Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
I B I S 136: 397-41 1 ~ THE FIRST ALFRED NEWTON LECTURE Presented at the “Bird Conservation in Action” conference, April 1994 Experiments on the limitation of bird breeding densities: a review I . NEWTON Institute of Terrestrial Ecology, Monks Wood, Abbots Ripton, Huntingdon, Cambridgeshire P E l 7 2LS, UK The breeding densities of birds could be limited by resources, such as food and nest sites, or they could be held at a lower level by natural enemies, such as predators and parasites. In this paper, I review the experimental evidence for each of these limiting factors affecting bird breeding densities. Of 18 experiments involving winter food provision (mostly on tits, Paridae). 11 led to increased breeding densities compared with control areas. Of four involving summer food depletion (all on forest insectivores),none led to decreased breeding densities. In experiments with Red Grouse Lagopus 1. scoticus, fertilizing areas of heather moor led to increased densities during a period of population increase but did not prevent a later decline. Of 32 studies on tree-cavity nesters, the provision of nestboxes led to increased breeding density in 30 (95%)studies, each involving one or more species of hole nesters. Of 15 experiments involving predator removal (mostly on ducks and gamebirds), at least 14led to increased hatching success, four out of eight led to increased post-breeding numbers, and six out of 11 led to increased breeding density. In one experiment, the removal of strongyle parasites from a Red Grouse population prevented a cyclic decline on five out of five occasions. Taken together, these experiments confirmed that all main potential external limiting factors have affected breeding density in one bird species or another. They also confirmed that the same species has been limited by one factor in certain areas or years and by another factor in different areas or years. contents. But the number alive at the start of the next breeding season might depend largely on overwinter survival, in turn dependent on food supply. In this case, increasing the winter food supply would have more intluence on breeding numbers than would decreasing the predation on nest contents. This paper is primarily about the limitation of breeding densities for two reasons. First, it is in the early stages of breeding each year that bird numbers reach their annual low, and it is upon the breeders that future additions to the population usually depend. Second, it is in the breeding season, when birds are most conspicuous and tied to fixed sites, that they can be counted most readily. In fact, most experiments have been concerned with breeding numbers. In addition to breeders, however, some bird populations may contain a large nonbreeding contingent, which may or may not be countable. No one is likely to dispute the fact that food supply or other resources could provide a ceiling to the numbers of any bird. The key question is, in practice, which species are limited by resources and which are held at a level lower than resources would permit by other factors, notably nat- As ever more land falls to intensive human use, many bird species are increasingly constrained in distribution and abundance by the progressive destruction and degradation of their habitats. For effective conservation, better understanding is needed of the factors that limit numbers within areas of remaining habitat. Only then can such areas be managed so as to sustain large populationsof desired species. In this paper, I shall be concerned with the factors that limit bird numbers within the available habitat, concentrating on the experimental evidence. The best we can hope to show by experiment is that some particular factor limits bird density at a particular place and time. It is necessary to define the area carefully, for while density may be limited in this area, some individuals may move out and survive elsewhere, so that total numbers are unaffected. The emigrants may or may not return at a later date. It is important to specify the time period, because bird numbers may be limited by different factors at different times and only the last-acting may be crucial in setting the eventual population level. For example, the number of birds present at the end of a breeding season might depend largely on breeding success, in turn dependent on predation of nest 397 398 I. NEWTON Model of limitation of breeding density Surplus Resource limlt Deficit I Non-breeding period I Breeding period Figure 1. Model showing seasonal changes in bird numbers in relation to the carrying capacity of the breeding habitat (thick line). In the lower curve (l), overwinter losses are severe and reduce numbers well below the level that the breeding habitat would support: in (2) overwinter losses are such that the numbers left in spring match those that the breeding habitat will support and in ( 3 ) overwinter losses are slight, leaving more birds than the breeding habitat will support and giving rise to a surplus of nonbreeders. In (1). breeding density is largely determined by overwinter losses, in (2) by both overwinter losses and the carrying capacity of the breeding habitat and in ( 3 ) by the carrying capacity of the breeding habitat. ural enemies such as predators, parasites or pathogens or human agency. Effective conservation of any species depends on knowing the limiting factors, for it is these external factors which must be altered if an increase in population is to be achieved. Any factor might be considered limiting if it prevents at population increase or causes a decline. In reality, no one factor is likely to account wholly for a given population level. During a period of food shortage, for example, some individuals may starve, while other nonstarving individuals may die from other causes such as predation. In such cases, the main limiting factor can be considered as the one which, once removed, will permit the greatest rise in numbers. Because populations may be influenced by many different factors, acting individually or in combination, it is often hard to tell the relative importance of each, except by experiment. Moreover, the crucial factor cannot always be deduced from knowledge of mortality causes, even if such information were available. Imagine that the density of a territorial species was limited by some aspect of habitat quality, so that surplus individuals were forced into suboptimal habitat where they were eaten by predators (the ‘doomed surplus’ model of Errington [1946]). From a study of mortality, one would conclude that predators limited numbers because virtually all the mortality occurred through predation: however, the underlying limiting factor was habitat quality, which influenced the density of territories. To change breeding density in the long term would entail a change of habitat, not of predators. and in each case surplus individuals would be removed by predators or any other mortality agents available locally. The