tuck’s and Squires’ value. Although some of the differences between estimates may be due to variations in climate, food availability, adult activity budgets or the calorific value of the food taken, such empirical calculations of food consumption are clearly only able to provide an order of magnitude estimate for the seabird communities studied by these authors.

Several other methods are potentially available to give direct field measurements of food consumption or individual metabolism. These include the use of doubly labelled water 2H, l*O to determine the total energy budget over a period of time between injections and recapture for removal of a body- water sample (Lifson and McClintock, 1966). This requires the capture and recapture of an individual over a short period (usually about 24 h), assumes that the bird’s behaviour is not affected by capture and injection, and involves costly isotopes and technically complicated laboratory analyses of samples. Nevertheless it has been applied very successfully to investigating the free- living energetics of swallows and martins (Bryant, 1979; Hails and Bryant 1979) and will no doubt be an important tool in studies of seabird energetics in the near future.

Injection of radioisotopes of elements whose excretion rate is correlated with the rate of metabolism (Odum, 1961) may also be used in future studies, but the methodology for this is not yet fully developed (Gessaman, 1973). Heart rate biotelemetry may also eventually be of use in estimating free- living metabolism or costs of specific behaviour in the daily activity budget. It has been used to study the diving and flying respiration of unrestrained birds (Butler, 1980) and to estimate metabolic costs of redshank Tringa totanus (L.) behaviour in laboratory conditions (Ferns et a[., 1980), although its application to metabolic studies is hindered by individual variations in the relationship between heart rate and oxygen consumption (Ferns et al., 1980), changes in heart stroke volume or oxygen content of blood independent of heart rate (Butler et al., 1977) and changes in heart rate induced by “emo- tional” stresses during periods when the metabolic rate may remain un- changed (Ball and Amlaner, 1980).

Nevertheless, present studies to validate the use of bioenergetic equations include only those making empirical measurements of food consumption, which we have seen to be of rather uncertain, and apparently low accuracy.

B. Bioenergetics Equations

As direct methods of measuring food consumption of free-living birds are often unsatisfactory, indirect methods must be employed. Values obtained for caged birds under controlled conditions can then be projected to free-

230 R. W. FURNESS

living populations with known activity budgets in measured environmental conditions. The use of bioenergetics equations is reviewed by Kendeigh et al. (1977) and the application of bioenergetics modelling in estimating the potential ecosystem impact of granivorous passerines is reviewed by Wiens and Dyer (1977). The aim of this approach is to obtain the values for the numerous parameters of breeding biology required for modelling from the extensive literature on seabird ecology and to determine those not already known from field studies. These input parameters are then coupled with bioenergetics equations, generalized for all non-passerine species in relation to body mass, to calculate the daily energy budget of the seabird populations under natural conditions.

Basal metabolism (the rate of energy utilization by animal tissues at rest and unstimulated by food assimilation or digestion or by low or high temperature) can be strictly defined. It is therefore a physiologically useful measure, but it cannot be precisely measured in higher animals, for which the term “standard metabolism” (Krogh, 1916) is used to refer to “basal metabolism” in a less strict sense. In the field, animals rarely, if ever, exist at their standard metabolic rate, since they are usually digesting food and are not at complete rest. For this reason standard metabolism is not appropriate for bioenergetics modelling. A more useful measure is “existence metabolism”. This is defined as the rate of energy utilization by caged birds able to undertake limited locomotor activity (but not flight) and which are maintaining a constant mass and not undergoing reproduction, moult, growth or migratory rest- lessness. Kendeigh (1 970) gave logarithmic allometric equations for existence metabolism,of 18 species of bird; 13 passerines and 5 non-passerines. Some of these were determined for both males and females, increasing the number of data points for regression analysis where the species are sexually dimorphic in body mass. He showed that the relationship differed significantly between passerines and non-passerines at 30°C, and between long and short photoperiods, but that species within each grouping did not differ significantly from the common regression derived for 0°C or 30°C. The implication from this is that the existence metabolism of any seabird species can be calculated from a knowledge of its mass and interpolation between the values obtained at these two ambient temperatures.

Wiens and Scott (1975) based their simulation of Oregon seabird energetics on Kendeigh’s equation derived from studies of five non-passerines (a duck, a goose and three pheasant species). Furness (1978b) also used this equation as the basis of his simulation of the energy budget of a Shetland seabird community, but pointed out the wide confidence interval associated with this equation, such that it provided the greatest single source of error in the entire model.

Fortunately, since 1970, extensive studies of existence metabolism have

COMPETITION BETWEEN FISHERIES AND SEABIRDS 23 1

been carried out with a wide variety of species. Kendeigh et al. (1977) con- firmed the differences between passerines and non-passerines and between photoperiods, and found that the regressions were identical for species in almost all orders of non-passerines. Based on 40 or more species of non- passerines the regressions have much smaller standard errors than those used by Wiens and Scott (1975) or Furness (1978b). For a typical seabird this would be about 1 rather than the 15-25 % resulting from the Kendeigh (1970) equations. In addition to existence metabolism, Kendeigh et al. (1977) give equations

for the calculation of the energy costs or savings of weight change, insolation, huddling, wind and rain, gliding and flapping flight, swimming, running, migration, egg-laying, incubation, brooding, moulting, chick growth and existence. For several of these categories the energy costs or savings are negligible for seabirds in relation to the overall energy budget.

Apart from extremely exceptional cases, such as the creching (huddling together) of Antarctic penguins, the only variables likely to make substantial contributions to the population energy budget are foraging activity (usually flapping or gliding flight, surface or underwater swimming), chick daily energy budget, adult moult costs and egg production costs.

Kendeigh et al. (1977) use empirical values of energy expenditure during sustained horizontal flapping flight to provide an equation for non-passerines (excluding aerial feeders) based on 11 species, which has a standard error of about 2 4 % depending on the body mass of the species. Energy costs of gliding flight and surface swimming have been determined for a small number of species and tend to be about twice resting metabolism (Prange and Schmidt-Nielsen, 1970; Baudinette and Schmidt-Nielsen, 1974). No data are available for the costs of swimming or flying under water, a feeding technique widely employed by penguins, auks, diving petrels, shearwaters, cormorants and divers. This provides a major source of uncertainty in the calculation of foraging costs of adults for seabird communities consisting of substantial numbers of these species. For lack of better data I have assumed that the costs of underwater swimming will approximate to the costs of sustained flight (Furness and Cooper, in press). Given the uncertainties over rhe allocation of foraging time to different activities, such an assumption is unlikely to provide a major source of error in the modelling process.

Chick daily energy budgets have been computed by Kendeigh et al. (1977) from empirical data derived from detailed studies of house sparrows Passer domesticus (L.) and black-bellied tree ducks Dendrocygna autumnalis (L.) and supported by less detailed studies on a number of other species. Regres- sion of the daily energy budget on body mass of the young birds (Fig. 1) shows good agreement in spite of the wide variety of species and modes of development. Adult moult costs depend on the mass of feathers replaced.

232 R. W. FURNESS

5 10 50 100 500 1000

Weight ( g )

FIG. 1. Regression of the daily energy budget on weight of young birds of a variety of species (from Kendeigh et al., 1977).

Plumage mass is proportional to body mass to the power 0.96 (TurEek, 1966). Feather replacement in the house sparrow costs 185 kcals/bird (Kendeigh et al., 1977) so the cost of moult may be approximated by the general equation:

Moult cost = 8.3 Wo’B6 (kcals)

where W is the body mass in grams. This is the only equation available for estimating seabird moult costs and is likely to be rather inaccurate since it is based on the study of only one passerine species. However, moult costs are a very small part of a seabird population energy budget (Furness and Cooper, in press) so this approximation is adequate. The cost of egg production is calculated from the fresh mass of the egg at laying, the calorific value of the egg (averaging 1.3 kcals/g wet mass) (King, 1973; Schreiber and Lawrence, 1976) and the efficiency of egg production from body reserves, taken to be 73 % (El-Wailly, 1966; King, 1973).

While it is desirable to have more precise knowledge of swimming and moult costs in seabirds, the accuracy of a simulation model is limited not so much by these bioenergetics equations, but mainly by imprecisions in the estimates of seabird population sizes and the foraging activity budgets of adults.

Energy Demand

FIG. 2. Compartmental diagram of population bioenergetics model. Rectangular boxes indicate state variables; five-sided boxes, computational controls; circles, input variables. Solid arrows indicate flows of materials or energy or changes of state; dashed arrows indicate controls or computational transfers. Input variables, CS: clutch size; HS: hatching success; FS: fledging success; PFS: post-fledging survival; JDR: juvenile daily mortality rate; WM: winter mortality; PPBF: proportion double- brooded; PS: population size at start; PBD: post-breeding dispersal ; ADR: adult daily mortality rate; PE: population size at end; AMW: adult mean weight; TEMPC: ambient temperature; HMW: hatching weight; FW: fledging weight; K: growth rate of chicks (from Wiens and Innis, 1974).

R. W. FURNESS

( a )

Spring Summer Autumn

Date

FIG. 3. Assumed seasonal patterns used in bioenergetics model for Shetland seabird populations; absolute dates and numbers vary from species to species (from Furness, 1977b).

C . Input Parameters, Model Sensitivity and Output Accuracy

Wiens and Innis (1974) and Wiens and Scott (1975) used a population submodel based on breeding biology parameters to compute the population

production

dally

chicks

W FIG. 4. Population bioenergetics model for Shetland seabird communities: input variables A, B, C parameters for

growth equation, T: temperature; F.U.E. : food utilization efficiency (from Furness, 1977b). logistic chick

236 R. W. FURNESS

age structure and density on each day of the year. This was then integrated with an energy submodel, consisting of growth and bioenergetics equations, to determine the population energy demand (Fig. 2). They subjected their model to a sensitivity analysis (Smith, 1970) to test its “robustness” (Levins, 1966). Their conclusion was that an alteration of most input parameters had a correspondingly smaller, often negligible, influence on the model output estimates of total breeding season energy demands. Furness (1978b) constructed a simple model of the seasonal patterns in numbers of breeders, non-breeders and fledglings of each breeding seabird species in the vicinity of Foula, Shetland and their foraging activity budgets and distributions of egg laying (Fig. 3). This was integrated with an energy submodel (Fig. 4) similar to that used by Wiens and Scott (1975).

The sensitivity of the model was explored by altering each input or bioenergetics parameter value in turn by 1 x, and recording the percentage change in the output estimate of total population energy requirement. Parameter sensitivity values were defined as the percentage change in the output value resulting from the 1 change of an input value. Almost all parameters had sensitivity values of considerably less than one, but their exact magnitude depended to some extent on the biology of the species. The great skua Catharacta skua Brunnich is a seabird with a fairly high body mass, which spends only a few hours foraging each day during its five- month breeding season. The Arctic tern Sterna paradisaea Pontopp. nests at the same colony, but has a small body mass, spends only three months at the breeding site and at least one adult of each pair is foraging throughout most of the daylight period. In the case of the great skua the model is most sensitive to’the estimation of existence metabolism, numbers of individuals in the population and food utilization efficiency. For the Arctic tern the model is most sensitive to the activity budgets of the adults, numbers in the population and food utilization efficiency. Determination of sensitivity values should be carried out whenever this type of simulation modelling is undertaken as it indicates which parameters must be precisely known to give output results with small standard errors. As the majority of sensitivity values are small, the standard errors of the few parameters with large sensitivity values will primarily determine the precision of the output results. Furness (1978b) extended this analysis by using a Monte Carlo technique. A computer function was employed to generate a random value for each parameter, with a specified normal distribution, using the known mean and standard deviation for each parameter. The errors in estimated parameter values were assumed to be uncorrelated and the generated set of parameter values was then input into the model and population energy requirements calculated. The Monte Carlo analysis comprised 300 runs of the model for each species, each run using a unique set of normally distributed, randomly generated parameter values.

TABLE I. RESULTS OF A MONTE CARL0 SIMULATION ANALYSIS OF THE PRECISION OF OUTPUT ESTIMATES OF ENERGY REQUIREMENTS

FROM A BIOENERGETICS MODEL (MEANS AND STANDARD DEVIATIONS IN KCALS X 104/YEAR) (FROM FURNESS, 1978b)

Energy requirement estimate: Great skua Arctic tern

Parameter for which energy requirement Standard Coefficient Standard Coefficient

was estimated Mean deviation of variation Mean deviation of variation

Breeders : Existence Activity Egg production Total

Activity Total

Growth

Nonbreeders : Existence

Chick : Existence

Entire. population

18 806 6770

115 25 692

2144 558

2700 2111 843

31 345

5735 2465

11 7394 654 210 792 613 86

8800

30.5 36.4 9.6

28.8 30.5 37.6 29.3 29.0 10.2

28.1

4563 5178

47 9790 202 224 426 837 168

11 224

1271 2027

6 2839

80 119 182 242 27

3180

27.9 39.1 12-8 29.0 39-6 53.1 42.7 28.9 16.1

28.3

238 R. W. FURNESS

Using Kendeigh (1970) equations the output had a 95 % confidence interval for the total population energy requirement of each species of & 50 % of the mean. Using the same model but with the bioenergetics equations replaced by those in Kendeigh et al. (1977) the 95 % confidence interval is reduced to &30 % of the mean (Furness, 1982) for most of the seabird species breeding in Shetland colonies. Further inprovements in model precision are limited by the parameters population size, food utilization efficiency and adult activity budget: As seabird populations can rarely be estimated to an accuracy better than &20 % (Harris, 1976) it is pointless to attempt to refine the model further, unless the aim is to examine seasonal patterns of energy expenditure or the proportions used in different activities or by different parts of the population.

Furness (1978b) found that adult existence requirements were at least equal to, and often much greater than requirements for foraging acitivity (Table I).

( 0 )

Arctic tern

( b ) Great sku0

Month

FIG. 5 . Model output estimates of the daily energy requirements of populations of Arctic terns and great skuas: upper solid line, total population requirement; dashed line, breeding adults; dotted line, nonbreeders; lower solid line, chicks (from Furness, 1977b).

TABLE 11. SALDANHA BAY SEABIRD POPULATION ENERGY REQUIREMENTS FOR ADULT EXISTENCE, ADDITIONAL COSTS OF FORAGING, MOULT, EGG PRODUCTION AND CHICK EXISTENCE PLUS GROWTH. VALUES ARE ANNUAL TOTALS BEFORE ALLOWANCE HAS BEEN

MADE FOR DIGESTIVE EFFICIENCY (FROM FURNESS A N D COOPER, IN PRESS)

Jackass penguin Cape gannet Category

Population Percentage Population Percentage requirement of total requirement of total (kJ x 108) (kT x 108)

Adult existence 229.4 71.4 101-9 50.7 Adult foraging 62.4 19.4 72.7 36-1 Chick daily budget 24.6 7.6 22.2 11.0 Adult moult 4.9 1-5 4.3 2-1 Egg production 0.4 0.1 0.1 0. I

~~

Cape cormorant

Population Percentage requirement of total

(kJ x 108)

72.1 63.6 32.3 28-5 6.2 5.4 2.8 2.5 0.05 0.0

Population total 321.7 100 % 201.2 100% 113.4 1 0 0 %

240 R. W. FURNESS

Furthermore, the energy requirements of Shetland seabird chicks or nonbreeders were very small in comparison to the requirements of the breeding adults, even around the middle of the breeding season when numbers of nonbreeders at the islands reach a maximum and chick food requirements peak (Fig. 5). Using a slightly altered version of the model in a study of the energy requirements of seabird populations in the Saldanha fishery area of South Africa (Furness and Cooper, in press) the same pattern was found. Adult existence accounted for 50-70% of the total population annual energy requirement, while the costs of moult and egg production represented less than 3 % for each species (Table 11).

Using a bioenergetics model the annual energy requirement of a seabird population can usually be estimated with a precision of about f30% of the mean, given the detailed data that exist on the breeding biology of most seabird species. The main limitations to this are inadequate census data or a lack of knowledge of the budgets of adult foraging activity. The three seabird communities examined to date give similar results in terms of their impact on food supplies. Wiens and Scott (1975) estimated that the seabirds of coastal Oregon consume 22% of the annual fish production. Furness (1978b) estimated that the seabirds of Foula, Shetland consume the equivalent of 29% of the fish production within a 45 km radius of the colony (the neighbouring major colonies are approximately 70 km away) and Furness and Cooper (in press) estimated that the Saldanha seabird populations of the mid-1970s consumed 13 000 tonnes of fish each year, equivalent to 24% of the annual catch by commercial pelagic fisheries between 1971 and 1976, and representing an annual cropping of 20% of the South African anchovy Engraulis capensis Gilchrist biomass in the Saldanha fishery area. These high consumption rates may exaggerate the role of seabirds as marine predators, since the three communities studied were chosen partly because of their large seabird populations. Nevertheless, it is clear that, at least in some marine ecosystems, seabird predation is quantitatively important and may potentially compete with fishing interests, while changes in fish stocks are likely to have a direct effect on seabird population biology.

111. Changes in Marine Ecosystems and Seabird Populations

A. British Columbia

Robertson (1972) investigated the relationship between fish-eating birds and stocks of the Pacific herring Clupea pallasii Cuvier et Valenciennes in the Gulf of British Columbia. He assumed that the daily food intake of the


seabirds averaged 18-20% of body mass, based on values obtained by Spaans (1971) when feeding captive adult herring gulls adlibitum on herring, by Heinroth and Heinroth (1928) for grebes and by Madsen and Sparck (1950) and Skokova (1962) for cormorants. Bioenergetics considerations would suggest that this value is appropriate for the existence requirements of a 1000 g seabird at 15'22, but would be too low by a factor of two for a 100 g seabird at 5°C. In addition, foraging activity costs need to be taken into account and these may represent an addition of 20-100% of the existence costs (Furness, 1978b; Furness and Cooper, in press). The main avian predators of the Pacific herring in the area around the Gulf Islands are western grebes Aechmophorus occidentalis (Lawrence), Brandt's cormorants Phalacrocorax penicillatus (Brandt), glaucous-winged gulls Larus glaucescens Naumann, black-throated divers Gavia arctica (L.) and common guillemots. These are all large species, averaging from 1000 to 2500 g, and their activity requirements in addition to existence are likely to be relatively small, particularly as they occur in this region principally outside the breeding season so do not have to travel between feeding areas and breeding colonies. The assumption of a daily intake of 20 % of body mass is therefore not likely to be far from the truth, and actual consumption is unlikely to be more than twice this amount.

Robertson's (1972) study is particularly interesting because, in contrast to all the other investigations, it examines an area where numbers of breeding fish-eating birds are low and almost all the impact on fish stocks occurs as a result of predation by wintering populations. The herring stock migrates to the west coast of Vancouver Island during the summer and autumn (Taylor, 1964) which probably explains the small numbers of fish-eating birds breeding on the Gulf Islands. The herring stock was overfished in the 1950s and early 1960s, with an average annual catch of 45 000 tonnes in this area. Stock depletion caused the closure of the fishery in 1967. A recovery began to take place, with spawning in 1970 exceeding the 1940-64 average. However, the 1971-72 adult stock was estimated at 26 000 to 41 000 tonnes, which is still perhaps less than half of the stock size of the 1950s (Robertson, 1972).