important point is that the factors that cause most mortality in a bird population are not necessarily those IBIS 136 which ultimately determine the population level. In any case, assessing the causes of mortality in a bird population is itself not straightforward. A bird weakened by food shortage may succumb to disease, but just before its death it may fall victim to a predator. For this bird, food shortage is the underlying (ultimate) cause of death, while disease or predation is the immediate (proximate) cause. The importance of experiments in clarifying the picture is obvious. My aim in this paper is to review the field experiments that have been done to test the effects on bird breeding densities of various external limiting factors, namely food supply. nest sites, predators and parasites. I have excluded experiments that have examined the role of territorial or other social interactions in limiting density, which have been reviewed elsewhere (Newton 1992), and also experiments on the role of food supply in influencing laying dates and clutch sizes. EXPERIMENTAL DESIGN Most experiments on population limitation in birds fit within the same conceptual framework (Fig. 1).This is true whether the population is resident, with individuals remaining in the same area year-round, or migratory, with individuals spending the nonbreeding and breeding periods in different areas. Numbers are highest at the end of one breeding season and then decline to the start of the next. But within the nesting habitat. breeding density can be limited, as birds compete for territorial space or nest sites. Three scenarios can be envisaged: (1) In some populations, overwinter losses-from whatever limiting factors operate in winter-may be so great that, by the start of breeding, the remaining birds are too few to occupy the nesting habitat fully, so that practically all individuals of appropriate age and condition could breed. In this case, breeding density is limited by whatever factors act to reduce numbers in winter, and manipulation of these winter factors would be needed to produce an increase in breeding numbers. (2) In other populations, overwinter losses may reduce numbers to more or less the level that the nesting habitat will support, so that again almost all potential breeders left at the end of winter could breed. In this case, manipulation of both the factors influencing winter losses and the carrying capacity of the nesting habitat would be needed to produce an increase in breeding numbers. ( 3 ) In yet other populations, overwinter losses may be so light that, after the nesting habitat is fully occupied, a surplus of potential breeders is left over which move on to breed elsewhere or form a nonbreeding component. In this case, manipulation of the factors influencing the carrying capacity of the nesting habitat would be needed to permit a rise in breeding numbers. Although, in this model, I have viewed the carrying capacity of the breeding habitat at the time of settlement as 1994 399 LIMITATION OF BIRD BREEDING DENSITIES Table 1. EKects of winterfood provision on the breeding density of various birds Species Great Tit Parus major Blue Tit Parus caeruleus Willow Tit Parus montanus Crested Tit Parus cristatus Coal Tit Parus ater Black-capped Chickadee Parus atricapillus Tufted Titmouse Parus bicolor Nuthatch Sitta europaea Song Sparrow Melospiza melodia Carrion Crow Corvus corone Red Grouse Lagopus 1. scoticus Black Grouse Tetrao tetrix Region Grade’ (years) Increase in density Reference Germany Finland England Netherlands Sweden Germany England Sweden Sweden 3 (1) 1(3) 2 (1) 2 (5) 3 (2) 3 (1) 2 (1) 3 (2) 3 (1) Yes No Yes Berndt et al. 1964 Haartman 1973 Krebs 1971 van Balen 1980 Kallander 1981 Berndt & Frantzen 1964 Krebs 1971 Kallander 1981 Jansson et aZ. 1981 Sweden 3 (1) Yes Jansson et al. 1981 Scotland 2 (1) No Pennsylvania 10) Yes A.J. Deadman (unpubl. PhD thesis, University of Aberdeen. 1973) Samson & Lewis 1979 Pennsylvania 1 (1) No Samson & Lewis 1979 Sweden 3 (2) Yes Enoksson & Nilsson 1983 British Columbia 2 (1) Yes Smith et al. 1980 Scotland 1(1) No Yom-Tov 1974 Scotland 2 (1) YesZ Watson & Moss 1979 Sweden 3 (4) No Willebrand 1988 Yes No Yes Yes No Yes * See Table 5. In certain conditions only, see text. fixed, for some species this is probably an oversimplification. The numbers of birds that can settle in a given area of nesting habitat may be variable within limits, depending on the number of potential settlers available, with more individuals ‘squeezing in’ when the number of contenders is high. In this case, the numbers of birds left over at the end of winter could have some influence on breeding density, even in the third situation described above. In practice this means that under severe competition for space, some individuals might accept smaller territories or inferior nest sites, facilitating higher density (Davies 1978. Newton 1992). As another modification, numbers may fit the carrying capacity of the nesting habitat at the start of breeding but then fall below it, as a result of mortality or as carrying capacity rises with the seasonal improvement in conditions. Eventually, the production of young ensures a rise in density to give another, post-breeding peak. Finally. any given population may fit one of the patterns described above in some years and areas and a different pattern in other years or areas. In some of the experiments reported below, the experimental treatment covered the whole year, while in others it covered the nonbreeding period only or the breeding pe- riod only. The commonest type of experiment involved a single area in which breeding numbers were monitored for some years, then the treatment was applied and numbers were monitored for several further years, giving a simple before-and-after comparison (grade 1 in Tables 1-5). The weakness of this procedure is that. without a control, one cannot be certain that a change in breeding numbers was due to the treatment and not to some other unknown factor which changed at the same time. The second type of experiment involved two areas, with the treatment applied in one area while the other area served as a control (grade 2). If breeding numbers changed more in the experimental than in the control area, this was taken as evidence that the treatment influenced numbers. Ideally, the two areas should be far enough apart that treatment in the experimental area does not affect birds in the control area, which, if the areas were close, could move back and forth. The third type of experiment involved some replication, either by reversing the treatments, so that after a time the experimental area became the control and vice versa, or by the simultaneous use of several experimental and several control areas (grade 3 ) . Such replication strengthens the findings because it in- $00 IBIS I . NEWTON 136 Table 2. Limitation of breeding density in birds nesting in tree cavities as shown by the experimental addition or subtraction of nest sites. Most studies involved