From winter surveys of numbers and analysis of stomach contents Robert- son estimated that the seabirds consumed 9-6 tonnes of herring each day (TableIII), or 1760 tonnes between 1 November 1971 and 31 March 1972. He estimated that summer consumption might add a further 440 tonnes, giving an annual consumption of 2200 tonnes in the year 1971-72. This includes consumption of juvenile herring, and from consideration of fish sizes in stomach samples it would appear that about 1800 tonnes of the herring eaten by the seabirds in the Gulf Islands area are adult stock. This amounts t o 4-7 % of the estimated adult Pacific herring stock of this area. This calcula-

TABLE 111. ESTIMATION OF DAILY HERRING CONSUMPTION BY FISH-EATING BWS IN THE GULF ISLANDS, BRITISH C~LUMBIA (FROM ROBERTSON, 1972)

Parameter

~ ~ ~ ~ ~~~

Black-throated Western Brandt’s Glaucous-winged Common gull guillemot diver grebe cormorant

Body weight (8) 2450 1425 2375 1075 1045 Mean daily food intake (g) 490 285 475 193 209 Herring in diet (%) 100 45 75 so 85 Weight herring/day/bird (g) 490 128 356 96 178 Winter population 5537 19 369 7527 7381 4406 Herring intake/day (kg) 2713 2484 2680 709 784


tion is rather crude, but it is unlikely that seabird predation accounts for more than twice this amount, so no more than 14 % of the adult herring stock is taken by seabirds around the Gulf Islands. However, the stock is also subject to seabird predation during the summer and autumn when it is off the west coast of Vancouver Island, and the extent of this has not been quantified.

One explanation for the rather low predation pressure by seabirds is found in the history of human exploitation of the Pacific herring stock. The recovery of the stock after excessive overfishing has probably occurred more rapidly than the seabirds have been able to respond to the improved food supply. The seabird numbers wintering around the Gulf Islands in 1971-72 were certainly higher than in 1963-64. Counts between November and March 1971-72 were double those made in the same area and months in 1963-64 (Edwards, 1964; Robertson, 1972). The intense fishing in 1963-64 will have left little of the adult herring stock, while in 1971-72 the entire (increasing) stock will have been available for the seabirds.

B. California Current

The Farallon Islands lie at the edge of the continental shelf off central California and hold important concentrations of breeding seabirds. Their populations have been documented in some detail as far back as the 1850s, and this section is based on the detailed description and analysis of their population histories presented by Ainley and Lewis (1974). The population histories provide evidence of the influences of human persecution and exploitation of seabirds, oil pollution, the effects of a major oceanographic change and, at a later date, the complete disappearance of a major fish population (an important prey species for some of the seabirds) as a result of overfishing and a change in ecosystem structure.

Five species show the same pattern of population change. These are Brandt’s cormorant, pelagic cormorant Phalacrocorax pelagicus (Pallas), western gull Lams occidentalis Audubon, common guillemot and pigeon guillemot Cepphus coZumba (Pallas). They all declined in numbers during the last half of the 1800s and have recovered in recent years (Figs 6 and 7). The declines were clearly caused by prolonged low reproductive success due to disruption of nesting by human activities or prolonged high mortality from oil pollution, while the subsequent recovery was allowed by protection measures and a reduction in oil pollution (Ainley and Lewis, 1974). Several species showed different patterns. Throughout the period from 1850 to the present ashy storm petrels Oceanodroma homochroa (Coues) and Leach’s storm petrels 0. leucorhoa (Vieillot) have nested on the islands in populations of a few thousand and many hundreds of individuals, respectively. As

500 000

400 000

300 000

e 50000

r

100 - 90 - - p 80-

3 70-

5 60-

50-

b 40- ’ 30- 20-

(D

t

L 0 n

z 2 5 - 20 - 1 5 -

10-

5 -

E,

0

180070 8090 1900 10 20 30 40 50 60 70

Yeor

FIG. 6. Population changes of Farallon Island seabirds: 0 common guillemot; pigeon guillemot; A western gull; V Brandt’s cormorant (from Ainley and

Lewis, 1974).

Year

FIG. 7. Population changes of Farallon Island seabirds : Cassin’s auklet (from Ainley and Lewis, 1974).


storm petrels are nocturnal and secretive, nesting in crevices or in deep and narrow burrows, their populations appear to have been unaffected by human activities on the islands and their populations have not changed noticeably. Cassin’s auklets Ptychorumphus aleuticus (Pallas) increased rapidly in number during the late 1800s and have since remained in high numbers (Fig. 8). Double-crested cormorants Phalucrocorux auritus (Lesson) and tufted puffin Lunda cirrhuta (Pallas) populations declined due to disturbance and oil pollution, but unlike the other species they have failed to recover, and they have remained stable at low numbers for many years (Fig. 8).

1860 70 80 90 1900 10 20 30 40 50 60 70

Year

FIG. 8. Population changes of Farallon Island seabirds: 0 tufted puffin; A double-crested cormorant (from Ainley and Lewis, 1974).

Ainley and Lewis (1974) suggest that the most plausible explanation for the dramatic increase in the Cassin’s auklet population between 1870 and 1900 is that it resulted from oceanographic changes along the California coast. The California current flows south along the coast carrying cold nutrient-rich water. During spring and early summer the current is increased by strong north-west winds which further encourage upwelling of cold nutrient-rich water. Warm, subtropical nutrient-depleted oceanic water that would otherwise be present is displaced by the California current. From time to time circulation in the California current is altered with the result that warm waters move unusually far north for extended periods. Such periods of warm water incursions can be seen from temperature records and from records of warm-water animals found much farther north than normally recorded (Hubbs, 1948; Robinson, 1965). A warm-water period of unusually long duration occurred from before 1853 to the 1870s, and allowed many sessile warm-water species to penetrate farther north than recorded since (Hubbs, 1948). This warm water incursion affected the Farallon Islands

246 R. W. FURNESS

as well (Ainley and Lewis, 1974). Two similar warm-water periods of short duration have been recorded recently. During both, the breeding success of Cassin’s auklets was significantly lower than normal, with 0.27 and 0.58 chicks fledged per pair compared to 0.71 and 0.62 chicks fledged per pair in normal years of colder sea temperatures. As these auklets feed on zooplankton, particularly euphausiids (Manuwal, 1972), Ainley and Lewis (1974) suggest that the very much lower productivity of the warm water (Aron, 1960) results in reduced food resources for the auklets and hence poor breeding. They infer from this that the extended period of warm water up t o the 1870s will have led to a population decline of Cassin’s auklets as a result of prolonged sub-optimal breeding. The documented increase after the 1870s was then a response to a return to the high productivity of the cold- water current which re-established itself about this time. As no other sea birds on the Farallon Islands compete with Cassin’s auklet for food or nesting sites the populations of the other species did not suffer direct competitive effects as a result of the changes in Cassin’s auklet numbers, although they may have been influenced to a slight extent by the lower productivity of the warm water. Since they feed at a higher trophic level than Cassin’s auklet this influence will be damped down to some extent by transfers of energy through the ecosystem.

The double-crested cormorant and tufted puffin populations contrast with those of the other seabirds by their failure to recover after human exploitation ceased around 192MO. This pattern is found at other seabird communities on the California coast, and is also shared by the Steller’s sea lion Eumatopias jubatus (Schreber) which has declined in numbers from 1940 to the present. In contrast, the Californian sea lion Zalophus californianus (Lesson), the northern elephant seal Mirounga angustirostris (Giel) and the rhinoceros auklet Cerorhinca monocerata (Pallas) have recolonized islands off the California coast and greatly increased in numbers. These differences cannot be accounted for by changes in sea temperature, but appear to be related to the loss during the 1940s of the Pacific sardine Sardinops caerulea (Girard) population. This loss appears to have been due to a combination of overfishing and environmental stress, the latter caused ly an extended period of cold water (Clark and Marr, 1955; Frey, 1971). Ainley and Lewis (1974) present evidence to support their argument that the double-crested cormorant, tufted puffin and Steller’s sea lion were heavily dependent on the Pacific sardine, and less well adapted than the other species to feed on the rather smaller ecological replacement, the northern anchovy Engraulis mordax Girard. These three predators are the largest species of cormorant, puffin and otariid found in the region, and it is a general rule that larger species utilize larger prey (MacArthur and Levins, 1964). Certainly the available information on diet does support this argument (Ainley and Lewis, 1974).

COMPETITION BETWEEN FISHERIES A N D SEABIRDS 241

This case history provides a good example of the way in which seabird numbers can be reduced, held below the carrying capacity set by food availability, or even brought to local extinction by human exploitation, disturbance or pollution. It also shows that changes in productivity or species stock sizes induced either by environmental fluctuation or by human fishery pressures can have an important controlling influence on the absolute size of some seabird populations, or may tip the balance of competition in favour of one species or another. A further increase in the food supply of a seabird population held below the environment’s carrying capacity by human persecution is unlikely to result in an increase in the numbers of that species (a “more superabundant” food supply is no improvement over a “superabundant” food supply). It is probably reasonable to infer that the increase in the population of Cassin’s auklet between 1870 and 1900 was a direct result of improved food supply and that the numbers present before 1870 were limited by food availability. We will return to the possible methods by which the population sizes are limited by food at a later stage.

C. South Africa

The Benguela current system is of particular interest because it supports the important South West and South African purse-seine fisheries and large numbers of coastal breeding seabirds utilize the same fish stocks. Relation- ships between the seabirds, fish stocks and pelagic fisheries are discussed in detail by Frost et al. (1976), Crawford and Shelton (1978), Crawford (1979), Furness and Cooper (in press). The seabirds nest and roost on small and generally flat offshore islands and man-made platforms, and have provided an annual harvest of guano for over 100 years. The seabird communities are dominated numerically, and even more so in terms of biomass, by three large diurnal species, the jackass penguin Spheniscus demersus (L.), the cape cormorant Phalacrocorax capensis Licht, and the Cape gannet Sula capensis Licht. The populations of these species are largely protected in order to maximize the harvest of guano, although egg collecting occurred at some penguin colonies in the past, and may have been sufficiently intense to cause minor perturbations in the size of the breeding populations (Siegfried and Crawford, 1978). Studies of the diets of these seabirds show that all three are to a large extent predatory on shoaling pelagic fish (Table IV). Their main prey are the South African pilchard Sardinops ocellata Gilchrist and the South African anchovy. Round herring Etrumeus teres (de Kay) and horse mackerel Trachurus trachurus L. are also taken, but in much smaller numbers, while other fish, crustaceans and cephalopods are of minor importance.

The South West African purse-seine fishery is dominated by the pilchard. Before the mid-1960s this was the only species caught, but stock depletion

TABLE LV. DIETS OF THE CAPE GANNET, JACKASS PENGUIN AND CAPE CORMORANT

Species Cape gannet Jackass penguin Cape cormorant Period 53/54 54/55 54/56 57/58 77/78 53/54 54/55 57/58 77/78 Source of data Davies, Davies, Rand. Matthews, Cooper, Davies, Davies, Matthews. Cooper, Davies, Davies. Rand, Matthews, Cooper.

% by frequency:

53/54 54/55 54/56 57/58 77/78

1955 1956 1959 1961 1979 1955 1956 1961 1979 1955 1956 1960 1961 1979

Pi 1 chard 44 Anchovy 30 Horse

mackerel 18 Mackerel 6 Round herring 0 Other fish 1 Cephalopods 0 Crustaceansand

polychaetes 0

Study area Saldanha (fishery area) Bay

62 19 26 25

I t 30 0 4 0 0 1 10 0 12

0 0

St Hout Helena and Bay Saldanha

85 0

10 0 0 5 0

0

Walvis

SWA Bay

13 39

1 I 1

44 1

0

Saldanha Bay

37 2

23 20 0

10 0

8

Saldanha Bay

49 44

2 2 0 1 2

1

St Helena Bay

a3

0

6 0 0 0

I 1

0

Walvis

SWA Bay

1 79

0 0

10 0

10

0

Saldanha Bay

36 32

21 0 0 6 0

5

Saldanha Bay

44 19

14 0 0

18 0

5

St Helena Bay

I5 12

16 1 0

45 0

9

Hout and Saldanha

~

76 0

18

0 0 0 1

3

Walvis

SWA Bay

~~

5 55

I 0 25 I2 2

0

Saldanha Bay


became severe in the late 1960s and the pilchard stock biomass was reduced by 50% in three years (1967-70). After this, increased catches of anchovy and horse mackerel were taken and catches of pilchards fell considerably, as did the total catch (Crawford and Shelton, 1978). The South African purse- seine fishery is also largely based on pilchard and anchovy, although horse mackerel and mackerel Scornber japonicus Houttuyn have been caught in large quantities at certain stages during the history of the fishery, and round herring and a lantern-fish Lampanyctodes hectoris (Gunther) have provided small catches since the late 1960s. Heavy fishing in the 1950s and early 1960s led to a reduction in landings and the introduction of a smaller meshed net between 1963 and 1965, after which pilchard and horse mackerel stocks were considerably reduced by fishing juvenile stock, and anchovy became the main contributor to the fishery (Crawford and Shelton, 1978).

The distributions of breeding penguins, cormorants and gannets between colonies show differences between the species which can be related to food and species biology. Although all three species can eat fully grown pilchards, Cape cormorants select slightly smaller fish than taken by Cape gannets, and this may explain their tendency to concentrate at the northern extremities of both the South West and South African fishing grounds where recruitment of pilchards occurs (Crawford and Shelton, 1978). Flightlessness limits the feeding range of the jackass penguin. Frost et al. (1976) estimate that the theoretical maximum foraging area of a breeding jackass penguin is no more than 1500 km2. Siegfried et al. (1975) found that nearly 80% of jackass penguins at sea were within 12.5 km of the nearest mainland, while Dunnet (1977) found that 98 % of those he saw (473) on a transect line out from the shore were within 4 km of land. Frost et al. (1976) suggest that due to their limited foraging range jackass penguins can only breed in areas where the temporal and spatial pattern of prey distribution is both highly predictable and favourable. Most of the jackass penguin population breeds on Dassen Island and on the Saldanha Bay islands, which places them in the centre of the South African purse-seine fishery, which being characteristically multispecies and especially so in the vicinity of these islands, indicates a stable food resource (Crawford and Shelton, 1978). As the diets and nest site requirements of these three seabirds show considerable overlap it appears that the relative numbers of each species nesting at each of the colonies along the coast will be determined by these small differences in ecology which lead to slight competitive advantages for one of the three species. The cormorants, gannets and penguins can be viewed as sharing out the fish resource according to their relative competitive abilities.

As the three seabird species can be considered to be in the same trophic niche, estimation of the energy requirements of their combined populations gives a measure of the impact of the seabird community on stocks of pelagic

250 R. W. FURNESS

fish. Furness and Cooper (in press) used a bioenergetics model as described earlier. This was applied to the seabird populations of the Saldanha fishery area as the most detailed data on numbers, diets, feeding and breeding biology were available for these colonies. Numbers of breeding adults were obtained from colony counts published by Rand (1963), Frost et al. (1976), Crawford and Shelton (1978), Cooper (1979) and Crawford, Shelton and Cooper (in press). Numbers of immature age classes were determined by constructing a life table for a stable population using available or estimated values of adult survival, age at first breeding, clutch size, hatching and fledging success. Details of the input parameters are given in Furness and Cooper (in press). A sensitivity analysis indicated that in this model, population energy estimates were particularly sensitive to errors in estimates of seabird population size and rather less to errors in the hours spent in flapping flight or swimming underwater. The likely errors in other parameters or the model equations themselves all contribute relatively little to the total output error.

TABLE v. ANNUAL ENERGY CONSUMPTION BY SEABIRD POPULATIONS IN THE

SALDANHA FISHING GROUND, SOUTH AFRICA (FROM FURNESS AND COOPER, IN PRESS)

Annual energy consumption (kJ x lo8)

Saldanha Bay Dassen Island Saldanha fishing Species

Islands ground (total)

Jackass penguin Cape gannet Cape cormorant

Total

402.1 424.9 827,O 251.5 0.0 251.5 141.8 94.4 236.2

795.4 5 19.3 1314.7

Estimated annual energy costs of adult existence, feeding activity (additional to existence costs), moult, egg production and chirk daily energy budget (growth plus existence) were calculated for the populations of each species in Saldanha Bay (Table 11) and Dassen Island. Adult existence costs represented 51-70% of the total population budget, while costs of moult and egg production accounted for less than 2 % of any species’ total (Table 11). Total population annual energy requirements need to be increased by a factor of 1 *25 to allow for a digestive efficiency of 80 % (Table V). The resulting annual consumption by each population can be converted to tomes of each fish species from a knowledge of diets. Diets and consumption figures are given in Table VI. Pelagic fish species, particularly anchovy, predominate.

TABLE VI. DIET AND CONSUMPTION OF FISH BY SEABIRD POPULATIONS IN THE SALDANHA FISHING GROUND, SOUTH AFRICA (FROM

BY THE DEPARTMENT OF INDUSTRIES, SEA FISHERY BRANCH FURNESS AND COOPER, IN PRESS); DIETARY INFORMATION DIFFERS FROM TABLE I11 AS IT INCLUDES UNPUBLISHED DATA COLLATED

Jackass penguin Cape gannet Cape cormorant Total

Diet Tonnes Diet

weight) per year weight) (% by consumed (% by

Total 10 338

Anchovy 80 8270 60 Round herring 10 1034 0 Pilchard 5 517 15 “Other species” 5 517 25

Tonnes consumed per year

3144

1887 0

472 786

Diet

weight) (% by

Tonnes Tonnes consumed

per year per year 2953 16 435

consumed

55 30

5 10

1624 1 1 781 886 1920 148 1137 295 1598

252 R. W. FURNESS

The category “other species” consists largely of demersal species scavenged from trawlers by Cape gannets (Sinclair, 1978) and squid eaten by jackass penguins. These fish consumption statistics apply to the seabird populations of the early 1970s, when most of the censuses and biological studies were carried out. It seems reasonable to compare the fish consumption statistics (Table VI) with the pelagic fish catches and stocks of the Saldanha fishery area during these years, since relatively few of the seabirds appear to travel beyond the limits of this region to feed (Furness and Cooper, in press). The maximum lengths of fish recorded from stomach samples of the Cape gannet, jackass penguin and Cape cormorant respectively are 35 cm (Rand, 1959), 57 cm (Matthews, 1961), and 26 cm (Davies, 1956). Very few pelagic fish in South African waters exceed these sizes (Table 5 in Crawford and Shelton, 1978; Crawford et a/ . , 1978) so that the majority of the pelagic fish are suitable for consumption by these seabirds. The total annual consumption of ca. 13 000 tonnes of fish by the Saldanhaseabirdsdoes represent asignificant loss to the pelagic fishery. Almost all consumption comprises pelagic species, particularly anchovy (ca. 10 000 tonnes). As the commercial fishery has been working at or above the maximum sustainable yield for most of the pelagic species (Baird, 1975; Centurier-Harris, 1977; Crawford, 1979; Newman et al., 1978; Stander and LeRoux, 1968) a reduction in consumption by seabirds would lead to an increased fishing yield. Between 1971 and 1976 catches of pelagic fish in the Saldanha fishery area varied between 12 100 and 85 600, averaging 55 000 tonnes (Crawford, 1979) so that the seabird consumption in the same period equalled the fish landings in one year and averaged 24% of the mean catch. It would be naive to assume that the commercial catch would increase by 24% on removal of all seabirds, since some of the “surplus” stock created would be consumed by other natural predators (for example snoek Thyrsites awn (Euphrasen) and fur seals Arctocephah pusilius (Peters)), but most of the fish consumed by the seabirds would be of a size which had recruited into the catchable part of the stock.