simple before-and-after comparisons, but those marked included control areas and those marked **alsoincluded replication Species Area Nestboxes added, leading to an increase in breeding activity Starling New Zealand Sturnus vulgaris House Sparrow New Zealand Passer domesticus Tree Sparrow England Passer montanus Tree Sparrow Germany Germany Tree sparrow Germany Tree Sparrow Arizona Western Bluebird Sialia rnexicana" Mountain Bluebird Sialia currucoides Mountain Bluebird Eastern Bluebird Sialia sialis Rcdstart Phoenicurus phoenicurus Pygmy Nuthatch Sitta pygmaea** Mountain Chickadee Parus gambeli Great Tit Parus major Varied Tit Parus varius Blue Tit Parus cueruleus* Pied flycatcher Ficedula hypoleuca Pied Flycatcher Habitat Reference Farmland Coleman 1974 Farmland Coleman 1974 Farmland Boyd 1932 Creutz 1949 Dornbusch 1973 Gauhll984 Brawn & Balda 1988 Manitoba Farmland Pine plantation Farmland Ponderosa pine forest, lightly managed Parkland and meadows Manitoba Manitoba Mainly farmland Mainly farmland Miller et al. 1970 Miller et al. 1970 Finland Birch forest, lightly managed Pine forest, lightly managed rarvinen 1978 Arizona California Cutforth 1968 Brawn & Balda 1988 Conifer forest, lightly managed Second-growth forest Dahlsten & Copper 1979 Japan Second-growth forest Higuchi 1978 Britain East & Perrins 1988 Higuchi 1978 Pied Flycatcher* Northern Sweden Pied Flycatcher Southern Sweden Pied Flycatcher Southern Sweden Collared Flycatcher Ficedula albicollis** Violet-green Swallow Tachycineta thalassina** Tree Swallow Iridoprocne bicolor Sweden Ontario Deciduous woodland, unmanaged Deciduous woodland, managed Birch forest, lightly managed Subalpine Birch forest, natural Broad-leaved woodland, managed Broad-leaved woodland, managed Deciduous woodland. managed Pine forest, lightly managed Mixed. rural Purplc Martin Progne subis Ash-throated Flycatcher Myiarchus cineraseens* Green-mmped Parrotlet Forpus passerinus' European Kestrel Falco tinnunculus European Kestrel Ontario Mixed, rural Holroyd 1975 Arizona Brush 1983 Venezuela Riparian, lightly managed Farmland Netherlands Grassland Cavk 1968 Britain Grassland Village 1983 Britain Finland Arizona Currie & Bamford 1982 Jarvinen 1978 Enemar & Sjostrand 1972 Enemar & Sjostrand 1972 Enemar & Sjostrand 1972 Gustafsson 1988 Brawn & Balda 1988 Holroyd 1975 Beissinger & Bucher 1992 1994 401 LIMITATION OF BIRD BREEDING DENSITIES Table 2. Continued Species Area Habitat ~ ~ American Kestrel Wisconsin Farmland Ealco sparverius Goldeneye Scotland Forest and lakes, managed Bucephala clangula Goldeneye Sweden Forest and lakes Goldeneye Finland Forest and lakes Goldeneye Canada Forest and lakes Barrow’s Goldeneye British Columbia Forest and lakes Bucephala islandica Wood duck Massachusetts Forest and lakes Aix sponsa Wood Duck New York Forest and lakes Wood Duck California Marsh Nestboxes removed or natural holes blocked. leading to a decline in breeding density Mountain Bluebird California Pine-fir forest, lightly managed Great Tit* Britain Deciduous woodland. unmanaged Ash-throated Flycatcher Arizona Riparian, lightly managed Reference ~~ Hamerstrom et al. 1973 Dennis & Dow 1984 Sibn 1951 Eriksson 1982 Johnson 1967 Savard 1988 McLaughIin & Grice 1952 Haramis & Thompson 1985 Jones & Leopold 1967 Raphael & White 1984 East & Penins 1988 Brush 1983 Note: In addition to the experiments listed above, in two others the addition of nestboxes led to no increase in breeding density: European Kestrel in an area of farmland in England (Viiage 1990) and BuWehead Bucephala albeola in an area of mainly parkland in British Columbia (Gauthier & Smith 1987). In one other experiment in oak-pine woodland in California. the blocking of a proportion of nest holes led to no reduction in the breeding density of songbirds (Waters et al. 1990). creases the likelihood that any response observed in the study species resulted from the treatment and not from some other features peculiar to the area or time period. Statistically, the more the experiment is replicated, the stronger the conclusions. The most substantial feeding experiment, which covered several successive winters, was that by van Balen (1980) in the Netherlands. It was based in two areas of similar woodland 7 km apart, in one of which food was provided while the other acted as a control. Before winter feeding began, EXPERIMENTS INVOLVING FOOD Table 3. Efects orpredator removal on tetraonids on two islands. northern Sweden (from Marcstrom et al. 1988) Circumstantial evidence that bird numbers are influenced by food supply derives mainly from the observation that, within species, changes in bird density from year to year, or from one place to another, often correlate with temporal and spatial changes in food supplies (Newton 1980). Not all bird species are amenable to food provision experiments. Ideally, they should live at high density in the same locality year round, so that the effects of feeding on their subsequent numbers can be assessed readily, and they must eat foods which can be provided readily. These three constraints eliminate most species as easy subjects for experiment, and in any case care must be taken to ensure than any food provided is nutritionally appropriate and adequate. Most published experiments concern tits (Paridae) which were given seeds in winter, and their subsequent breeding numbers in the same area were then measured from the number of nests in nest boxes, supplied in excess. Predators preserved Productivity Brood size in August % of hens with broods Young per hen * 3.3 0.1 59% 1.9 Predators removed 5.5 f 0.1 77% 4.2 Under predator removal, Yo increase in counts of Numbers in breeding season Lek counts Transect counts Capercaillie Tetrao urogallus Black Grouse Tetrao tetriz 174 56 168 80 402 IBIS I. NEWTON Table 4. Efects on avian prey of various predator-removal studies: a summary of experimental findings (B = birds, M measured) = 136 mammals, ? = not Experimental findings for prey Prey species Location Increased Inpost- creased Pred- Grade In- breed- breedof creased ing ing ators re- experi- nest num- nummoved ment’ success bers bers Reference Ruffed Grouse Bonasa umbellus Ruffed Grouse Connecticut Hill, New York Valcour Island. New York MB 3 Yes No No Bump et a/. 1947 MB 1 Yes No No King-necked Pheasant Phasianus colchicus Ring-necked Pheasant Wild Turkey Meleagris gallapavo & Northern Bobwhite Colinus virginianus Grey Partridge Per& perdir Black Grouse Tetrao tetrix & Capercaillie Tetrao urogallus Ptarmigan Lagopus lagopus & Black Grouse Various ducks Various ducks Various ducks Various ducks Simulated ground nests Simulated ground nests White-winged Dove Zenaida asiatica Southern Minnesota MBz 2 Yes’ No No Bump et al. 1947 Crissey & Darrow 1949 Chesness et al. 1968 South Dakota M M 3 2 ? Yes > Texas Yes Yes Yes Trautman et al. 1974 Beasom 1974 England MBL 3 Yes Yes Yes Tapper et a!. 1990. 1991 Sweden M 3 Yes’ Yes3 Yes’ Marcstrom et al. 1988 Sweden B4 2 Yes’ No No Parker 1984 Minnesota South Dakota South Dakota North Dakota Manitoba North Dakota MB M M 3 ? ? ? M6 M M B’ Yes Yes’ Yes Yes’ No 2 2 > ? 2 2 2 Yes’ - __ Yes’ Yes - - ? Yes Baker et al. 1968 Duebbert & Kantrud 1974 Duebbert & Lokemoen 1980 Greenwood 1986 Lynch 1972 Schrank 1972 Blankenship 1966 Texas ’ Sce ‘I’able 5. ’ Excludes raptors, which were left unharmed. 3 Yes Yes ‘ Supported by statistical analysis. t Corvids only. Great-tailed Grackles Cassidix mexicanus only. ’~Striped skunks only. Great Tit Parus major breeding numbers in the two areas were similar and fluctuated more or less in parallel from year to year (Fig. 2). After feeding began, breeding densities in the two areas diverged and no longer fluctuated in parallel. Over several years, breeding numbers in the experimental area averaged 40% higher than those in the control area. The impact of artificial feeding varied between years depending on the Beech Fagus silvatica seed crop, which was the most variable major component in the natural food supply. Comparison of numbers in the experimental area in the years before and after food provision indicated that the extra feeding doubled the population in poor Beech years but made little difference in good Beech years (Fig. 2). The fact that the effects of food provision were most marked in years when natural food was scarce provided further evidence that winter food supplies influenced local breeding densities. On this evidence. the Great Tits in this experiment fitted the lower curve (1)in Figure 1. in that variation in winter conditions were mainly responsible for variations in breeding density. 1994 403 LIMITATION OF BIRD BREEDIYG DENSITIES Table 5. Summary of experiments on the limitation of breeding density in birds Appropriate change in breeding density occurred Grade of experiment' Food provision (winter) Nest-site provision Predator reductionz Parasite reduction Totals 1 2 3 Total (Yo) 4 6 8 18 23 1 0 28 5 6 4 32 0 4 1 17 17 11 1 62 ll(61) 30(95) 6(55) 71 8 8 69 0 0 70 0 0 800 0 . A T . I 0 -1 . 0 0 , , -7 - 8 Temperature ("C) -9 I -2 -3 - 4 l(100) 48(77) . 0 -5 , -6 , , - -10 -11 Figure 3. Numbers of breeding Great Tits in a study area near Turku, Finland, in relation to winter temperature (the average temperature in the months December-February). Open circles denote normal winters and solid circles, winters when much artificial food was provided. From Haartman (1973). Grade 1 = simple before-and-after comparison in the same area: grade 2 = simultaneous comparison between experimental and control area; grade 3 = reversal of treatments between experimental and control areas or replication of experimental and control areas (for further details. see text). 2 A total of 16 experiments included only 11 which examined the effects of predator removal on breeding density. area and year, sometimes one species responded while the other did not (e.g. Krebs 1971). Where a response to food provision occurred, breeding density increased by up to around 100°/ocompared with that in control areas, and where the birds were ringed, the increase in breeding density resulted from some combination of increased local survival and increased immigration, affecting mainly first-year birds (Krebs 1971. vanBalen 1980, Janssonet al. 1981).Similarly, in the Song Sparrow Melospiza melodia on Mandarte Island, western Canada, seed provision in one winter led to improved survival, especially of juveniles, and to increased breeding density (38%) (Smith et al. 1980). In an experiment with Carrion Crows Corvus corone. extra food (hen eggs and chicks) was put out throughout one winter and spring to investigate whether territories would shrink and extra pairs would settle (Yom-Tov 1974). This did not happen, although the food was taken and nest sites were Further north in Europe, Great Tit breeding densities were greatly influenced by the severity of winter, with the lowest densities following the coldest winters (Fig. 3). When food was provided artificially in three winters, the same relationship held with weather but breeding densities were correspondingly higher (Haartman 1973). Other experiments on tits, covering one or two winters, gave variable results. In some areas, two species had access to the supplementary food, and if each of these is counted as a separate experiment, then, compared with control areas, 8 out of 13winter feeding experiments led to obvious increases in the subsequent breeding density (Table 1).The same species responded to food provision in one area but not another, and in the same 180- 180160 120 U g 100L n 0 L 80- 5 40- a n = / 80- ' supplementary feedlng 20- 0 1 1 1 80 62 84 1 86 Year 1 68 . 1 1 70 72 40 20- ° 1 1 1 1 1 1 , 1 10 20 30 40 50 80 70 80 Beech crop index Figure 2. Left: Numbers of Great Tit pairs in two woodland areas in the Netherlands, before and after the provision of extra food in one area (from winter 1966-1967). Right: Numbers in the experimental area in years before and after the provision of extra food, shown in relation to the Beech crop. Redrawn from van Balen (1980). 404 I. NEWTON present in excess. So this experiment went against the idea that food limited breeding density in Carrion Crows in a direct and simple way. However, one might not expect a sudden increase in a bird such as this, in which individuals are long lived and territories normally remain stable for years, despite some fluctuation in natural food supply. The result might have been different if food had been provided over a longer period (and not just in one winter and spring) or if the existing birds had been removed to enable different ones to settle. That food could act at the level of the individual territory in shorter-lived species was shown in an experiment in which Sunflower Helianthus sparsijolius seeds were provided in winter to Wood Nuthatches Sitta europaea, which then took smaller territories than in another winter, so that density was increased (Enoksson & Nilsson 1983). In Red Grouse Lagopus 1. scoticus, burning or fertilizing areas of heather (the food plant) promoted an increase in breeding density 1year later over the previous year, as well as over that in control areas nearby (Miller et al. 1970). This occurred when the experiment was started in years with low or moderate densities, but in a later trial, fertilizing an area failed to halt a big decline (Watson & Moss 1979). The Red Grouse in the area concerned underwent marked fluctuations, and these experiments suggested that the birds could respond to an improved food supply during a period of increase but not during a period of decline. Artificial feeding of Black Grouse Tetrao tetrix in winter started in Finland as a widespread management procedure in an attempt to stop their decline in numbers. The birds were provided with oats at lekking sites throughout the winter. Although the birds took the food, long-term population decline continued, as did shorter-term cycles in numbers. The only obvious effect of the feeding was change in the local distribution of the grouse, which were attracted from wide areas to leks where food was