Although some South African pelagic fish species show distinct migrations in relation to age, and occur more frequently in rarticular coastal areas, the seasonal pattern in commercial catch per unit effort is one reflecting a relatively constant resource within and between areas (Crawford, 1979), as might be expected from the fact that some seabirds are breeding in all months of the year. This appears to be particularly evident in the Saldanha fishery area. Between 1971 and 1976 the South African mixed-species pelagic fishery landings comprised 59 % anchovy, 22 % pilchard, 11 % mackerel, 4 % round herring and 4 % other species (in terms of biomass). Assuming that the frequencies of these species are the same in the Saldanha fishery area as for the whole South African fishery, one might expect the


consumption by seabirds to reflect these proportions, since their diets often appear to reflect relative abundance of pelagic species (Jarvis, 1970; Crawford and Shelton, 1978). About 13 % of the South African pelagic stock is in the Saldanha fishery area (Furness and Cooper, in press). Virtual population analysis (VPA) indicates an average stock of 50700 tonnes of anchovy in the Saldanha fishery area between 1971 and 1976. An average consumption of 10 109 tonnes by the seabirds represents an annual cropping of 20% of the anchovy biomass by the seabirds. VPA for pilchards suggests a biomass of 29000 tonnes in the Saldanha fishery area. Seabird consumption of 1072 tonnes represents a predation of 4 % of this stock. VPA for round herring suggests a biomass of 8060 tonnes, of which seabirds consume 1789 tonnes, or 22% of this estimated stock. Predation on the stocks of horse mackerel, mackerel and lantern-fish is negligible. The first two of these are mesopelagic for most of their life, and so unavailable to the seabirds, while the last is only present in small numbers (Crawford, 1979).

The apparent impact of seabird predation on each fish stock appears to differ quite widely. Possibly pilchard are less available to the seabirds than are anchovy and round herring, but it is likely that the main cause of this apparent difference is the relative abundance of the fish stocks in the Saldanha area compared to other South African fishing areas. If pilchard are relatively scarce and anchovy and round herring relatively more abundant in the Saldanha fishery area, then the seabirds may not be selecting between species, but inflicting a predation of slightly less than 20 % on the stocks of all three of these species. However, there is good evidence that seabirds breed more successfully when able to select a diet with a high calorific value (Harris and Hislop, 1978), and tend to feed their young on a diet with a higher calorific value than taken by breeding adults or immatures (Furness and Hislop, 1981), so that selection is likely to take place when food availability allows.

The seabird community of the early 1970s in the Saldanha fishery area removed a large part of the pelagic fish biomass each year, so that it would be reasonable to expect changes in fish stocks to affect seabird numbers in a direct and detectable way. Crawford and Shelton (1978) examined the history of seabird populations in South West and South Africa by looking to see if guano yields could be used as an index of seabird population size. Early heavy exploitation of guano deposits meant that by 1845 practically all accumula- tions had been removed (Jarvis, 1970) so that annual yields since the end of last century approximate to the quantity deposited in the previous 12 months. They found a good correlation between guano production and known seabird population sizes on a number of islands for which accurate census data had been obtained in more than one year (Fig. 9), and inferred that the guano yield is largely determined by the number of breeding pairs of the major seabird species. Hence guano yield can be used as an index of the changes in

254 R. W. FURNESS

2000 - I000 -

numbers of breeding pairs, although the changes recorded may be due to absolute population changes or to changes in the proportion of the population which breeds in any one year. Crawford and Shelton (1978) went on to compare the seabird population changes measured from guano yields with the available estimates of fish stock abundance, obtained from catch, catch per unit effort or virtual population analyses.

iChabD0 Island P068~mon Illland

300

100--200

100

u 0

D a

g 400

2 300

200

I00

- Lombrts Boy laland Oassn Islond

I 2 0 I00 8 0 6 0 40 20

60 50 40 30 20 10

40 3

36

32 5 r

38 ?

34 a

2 s 140 120

80 v)

60 40

40

30

20

10

100 5

1956 1967 1970 1972 1936 38 41 59 39 40 56

Years

FIG. 9. Relationship between seabird population size A and guano production on certain South West and South African islands (from Crawford and Shelton, 1978).

Up to 240 000 Cape cormorants breed on Bird Rock platform in the Walvis Bay area of the South West African purse-seine fishery, and this species represents 98 % of the breeding seabirds of the bay (Berry, 1975). The annual guano harvest, the total pilchard catch at Walvis Bay and the pilchard catch per unit effort are shown in Fig. 10. Crawford and Shelton (1978) point out that the pilchard catch rose rapidly to a peak of 1.3 million tonnes in 1968


I O -

8-

6-

4-

2 -

0-

which was well in excess of the maximum sustainable yield, estimated at 0-8 million tonnes (Newman, 1970). The overfishing resulted in a decrease in pilchard abundance after 1966, demonstrated by the catch per unit effort data. From 1941 to 1966 the guano production varied between 500 and 1000 tonnes, with no consistent pattern or trend, but it dropped rapidly after 1966, to under 300 tonnes in 1970, indicating a reduction in breeding seabird numbers which occurred simultaneously with the decline in pilchard stock. Presumably the reduction of the pilchard stock to one third of its earlier level (as indicated by the threefold reduction in catch per unit effort data) resulted in either emigration of adult Cape cormorants, breeding failure, or both. After 1970 guano production and pilchard catch per unit effort data do not correlate well. Crawford and Shelton (1978) suggest that this resulted from the introduction of a quota system which directed fishing effort away from the pilchard, and also an increase in abundance of horse mackerel which may have helped the seabird population to recover even though pilchard biomass remained low.

A

- 1.2

- 1.0

-08

-0 6

-0 4

-0 2

. '00 A

-A

u- 0

c 0

0

D

.- 4-

: e O

(3

1 1.4

Year

FIG. 10. Relationship between guano production on Bird Rock platform (as an index of seabird population) (solid line), total pilchard catch off Walvis Bay (A) and catch per unit effort (0) (from Crawford and Shelton, 1978).

Adult pilchard provide the main food source for Cape gannets breeding on Ichaboe Island and form the bulk of the catches landed at Luderitz, South West Africa. Guano production on Ichaboe Island correlates well with estimates of the biomass of the adult pilchard stock (r = 0.62, n = 14, p < 0.025) and with the Liideritz pilchard catch (r = 0.92, n = 12, p < 0.005). The close relationship (Fig. 11) again demonstrates the dependence of seabird

6 I / -

OL, , , , , , , . , , , . L 0 I965 I970 I975

Year

FIG. 1 1. Relationship between guano production (0) on Ichaboe Island of South African pilchard aged three or older (m) (from Crawford 1978).

c

I C 0 c

Y- 0

C

0

-0

.- +

h 0

s (3

and biomass and Shelton,

Y-

Year

FIG. 12. Relationship between guano production (0) on Bird Island, Lambert’s Bay and biomass of 0-year old South African pilchard (m) (from Crawford and Shelton, 1978).


populations on their food resource. Juvenile pilchard recruit to the South African fishery in the vicinity of Lambert’s Bay. Biomass estimates of the 0-group and guano production at Lambert’s Bay are shown in Fig. 12. Again, the correlation is striking ( r = 0.78, n = 20, p < 0.005), and implies that the seabirds in this area are highly dependent on pilchards (unlike those in Saldanha Bay, discussed earlier, where predation is mainly directed at the anchovy stock). In fact the close correlation between breeding seabird populations and fish stocks can only be demonstrated in areas where a single stock provides the seabirds with their food. In areas such as Saldanha Bay where the fishery and the seabird diet is multispecies, guano production could not be linked to trends in any one fish species in isolation.

Crawford and Shelton (1978) also demonstrate that the response of predatory fish, which are also dependent on the pelagic stocks for their food, is closely similar to that of the seabirds. Populations of both are highly correlated with the abundance of the pelagic stocks. For example, the catch per unit effort of snoek for the area west of Cape Point between 1898 and 1905 shows a high correlation with the guano production by seabirds breeding at Lambert’s Bay (r = 0.80, n = 7, p < 0.025).

0.6 1

ul 0

8 L 0

- 10 - - 9 g

C

- 8 g .c 0

-0 a,

- 7 v)

-6

- 5 5

- 4 ’u

A ’ (1

3

C 0

3

- 3

1900 1910 1920 1930 1940 1950 1960 1970

Year

FIG. 13. Annual guano and penguin egg (0) harvests at Dassen Island, South Africa (from Siegfried and Crawford, 1978).

The combined guano yield from all islands off the South African coast shows that large fluctuations were characteristic before the pelagic fishery began in 1943. Between 1943 and 1961 the guano yield remained fairly

258 R. W. FURNESS

stable at about 1800 tonnes/year, possibly because the fluctuations of pilchard and horse mackerel stocks tend to be out of phase so conferring some stability (Centurier-Harris et al., 1977). Depletion of the pilchard and horse mackerel stock led to a reduction in guano yield after 1961 to under 1000 tonnes/year in the early 1970s. Centurier-Harris (1977) suggested that the reduction in pilchard biomass was not followed by an increase in the anchovy stock, so it would appear that the seabirds declined in numbers in response to the reduced fish availability. Crawford and Shelton (1978) discuss the use of the guano harvest data to explore the dynamics of the fish stocks during the period before commercial exploitation. Although the fish stocks of South West and South Africa are quite distinct, they found that the annual guano yield from South African islands correlated well with that from South West Africa ( I = 0.80, n = 70, p < 0.005) and infer from this that the two marine ecosystems are governed by a single factor, possibly some aspect of climate. They also speculate from the data that the guano production peaks at approximately 30-year intervals may reflect a regular cycle of fish abundance, similar to the 40-year fluctuation in abundance noted for the catch of Japanese sardines Sardiiiops melanosticta (Cuvier).

Dassen Island has clearly been intensely exploited for penguin eggs over a period of many years. In 1956 the breeding population was estimated to be 72 500 pairs (Rand, 1963) and the following autumn 98 640 eggs were har- vested, or 1.36 eggs per pair. The mean clutch size is only 1.8 eggs (Furness and Cooper, in press) although replacement clutches are likely to be laid when fresh eggs are removed. Siegfried and Crawford (1978) demonstrated a good fit between egg and guano yields from this island (Fig. 13), at least between 1920 and 1961. A linear regression gave a significant positive correlation ( r = 0.44, n = 39, p < 0.01) confirming that the falling egg harvest is related to a decline in the penguin population. They also detected a 15-year cycle in both guano yield and egg harvest. As this is unlikely to be related to demand or selling price it seems probable that an environmental factor is responsible. However, Siegfried and Crawford (1978) offer an alternative explanation, noting that troughs in guano production may tend to occur about five years after peaks in egg crop. They suggest that heavy exploitation of jackass penguin eggs could result in poor recruitment of penguins to the breeding population a few years later and a consequent population decline. Conversely, reduced exploitation would allow population recovery. While this explanation may be plausible, it seems improbable as the laws of supply and demand would tend to result in the cropping of eggs being more intense when penguin numbers were low. Siegfried and Craw- ford’s feed-back explanation for the cycle would require the reverse of this. Further, the 30-year cycle suggested by Crawford and Shelton (1978) may in fact be a 15-year cycle as they recorded peaks in guano production in


1927 and 1957 and these approximate to alternate peaks shown by the Dassen Island penguins (Fig. 13). The coincidence of the peaks in the cycles found in numbers of eggs cropped at Dassen Island, the guano yield on Dassen Island, the guano yields from all South African islands and from South West African islands suggests that some 15- or 30-year environmental cycle drives the marine ecosystems of the region. Then seabird numbers depend both on the stage in the environmental cycle and on the state of the fish stocks as determined by the pressures from the purse-seine fisheries.

The studies of South West and South African seabird populations in relation to fisheries show that changes in the composition of the fish stocks induced by overfishing have affected the seabird communities, by reducing their numbers, and by altering the competitive balance between species, particularly against the jackass penguin. The Cape gannet has fared relatively well because it has the ability to forage over a greater range, and has also learnt to exploit a new food source by scavenging from deep sea trawlers. The close correlation between populations of unexploited seabirds and the exploited jackass penguins on Dassen Island in relation to long term cycles in the marine ecosystem, which alter the availability of food, suggests that the numbers of seabirds are controlled primarily by food abundance, and that the level of egg collecting at Dassen Island was not having any serious influence on the mean population size of the penguin colony. The close relationship between seabird numbers, guano production and fish stocks allows seabirds to be used as an index of the state of the fish stocks. For this reason seabird populations are being monitored by the South African Department of Sea Fisheries. It also means that the seabird community is vulnerable to perturbations in food availability generated by fishery practices. In this respect concern has been expressed for the future of the jackass penguin population, which must now be regarded as a threatened species, largely as a result of its inability to adapt to the changes in fish stock distribution and predictability caused by the fishing industry (Frost et al., 1976).

D. Peru current

Coastal Peru is one of the main regions of upwelling in the world (Menzel et al., 1971). The coast is bathed by the cool nutrient-rich waters of the Humboldt current, also known as the Antarctic (sic), or Peru current. This is analagous to the California current in the northern hemisphere. Secondary upwelling of a more local nature is generated by wind, often about 100 km from the shore. The high primary productivity tends to comprise large, colonial phytoplankton, which are eaten directly by the Peruvian anchovy Engraulis ringens (Jenyns) (Ryther, 1969). The Peruvian anchovy is the basis of a commercial fishery which in the 1960s became the largest single-species

260 R. W. FURNESS

fishery in the world, producing 10 million tonnes during the fishing year 1968-69. It is also the staple diet of an enormous seabird community, consisting of millions of Peruvian cormorants Phalacrocorax bougainvillii (Lesson), also called guanays, and Peruvian boobies Sula variegata (Von Tschudi), also called piqueros, together with smaller but considerable numbers of brown pelicans Pelecanus occidentalis L., penguins, burrowing petrels, gulls and terns.

The upwelling intensity varies with the season, tending to be most intense in winter. From time to time the upwelling weakens and warm oceanic nutrient-poor water and a warm coastal counter-current are allowed to displace the denser, colder water. This phenomenon tends to occur annually at the middle of the austral summer, and so has been associated with the celebration of Christmas through its name El Niiio (“The Child”). In most years this perturbation is small, but occasionally it is very pronounced and has catastrophic effects on the ecosystem. The rise in water temperature results in the anchovy shoals dispersing and becoming unavailable to both man and seabirds. Vogt (1942) suggests that the fish move southwards to seek cooler water, while Schweigger (1940) and Fiedler et al. (1943) consider that they simply move into deeper, cooler water. Jordan and Fuentes (1966) also consider that the anchovies remain in the same area but in deep water in fragmented groups which move up to the surface only at night.

The disappearance of the anchovies causes catastrophic mortality (Table VII) among the guano birds which depend so heavily on them. In fact recurrent disasters on this scale are unknown anywhere else in the world. The catastrophes have occurred on a semi-regular basis for thousands of years and must have exerted tremendous selection pressure on the seabird populations, favouring the ability rapidly to increase in numbers after each crash. For this reason the guanay and piquero have large clutches in comparison to related species, may attempt to breed more than once within one year, and reach sexual maturity at an unusually early age (Nelson, 1978). These characteristics must have been particularly strongly selected because food becomes superabundant in the period following each crash. In other words the seabirds, even young inexperienced adults, are able to raise extra large broods in times of population recovery because the food supply per bird is much greater than for a population which has reached an equilibrium with the environment. In these more stable ecosystems such as the coasts of South West and South Africa, we have seen that seabird numbers follow changes in fish stock abundance. In this respect, the Peruvian seabird communities are most unusual in being strongly r-selected. In general seabird populations show most of the characteristics of K-selected species (Mac- Arthur and Wilson, 1976). This also implies that their utilization of the anchovy stock will vary from their taking a very small proportion in years

TABLE VII. MAGNITUDE OF "CRASHES" OF PERUVIAN GUANO BIRDS AND RECOVERIES BETWEEN 1917 AND 1976 (FROM NELSON,

(NELSON, 1978) 1978 AND VALDIVIA, 1978). FIGURES FOR 1917 TO 1954 ARE BASED ON GUANO YIELDS AND ARE PROBABLY UNDERESTIMATES

Population in Population in Lowest population Highest population Year of year before year after reached after reached before next Year of crash crash (millions) crash (millions) crash (millions) crash (millions) maximum

1917 3.9 1925 7.7 193941 9.4 1957-58 22.0 1965 14-8 1972 6.0

4.3 3.3 5.6 5.6 3.8 3.8

11.1 10.1 4-0 4.0 2.2 1.8

7.7 10.0 27.7 18.1 6.0 3.0

(1 924) (1937) (1 955) (19631 (1 972)

(1 976)

262 R. W. FURNESS

immediately after a crash, to taking a much larger proportion when their populations have built up to a peak just before the next crash occurs. Where the “surplus” anchovies go during years of small seabird population is unclear. It is also unclear whether the seabirds ever used to reach a limit imposed by food availability before the next catastrophe arrived, or whether their populations would continue to increase under conditions of superabundant food beyond the maxima actually recorded. In other words the seabirds may or may not be able to respond in a flexible manner to a reduction in food abundance generated by commercial fishing during a period in which their population was attempting to recover from a previous catastrophic El Nifio. A catastrophic El Nifio tends to occur every seven years or so, although not with a regular periodicity, and it may be that this period is too short, even for seabirds adapted to rapid increase, to allow their numbers to reach food-limited equilibrium.

Details of crashes in guano bird populations documented since 1618 are reviewed by Nelson (1978). Here I shall just outline the features of interest. The time elapsing between the onset of bad conditions around Christmas and mass seabird mortality can be quite variable. In some years millions of guano birds die or emigrate within a few days. In some years, as in 1938-39, the birds may show no ill effects for two or three months. Either way, the seabirds obviously find it impossible to cope with the changed behaviour of the anchovies. They abandon their breeding activities even if they have well- grown young in the nest, and die or emigrate in millions, the survivors returning after a variable period of absence. The migration tends to be southwards, towards areas of cooler water.

According to Nelson (1978) the piquero, guanay and pelican often share breeding islands, but their colonies do not intermingle. Each species forms its own clearly demarcated congregation. As a result interspecific competition for next sites is minimal, although most areas traditionally occupied by one species appear suitable for either of the other two were they able to move into them. Nelson (1978) feels that interspecific competition for nest sites has not been of any consequence over the last 100-year history of the Peruvian seabird communities since all colonies have vacant areas which appear to be suitable for breeding and could potentially be colonized by any of the three seabird species. Then neither total numbers nor species composition of the Peruvian guano seabird communities appears to be restricted in any way by nest site availability or quality. However, the species do differ in various aspects of their biology. The guanay cannot dive as deeply as the piquero, but neither the piquero nor the pelican is as tolerant of human disturbance as the guanay. The result is that the three species will differ in competitive ability according to the environmental (in the widest sense) conditions. Hutchinson (1950) and Nelson (1978) review the changes in distribution


- 20 - 16-

1 2 -

8 -

4 -

and relative numbers of guanays, pelicans and piqueros. They provide convincing evidence that the community was dominated by piqueros and pelicans in the period immediately before human exploitation of the guano crop began in the middle of last century. Since then, apparently as a response to human disturbance, the guanay has come to replace the piquero as the numerically dominant species. Interestingly, the competitive balance has now been tipped back in favour of the piquero by the development of the commercial fishery for the anchovy. The piquero appears to be relatively less seriously affected by the reduced food abundance and the direct mortality caused by birds tangling and drowning in nets. Thus the species composition of the Peruvian seabird communities has been controlled by the indirect effects of man for over 130 years.

51 59 1911 1917 1923 1925 1932 1939 1941 1949 50153 1958/ 1962

u)

B n ._ Ic 0

O J , I I I I I I I I I I

1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960

Year

FIG. 14. Fluctuations in numbers of guano birds off Peru, determined from guano yields assuming 15.9 kg deposited/bird/year. Years of population crashes shown by arrows (from Nelson, 1978).

If we examine the changes in numbers of all Peruvian seabirds over this century we can see that the introduction of the huge anchovy fishery had a devastating effect on the total guano bird numbers, and their importance in the overall ecosystem. AS in South West and South Africa, seabird numbers can be assessed from the annual guano yield. On average, one bird deposits a harvestable 15-9 kg of guano/year (Jordan and Fuentes, 1966; Jordan, 1967). Between 1909 and 1962 numbers of guano birds estimated from guano yield fluctuated widely, recovering rapidly from each crash induced by exceptional warm water incursions (Fig. 14). The generally increasing trend in the early part of the century can be explained as a recovery of the birds from exces- sively intensive guano cropping disturbance and direct persecution of adult birds in an early period of uncontrolled exploitation, while from 1909 the birds received the protection of the Guano Administration. Counts of guano birds between 1955 and 1968 were compared with annual anchovy catch

264 R. W. FURNESS

data (Fig. 15) by Jordan and Fuentes (1966), Schaefer (1970) and Santander (1980). This figure indicates that the guano birds did not recover quite as rapidly as normal after the 1957-58 crash, and this resulted in the numbers falling to an all time low since 1915 in 1965-66, and falling again in 1972.

* 0

38 - 36-

34- 32- 30-

- 5 = E 1

195556 57 58 59 60 61 62 63 64 6566

Year

FIG. 15. Relationship between Peruvian guano bird numbers (0) and the anchoveta catch (a (from Nelson, 1978).

During the small 1963 crash it was found that pelicans suffered the highest mortality rate and piqueros the lowest (Table VIII) (Jordan, 1964). In 1965, as in 1963, the guanay suffered a higher mortality rate than the piquero, and mortality occurred in two waves, the first largely affecting young birds and the second the less susceptible adults. The total population was reduced from 17 million to 3-4 million birds; a mortality of 7682 %. After the crash, numbers hardly recovered, remaining well below 5 million individuals in 196667 and 1967-68 (Schaefer, 1970). A further El Nifio occurred in 1972, reducing the population even further (2.5 million birds in 1974 (Tovar, 1974)), and again largely affecting the guanay rather than the piquero. Apparently the surviving seabird populations again failed to recover after this crash. Thus, since the establishment of the anchovy fishery, the dynamics of the Peruvian guano seabird populations have changed. Instead of rapidly increasing by raising large broods at least once each year, they failed to respond to the reduced competition brought about by their reduction in


numbers. The reason for this seems to be that the anchovy fishery has taken up the superabundance of food on which the Peruvian guano seabirds depended in order for them to cope with the recurring crashes induced by ocean perturbations.

TABLE VIII. DIFFERENCES BETWEEN THE THREE MAIN GUANO SEABIRDS OF PERU IN THE EXTENT OF THEIR MORTALITY IN THE 1963 “CRASH” (FROM JORDAN, 1964 AS

QUOTED BY NELSON, 1978)

Species Total number Percentage of

of corpses deaths falling found to each species

Peruvian cormorant (guanay) 6566 73.4 Brown pelican 1973 22.1 Peruvian booby (piquero) 375 4.5

Percentage which each species

contributed to total population

before the “crash”

82.4 2.3 15.3

Schaefer (1970) analysed the apparent abundance of the Peruvian anchovy using catch-per-unit-effort data from the commercial fishery from 1960 to 1968, combined with an estimate of the harvest taken by the guano birds. Jordan and Fuentes (1966) estimated from field data that each guano seabird consumed on average 430 g of anchovylday. Bioenergetics considerations would indicate that a Peruvian cormorant (ca. 2000 g) at about 20°C ambient temperature would require an intake of about 200 g of anchovy for existence metabolism, and no more than twice this amount to cover the additional costs of foraging activity, moult and egg production. However, the costs of chick daily energy budgets has to be taken into account as well, and these birds have large broods of rapidly growing young, so a total of 430g of anchovy/guanay/day for all costs is probably not unreasonable. The value is probably too high for an adult piquero (1300-15OOg) but as this species forms only 15-30 % of the guano seabird total (Nelson, 1978) it is an acceptable figure to use, and not likely to overestimate by much the total food consumption of the seabird community from 1961-68. From this value Schaefer (1970) estimated that in 1961-65 the 17 million guano seabirds consumed 2.6 million tonnes of anchovyfyear, while after the 1965 crash the population of 4.5 million seabirds consumed 0.7 million tonnes/year from 1965 to 1968. These figures can be compared with annual fisheries data to give the combined catch and effort for the fishery by men and birds (Table IX). Considering the commercial fishery alone, close examination of the relationship between effort by anchovy fishermen and stock abundance

TABLE IX. DATA CONCERNING CATCH AND EFFORT BY THE COMBINED FISHERY BY MEN AND SEABIRDS ON THE PERUVIAN ANCHOVETA (FROM SCHAEFER, 1970)

Fishing Year

1960/61 1961162 1962163 1963164 1964165 1965166 1966167 1967168

Catch by Catch per Fishermen's Fishermen effort effort

(10' tonnes) (tonnes/trip) (1000 trips)

Adult bird population (1 O8 birds)

Catch by birds

(10' tonnes)

Combined catch

(loo tonnes)

3-93 5-50 6.91 8.01 8.04 8.10 8.24 9.82

0-55 0.60 0.48 0.38 0-38 0.36 0.44 0.47

7134 9129

14 447 21 285 21 374 22 741 18 948 20 800

12 0 17.0 18.0 15-0 17.3 4.3 4.8 4.5

1.88 2.67 2-83 2.36 2-72 068 0.75 0.71

5-81 8-17 9.74

10-37 10.76 8.77 8-99

1053

Corn bined effort

( 1 OOO trips)

10 544 13 549 20 377 27 580 28 617 24 635 20 667 22 309

COMPETITION BETWEEN FISHERIES A N D SEABIRDS 267

(catch-per-unit-effort) shows that the relationship changed between 1965 and 1966. The maximum sustainable yield during the period of high seabird numbers can be estimated using a least squares regression by assuming that the fishery was in a steady state. This gives an estimated maximum sustainable yield of 8-05 million tonnes/year (Schaefer, 1970). In fact it was an expanding fishery in the 1960s so this provides too high an estimate of the maximum sustainable yield. A better method is to plot the catch-per-unit-effort against the average effort experienced during the life-span of a cohort in the fishery (Gulland, 1961). In the case of the Peruvian anchovy this is two years. Gulland’s method estimates the maximum sustainable yield between 1960 and 1966 (Fig. 16) at 7.5 million tonneslyear (Schaefer, 1970).

Rshing effort (thousand boot-ton mnth)

FIG. 16. Relationship between commercial fishing effort and anchoveta abundance (CPUE) for the seasons 1960-66 (from Schaefer, 1967).

Repeating the analysis of catch-per-unit-effort against effort and including data for the three seasons following the seabird crash (Fig. 17) shows that the estimated average annual maximum sustainable yield has increased to 8.5 million tonnes. The apparent abundance of the anchovy stock in the years after the seabird population crash also exceeds expectation. The 1966-67 and 1967-68 points fall well above the regression line (Fig. 17) suggesting that the anchovy stock available to the commercial fishery increased between 1965 and 1966. Schaefer (1970) points out that recruitment to the exploitable stock also increased at this time, and suggests that this resulted not from any environmental change in the ocean, but directly as a consequence of the lower

268 R. W. FURNESS

predation by the guano seabirds. Schaefer (1970) was able to demonstrate this convincingly by plotting the relationship between effort by commercial fishermen and birds combined against anchovy abundance from 1960-61 to 1967-68. From this (Fig. 18) the combined average annual maximum sustainable yield can be estimated at 9.9 million tonnes (Gulland’s method). Schaefer (1970) interpreted this important result as follows. First, the points in Fig. 18 show less variation about the line of best fit than do the corresponding points in Fig. 17. The standard-error-of-estimate for the data in Fig. 17 is 0.050 while it is only 0.029 for Fig. 18, so including the influence of seabird predation improves the precision of the estimate of maximum sustainable yield. Unexplained fluctuations in Fig. 17 may therefore be attributed in large part to variations in seabird predation before and after their population crash. Secondly, the exact location of each point in relation to the estimated line of equilibrium corresponds better in Fig. 18 to what is predicted by fishing theory. When effort is increasing the points should fall above the line, while they should fall below the line when fishing effort is decreasing (Schaefer, 1954).

r

al 0 3-

= 0 .2 -

a 0 I -

c ._ c

L al

S

g 0 0 , I . . , , I , I , , 0 0 5 10 15 20 25 30 35 40 45 50 55

Effort (millions of trips)

FIG. 17. Relationship between commercial fishing effort and anchoveta abundance for the seasons 1960 to 1968 (from Schaefer, 1970).

Schaefer (1970) concluded from his analysis that the changes in abundance of the Peruvian anchovy stock could be largely explained by the effects of the combined fishery of men and seabirds. He found that a commercial fishery of 7.5 million tonneslyear corresponded to a fishing mortality of l.O/year (Schaefer, 1967) and this equalled the natural mortality. This means that the consumption of 2.5 million tonnes by the 16 million guano seabirds between 1961 and 1965 was equivalent to a mortality caused by sea birds with a coefficient of 0-331year and to mortality caused by other predators with a coefficient of 0*67/year. In other words the seabirds were consuming 17 % of the anchovy stock in these years, a value nearly as high as the pre-


dation intensities (20-27 %) estimated for the seabird communities in Saldanha Bay (South Africa), Foula (Shetland) and coastal Oregon. The similarity suggests that the Peruvian guano seabirds may possibly have reached, or been only a little below, a food-limited population ceiling in the early 1960s. Certainly their consumption was considerable and likely to have been sufficient to result in depletion of food resources with consequent increased competition. In fact the increasing commercial fishery appears to have led to a reduction in the breeding success of the guano seabirds over the years 1962 to I965 (Table X) (Nelson, 1978). This will be discussed further in a later section (p. 296).

0.8,

\

0 5 10 15 20 25 30 35 40 45 50 55 60

Effort (millions of trips)

FIG. 18. Relationship between combined effort by fishermen and seabirds and anchoveta abundance 1960 to 1968 (from SchaefeI, 1970).

The reduction of seabird populations to 4.5 million individuals in the years 1966 to 1968 (Nelson, 1978) reduced their consumption of anchovies to 0-7 million tonneslyear, a mortality coefficient of 0*09/year, allowing the commercial fishery to increase its harvest to 9.3 million tonnes/year, giving a fishing mortality coefficient of 1*24/year. It is clear from Schaefer's analysis that a deliberate reduction of seabird numbers would maximize the anchovy yield to man, and similarly, a recovery of the bird populations would neces- sitate a corresponding decrease in the catch by the fishermen.

E. The Southern Ocean

So far the ecosystems I have examined have been very simple ones. In most the primary production is high, the fish stocks are dominated by one or two small shoaling pelagic species, and these provide all but an insignificant amount of the food of the local seabird community. Further, each seabird community I discussed is dominated by a small number of large, diurnal

TABLE X. REPRODUCTIVE SUCCESS OF PERWIAN GUANO SEABIRDS IN YEARS IMMEDIATELY BEFORE AND AFIER A MINOR “CRASH”

(IN 1963) AND DURING THE BUILD-UP OF THE ANCHOVETA FISHERY (JORDAN AND FUENTES, 1966; QUOTED IN NELSON, 1978)

Number of Number of Reproductive Anchovy catch Before or adults chicks success in that year

Year after “crash” (millions) (millions) ( O 4 (lo6 tonnes)

1961/62 Before 1962163 Before

17.0 18.1

11.6 12-8

68 70

5.50 6.91

1963164 After 15.0 6.0 40 8-0 1 1964165 After 17.3 8.6 49 8-04

COMPETITION BETWEEN FISHERIES AND SEABIRDS 27 I

species for which we have a detailed knowledge of breeding biology and ecology. In each ecosystem an important fishery exists in potential or actual competition with the seabirds for the same food resource, and effects of changes in the ecosystem have been clearly displayed. I hope these patterns have been sufficiently convincing to allow me to apply the same principles by analogy to a more complicated marine ecosystem about which we know rather less. The Southern Ocean ecosystem has shown changes which we must explain in a more speculative way, assuming that it follows the same responses of seabirds to competitors on the same trophic level and to perturbations in food supplies.

Krill, Euphuusia superba Dana, is the food source for many species and dominates the second trophic level in the Southern Ocean. It supports a complex marine ecosystem consisting of fish, small cephalopods, baleen whales, seals and seabirds as direct consumers of krill. These in turn support sperm whales Physeter catodon L., seals, larger cephalopods and seabirds in the fourth trophic level. Energy flow through the ecosystem is predomin- antly determined by the stocks of baleen whales, crabeater seals Lobodon carcinophugus Jacquinot and Pucherhan, squid and penguins (Fig. 19). In recent years there has been an upsurge of interest in the marine ecosystem of the Southern Ocean due largely to the realization that krill is a highly pro- ductive resource providing the basis for the development of a very large commercial fishery. Krill production probably exceeds the current total world fish catch by a factor of three or more (Everson and Ward, 1980) so there is a strong stimulus to develop catching techniques, and considerable progress is being made in this (Nemoto and Nasu, 1975; Everson, 1978). A further stimulus to initiate krill fishing comes from the suggestion that depletion of baleen whale stocks, caused by overfishing throughout the twentieth century, will have resulted in a krill surplus being available for harvesting. Estimates vary, but it is generally considered that some 150 million tonnes of krill would have been consumed annually by the whale stocks in earlier years (Laws, 1977). In fact a large surplus probably does not exist at the present time. It appears that the seal and seabird populations have responded to the decline in whale biomass by increasing in numbers to take up much of the krill surplus. Although no data are available, it seems likely that fish and cephalopod populations have responded similarly. Both ecological theory (May et al., 1979) and biological common sense indicate that extensive harvesting of krill would be at the expense of other elements in the ecosystem.

Laws (1960, 1977) notes that whales show segregation in Antarctic waters. Within species the migration of different classes is staggered in relation to size and feeding requirements. Larger individuals tend to reach higher latitudes and pregnant females arrive before lactating ones. In addition, the

Killer whale

Sperm whale

I__

Baleen whales

I Krill

FIG. 19. Generalized food web for Antarctic marine ecosystems showing major routes of energy flow.


larger and older whales arrive first and tend to occupy central, presumably optimal, parts of the feeding grounds, while later arrivals are displaced to peripheral areas. Laws (1977) interprets this as indicative of competition for food and he suggests that this may have been limiting to whale stocks. Since stock reduction there is evidence that food resources have improved for individual whales. Pregnancy rates of blue whales Balaenoptera musculus L. and fin whales B . physalus (L). have increased (Mackintosh, 1942; Laws, 1961). The pregnancy rate of sei whales B. borealis Lesson also increased, but this preceded the large-scale exploitation of this species, indicating that it was not a response to changes in the social structure or density-dependent behaviour of the population as a result of exploitation, but supporting the argument that it resulted directly from the effect of whaling on the food supply for baleen whales in general (Gambell, 1973). Laws (1 962) found that the mean age sexual maturity in female fin whales had decreased between 1945 and 1956 and this correlated with an increased rate of growth in body size which he related to increased food availability. Lockyer (1972) showed that age of sexual maturity had decreased from 10 years in both sexes between 1910 and 1930 to 5 or 6 years in the 1960s, and that the body size at puberty had remained constant, confirming that the change was due to a more rapid body growth rate.

The reduced whale stocks also appear to have improved krill availability for seals and seabirds. Crabeater seals age at sexual maturity decreased from about four years between 1945 to 1955 to three years by 1965 and 2.5 years by 1970 (Laws, 1977) and this suggests that the population is increasing in response to increased krill availability. Most southern fur seal populations are increasing, partly in response to a cemztion of commercial sealing early this century (Laws, 1973). The most rapidly increasing species is Arcto- cephalus gazella Peters, while the species north of the Antarctic convergence have only increased slowly. Unlike the other species, Agaze l la feeds on krill, and the greatest population increases have occurred in the Scotia Arc region where the distribution overlaps the baleen whale feeding distribution (Laws, 1977). Many Antarctic seabird populations are also increasing, again as a response to increased krill availability. King penguins Aptenodytes patagonica J. F. Miller were heavily exploited in the nineteenth century so their present increases may be partly a recovery from this. However, most Antarctic seabirds were not exploited, so their increases are not due to a history of early persecution and subsequent protection. Further, the species showing the greatest rates of increase are those most dependent on krill, while squid feeders are increasing rather slowly (Conroy, 1975; Croxall and Kirkwood, 1979; Croxall and Prince, 1979).

The changes in the demographic parameters in seal and whale populations and the increases shown by populations of seabirds and seals, and particu-

274 R. W. FURNESS

larly the greater response of the krill feeders, suggest that the populations of whales, seals and seabirds were all held at equilibrium sizes by competition for krill during the period before exploitation of the whale stocks. Reduction of whale stocks has improved krill availability and so upset the competitive balance between species. May et al. (1979) described the response of a model theoretical ecosystem, containing interacting seals, baleen whales and krill, to three different harvesting regimes and using the present situation as a starting point for their simulation of the dynamic response of the three populations. Figure 20a shows the effect of stopping all whaling. Figure 20b shows the effect of maintaining whaling and also initiating a krill harvest. Figure 20c shows the effect of harvesting krill but stopping all whaling. Seabirds were not considered in their model, but the response of seabirds would be qualitatively the same as, and slightly more rapid than that of the seals. All three scenarios would result in a reduction of seabird and seal populations. The relative reduction would depend on the relative competitive abilities of each species, but would be greatest where krill exploitation and whale protection were instigated together. At present this seems a likely course of action.

I Time

’.. Seals

Whole5 -___ ___. _ _ _ _ _ L

Time

i Seals

Krill _ _ - - - Wholes __- -

I , , . . . . Time

FIG. 20. Models of the behaviour of populations of krill, baleen whales and seals under three different hypothetical harvesting regimes : a; after cessation of whaling; b, harvesting of whales and krill; c, harvesting of krill but no exploitation of whales (from May et al., 1979).

Much consideration has recently been given to quantifying the direct and indirect consumption of krill by whales, seals and seabirds, in order to determine whether all the krill “surplus” has been taken up by seals and seabirds, and to find out the relative importance of the three groups of consumers in terms of krill consumption. Crude estimates of biomass show that for the Southern Ocean as a whole the present biomass of whales still exceeds that of seals which exceeds that of seabirds. However the energy requirements of seabirds greatly exceed those for the same biomass of seals or whales, so consideration of biomass is misleading.