provided (Majakangas 1987). The procedure also was evaluated experimentally in Sweden, where Black Grouse were fed at two leks for 4 years, while those at three other leks in the same area were left as controls. No differences in body-weights or survival were noted between birds at the two types of lek (Willebrand 1988). This last experiment was in an area of abundant natural winter food (Birch), but the conclusion from both studies was that Black Grouse numbers were not limited by winter food. This was in line with other studies which pointed to habitat deterioration as causing long-term decline and to predation as the main year-by-year limiting factor (Angelstam et al. 1984, Marcstrom et al. 1988). Turning to manipulation of summer food supplies, a problem hangs over the experiment of Franzblau & Collins (1980), who placed extra food (Tenebrio larvae) in five territories of Rufous-sided Towhees Pipilo erythrophthalmus montanus part way through the breeding season. No change was detected in the sizes of these territories before and after food provision or in comparison with five control territories where no food was provided. It may have been unreasonable to expect a change in territory size during the course of a breeding season, when the adult population is normally static. IBIS 136 Experiments involving reduction of food supplies caused by the use of insecticides have been tried in summer, as an adjunct to forest-insect control programmes. The chemicals most often used in these programmes include DDT. carbaryl and fenitrothion, applied in a wide range of dosages and spraying regimes. In some studies, no effects of spraying were apparent on local bird breeding densities. while in others the reported declines may have been due more to direct poisoning of birds by the pesticide than to reductions in their food supplies (Pearce & Peakall 1977, Peakall & Bart 1983, Spray et al. 1987). However, in the study by Cooper et al. (1990) in West Virginia, the chemical used was diflubenzuron, a growth regulator which kills insect larvae but is nontoxic to birds. These authors counted the songbirds present in three sprayed 59-ha plots for comparison with those in three similar-sized unsprayed plots. Plots were separated by at least 150 m to minimize spray drift. Spraying was done in early May, and bird counts followed from late May to mid-July. No significant differences in the numbers of 21 common species. or in their combined numbers, were found between treated and untreated plots. Many species were shown to turn to alternative foods on the sprayed plots, and one species (Red-eyed Vireo Vireo olivaceous) studied in detail was found to forage over larger areas than in unsprayed plots. The findings suggested that breeding density was not affected in the same year by reduced summer food supply, but the birds had settled by then and alternative food sources were available. Similar results were obtained in other studies involving the same chemical (Richmond et al. 1979. DeReede 1982. Stribling & Smith 1987). The general conclusion from studies involving diflubenzuron was that breeding densities of insectivorous species. after the time of settlement, were not affected by reductions in the larval insect component of the diet. To summarize, some species in feeding experiments showed clear evidence that breeding numbers were influenced by winter food supplies, at least in some years or areas. Others showed no evidence that they were influenced by food supplies in the years and areas concerned, but in some cases the food may have been provided for too short a period or at the wrong time of year. Reduction of insectivorous food supplies in summer spray programmes caused no reduction in the breeding densities of local songbirds. EXPERIMENTS INVOLVING NEST SITES While food supply could potentially limit the numbers of all birds, in some species breeding density is often held at a level lower than food would permit by shortage of some other resource, notably nest sites. Limitation by nest sites is evident mainly in species which use special sites, such as cliffs or tree cavities. The evidence is partly correlative, in that spatial or temporal variations in the breeding density of such species often parallel spatial or temporal variations in nestsite availability, and breeding pairs are absent from areas which lack nest sites but which seem suitable in other re- 1994 LIMITATION OF BIRD BREEDING DENSITIES spects (nonbreeders may live there). Those cliff nesters, such as some hirundines, which have taken to nesting on buildings have thereby spread over huge areas from which they were formerly absent. Most experiments on nest-site provision have involved species which normally nest in tree cavities but cannot excavate their own holes. Such species readily accept appropriately designed nestboxes. In any experiment involving nest sites, one has to be sure that the acceptance of new sites has allowed extra pairs to nest and has not merely entailed shifts of existing pairs from other sites. The process of density limitation depends on defence of the nest site, or of the territory containing it, by the pair in occupation, so that other pairs are excluded. Experimental demonstrations of nest-site shortages provide some of the clearest examples of resource limitation in birds. They also show the action of successive limiting factors, for once the shortage of nest sites has been rectified, numbers usually increase to a new higher level, at which they may be assumed to be limited by other factors. Some studies have examined the response of the entire cavity-nesting guild to nest-site provision or removal, enabling species differences in the degree of site limitation to be assessed. In Ponderosa Pine Pinus ponderosa forest in Arizona, Brawn & Balda (1988) studied hole-nesting birds on five 8-ha plots of different structure. On three plots they counted all hole-nesting birds for 4 years, then added nestboxes and continued the counts for another 4 years. The other two plots were left as controls: no boxes were provided, but counts were continued throughout (Fig. 2). The degree to which sites were limiting varied between species and between plots, depending partly on the number of dead snags present. Overall breeding densities (all hole nesters combined) increased significantly after box provision on the two plots with fewest natural sites. Increases occurred progressively for 4 years (when the study ended), giving in the fourth year overall densities 230% and 760% higher than previously. On these plots, more than 90% of the hole nesters found were in boxes. Three species responded strongly, so were most clearly limited by nest-site availability before boxes were installed, namely the Violet-green Swallow Tachicinata thalassina. Pygmy Nuthatch Sitta pygmaea and Western Bluebird Sialia rnexicana. The last increased from 20 pairs per km2 before box provision to 78-88 pairs per km2 after box provision. Three other hole nesters did not respond significantly. In these