Producing a detailed bioenergetics model for Antarctic or Subantarctic seabird communities is not yet possible because the population sizes are not known and knowledge of the diets and activity budgets of many, particularly the nocturnal species, is very limited. Even the species which are relatively amenable to census can produce very different population estimates. For example, Williams et al. (1979) estimated the world population of the macaroni penguin Eudyptes chrysolophus (Brandt) to be about 4 million breeding pairs, but Croxall and Prince (1979) gave a figure of 5.4 million pairs on South Georgia alone, and stated that their census may actually be an underestimate of the population by as much as 50%. Present sizes and age compositions of seal populations are slightly better known, and fairly reliable data are available for whales. Assessing the sizes of the populations as they were in 1900 is rather more difficult. It can be done from life table data for exploited whales and fur seals and, by extrapolation for crabeater seals, assuming that the demographic parameters have remained constant apart from the known changes in age of puberty. Rates of increase of seabird populations are not known in most cases, and only recent data are reliable for those species which have been carefully studied. For these reasons it is difficult to calculate an accurate figure for the krill “surplus”. Mackintosh (1973) estimated that baleen whale biomass had decreased by 36.5 million tonnes since 1900, from 43 million tonnes to 6.5 million tonnes. Lockyer (1972) estimated that whales consume 3.5 % of their body weight per day over a feeding season of about 120 days. This indicates a consumption of 180 million tonnes of krill in 1900 and 28 million tonnes now, suggesting a krill surplus resulting from whale stock depletion of about 152 million tomes/ year. Some of this is obviously now taken up by seals, cephalopods, fish and seabirds. Laws (1977) suggested that baleen whales probably remain the most important vertebrate group in the Antarctic ecosystem, and that seabird population energy requirements are unlikely to be important in comparison, although he based this opinion on biomass considerations alone.

Everson (1977) attempted to assess krill consumption by southern ocean seabird stocks by multiplying population size estimates by estimated daily food intake data from field studies. His calculation suggested that the total stocks of seabirds (487000 tonnes, mainly penguins) in the Antarctic eat 15-20 million tonnes of krill, 6-8 million tonnes of squid and 6-8 million tonnes of fish/year, or approximatly 1.4 x 1014 kJ/year. Using quite independent data, Prevost (unpublished) and Mougin and Prevost (1980) combined estimates of the Southern Ocean’s seabird populations, obtained from breeding totals multiplied by species-specific constants to allow for pre- breeding age classes, with the bioenergetics equation for existence metabolism a t 0°C (Kendeigh, 1970), multiplied by two to allow for activity and the energy requirements of chicks. He obtained a total food consumption of

276 R. W. FURNESS

38.7 million tonneslyear, or approximately 1.5 x lOI4 kJ/year. Grenfell and Lawton (1979) also used a metabolic approach to the problem, but again used completely independent data and bioenergetics equations. They obtained current population estimates and body weights for each species in the Southern Ocean (Stonehouse, 1967 ; British Antarctic Survey, unpublished data) and applied these to a bioenergetics equation for resting metabolic rate:

log M = log 74.3 + 0.744 log W k 0-074 (King and Farner, 1961)

where A4 is the metabolic rate in kcals/day and W is body weight in kg. In addition they allowed for a production to assimilation ratio (P/A) of 1 a29 % (S.E. 0.03, n = 9; Humphreys, 1979) and assimilation efficiency of 80% (Lawton, 1970; Furness, 1978b). Their results indicate an annual resting metabolic requirement of 4.8 x 1013 kJ. Assuming that the field metabolic rate is approximately three times the resting metabolic rate (King, 1974) this gives an annual total energy requirement by the Southern Ocean seabirds of 1.4 x 1014 kJ, in close agreement with the other two, quite independent, estimates.

Grenfell and Lawton (1979) extend this analysis by computing the food consumption in terms of krill equivalents. They assume that fish and cephalopods eaten by whales, seals and seabirds feed on krill. Using standard population energetics relationships, the indirect consumption of krill by secondary consumers (C,) is given by:

C, = PP/O.8(P/Af

where P, is production of the primary consumer and:

for squid P/A = 25% (S.E. = 3.7, n = 73 (Humphreys, 1979)) for fish

Then the annual consumption (direct and indirect) of krill by seabirds amounts to 100 to 140 million tonnes. This nearly equals Laws (1977) estimate of krill consumption by baleen whales before exploitation (1 80 million tonnesl year) and greatly exceeds their current consumption (28 million tonnes/year), against his expectation.

Grenfell and Lawton (1979) used population estimates and demographic parameters for each whale and seal species, combined with energetics equations, to estimate direct and indirect consumption of krill by these groups. They used the relationship between body weight and resting metabolic rate provided by Lockyer (1976) derived for whales and seals:

RMR (Kcals/day) = 126.2 W0'77 (S.E. 0.035), where W is weight in kg.

They assumed that field metabolic rate equals 1.5 to 3, and probably two times resting metabolic rate (RMR). Again they allow for production, assum-

PIA = 9.8% (S.E. = 0.89, n = 22 (Humphreys, 1979))

TABLE XI. ESTIMATED KRILL CONSUMPTION (DIRECT AND INDIRECT) BY WHALE, SEAL AND SEABIRD POPULATIONS IN THE SOUTHERN OCEAN IN 1900 AND THE PRESENT ASSUMING FMR/RMR = 2 AND P/A = 30% FOR BALEEN WHALES, AND THAT SPERM WHALES

ARE SECONDARY PREDATORS ON KRILL (SEE TEXT) (FROM GRENFELL AND LAWTON, 1979)

Predator Present consumption 1900 consumption

of krill of krill Difference ( lo6 tonnes) (1 0 t onnes) (lo8 tonnes)

Baleen whales Sperm whales Seals Seabirds

Total (excluding squid and fish)

44 16

185 120

365

304 33

min. SO? min. 30?

- 260 - 17

+100? +90?

min. 447 at least -8O?

278 R. W. FURNESS

ing the general value of 3.14% for P / A for larger mammals as an annual average (Humphreys, 1979), but allowing for a greater production efficiency during feeding in the Antarctic when depositing fat (as done particularly by baleen whales), again assuming that typical values for homeotherms apply (Calow, 1977; Morowitz, 1968). Overall they suggest a P / A of 30% for baleen whales as the best estimate for this parameter. As with their calculation for seabirds, assimilation efficiency is taken to be 80% and conversions are made to allow for indirect krill consumption in the form of squid or fish.

The total krill consumption (direct and indirect) of baleen whales, sperm whales, seals and seabirds is as given in Table XI. Krill consumption by seals and seabirds in 1900 can only be guessed at. If, as seems likely, seal stocks have more than doubled since 1900 then the total consumption in 1900 was probably less than 100 million tonnes, and 80 million tonnes is perhaps a reasonable guess. Croxall and Prince (1979) show that some populations of seabirds are increasing quite rapidly. Adelie penguins Pygoscelis udeliae (Hombron and Jacquinot) at 2 to 3% per year on parts of Signy Island, chinstrap penguins P . anturcticu (Forster) at an average of 8 % per year at 11 colonies on Signy Island over periods within the years 1947 to 1979, and macaroni penguins at up to 9 % per year at South Georgia. If even much lower rates of increase apply to other colonies, other species, and have per- sisted for some time, then the Antarctic seabird numbers could have been doubling every few decades this century. In that case their consumption of krill in 1900 may have been very much less than, possibly no more than a quarter of the present level. That seabirds seem to be showing a higher rate of increase than seals would be predicted by the modelling approach employed by May et al. (1979), so the suggested trends are compatible with theoretical expectation.

The consumption of krill by whales, seals and seabirds in 1900 and at present still suggests that seals and seabirds have not taken up all the surplus provided by whale stock reduction. This might be expected, as cephalopod and fish stocks are also likely to have increased in response to greater krill availability. The calculations do show that present consumption by seabirds is considerable, and that they are ecologically more important in terms of energy flow than the depleted stocks of whales.

Grenfell and Lawton’s estimates of krill consumption by whales are rather higher than Law’s. The calculations were based on similar stock data, but Grenfell and Lawton took into account indirect predation through the food chain as a result of squid and fish consumption. They also allowed for production of body mass during the 120 days of feeding, while Law’s method simply multiplied numbers by a percentage of body weight to allow for average daily food intake. Grenfell and Lawton’s results probably are more reliable, but the importance of their calculation is that they used similar


approaches for whales, seals and seabirds, so the relative importance of each group can be assessed with confidence.

Although a complicated ecosystem, the Southern Ocean shows the same pattern as the earlier case studies. There is good evidence to suggest that in 1900 when stocks were stable they were limited by food, but exploitation of whales has improved food availability allowing growth of seal and seabird populations. Harvesting of krill would clearly result in reductions in populations in response to reduced krill availability. Modelling would predict that shorter-lived seabirds would show a rapid response, seals and longer- lived seabirds a slightly slower one, and whales the slowest of all. Nevertheless it is difficult to guess how the competitive balance between whales, seals and seabirds is likely to be affected; as May el al. (1979) put it:

“under heavy exploitation of krill the ecosystem would lurch toward some new equilibrium, in which baleen whale populations may well lie below their current levels” and also “multispecies ecosystems will often manifest complex “catastrophic” behaviour. This transformation will not usually be continuously reversible . . . such changes are seldom, if ever, predictable in a quantitative sense”. Seabird populations in the Antarctic are going to be affected by any

changes in whale exploitation or krill harvesting. One might hazard a guess that species with limited foraging ranges, specialized diets, or a high dependence on krill might be most severely affected, but this will be speculation until we understand more about the nature of the competition between seabirds and whales in the southern hemisphere. Before whale exploitation it is obvious that seabirds in the Southern Ocean were much less important in terms of krill consumption than whales, and the importance of seabirds in the Southern Ocean was much less than in the Peruvian, South African or west Pacific ecosystems.

F. The North Sea

Although the seabirds and fish stocks of the North Sea fishery area have been studied in greater detail and for a longer time than studies anywhere else in the world I have left this ecosystem until last among the case studies. There are two reasons for this.

First, it is a complicated ecosystem; the number and variety of seabird species is very great; some are diurnal and others nocturnal; the diets of some are largely unknown, while the detailed studies of others have shown that there is tremendous variation between different colonies. Although immense quantities of data have been gathered concerning the biology of whitefish, herring CIupea harengus L. and mackerel Scomber scombrus L. in the North Sea, the main food of most seabirds consists of the small food

280 R. W . FURNESS

fish such as sprats Sprattus sprattus (L.) and particularly sandeels Ammodytes marinus Raitt. These fish were of little commercial significance, except as food for other fish species, until the rapid development of industrial fishing in the North Sea began in the 1950s, and it has only been in the last few years that detailed investigations of their stocks and biology have been made. Modelling of the fish consumption by North Sea seabird communities is, therefore, rather difficult compared to the simple ecosystems of Peru and South Africa. The complex nature of the North Sea food web also makes the interpretation of perturbation effects difficult and as yet the theory of multispecies harvesting regimes is not sufficiently advanced to allow precise predictions to be made (May et al., 1979).

Secondly, many North Sea seabird populations were subject to heavy exploitation up to, and in a few cases beyond, the end of the nineteenth century. While in most parts of the world, particularly areas like Peru, the Pacific west coast, and South Africa, it is generally held that seabird numbers are regulated in relation to food abundance, in Britain a proportion, and possibly a majority, of seabird biologists believe that breeding seabird populations in the North Sea area have been limited by density-independent and food-independent effects of human exploitation. They argue that the rapid increase in many British seabird populations during the last 80 years has been the direct result of the protection now afforded them. It is a matter of dispute whether the present seabird populations, many of which are now much larger and more widely distributed than at any time in the documented or archaeological past (Fisher, 1952; Fisher and Lockley, 1954) have increased only because of protection. The increase in numbers may be explained by man’s fishing activities which have caused increases in the numbers of small fish, changed the species composition of the fish stocks and provided extra food in the form of offal and discards (see p. 284).

Detailed surveys of breeding seabirds in Britain and Ireland have provided accurate information on the sizes and rates of increase of populations of many species. In Britain, fulmars bred only on St Kilda before 1878, and archaeological evidence indicates that they were common there for at least 900 years (Lockwood, 1954; Fisher, 1966). Censuses in 1931, 1939 and at ten-year intervals thereafter, together with other irregular counts, show that the population on St Kilda has increased very little over the last 50 years. It has less than doubled in numbers (Harris and Murray, 1978) in spite of being freed from the intensive harvesting practised by the St Kildan community before they were evacuated in 1930. Mackenzie (1905) estimated annual harvests of 12 000 young fulmars from 20 000 hatched, and Clarke (1912) was told that 9600 were taken in 1910. In contrast, fulmars have increased rapidly in other parts of Britain since the colonization of Shetland in 1878. This spread appears to have originated not from St Kilda, but from


the Arctic, and may be correlated with the increase of food available from trawlers (Fisher, 1952, 1966), although this explanation has been challenged (Wynne-Edwards, 1962 ; Salmonsen, 1965 ; Bourne, 1966). The increase, averaging about 7 % per year, has resulted in the fulmar becoming one of the most numerous seabirds in the North Sea area (Cramp et al., 1974) and owing to its relatively high biomass and presence throughout most of the year, it is one of the main avian consumers (Furness, 1978b). Its inability to increase much in numbers when harvesting ceased on St Kilda suggests that there the population is limited by food or nest sites. In other areas in Britain and Ireland its numbers continue to increase, though now more slowly than before (Mudge, 1979), indicating that some environmental check is beginning to act.

A census of kittiwakes Rissa tridactyla (L.) has also been taken at ten-year intervals, since 1959, and these data together with a survey of historical information (Coulson, 1963) indicate that the population began to increase about 1900 and grew at 3 to 4 % per year up to 1969. Then the rate of increase fell to only 1 % per year between 1969 and 1979, with numhers declining slightly in some areas (Coulson, 1980). An environmental check appears to be starting to act on kittiwake numbers.

The numbers of the gannet, Sula bassana (L.), have been better known than for any other seabird. The population in Britain and Ireland increased from 47 000 pairs in 1909 to 54 500 pairs in 1939 and 140 500 pairs in 1969. The population rate of increase has apparently increased from an average of less than 1 %/year between 1909 and 1939 to about 3 %/year between 1939 and 1969, suggesting that the main improvement in environmental conditions for gannets occurred several decades after extensive human exploitation had ceased. Detailed examination of the history of colonies (Nelson, 1978) also indicates differences between areas in the timing of population increases which it is impossible to relate directly to patterns of human exploitation. The rates of increase of individual colonies appear to be determined to a large extent by the relative attractiveness of colonies to young recruit birds which may visit several areas before deciding where to establish a nesting site.

Colonies of the herring gull, Larus argentatus (Pontopp.), in many parts of Britain have shown rates of increase of 12-13%/year since at least 1930 (Chabrzyk and Coulson, 1976), although rates of increase have been much lower in some areas, such as west Scotland, Orkney and Shetland (Cramp et al., 1974), and several Shetland colonies have recently decreased in size (Furness, unpublished). Similarly, lesser black-backed gull L. fuscus L., great black-backed gull L. marinus L., common gull L. canus L. and black- headed gull L. ridibundus L. populations all appear to have been increasing in most of Britain and Ireland during the last 80 years, although a few

282 R. W. FURNESS

colonies of each species, particularly in north Scotland, have shown the opposite trend.

Numbers of most tern species seem to have increased since the beginning of the century except where human disturbance of breeding beaches has caused serious local declines (Cramp et al., 1974; Lloyd et al., 1975). The great skua probably colonized Britain in the mid-eighteenth century (Furness, 1977b) and it was strictly protected as a breeding species because it defended inland areas against sea eagles Haliaeetus albicilla (L.) which would otherwise be able to attack young lambs (Low, 1879). Numbers increased until egg and skin collection began shortly after 1800. Then the history of each colony depended on the extent of exploitation or the protection afforded. By 1900, numbers in Britain had been reduced to about 40 pairs in only four localities (Furness, 1977b). After 1900 several new colonies were founded and most increased at about 7%/year. Since 1970 the rate of increase of the British population has fallen, and several colonies are no longer increasing in numbers (Furness, 1977b, 1981, unpublished). The numbers of arctic skua Stercorarius parasiticus (L.) have also increased in Britain, although trends differ considerably between colonies (B. L. Furness, 1980), possibly partly as a result of competitive interactions with the larger great skua and predator- prey interactions with arctic terns (Furness, 1977a, 1978a). Populations of shags Phalacrocorax aristotelis (L.) have increased considerably, and cormorants P . carbo L. to a lesser extent (Cramp et al., 1974).

Numbers of puffins Fratercula arctica (L.) are extremely difficult to check, and no reliable figures exist for many British colonies even now, It is clear that most are at present increasing in numbers, although declines were re- ported for many areas earlier this century and the long term trend is obscure (Harris, 1976). Guillemot and razorbill Alca rorda L. populations are also extremely difficult to count and so it is not possible to detect trends in many colonies, but some are certainly increasing quite rapidly (Harris, 1976; Furness, 1981). Suitable methods for making a census of Manx shearwaters Pufinus pufinus (Brunn.), British storm petrels Hydrobates pelagicus (L.), Leach’s petrels and black guillemots Cepphus grylle (L.) have yet to be worked out or put into practice, and nothing is known about changes in the sizes of their populations.

The general pattern which may be seen to emerge is that most seabird populations in Britain and Ireland have been increasing for the last 80 years or so, although several are now showing signs of reaching a population ceiling. Rates of increase have probably been highest in the gulls, skuas and fulmar and, for many species, have been greater in Shetland and east Britain than in west or south Britain. Can these changes be related to food? As an initial approach to this question it would be useful to get an indication of the quantity of fish consumed by seabirds and the quantity available.

Seabirds / Seals

carnivores

herbivores 170

Primary production

FIG. 21. A North Sea food web based on major groups of organisms and inserting values for yearly production (kcals/me/year) (from Steele, 1974).

284 R. W. FURNESS

Steele (1974) constructed an energy web for the North Sea based on measured values of primary production, fish yields and mortality statistics. Using data for the years 1965 to 1969, and assuming that the fish catches over that period represented sustainable yields, he obtained an estimate of 2.0 x log tonnes and 0-9 x lo6 tonnes for the annual yields of pelagic and demersal species to man. He assumed that mortality due to fishing represented 80% of total mortality for demersal fish and 50% for pelagic fish, giving a total annual production of 1.3 x lo6 tonnes and 4.0 x lo6 tonnes for demersal and pelagic fish respectively. From an estimated area for the North Sea of 0.5 x lo6 km2 he obtained an estimated annual production of 2.6 and 8-0 kcals/m2 for demersal and pelagic fish. He estimated primary production to be 900 kcals/m2/year (Steele, 1974, pp. 15-19), giving quantitative values to the base and top of the food web, and allowing intermediate steps to be interpolated to give acceptable values for ecological efficiencies of energy transfer (Fig. 21). It is clear from Steele’s analysis that the efficiency of the North Sea ecosystem must be high. He points out

“transfer efficiencies around 20% appear to be required of the pelagic herbivores and also, possibly, of the benthic infauna that feed on faecal material. The numbers could be rearranged in various ways, but this would not alter one conclusion-that the yield of commercial fish is high in terms of the food web on which it is based”.

Steele assumed that all zooplankton production was utilized by pelagic fish or by benthic invertebrates through the decomposer chain of the web. In other words any significant consumption of zooplankton by seabirds would require even greater ecological efficiencies of the food web. Small petrels, kittiwakes and fulmars do consume zooplankton to some extent, but the relative importance of fish and zooplankton in their diets has yet to be assessed in any detail. Steele assumed that 50% of pelagic fish production was consumed by demersal fish since fisheries data show that fishing mortality is, on average, about 50% of total mortality of pelagic stocks. In fact this 50% must be shared by demersal fish, seabirds and marine mammals. From Steele’s energy web it would appear that 4 kcals/m2/year of pelagic fish production is available to be shared in this way. Similarly, the yield of 2 kcals/m2/year of demersal fish results in food becoming available to seabirds as offal and discarded undersize fish, Perhaps 5 % of the demersal fish catch (0-1 kcals/m2/year) is made available to seabirds in the form of viscera, liver and roes removed and deposited at sea (Bailey and Hislop, 1978). These authors suggest that it is now normal for a similar quantity of undersized fish to be discarded. The volume of discarded whitefish from an area of sea around Shetland (fishery rectangles 32 and 33) averaged 165 tonnes/month between May and August in 1975 and 1976 (Furness and Hislop, 1981). As this sea area measures about 6 x 109/m2 the discard volume represents an


average of 0.028 kcals/m2/month over the period May to August. The whitefish will be growing rapidly at this time of year and recruitment into fishable size classes is likely to result in considerable discarding. The total discard of about 0.1 kcals/m2 for the period May to August inclusive is probably a major part of the annual discard total, and this is the time of year when energy demands of seabird populations will be highest owing to their breeding activities. These energetic considerations suggest that the maximum amounts of fish food available to seabirds will be about 0.1 kcals/ m2/year of discards, 0.1 kcals/m2/year of offal and 4.0 kcals/m2/year of pelagic fish not caught by fisheries. Clearly the quantity of pelagic fish is much greater than offal or discards, although only a small proportion may actually be available to seabirds.

TABLE XII. ENERGY REQUIREMENTS OF SEABIRD POPULATIONS OF FOULA, SHETLAND

COLONY (FROM FURNESS, 1978b) OVER THE PERIOD DURING WHICH THE BIRDS ARE PRESENT IN THE VICINITY OF THE

Species

Maximum number of

Number of nonbreeding Population energy breeding birds in requirement

individuals colony area (Kcals x 106/year)

Fulmar Guillemot Shag Puffin Kittiwake Great skua Razorbill Arctic tern Storm petrel Great black-backed gull Herring gull Gannet Black guillemot Arctic skua Manx shearwater Common gull Lesser black-backed gull Leach’s petrel

40 000 40 000

6700 60 000 11 140

6000 6000

11 300 6000

44 46 0

240 600 100 20 4

60

18 000 15 000

2000 20 000

1000 2000 2000 1500 3000 500 500 500 80

200 100 20 20 40

4803 2675 1943 1614 426 331 269 120 54 47 25 18 15 11 2

<1 <1 t l

Furness (1978b) modelled the energy requirements of one of the large Shetland seabird communities. Using a “Monte Carlo” technique he obtained estimates for the mean and standard error of the population energy requirements for each seabird species over a year. The 95 % confidence intervals for

286 R. W. FLJRNESS

each species ranged over & 50 % of the mean estimate, mainly as a result of imprecisions in the equations used to estimate existence requirements. Reanalysis using the improved equations of Kendeigh et ul. (1977) reduces the confidence interval to 5 3 0 %, but the mean estimates for each species remain much the same (Table XII). Four species, i.e. fulmar, guillemot, shag and puffin, are responsible for 89% of the annual total energy requirements of the 18 species community. In this community the species thought to feed to a significant extent on discards (gulls, great skuas, gannets) contribute little to the overall budget. Consumption of offal and zooplankton by fulmars may reduce the total pelagic fish consumption by the community, since the fulmar has the greatest species food requirement, but we do not know whether offal or zooplankton form an important part of its diet. In Shetland, fulmar chicks appear to be fed largely on sandeels (Furness, unpublished), although no systematic or quantitative studies of diet have been carried out. The total energy requirement of the seabird community while in the vicinity of the colony is 1.2 x 1O1O kcals/year. This can be related to pelagic fish availability if the area over which the seabirds forage is defined. No direct studies of foraging distances have yet been made. Ideally radio-telemetry could be used to determine ranges of individuals throughout the breeding season. At present there are technical difficulties; foraging ranges may result in birds travelling over the radio horizon, while triangulation to obtain position would require an extensive baseline with receiving stations perhaps 40 km apart. An opposing requirement is that the total weight of transmitter and batteries should not impede normal behaviour of breeding adult seabirds. Transect counts made radially from colonies tend to show high concentrations of seabirds close to the colony and a low patchy distribution farther away. Interpreting such counts is difficult as it is not clear which birds are foraging adults and which are immatures, failed breeders or non-breeding adults not associated with the colony. Few transect studies have been made, and inter- pretations of results may differ widely (Cody, 1973; Bedard, 1976). However they do indicate that in most situations around Britain, breeding arctic terns, shags, black guillemots and arctic skuas feed within a few kilometres of the colony, while auks, great skuas and gulls travel farther, but usually remain within sight of the colony. Transects do not seem to give meaningful results for fulmars, kittiwakes, gannets or small petrels, suggesting that these species may range over much greater distances.

An indirect indication of the maximum foraging range can be obtained from the time spent away from the nest by each parent between chick feeds. Assuming that all this time is spent in flight in a straight line and that feeding itself takes a negligible time, the maximum potential feeding range can be calculated from a knowledge of flight speeds for each species (Pearson, 1968). In practice none of these assumptions is likely to hold, so the actual maximum


foraging range will be less, probably considerably less, than that calculated in this way, but the calculation does set a useful upper limit to the potential feeding range for each species. Working with this method on the seabirds of the Farne Islands, Northumberland, Pearson (1968) suggested that the maximum potential feeding range for species at that colony was less than 80 km. In Shetland it is likely that a few fulmars, gannets and perhaps small petrels travel even greater distances, but probably most forage well within this range, and probably mainly within 50 km of the colony. The main Shetland seabird communities are approximately 70 km apart, and placed strategically at the north, south, east and west corners of Shetland (Fig. 22).

FIG. 22. Major seabird colonies in Shetland and radii of 45 km around each showing the likely core foraging areas of the seabirds from each breeding colony (from Furness, 1977b).

288 R. W. FURNESS

Average annual North Sea catch of 0. ' a11 fish species 1900-1939

. 0

0

0 . 0

-0.

Year

FIG. 23. Annual landings of sandeels caught in the North Sea (from K. Warburton, personal communication).

If birds travel more than 45 km from one colony to feed they are likely to enter the feeding zone of one of the adjacent colonies. As all four are of . . 1-- _I-_ I - , * L ---- :-- ------ :.:-- *,. ___^ ,,+,,+\ I4.A"

travelling outside a 45 km radius of their colony will to some extent be compensated for by others travelling into the area, so for these reasons I chose to compare the food consumption estimate with the pelagic fish production within a 45 km radius of the colony. This area of 4700 km2 of sea would produce 4-2 x 1O1O kcals of zooplankton-consuming fish per year if typical for the North Sea as a whole. Consumption of 1.2 x 1O1O kcals/year by seabirds would be equivalent to 28% of the fish production. As Steele (1974) indicated that, on average, 50% of pelagic fish production is taken

, I " - - - - -- -- -.'

other predatory fish, indicating that the relationship between seabird consumption, pelagic fisheries and predatory fish is tight, with little scope for an


increase in one without a concomitant decrease in another. For this reason the rapid growth of a sandeel fishery for industrial purposes (Fig. 23) and the wider growth of industrial fishing in the North Sea as a whole (Fig. 24) are threats to seabird populations. These fisheries can only increase at the expense of the food supplies for the demersal stocks or seabirds, and it is possible that they have already reached a stage where food availability to seabirds has declined sufficiently to result in a reduction or reversal in their population growth rates.

c In

E E L 0

P P a f d

In

.- F z 4

2000 J

1000-

500 -

I O O -

50 -

20 - 1 0 9 . . . . . . . . . . . . . . . . . . . . . . . . i

1950 55 60 65 75

Year

FIG. 24. Total landings of industrial fisheries in the North Sea (from Hempel, 1978).

Does the history of North Sea fisheries indicate why seabird populations were able to increase over the past 80 years? We have seen that seabirds appear to consume an important quantity of sandeel and other “food-fish’’ production. Have sandeels and other “food-fish” become more abundant or available to seabirds? The introduction of steam trawling and power winches between 1870 and 1900, together with the development of the otter trawl towards the end of the nineteenth century greatly increased fishing power. Very soon evidence of overexploitation of whitefish stocks came to light and the International Council for the Exploration of the Sea was set up in 1902 as a result, with the aim of monitoring fish stocks and the effects of fishing. Because of the extensive improvements since 1900, such as improvements to the otter trawl, increased vessel power and size, introduction of various location and fish detecting devices and the development of purse-seine nets, it is not possible to standardize fishing effort over long periods of time. The

290 R. W. FURNESS

data for the earliest years are also less reliable than those obtained since 1945. However, Lundbeck (1959, 1960, 1962) found that major changes in whitefish abundance had taken place in all areas of the North Sea and indicated effects of considerable growth overfishing. Most severe was the reduction in whitefish biomass in the southern North Sea (Fig. 25) where it is clear that overfishing at the turn of the century reduced the stock biomass by 70%. A partial recovery took place during the First World War, but the stock was further reduced to only 15 % of its 1887 biomass by 1936. A second recovery occurred during the Second World War, but this was quickly reversed after 1945. Thus a major reduction in whitefish stock biomass took place around 1890 to 1900, just when the exploitation of seabird populations was tending to cease. The increase of many seabird populations dates from about 1900, suggesting that it could be a response to the reduced predation on “food-fish” by whitefish, making more food available to the seabirds. Protection of seabirds may have accelerated this process, but it seems likely that the increased food availability would have been necessary to allow most populations, except those reduced near to extinction, to increase.

7 0 -

I

60- c 0 c u

.- a 50- . > 0

- 40-

e k a 30- S 0

0 c

” 2 0 -

I O -

o ! 1 I I I I I

1885 95 05 15 ’ 25 35 45 90 1900 10 20 30 40

Yeor

FIG. 25. Catch per unit effort of a standardized German trawler in the southern North Sea (from Lundbeck, 1962).

COMPETITION BETWEHN FISHERIES AND SEABIRDS 29 1

Whitefish also responded to the improved food availability as a result of stock reduction. Haddock Melanogrammus aeglefinus (L.) growth rate increased (Jones and Hislop, 1978) as did that of whiting Merlangius mer- langus (L.) (Daan, 1975). As a result of increased growth rates both species reached reproductive condition at an earlier age. Cod Gadus morhua L. showed no change in growth rate, but the age of maturity did decrease so that cod reached reproductive age at a smaller size (Daan, 1978). All these changes can be ascribed to greater food ,availability per fish as a result of reduced competition with other whitefish.

Stocks of herring and mackerel appear not to have been seriously depleted until the 1950s or 1960s. The adult biomass of North Sea herring remained around 2-5 x log tonnes until 1965 when purse-seining rapidly depleted the stock to one tenth of its original level, at which stage growth rate suddenly increased and partly compensated for the reduction in stock (Burd, 1978). When herring became unprofitable the purse-seine fishermen turned to mackerel and the stock of 2 x lo6 tonnes before 1965 was reduced to about one tenth of this in only four years of fishing (Hamre, 1978). The fisheries for adult herring and mackerel before 1960 will have been directly beneficial to seabirds by reducing the average size of fish in the populations without greatly reducing stock biomass, so that a higher proportion of the stock will have been in the size range suitable for seabird consumption. After 1960 the reductions in stocks will not have been directly beneficial to seabirds, but as most species feed more on sandeels than on herring or mackerel there was probably an indirect beneficial effect. Using a complicated model, Andersen and Ursin (1977) found that the reduction in stocks of herring and mackerel is likely to have led to increases in the populations of their ecological competitors, sandeels, sprats and Norway pout Trisopterus esmarkii (Nilsson). Evidence that such increases have taken place is not readily available as sandeels were of no commercial interest until recently. Stock sizes were, and still are, largely unknown, although Sherman et al. (1981) demonstrated increases on both sides of the Atlantic. Catch-per-unit effort data of Norway pout by Scottish research vessels show that the stock of this species has also greatly increased since the mid-1950s from a level which had previously been fairly constant from 1925 to 1955 (Richards et al., 1978). Andersen and Ursin’s model also predicted that whitefish stocks would increase as a result of decreased predation on their larvae by herring and mackerel. Such increases have occurred, and are very difficult to explain except in terms of such an ecosystem interaction (Hempel, 1978). As a corollary of their recovery the growth rates of whitefish have fallen again, suggesting that the superabundance of food generated by their stock depletion no longer exists (Hempel, 1978). Thus the overfishing of herring and mackerel may have improved sandeel availability to seabirds in the short term, but the partial recovery of

292 R. W. FURNESS

whitefish stocks will have taken up part of the sandeel surplus, while the increasing fishery for sandeels, particularly around seabird colonies in north Britain, is currently removing an increasing part of the sandeel stocks.

IV. Influences of Food on Seabird Population Ecology

The case studies in the previous section indicate that seabird communities may consume a considerable proportion of the production of lower trophic levels in marine ecosystems. Increases or decreases in the amount of food available to seabirds, as a result of changes in ecosystem structure, often lead to closely coupled changes in seabird biomass and energy consumption. This is clearly evidence that the size of the seabird community is determined by the amount of energy available, but this trivial deduction is often obscured by the fact that seabird communities comprise a variety of species, some abundant and some rare. Reasons for the particular size of a population, or rate of change in size of populations, are often difficult to discover. Seabird population ecology can be examined at two discrete levels. At one level, consideration of seabird biology has led to a number of theories concerning the limitation and regulation of population sizes (Salomonsen, 1955; Lack, 1954, 1966; Wynne-Edwards, 1962; Ashmole, 1963, 1971 ; Diamond, 1978). By comparing the ecology of species members of seabird communities it has been possible to make a number of deductions about the role of interspecific competition in the limitation of sizes of populations within the overall community (Ashmole, 1971 ; Ashmole and Ashmole, 1976; Ainley, 1977; Belopolskii: 1961 ; Cody, 1973; Croxall and Prince, 1980; Diamond, 1978; Pearson, 1968). At another level, studies of aspects of seabird breeding biology or population dynamics, usually of a single species considered in isolation, can shed light on mechanisms whereby food limitation may act in a density-dependent way to regulate population size. These approaches will be considered in turn.

A. Evidence from Studies of Community Structure

The competitive exclusion principle (Gause’s hypothesis) predicts that significant ecological isolating mechanisms will exist between all members of a multispecies seabird community. As a result of these, variations in conditions from place to place will lead to differences in the relative fitness of species within communities and thus to changes in sizes of species populations. Inter- specific competition for food would be expected to c,ause divergent evolution or character displacement, and considerable overlap in food and feeding ecology of species within seabird communities has been cited as evidence that


food must be “superabundant” and not a limiting factor in terms of seabird population sizes (e.g. Salomonsen, 1955; Beck, 1970). Some studies have shown that interspecific competition for nest sites is far greater than for food, and this may also be taken to suggest that populations are not food-limited (Lack, 1934; Belopolskii, 1961 ; Bedard, 1969; Williams, 1974). However, many seabird communities exist in areas where suitable nest sites are not in short supply. In such areas there is little or no evidence for interspecific competition for habitats or nest sites, and ecological isolation is purely in terms of segregation of food by temporal, spatial, behavioural or dietary separation. At the sub-Antarctic island of South Georgia there are 25 breeding species of seabirds. Although nest site preferences differ according to species morphology and ecology, there are no differences between species which can be attributed to interspecific competition, there is no evidence of competition for nest sites within species, and many areas of fully suitable breeding habitat are unexploited (Croxall and Prince, 1980). Similarly, nest sites are apparently not in short supply on Christmas Island (Ashmole, 1963) where ecological segregation is largely by diet or feeding range (Ashmole and Ashmole, 1967). Diamond (1978) points out that the amount of food available to tropical seabirds is best estimated by a measure of feeding area alone, as these communities consist of surface feeding species only. From this he predicts that if populations are limited by food availability then populations of pelagic feeding species should outnumber populations of inshore feeding species, and populations should be greater for species which migrate rather than remain resident throughout the year (migration being equivalent to increasing the feeding area). Both predictions are supported by the data he presents, suggesting that the sizes of tropical seabird populations are indeed food-limited.

In Shetland, as in many higher latitude areas, there is considerable competition between species for nest sites in optimal localities, although nest site preferences clearly differ much more between species than does dietary composition. Records of nesting areas or nest sites being usurped are frequent. On Foula, Shetland, the most common recorded examples are fulmars, shags and guillemots displacing kittiwakes (B. L. Furness, 1979), great skuas displacing Arctic skuas ; shags and gannets displacing guillemots ; and guillemots displacing razorbills (Table XIII). Clearly optimal nest sites are in limited supply, but this does not necessarily mean that the seabird community is not food-limited. On Foula there are no records of any species taking over nest sites of fulmars, storm petrels, gannets, great skuas or great black-backed gulls. Further, there are more nest sites or areas which have been used over the years and have remained available than there are breeding pairs of each gull species, Arctic terns and black guillemots. The same is probably true, although not possible to assess, for many other species. Thus numbers of

294 R. W. FURNESS

many species could increase without necessarily displacing other birds. The species composition and relative abundance of Shetland seabird communities varies greatly between islands although their diets, and apparently the availability of fish stocks, are fairly homogeneous. The island of Noss has large cliffs suitable for nesting gannets, guillemots and kittiwakes, which dominate its seabird community. Foula cliffs are mainly occupied only by fulmars owing to their sheer nature and lack of ledges, while the community is dominated by 'boulder nesting birds, chiefly shags, guillemots, puffins and more fulmars. It may be that lack of suitable nesting habitat limits numbers of particular seabird species in particular colonies, but this will simply give species not so restricted by nest site limitations a competitive advantage, allowing them to increase to the limit of the community size set by food availability. In such situations one might expect species limited by habitat availability to fluctuate least in numbers in response to changes in food availability. This prediction has yet to be tested.

TABLE XIII. NEST SITES USED BY SEABIRDS BREEDING AT FOULA, SHETLAND AND SPECIES OBSERVED TO USURP THEIR NEST SITES (FROM FURNESS, UNPUBLISHED DATA)

Species regularly Preferred nest Alternative nest taking over nest

Species site sites sites

Red-throated diver Loch side Shag Boulderfield Fulmar Sheer cliff Storm petrel Boulder scree Leach's petrel Grass bank Manx shearwater Grass bank Gannet Wide ledges Great skua Moor Arctic skua Moor Herring gull Rocky shore Lesser black-b. gull Rocky shore Great black-b. gull Rocky shore Common gull Moor Kittiwake Cliff

Arctic tern Moor Guillemot Cliff Razorbill Boulder field

Black guillemot Boulder beach Puffin Grass bank

Fulmar Cliff Fulmar, guillemot Boulderfield, inland Walls, grass bank

Puffin Puffin

Stack top

Grass

Moor

Boulderfield

Rocky shore Boulderfield Cliff fissures

Boulderfield, fissures

Great skua Great black-b. gull Great black-b. gull

Arctic skua Fulmar, shag,

guillemot Arctic skua, gulls Fulmar, shag, gannet Puffin, guillemot,

shag Puffin Fulmar


There is considerable uncertainty as to the relative importance of differences in timing of breeding (temporal isolation of peak food demands), feeding range (spatial isolation), feeding method (behavioural isolation) and dietary differences resulting from differences in morphological adaptations as species-isolating mechanisms. Cody (1 973) argued that differences in feeding range, resulting from direct competitive displacement, provide the main species-isolating mechanism in Atlantic and Pacific communities of auks. Few studies of the zonal feeding distribution of seabirds from colonies have been attempted so that the importance of spatial segregation is difficult to assess. Bedard (1976) re-examined Cody’s data and argued that it failed to display the zonation claimed by Cody; rather the differences between species were largely attributable to morphological adaptations of the feeding appara- tus of each species. Croxall and Prince (1980) found some evidence for species isolation through adaptations in the winter breeding seasons of wandering albatross Diomedia exulans L. and king penguin and out-of-phase breeding in the species pairs dove prion Pachyptila desolata (Gmelin) and blue petrel Halobaena caerulea (Gmelin) and the common and South Georgia diving petrels Pelecanoides urinatrix exsul Salvin and P. georgicus Murphy and Harper. They concluded, however, that differences in food and feeding ecology were of most importance within the community. The ecological differences between species are of considerable importance when food availability alters, either over a short or long time scale. Croxall and Prince (1980) also discuss some implications of the difference in diet of grey-headed albatrosses Diomedia chrysostoma Forster which feed primarily on squid, and black- browed albatrosses D. melanophris Temminck which feed mainly on krill and to a small extent on squid. The low nutritive value of squid is probably the reason why successful grey-headed albatrosses are unable to regain breeding condition in time to lay the following season, whereas black-browed albatrosses can breed each year. However, in 1977-78 when krill was abnor- mally scarce around South Georgia, grey-headed albatrosses had an unusually successful breeding season, but black-browed albatrosses were unable to switch sufficiently to feeding on squid, and their breeding success was very low. Croxall and Prince speculate that the failure of the black-browed albatrosses may have been due to their inability to compete with the pre- dominantly squid-eating grey-headed albatrosses.