same years, no sigruiicant density changes occurred on the two control plots where no boxes were provided. Similarly. in the third experimental plot, which had the most natural sites, densities remained unchanged, and only 3Oyo of the hole nesters used boxes. On this plot, cavities seemed already surplus to needs, and the occurrence of some box nests was attributed to birds switching from natural sites. Different responses of species to nestbox provision might have depended on the degree to which boxes were acceptable substitutes for natural sites, as well as reflecting different degrees of limitation and competitive abilities. In the presence of limited nest sites, dominant species 405 may obtain enough sites, and subordinate ones increasingly obtain nest sites as more become available and the needs of the dominants are satisfied. The role of competition for nest sites in limiting the breeding densities of other hole nesters has been shown experimentally by Dhondt & Eyckerman (1980).van Balen et al. (1982) and Gustafsson (1988). The results from 32 studies on nestbox provision are summarized in Table 2. Most were simple before-and-after studies, but some included control areas and replication. They included a range of 25 species, from songbirds to ducks and raptors. In 92Yo (30 of 32) of these studies, breeding density of one or more species increased after box provision, suggesting that limitation by shortage of nest sites is widespread among hole-nesting birds. However, many experiments were in young managed woods, deficient in old and dead trees, so may not be typical of all woods. Moreover, studies which did not lead to increased breeding density were perhaps less likely to be published. It is uncertain, therefore, how typical of forest in general these findings might be. Nonetheless, increases of two- to four-fold in breeding density were observed commonly and increases of 5-20-fold occasionally. As expected, the increases were more marked in species that defended little more than the nest site, such as Pied Flycatcher Ficedula hypoleuca and Tree Sparrow Passer montanus, than in species such as tits, which held large territories. In Finland, Haartman (1971) recorded that following box provision Pied Flycatchers could increase from virtual absence to densities of 2000 pairs per km2, greater than all other species together in those areas. As indicated above, nest sites are not always limiting for hole-nesting birds, especially in mature forest with abundant natural sites. In an oak-pine forest in California, Waters et al. (11990) found 176 cavities on a 37-ha plot, a minimum estimate of the total present. In 1 year they blocked 67 of these cavities and in another year 106, yet in neither year did the density of hole-nesting birds (or nests found) decline more than in a nearby control plot. Evidently cavities were surplus to needs in this forest. Several studies (notably on Pied Flycatcher) showed that numbers increased abruptly the year after boxes were installed, implying that large numbers of potential occupants were generally available. Without the boxes, these birds presumably would not have bred but would have remained as part of a large nonbreeding contingent unable to breed because of shortage of sites. In these species, limitation of breeding density followed scenario (3) in Figure 1, because in any 1 year more potential breeders had survived the winter than could be supported by the nesting habitat. Other studies showed that, following box provision, numbers increased slowly over several years (as in the Brawn & Balda [1988] study), implying that no more than a small surplus was present in any 1year and that reproduction or continued immigration contributed to the longer-term increase. Whatever the species and its rate of increase, numbers eventually levelled off, regardless of the number of extra boxes provided. The implication was that, once the shortage of nest sites was rectified, another factor took over to limit 4 06 I. NEWTON numbers at a higher level. Evidence for various tits and raptors showed that this second factor was food supply (Newton 1979, Dhondt & Eyckerman 1980).In such species, then, densities in different areas were limited by either nest sites or food, whichever was in shortest supply. Hole nesters are not the only birds that have responded to the experimental provision of nest sites. Population increases of other species have been noted following the provision of baskets or artificial stick nests in trees (Viage 1990). rafts (as island substitutes, Crawford & Shelton [1978]). pylons (as tree substitutes, Steerihoffet at. [1993]) and buildings (as cliff substitutes, Newton 119791). It seems that shortage of nest sites can limit the distributions and breeding densities of a wide range of bird species. EXPERIMENTS INVOLVING PREDATORS Observational evidence suggests that predators can have the following effects on bird breeding densities. depending on circumstances (Newton 1993): (1) No obvious reduction in breeding density. This is apparent in some species which suffer considerable predation but are demonstrably limited by other factors (examples include the nestbox species just discussed). It is also apparent in some large species (such as certain swans) which, in present conditions, are virtually immune to predation and often have a large nonbreeding surplus of mature adults. (2) Hold breeding density at an approximate equilibrium well below the level that resources would permit (as in the Grey Partridge Perdix perdix in some areas, Potts [1980]). ( 3 ) Cause marked fluctuations in breeding density, either irregular or regular (cyclic),as suggested for some northern gallinaceous birds exposed to periodic heavy predation (Lack 1954, Keith 1963, Angelstam et ul. 1984, Keith & Rusch 1988). (4) Cause decline to extinction, as in some birds of oceanic islands exposed to introduced rats and cats (Atkinson 1985). Predator-removal experiments have not been concerned in distinguishing these different types of impact, only in testing whether an effect was apparent, in other words, in distinguishing (1)from the rest. Most concerned ground-nesting game birds and ducks, stimulated by the commercial interest of hunting. One of the best predator-removal studies was conducted on two forested islands in the Baltic Sea off northern Sweden (Marcstrom Pt al. 1988). These islands were large enough to sustain populations of several mammalian predators but also were joined to the mainland in winter by sea ice, enabling predators to move freely on and off. Mammalian predators (mainly Fox Vulpes vulpes and Marten Martes martes) were removed from one island but were left on the other. After 5 years, the treatments were reversed for four further years, so that the experimental island then became the control and vice versa, giving 9 years of study in all. The effects of predator removal were measured on four species of grouse, mainly Capercaillie Tetruo urogallus and Black Grouse. Where IBIS 136 predators were removed, more