B. Evidence from Single Species Studies

Food shortage could influence seabird population dynamics by reducing breeding success, increasing adult mortality or age of first breeding. Seabirds are generally long-lived, and there are few studies which have accurately determined adult survival rates, let alone annual variations or variations in

296 R. W. FURNESS

relation to colony size, population density or food availability. Coulson and Wooller (1976) showed that the annual survival rate of adult breeding kittiwakes fell as the colony studied increased in size and density, and they associated this change with increased competition, particularly for nest sites in the larger colony. Culling of herring gulls has been used to reduce populations in a number of British colonies. A consequence of culling on the Isle of May has been a reduction in the age of first breeding (Chabrzyk and Coulson, 1976; Duncan, 1978). Nesting of birds in sub-adult plumage (Duncan, 1978) is presumably made possible by the reduced competition resulting from the population reduction, although the relative roles of social behaviour and increased food per bird are unclear in this process. The hor- monal basis of such release is described by Carrick and Murray (1964), who cite the royal penguin Eudyptes chrysolophus schlegeli (Brandt) as an example. In this species failure to attain successful breeding status, and failure of the gonads to mature during the first eight years of life, are related to inadequate fat storage for incubation, which indicates poorer feeding at sea than the successful breeders enjoy. As soon as food supplies allow adequate fat storage, birds can recruit into the breeding population.

Food availability, mediated through its influence on social behaviour, may affect adult survival rates and age at first breeding, but more data are required to substantiate this. In contrast, numerous studies have indicated that food availability is one of the main determinants of breeding success in seabird populations. Harris and Hislop (1978) showed that puffins select larger fish and species of high calorific value to feed chicks. Chick growth and fledging weights were highest when the diet consisted mainly of sprats, while in years when young whiting were fed to chicks their growth was poorer. Food quantity and quality tended to be better at the Isle of May than at St Kilda. Rates of increase of the two colonies coincide with these differences in food availability, although there is little evidence that variations in fledging weight, resulting from food shortage, alter puffin survival or subsequent return to the colony (Harris, in press).

Food availability may also influence breeding success in an indirect way. Puffin breeding success is greater on sloping habitat than on flat habitat on Great Island, Newfoundland, as a result of the interaction between food availability and gull interference. Hungry chicks spend more time at the burrow entrance, exposing themselves to gull predation. Frequency of chick feeding was lower on flat habitat because gulls were able to steal fish from food-carrying adults more successfully there, so that chicks on the flat habitat were more likely to suffer food shortage (Nettleship, 1972). The complicated interaction between habitat, food availability and interspecific relations indicates the difficulty of relating seabird community dynamics to food supplies.


Brood size reduction is a strategy employed by a number of seabird species. This is clearly correlated with food availability, and optimizes fledgling production in relation to food supply (Hahn, 1981). Procter (1975) showed that the older sibling in broods of the south polar skua Catharacta maccormicki (Saunders) would attack and kill the younger if deprived of food for a period of time. Young (1963) found that it was exceptional for the younger chick to survive to fledging when food was in short supply. The closely related great skua in the North Atlantic regularly rears both chicks to fledging. The population is currently growing rapidly under favourable conditions provided by the whitefish industry and increased sandeel stocks (Furness and Hislop, 1981) so that chicks even in supernormal broods rarely go short of food. Haymes and Morris (1977) found that herring gulls at Lake Erie were able to rear supernormal broods without increased brood reduction or predation because they were able to make use of human-supplied artificial food sources in addition to their natural food supply, while Hunt (1972) found that chicks fed on garbage and fish waste grew faster than those in more isolated colonies where only natural foods were available. Hunt and Hunt (1976) found that glaucous-winged gull chick survival correlated closely with growth rate and both were determined largely by food availability, the main cause of chick mortality being attack by neighbouring adults. Thus a number of observational studies have shown the importance of food availability, directly or indirectly, in determining breeding success. Using a combination of central place foraging theory (Hamilton and Watt, 1970; Orians and Pearson, 1979) and a deterministic simulation model of guillemot feeding rate and chick growth, Ford et al. (in press) concluded that the breeding success of a guillemot population would fall steeply with a food density reduction of only 10-30%, while a reduction of food availability of 40% or more would lead to total reproductive failure. Overfishing of fish stocks can easily lead to a stock density reduction of this magnitude (Hempel, 1978) so that serious effects on seabird population dynamics could be expected.

V. Acknowledgements

I would like to thank Dr J. C. Coulson, Professor G: M. Dunnet and Pro- fessor V. C. Wynne-Edwards for fostering my interest in this subject and for their continued stimulating interest and encouragement. I am indebted to Drs R. J. M. Crawford, B. T. Grenfell, J. R. G. Hislop, J. H. Lawton, J. Prevost and K. Warburton for providing me with data and for answering requests for help and information. Drs J. C. Coulson, J. P. Croxall, M. P. Harris, J. B. Nelson and P. Monaghan kindly read and criticised parts of the manuscript. I am most grateful for their helpful comments.

298 R. W. FURNESS

VI. References

Ainley, D. G. (1977). Feeding methods of seabirds: a comparison of polar and tropical communities. In “Adaptations in Antarctic Ecosystems” (G. A. Llano, ed.), pp. 669-685. Gulf Publishing Co., Houston.

Ainley, D. G. and Lewis, T. J. (1974). The history of Farallon Island marine bird populations, 1854-1972. Condor 76, 4 3 2 4 6 .

Andersen, K. P. and Ursin, E. (1977). A multispecies extension to the Beverton and Holt theory of fishing, with accounts of phosphorus circulation and primary production. Meddelelser fra Danmarks Fiskeri- og Havunders0gelser 7 , 31 9-435.

Andrewartha, H. G. and Birch, L. C. (1954). “The Distribution and Abundance of Animals.” Univeristy of Chicago Press, Chicago.

Aron, W. (1960). The distribution of animals in the eastern North Pacific and its relationship to physical and chemical conditions. Journal of the Fisheries Research Board of Canada 19, 271-314.

Ashmole, N. P. (1963). The regulation of numbers of tropical oceanic birds. Ibis 103b, 458-473.

Ashmole, N. P. (1971). Seabird ecology and the marine environment. In “Avian Biology” (D. S. Farner and J. R. King, eds), Vol. I, pp. 112-286. Academic Press, London and New York.

Ashmole, N. P. and Ashmole, M. J. (1967). Comparative feeding ecology of sea birds of a tropical oceanic island. Bulletin of the Peabody Museum of Natural History 24, 1-131.

Bailey, R. S. and Hislop, J. R. G. (1978). The effects of fisheries on seabirds in the northeast Atlantic. Ibis 120, 104-105.

Baird, D. (1975). The South African mackerel: its biology and fishery. Sourh African Shipping News and Fishing Industry Review 30, 46-51.

Ball, N. J. and Arnlaner, C. J. (1980). Changing heart rates of herring gulls when approached by humans. In “A Handbook on Biotelemetry and Radio Tracking” (C. J. Amlaner and D. W. MacDonald, eds), pp. 589-594. Pergamon Press, Oxford.

Baudinette, R. V. and Schmidt-Nielsen, K. (1974). Energy cost of gliding flight in herring gulls. Nature, London 248, 83-84.

Beck, J. R. (1970). Breeding seasons and moult in some smaller Antarctic petrels. In “Antarctic Ecology” (M. W. Holdgate, ed.), Vol. I, pp. 542-550. Academic Press, London.

Bedard, J. (1969). Histoire naturelle du Gode, Alca torda L., dans le golfe Saint Laurent, Province de QuBbec, Canada. Etude de Service Canadien de la Faune 7. Ottawa.

Bedard, J. (1976). Coexistence, coevolution and convergent evolution in seabird communities: a comment. Ecology 57, 177-184.

Belopolskii, L. 0. (1961). Ecology of sea colony birds of the Barents Sea. Trans- lations of the Israel Programme for Scientific Translations, Jerusalem.

Berry, H. H. (1975). History of the guano platform on Bird Rock, Walvis Bay, South West Africa. Bokmakierie 27, 60-64.

Blaxter, J. H. S. and Hunter, J. R. (1982). The biology of the clupeoid fishes. Advances in Marine Biology 20, 1-223.

Bourne, W. R. P. (1966). The plumage of the fulmar of St Kilda in July. Bird Study

Bryant, D. M. (1969). Reproductive costs in the house martin (Delichon urbicu). 13,209-213.

Journal of Animal Ecology 48, 655-676.

COMPETITION BETWEEN ’FISHERIES AND SEABIRDS 299

Burd, A. C. (1978). Long term changes in North Sea herring stocks. Rapports et ProcPs- Verbaux des Reunions. Conseil International pour I’Exploration de la Mer

Butler, P. J. (1980). The use of radio telemetry in the studies of diving and flying in birds. In “A Handbook on Biotelemetry and Radio Tracking” (C. J. Amlaner and D. W. MacDonald, eds), pp. 569-578. Pergamon Press, Oxford.

Butler, P. J., West, N. H. and Jones, D. R. (1977). Respiratory and cardiovascular responses of the pigeon to sustained, level flight in a wind-tunnel. Journal of Experimental Biology 71, 7-26.

Calow, P. (1977). Conversion efficiencies in heterotrophic organisms. Biological Reviews 52, 385409.

Carrick, R. and Murray, M. D. (1964). Social factors in population regulation of the Silver Gull Larus novaehollandiae Stephens. CSIRO Wildlge Research 9,

Centurier-Harris, 0. M. (1977). Estimates of size and interaction of the South African anchovy and pilchard populations. MSc thesis, University of Cape Town.

Centurier-Harris, 0. M., Crawford, R. M. and Newman, G. G. (1977). Fluctuations in the mixed-species pelagic stocks of the western Cape (ICSEAF Division 1.6),

Chabrzyk, G. and Coulson, J. C. (1976). Survival and recruitment in the herring

Clarke, W. E. (1912). “Studies in Bird Migration”. Oliver and Boyd, Edinburgh. Clark, R. N. and Marr, J. C. (1955). Population dynamics of the Pacific Sardine.

California Cooperative Oceanography and Fishery Investigation, Progress Report

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

Conroy, J. W. H. (1975). Recent increases in penguin populations in the Antarctic and Subantarctic. In “The Biology of Penguins” (B. Stonehouse, ed.), pp. 321-336. Macmillan, London.

Cooper, J. (1979). The status of seabirds at the island in Saldanha Bay, and re- commendations for their management. Unpublished report to the Advisory Committee for Ecological Studies in the Langebaan-Saldanha Area, University of Cape Town.

Coulson, J. C. (1963). The status of the kittiwake in the British Isles. Bird Study 10,

Coulson, J. C. (1980). Kittiwakes 1979. British Trust for Ornithology News 111, 1. Coulson, J. C. and Wooller, R. D. (1976). Differential survival rates among breeding

kittiwake gulls Rissa tridactyla (L.). Journal of Animal Ecology 45, 205-213. Cramp, S., Bourne, W. R. P. and Saunders, D. (1974). “The Sea-birds of Britain and

Ireland”. Collins, London. Crawford, R. J. M. (1979). Studies of the biology of the pelagic fish stocks of

South Africa. PhD thesis, University of Cape Town. Crawford, R. J. M. and Shelton, P. A. (1978). Pelagic fish and seabird inter-

relationships off the coasts of South West and South Africa. Biological Conserv- ation 14, 85-109

Crawford, R. J. M., Centurier-Harris, 0. M., Wingate, G. H. L. and Kriedemann, B. D. (1978). Revision of species and age composition of landings in the South African purse-seine fishery, 1950-1976. Fishery Bulletin of South Africa 10,69-88.

172, 137-153.

189-199.

1950-1976. ICSEAF/S.A.C./77/S.P.,4,1-28.

gull Larus argentatus. Journal of Animal Ecology 45, 187-203.

4911-28.

147-179.

300 R. W. FURNESS

Crawford, R. J. M., Shelton, P. A. and Cooper, J. (1982). Distribution, population size and conservation of the Cape gannet. Fishery Bulletin of South Africa (in press).

Croxall, J. P. and Kirkwood, E. D. (1979). The breeding distribution of penguins on the Antarctic Peninsula and islands of the Scotia Sea. British Antarctic Survey, Cam bridge.

Croxall, J. P. and Prince, P. A. (1979). Antarctic seabird and seal monitoring studies. Polar Record 19, 573-595.

Croxall, J. P. and Prince, P. A. (1980). Food, feeding ecology and ecological segregation of seabirds at South Georgia. Biological Journal of the Linnaean Society 14, 103-131.

Daan, N. (1975). Consumption and production in North Sea cod. Gadus morhua: an assessment of the ecological status of the stock. Netherlands Journal of Sea Research 9, 24-55.

Daan, N. (1978). Changes in cod stocks and cod fisheries in the North Sea. Rapports et ProcPs - Verbaux des Re‘unions. Conseil International pour I’Exploration de la Mer

Davies, D.H. ( 1955). The South African pilchard (Sardinopsocelfata) : bird predators, 1953-1954. Investigational Report, Division of Fisheries Union of South Africa 18,

Davies, D. H. (1956). The South African pilchard (Sardirtops ocellata) and maasbanker (Trachurus trachurus) : bird predators, 1954-1955. Investigational Report, Division of Fisheries Union of South Africa 23, 1 4 0 .

Diamond, A. W. (1978). Feeding strategies and population size in tropical seabirds. American Naturalist 12, 21 5-223.

Duncan, N. (1978). The effects of culling herring gulls (Larus argentatus) on recruitment and population dynamics. Journal of Applied Ecology 15, 697-713.

Dunnet, G. M. (1977). Observations of seabirds at sea off Cape Town, October 1975. Cormorant 2, 11-14.

Edwards, R. Y . (1964). Birds seen in Active Pass, British Columbia. Reporr of the Brirish Columbia Provincial Museum 1964, 19-23.

El-Wailly, A. J. (1966). Energy requirements for egg laying and incubation in the Zebra Finch. Taenopygia castanotis. Condor 68, 582-594.

Everson, I. (1977). The living resources of the Southern Ocean. Food and Agri- culture Organisation Southern Ocean Fisheries Survey Programme, Rome.

Everson, I. (1978). Antarctickrill (Euphausiasuperba) as an acoustictarget. Proceed- ings of the Conference on Acoustics in Fisheries 2.2, 26pp. University of Bath.

Everson, I. and Ward, P. (1980). Aspects of Scotia Sea zooplankton. Biological Journal of the Linnaean Society 14, 93-101.

Ferns, P. N., Macalpine-Levy, I. H. and Goss-Custard, J. D. (1980). Telemetry of heart rate as a possible method of estimating energy expenditure in the Red- shank Tringa totanus (L.). In “A Handbook on Biotelemetry and Radio Tracking” (C. J. Amlaner and D. W. MacDonald, eds), pp. 595-601. Pergamon Press, Oxford.

Fiedler, R. H., Jarvis, N. D. and Lobell, M. J. (1943). La pesca y las industrias pesqueras en el Peru. Lima.

Fisher, J. (1952). “The Fulmar”. Collins, London.

172,39-57.

1-32.

Fisher, J. (1966). The fulmar population of Britain and Ireland, 1959. Bird Study 13, 5-76.

Fisher, J. and Lockley, R. M. (1954). “Sea-birds.” Collins, London.


Ford, It. G., Wiens, J. A., Heinemann, D. and Hunt, G. L. (1982). Modelling the sensitivity of colonially breeding marine birds to oil spills: guillemot and kittiwake populations on the Pribilof Islands, Bering Sea. Journal of Applied Ecology (in press).

Frey, H.. W. (1971). California’s living marine resources and their utilization. Department of Fish and Game, California Resources Agency, San Francisco.

Frost, P. G. H., Siegfried, W. R. and Cooper, J. (1976). Conservation of the jackass penguin (Spheniscus demersus (L.)). Biological Conservation 9, 79-99.

Furness, B. L. (1979). The effects of great skua predation on the breeding biology of the kittiwake on Foula, Shetland. Scottish Birds 10, 289-296.

Furness, B. L. (1980). Territoriality and feeding behaviour in the Arctic skua (Stercorarius parasiticus). PhD thesis, University of Aberdeen.

Furness, R. W. (1977a). Effects of great skuas on Arctic skuas in Shetland. British Birds 70,96107.

Furness, R. W. (1977b). Studies on the breeding biology and population dynamics of the great skua Catharacta skua Briinnich. PhD thesis, University of Durham.

Furness, R. W. (1978a). Kleptoparasitism by great skuas (Cutharacta skuu Briinn.) and arctic skuas (Stercorarius parasiticus L.) at a Shetland seabird colony. Animal Behaviour 26, 1167-1 177.

Furness, R. W. (1978b). Energy requirements of seabird communities: a bioenergetics model. Journal of Animal Ecology 47, 39-53.

Furness, R. W. (1981a). Seabird populations of Foula. Scottish Birds 11, 237-253. Furness, R. W. (1982). Estimating the food requirements of seabird and seal

populations and their interactions with commercial fisheries and fish stocks. Proceedings of the Symposium on Sea and Shore Birds, Cape Town, 1979. African Seabird Group, in press.

Furness, R. W. and Cooper, J. (In press). Interactions between seabird populations and fish stocks of the Saldanha region, South Africa. Marine Ecology Progress Series.

Furness, R. W. and Hislop, J. R. G. (1981). Diets and feeding ecology of great skuas Catharacta skua during the breeding season in Shetland. Journal of Zoology (London) 195, 1-23.

Gambell, R. (1973). Some effects of exploitation on reproduction in whales. Journal of Reproduction and Fertility (Supplement) 19, 533-553.

Gessaman, J. A. (1973). “Ecological Energetics of Homeotherms.” Monograph Series Utah State University Press, Logan, 20, 1-155.

Grenfell, B. T. and Lawton, J. H. (1979). Estimates of the krill consumed by whales and other groups in the Southern Ocean: 1900 and the present. Un- published manuscript.

Gulland, J. A. (1961). Fishing and the stocks of fish at Iceland. Ministry of Agri- culture, Fish and Food (U.K.), Fishery Investigations, Series I I 23, 1-32.