young were produced and subsequent breeding numbers were higher than where predators were left (Table 3). The conclusion was that mammalian predation limited the breeding density of the game species concerned. On these islands, as on the mainland. the mammalian predators fed mainly on rodents and only secondarily on game birds. It was observed that predation on game birds was reduced and post-breeding populations were higher in years when voles were numerous. Where predators were left alone, grouse breeding success was correlated with vole abundance, as most young grouse were produced in the peak vole years, but no such relationship occurred where predators were removed. This was consistent with the view that predators turned more to grouse when voles were scarce and confirmed that predation was mainly responsible for synchronizing grouse productivity with vole abundance. as suggested by observational data (Angelstam et aI. 1984). However, the removal of Foxes and Martens had no significant effect on vole abundance itself during two 4-year cycles, suggesting that these predators did not drive the vole cycle. The findings from a total of 1 5 predator-removal studies involving various avian prey species are summarized in Table 4. Eight studies involved gallinaceous birds, six involved ducks (including two with simulated nests) and one involved doves. The parameters that were measured most commonly included nest success, post-breeding numbers (or ratio of large young to adults) and subsequent breeding numbers. In 14 studies in which nest success was measured. all showed an increase under predator removal; in eight stuhes in which post-breeding numbers were measured, four showed an increase and of 11 studies in which breeding numbers were measured, six showed an increase. Improved nest success was not always reflected in increased post-breeding numbers, and. similarly, an increase in post-breeding numbers was not always reHected in increased subsequent breeding numbers. Overall. it seems that, in about half the studies in Table 4, breeding density was limited by predation, even though all species experienced heavy predation at some stage of their lives. These studies were discussed in greater detail by Newton (1993). They conform to the first (lower line) scenario in Figure 1. Almost all the species studied were ground nesters, which as a group may have been more vulnerable to predation than were other birds which nest in safer sites. None of the species studied formed the main prey of the predators concerned, as the predators were generalists which fed mainly on other prey (such as voles or lagomorphs) and therefore were sustained mainly by other prey. These are precisely the circumstances in which marked effects on vulnerable subsidiary prey species might be expected, because the predators are buffered against decline in their subsidiary prey. Predation was influenced not only by the numbers and types of predators present but also by the availability of alternative prey (such as rodents) and by habitat features such as nesting cover. In all the experiments, the effects of predator removal invariably were short lived, and when control stopped, the 1994 LIMITATION O F B I R D BREEDING DENSITIES experimental areas were soon recolonized and predation rates reverted to normal. The maximum breeding densities achieved under predator removal were about twice as high as in areas where predators were left. These effects were similar to those found in food-provision experiments but were trivial compared with those achieved with invertebrates, in experiments or in biological control programmes, in which density differences of 100-fold or more followed from the addition or removal of predators (Sih et al. 1985). In several predator-removal studies, experimental and control areas were adjacent to one another, leaving them open to the effects of movements. The killing of predators in one area may have affected both predator and prey numbers over a wider area, including the control area, and the prey may have redistributed themselves each year, so that any potential gains from predator removal in one area might have been nullified by dispersal to the adjacent area so as to even out the densities. Only five of the studies were on islands or other well-separated areas, which would reduce the probIem of movements (Crissey & Darrow 1949, Trautman et al. 1974. Greenwood 1986. Marcstrom et al. 1988, Tapper et al. 1990). Another problem in some studies was that predator numbers were not monitored apart from the totals killed. It is hard to remove every last predator from any area, especially with continuing immigration, so unless their numbers in both treatment and control areas are monitored independently, it is impossible to assess the effectiveness of the removals. Because most experiments were short lived, they could not be expected to show long-term effects nor whether breeding numbers increased to the maximum level possible. Nor could they show whether predation regulated, rather than merely limited, prey population density. If the experiments had been continued longer, it is theoretically possible that the prey may have increased to the point at which predators could no longer reduce their numbers to former levels (the predator pit hypothesis). This was the outcome of a predator-removal study on rabbits, providing evidence for the existence of two equilibrium densities in the presence of predators (Pech et al. 1992). It is also possible that, if habitat quality (food and cover) had been changed instead of predator numbers, a bigger and more lasting rise in density might have occurred. EXPERIMENTS INVOLVING PARASITES AND PATHOGENS Empirical and modelling studies have suggested that parasites could have the same effects on host populations as predators have on their prey, leading to either (1) no obvious reduction in breeding density, (2) an equilibrium level lower than resources would permit, (3) marked fluctuations in abundance or (4) decline to extinction. Natural experiments, involving the accidental introduction of a disease or disease vector to a new area, have occurred from time to time, sometimes with devastating effects on the local avifauna, 407 with the best-documented examples involving Hawaiian buds (Warner 1968. van Riper et al. 1986). Only recently have attempts been made to remove disease organisms from wild populations by using chemicals to remove disease vectors from the local environment (Kissam et al. 1975) or parasites from individual hosts and host nests (Brown & Brown 1986. Hudson 1986, M~ller1990, Chapman & George 1991) and vaccination programmes to reduce disease impacts (Hudson & Dobson 1990). However, almost all the published work on birds so far has concerned effects on individual survival and breeding performance, and few studies have examined the effects of parasite removal on breeding numbers. One such study concerned the effects of the strongyle worm parasite Trichostrongylus tenuis on Red Grouse. Removal of this