Hahn, D. C. (1981). Asynchronous hatching in the laughing gull: cutting losses and reducing rivalry. Animal Behaviour 29, 421-427.

Hails, C. J. and Bryant, D. M. (1979). Reproductive energetics of a free-living bird. Journal of Animal Ecology 48, 471-482.

Hamilton, W. J. and Watt, K. E. F. (1970). Refuging. Annual Reviews in Ecology and Systematics 1, 263-286.

Hamre, J. (1978). The effect of recent changes in the North Sea mackerel fishery on stock and yield. Rapports et ProcPs- Verbaux des RPunions. Conseil International pour I’Exploration de la Mer 172, 197-210.

Harris, M. P. (1976). The seabirds of Shetland in 1974. Scottish Birds 9,37-68.

302 R. W. FURNESS

Harris, M. P. (in press). Seasonal variation of fledging weight in the puffin Fratercula arctica. Ibis 124, 100-103.

Harris, M. P. and Hislop, J. R. G. (1978). The food of young puffins, Fratercula arctica. Journal of Zoology (London) 185,213-236.

Harris, M. P. and Murray, S. (1978). “Birds of St Kilda.” Institute of Terrestrial Ecology, Cambridge.

Haymes, G. T. and Morris, R. D. (1977). Brood size manipulations in herring gulls. Canadian Journal of Zoology 55, 1762-1766.

Heinroth, 0. and.Heinroth, M. (1928). Die Vogel Mitteleuropas. Band 3. Bemiihler, Berlin.

Hempel, G. (1978). North Sea fisheries and fish stocks-a review of recent changes. Rapports et Proc2s- Verbaux des RPunions. Conseil International pour I’Exploration de la Mer 173, 145-167.

Hubbs, C. L. (1948). Changes in the fish fauna of western North America correlated with changes in ocean temperature. Journal of Marine Research 7,459-482.

Humphreys, W. F. (1979). Production and respiration in animal populations. Journal of Animal Ecology 48, 427-453.

Hunt, G. L. (1972). Herring gull population dynamics: the significance of man’s waste products. In Proceedings of the 15th International Ornithological Congress (Hague) (K. H. Voous, ed.), p. 652. Leiden.

Hunt, G. L. and Hunt, M. W. (1976). Gull chick survival: the significance of growth rates, timing of breeding and territory size. Ecology 57, 62-75.

Hutchinson, G. E. (1950). The biogeochemistry of vertebrate excretion. Bulletin of the American Museum of Natural History 96, 1-554.

Jarvis, M. J. F. (1970). Interactions between man and the South African gannet Sula capensis. Ostrich, Supplement 8, 497-513.

Jones, R. and Hislop, J. R. G. (1978). Changes in North Sea haddock and whiting. Rapports et Proc2s- Verbaux des Rkunions. Conseil Internationalpour I’Exploration de la Mer 172, 58-71.

Jordan, R. (1964). Las emigraciones y mortandad de aves guaneras en el Otono e invierno de 1963. Institute de Investigaciones de los Recursos Marinos Znforme 27,

Jordan, R. (1967). The predation of guano birds on the Peruvian anchovy (Engraulis ringens Jenyns). Report of California Cooperative Oceanic Fisheries Znvestigatiom

Jordan, R. and Fuentes, H. (1966). Las poblaciones de aves guaneras y su situation actual. Znstituto del Mar de Peru 10, 1-31.

Kaftanovski, Yu, M. (1951). Birds of the murre group of the eastern Atlantic. Studies of the fauna and flora of the USSR. Moscow Society of Naturalists 28,

Kendeigh, S . C. (1970). Energy requirements for existence in relation to size of bird. Condor 72, 60-65.

Kendeigh, S. C., Dol’nik, V. R. and Gavrilov, V. M. (1977). Avian energetics. In “Granivorous Birds in Ecosystems” (J. Pinowski and S. C. Kendeigh, eds), International Biological Programme Vol. 12, pp. 127402. Cambridge University Press, Cambridge.

King, J. R. (1973). Energetics of reproduction in birds. In “Breeding Biology of Birds” (D. S. Farner, ed.), pp. 78-107. National Academy of Science, Washington.

King, J. R. (1974). Seasonal allocation of time and energy resources in birds. In “Avian Energetics” (R. A. Paynter, ed.), Publications of the Nuttall Ornithological Club Number 15, pp. 4-85. Cambridge, Massachusetts.

1-31.

11,105-109.

1-170.


King. J. R. and Farner, D. S. (1961). Energy metabolism, thermoregulation and body temperature. In “Biology and Comparative Physiology of Birds” (A. J. Marshall, ed.), Vol. 2, pp. 215-288. Academic Press, New York and London.

Krogh, A. (1916). “The Respiratory Exchange of Animals and Man.” Longmans, London.

Lack, D. (1934). Habitat distribution in certain Icelandic birds. Journal of Animal

Lack, D. (1954). “The Natural Regulation of Animal Numbers.” Oxford University

Lack, D. (1966). “Population Studies of Birds.” Oxford University Press, Oxford. Laws, R. M. (1960). Problems of whale conservation. Transactions of the North

American Wildlife Conference 25, 304-3 19. Laws, R. M. (1961). Reproduction, growth and age of Southern Hemisphere fin

whales. Discovery Reports 31, 327-485. Laws, R. M. (1962). Some effects of whaling on the southern stocks of baleen

whales. In “The Exploitation of Natural Animal Populations” (E. D. LeCren and M. W. Holdgate, eds), pp. 137-158. Blackwell, Oxford.

Laws, R. M. (1973). Population increase of fur seals at South Georgia. Polar Record 16,856858.

Laws, R. M. (1977). The significance of vertebrates in the Antarctic marineecosystem. In “Adaptations within Antarctic Ecosystems” (G. A. Llano, ed.), pp. 41 1-438. Smithsonian Institution, Washington D.C.

Lawton, J. H. (1970). Feeding and food energy assimilation in larvae of the damselfly Pyrrhosoma nymphula (Sulz.) (Odonata: Zygoptera). Journal of Animal Ecology

Levins, R. (1966). The strategy of model building in population biology. American Scientist 54, 421-431.

Lifson, N. and McClintock, R. (1966). Theory of use of the turnover rates of body water for measuring energy and material balance. Journal of Theoretical Biology 12,4674.

Lloyd, C. S., Bibby, C. J. and Everett, M. J. (1975). Breeding terns in Britain and Ireland, 1969-74. British Birds 68, 221-237.

Lockwood, W. B. (1954). Linguistic notes on “Fulmar.” British Birds 47, 336-339. Lockyer, C. H. (1972). The age of sexual maturity of the southern fin whale

(Balaenoptera physalus) using annual layer counts in the ear plug. Journal du Conseil Permanent International pour I’Exploration de la Mer 34, 21iG294.

Lockyer, C. H. (1976). Growth and energy budgets of large baleen whales from the Southern Hemisphere. A.C.M.R.R. Scientific Consultation on Marine Mammals. F.A.O., Rome.

Ecology 3, 81-90.

Press, Oxford.

39,669-689.

Low, G. (1879). “A tour through Orkney and Schetland in 1774.” Kirkwall. Lundbeck, J. (1959). Biologisch-statistische Untersuchungen iiber die deutsche

Hochseefischerei IV,4 : Leistungsfahigkeit und Fangertrage der deutschen Fishdampferflotte 1885-1955. Berichte der Deutschen Wissenschaftlichen Kommis- sion fur Meeresforschung 15, 159-237.

Lundbeck, J. (1960). Mittlere Reiseertrage deutscher Fischdampfer 1887-1955 und Berechnung vergleichbarer Einheitsertrage. Mitteilungen Institut fur Seefischerei 10, 1-20.

Lundbeck, J. (1962). Biologisch-statistische Untersuchungen iiber die deutsche Hochseefischerei IV,5 : Die Dampfefischerei in der Nordsee. Berichte der Deutschen Wissenschaflichen Kommission fur Meeresforschung 16, 177-246.

304 R. W. FURNESS

MacArthur, R. H. and Levins, R. (1964). Competition, habitat selection and character displacement in a patchy environment. Proceedings of the National Academy of Sciences of the United States 51, 1207-1210.

MacArthur, R. H. and Wilson, E. 0. (1967). “The Theory of Island Biogeography.” Princeton University Press, Princeton.

MacKenzie, N. (1905). Notes on the birds of St Kilda. Annals of Scottish Natural History (1905), 75-80 and 141-153.

Mackintosh, N. A. (1942). The southern stocks of whalebone whales. Discovery Re- ports 22, 197-300.

Mackintosh, N. A. (1973). Distribution of postlarval krill in the Antarctic. Discovery Reports 36, 95-156.

Madsen, F. J. and Sparck, R. (1950). On the feedinghabit soft hesoutherncormorant (Phalacrocorax carbo sinensis Shaw). Danish Review of Game Biology 1,45-76.

Manuwal, D. A. (1972). The population ecology of the Cassin’s Auklet on Southeast Farallon Island, California. PhD thesis, University of California, Los Angeles.

Matthews, J. P. (1961). The pilchard of South West Africa, Sardinops ocellata and the maasbanker Trachurus trachurus. Bird predators, 1957-1958. Investiga- tional Report Marine Research Laboratory Administration of South West Africa

May, R. M., Beddington, J. R., Clark, C. W., Holt, S. J. and Laws, R. M. (1979). Management of multispecies fisheries. Science 205, 267-277.

Menzel, D. W., Ryther, J. H., Hulbert, E. M., Lorenzen, C. J. and Corwin, N. (1971). Production and utilisation of organic matter in Peru coastal current. Investigacion Pesquera, Spain 35, 43-59.

Morowitz, H. J. (1968). “Energy Flow in Biology.” Academic Press, London and New York.

Mougin, J. L. and Prevost, J. (1980). Evolution annuelle des effectifs et des biomasses des oiseaux Antarctiques. Revue d’Ecologie (Terre et la Vie) 34, 101-133.

Mudge, G. P. (1979). The cliff breeding seabirds of east Caithness in 1977. Scottish Birds 10, ?47-261.

Nelson, J. B. (1978). “The Sulidae: Gannets and Boobies.” Oxford University Press Oxford.

Nemoto, T. and Nasu, K. (1975). Present status of exploitation and biology of

3, 1-35.

krill in the Antarctic. Oceanology International Conference Papers, Brighton, 353-360.

Nettleship, D. N. (1972). Breeding success of the common puffin (Fratercula arctica L.) on different habitats at Great Island, Newfoundland. Ecological Monographs 42, 239-268.

Newman, G. G. (1970). Stock assessment of the pilchard Sardinops ocellata at Walvis Bay, South West Africa. Investigational Report, Division of Fisheries Union of South Africa 85, 1-13.

Newman, G. G., Crawford, R. J. M. and Centurier-Harris, 0. M. (1978). The effect of vessel characteristics and fishing aids on the fishing power of South African purse-seiners in ICSEAF Division 1.6. Collected Scientific Papers, International Commission for South East Atlantic Fisheries 5, 123-144.

Odum, E. P. (1961). Excretion rate of radioisotopes as indices of metabolic rates in nature: biological half life of zinc-65 in relation to temperature, food wnsump- tion, growth and reproduction in arthropods. Biological Bulletin 121, 371-372

Orians, G. H. and Pearson, N. E. (1979). On the theory of central place foraging. In “Analysis of Ecological Systems” (D. J. Horn, G. R. Stairs and R. D. Mitchell, eds), pp. 155-177. Ohio State University Press, Ohio.

COMPETITlON BETWEEfl FISHERIES AND SEABIRDS 305

Pearson, T. H. (1968). The feeding biology of sea-bird species breeding on the Farne Islands, Northumberland. Journal of Animal Ecology 37, 521-552.

Potts, G. R. (1969). The influence of eruptive movements, age, population size and other factors on the survival of the shag (Phalacrocorax aristotelis (L.)). Journal of Animal Ecology 38, 53-102.

Prange, H. D. and Schmidt-Nielsen, K. (1970). The metabolic cost of swimming in ducks. Journal of Experimental Biology 53, 763-777

Prevost, J. (unpublished). Population, biomass and energy requirements of Antarctic birds: attempted synthesis.

Procter, D. L. C. (1975). The problem of chick loss in the South Polar skua Cath- aracta maccormicki. Ibis 117, 452-459.

Rand, R. W. (1959). The biology of guano producing seabirds. The distribution, abundance and feeding habits of the Cape gannet, Morus capensis, off the southwestern coast of the Cape Province. Investigational Report, Division of Fisheries Union of South Africa 39, 1-36.

Rand, R. W. (1963). The biology of guano producing seabirds. 4. Compostion of colonies on the Cape Islands. Investigational Report, Division of Fisheries Union of South Africa 43, 1-32.

Richards, J., Armstrong, D. W., Hislop, J. R. G., Jermyn, A. S. and Nicholson, M. D. (1978). Trends in Scottish research vessel catches of various species in the North Sea, 1922-1971. Rapports et Proc&Verbaux des Riunions. Conseil International pour 1’Exploration de la Mer 172, 21 1-224.

Robertson, I. (1972). Studies on fish eating birds and their influence on stocks of the Pacific herring in the Gulf Islands of British Columbia. Herring Investiga- tions: Pacific Biology Station, Nanaimo British Columbia.

Robinson, M. K. (1965). Climatic implications derived from the comparison of bathythermograph (BT) data with two types of historic and modern sea surface data. California Cooperative Oceanic Fisheries Investigation Progress Reports

Ryther, J. H. (1969). Relationship of photosynthesis to fish production in the sea. Science 166, 72-76.

Salomonsen, F. (1955). The food production of the sea and the annual cycle of Faeroese marine birds. Oikos 6, 92-100.

Salomonsen, F. (1 965). The geographical variation of the fulmar (Fulmarus glacialis) and the zones of marine environment in the North Atlantic. Auk 82, 327-355.

Santander, H. (1980). The Peru current system. 2: Biological aspects. In “Proceedings of the Workshop on the Phenomenon known as ‘El Niiio’” pp. 217-227. UNESCO, Paris.

Schaefer, M. B. (1954). Some aspects of the dynamics of populationsimportant to the management of commercial marine fisheries. Inter-American Tropical Tuna Commission Bulletin 1, 27-56.

Schaefer, M. B. (1967). Dynamics of the fishery for the anchoveta, Engraulis ringens, off Peru. Boletin Instituto del Mar del Peru 1, 189-304.

Schaefer, M. B. (1970). Men, birds and anchovies in the Peru current-dynamic interactions. Transactions of the American Fisheries Society 9,461-467.

Schreiber, R. W. and Lawrence, J. M. (1976). Organic material and calories in laughing gull eggs. Auk 93,4652.

Schweigger, E. H. (1940). Studies of the Peru coastal current with reference to the extraordinary summer of 1939. Proceedings of the Sixth Pacific Science Congress

10, 141-152.

3, 177-197.

306 R. W. FURNESS

Sherman, K., Jones, C., Sullivan, L., Smith, W., Berrien, P. and Ejsymont, L. (1981). Congruent shifts in sand eel abundance in western and eastern North Atlantic ecosystems. Nature, London 291,486489.

Siegfried, W. R. and Crawford, R. J. M. (1978). Jackass penguins, eggs and guano: diminishing resources at Dassen Island. South African Journal of Science 74,

Siegfried, W. R., Frost, P. G. H., Kinahan, J. B. and Cooper, J. (1975). Social behaviour of Jackass penguins at sea. Zoological Africana 10, 87-100.

Sinclair, J C (1978). Birds of a trawling voyage. Bokmakierie 30, 12-16. Skokova, N. N. (1962). A quantitative study of the diet of fish-eating birds (in

Russian). Ornithologia 4,288-296. Smith, F. E. (1970). Analysis of temperate forest ecosystems. In “Analysis of

Ecosystems” (D. Reichle, ed.), pp. 7-1 8. Springer-Verlag, New York. Spaans, A. L. (1971). On the feeding ecology of the herring gull, Larus argentatus

Pont., in the northern part of the Netherlands. Ardea 59, 73-188. Stander, G. H. and LeRoux, P. J. (1968). Notes of fluctuations of the commercial

catch of the South African pilchard (Sardinops ocellata) 1950-1965. Investigational Report, Division of Fisheries Union of Sourh Africa 65, 1-14.

Steele, J. H. (1974). “The Structure of Marine Ecosystems.” Yale University Press, New haven.

Stonehouse, B. (1967). The general biology and thermal balance of penguins. Advances in Ecological Research 4, 131-196.

Swartz, L. G. (1966). Sea-cliff birds. In “Environment of the Cape Thomson Region, Alaska” (N. J. Wilimovsky and J. N. Wolf, eds), pp. 611-678. United States Atomic Energy Commission, Oak Ridge.

Taylor, F. H. C . (1964). Life history and present status of British Columbian herring stocks. Bulletin of the Fisheries Research Board of Canada 143, 1-81.

Tovar, H. (1974). Censos Graficos de Aves Guaneras para 10s Ciclos Reproductivos de 1969/70 a 1973/74. Informe Internal Instituto del Mar del Peru.

Tuck, L. M. and Squires, H. J. (1955). Food and feeding habits of Briinnich’s Murre (Uria lomvia lomvia) on Akpatok Island. Journal of the Fisheries Research Board of Canada 12,781-792.

TurEek, F. J. (1966). On plumage quantity in birds. Ekologia Polska Ser. A 14,

Uspenski, S. M. (1956). The bird Bazaars of Novaya Zemlya. Translations of Russian Game Reports Number 4. Queen’s Printer, Ottawa.

Valdivia, J. E. (1978). The anchoveta and El Nifio. Rapports et ProcPs-Verbaux des RCunions. Conseil International pour Z’Exploration de la Mer 173, 196-202.

Vesin, J. P., Leggett, W. C. and Able, K. W. (1981). Feeding ecology of capelin (Mullorus villosus) in the estuary and western gulf of St Lawrence and its multispecies implications. Canadian Journal of Fisheries and Aquatic Sciences 257-267.

Vogt, W. (1942). Aves guaneras. Boletin de la Cornpania administradora delguano 18,

Wiens, J. A. and Dyer, M. I. (1977). Assessing the potential impact of granivorous birds in ecosystems. Zn “Granivorous Birds in Ecosystems” (J. Pinowski and S. C. Kendeigh, eds), International Biological Programme Vol. 12, pp. 205-266. Cambridge University Press, Cambridge.

Wiens, J. A. and Innis, G. S. (1974). Estimation of energy flow in bird communities: a population bioenergetics model. Ecology 55, 730-746.

389-390.

61 7-634.

3-132.


Wiens, J. A. and Scott, J. M. (1975). Model estimation of energy flow in Oregon coastal seabird populations. Condor 77, 439-452.

Williams, A. J. (1974). Site preferences and interspecific competition among guillemots Uria aalge (L.) and Uria lomvia (L.) on Bear Island. Ornis Scandinavica 5 ,

Williams, A. J., Siegfried, W. R., Burger, A. E. and Berruti, A. (1979). The Prince Edward Islands: a sanctuary for seabirds in the Southern Ocean. Biological Comervation 15, 59-71.

Wynne-Edwards, V. C. (1962). “Animal Dispersion in Relation to Social Behaviour.” Oliver and Boyd, Edinburgh.

Young, E. C. (1963). The breeding behaviour of the South Polar Skua. Ibis 105,

1 13-1 21.

203-233.

[advances in marine biology] advances in marine biology volume 20 volume 20 || competition between...

Documents