parasite from individual grouse, using an anthelminthic drug, led to increases in both the breeding success and survival of individual grouse compared with untreated individuals (Hudson 1986, Hudson et al. 1992). Treatment of the majority of the grouse in the same way prevented a cyclic decline in numbers on five occasions on four different moors compared with trends in control areas (P. Hudson, pers. comm.). The conclusion was that treatment of this particular parasite at the level of the grouse population could prevent the periodic crashes in grouse numbers previously attributed to this parasite. Red Grouse in these areas thus conformed to the first (lower line) scenario in Figure 1,at least in the crash years. A second study concerned the effects of parasitic Brownheaded Cowbirds Molothrus ater on the rare Kirtland’s Warbler Dendroica kirtlandii, now restricted to a small area of Michigan (Rye1 1981, Mayfield 1983, Probst 1986, Weinrich 1988). A marked decline in warbler numbers between the 1960s and 1970s coincided with a period of heavy cowbird parasitism, when warbler breeding success was reduced to less than one young per nest. Following a period of cowbird removal, warbler breeding success increased three-fold and the decline in breeding numbers stopped. However, the warbler breeding population did not recover to its former levels. So while cowbird parasitism clearly affected warbler nest success, it remained uncertain whether it also affected breeding numbers. Indeed, other factors were suggested (Mayfield 1983). So this particular experiment was inconclusive with respect to breeding density. It is probably only a matter of time before other experiments reveal the effects of parasitism on breeding bird population levels. in addition to effects on individual performance. DISCUSSION The experiments discussed above are summarized in Table 5. Many were simple before-and-after studies, but 27% included a control area, and another 27% also involved reversal of treatments or replication. Despite their various designs, field experiments have served to confirm that, within areas of suitable habitat, the main potential limiting fac- 408 I. NEWTON tors-whether resources or natural enemies-have affected the breeding density of one bird species or another. They also have confirmed that the same species can be limited in breeding density by different factors in different areas or in different years. The provision of food or removal of predators Icd in extreme cases to a doubling of breeding densities compared with control areas, but provision of nestboxes often led to much bigger increases-5-20-fold in the most extreme cases. A n obvious drawback of experiments is that they are possible only on certain species which have food supplies, nest sites. predators or parasites that lend themselves to manipulation, are common enough for experiment and remain in the same area long enough to assess the effect of the manipulation. For each of the species studied, experiments were done on whatever limiting factor previous observation had suggested might have been important in the area concerned. A s a sample, then, the experiments which have been made were predisposed to give positive results, and they could give no indication of what would be the effect of manipulating some quite different factor in the species concerned (such as predation where breeding density appeared to be limited by nest sites). If species for each type of experiment had been selected from the available avifauna at random, rather than 0 1 1 the basis of prior knowledge, the number of positive results would probably be much less than the 77% recorded for the experiments reported here. In addition, all experiments so far have examined particular limiting factors individually. None has explored the possible interactions between different limiting factors, such as parasitism and food shortage; this is an obvious opening for future work. Because of thc small areas amenable to experiment, the response to an experiment could be measured only in terms of local breeding density, giving no indication of any possible wider impacts resulting from local manipulation. For example. if some local treatment had caused more birds to survive than could be accommodated in local nesting habitat, this could (through emigration) have led to increased breeding densities over areas larger than the study site. Conversely, if the treatment had favoured immigration, densities in surrounding areas might have been unaffected or deprcsscd as a result of thc experiment. The evidence provided by experiments is as good as we a r e likely to get, but even when a positive result was obtained. there is still the possibility that some unknown limiting factor changed in parallel to the one studied and caused thc change in numbers. And even where change in some particular factor did change numbers, it is not safe, of course, lime that all changes in numbers were due to this same factor. A given factor may be sufficient for some changes but not nccessary to explain them all. Despite these caveats, cxperiments are an improvement over observing natural changes because in experiments it is the observer who brings a bout the change. Experiments on the limitation of breeding density reveal little about the limitation of numbers at other times of year or about the limitation of total population size at the start IBIS 136 of breeding, consisting of breeders and nonbreeders. The main practical problem in most bird species is that, in contrast to breeders, nonbreeders are often nonterritorial and inconspicuous or flocking and range over much larger areas than breeders. This behaviour makes them hard to count accurately and to keep track of during experiments. However, whatever limits the numbers of breeders in a population must also indirectly set a ceiling on the numbers of nonbreeders. Thus, once all available breeding habitat is occupied, with no room for further settlers, an upper limit is set on both the number of breeders and the annual output of young. In due course, this must eventually limit the numbers of nonbreeders, which can increase only up to the point where the annual additions (from reproduction and exit from the breeding sector) match the annual losses (from mortality and entry to the breeding sector). The numbers of nonbreeders would then stabilize at that level (Newton 1992). The theoretical maximum ratio of nonbreeders to breeders was calculated by Brown (1969) for a range of bird species with different reproductive and mortality rates. In the most extreme examples, nonbreeders could outnumber breeders. The existence of a theoretical maximum for the numbers of nonbreeders (set by the numbers of breeders) does not of course exclude the possibility that environmental constraints might restrict their numbers below this level. The important point, however, is that whatever limits the breeding numbers of a species can also influence the total spring population of breeders and nonbreeders, either directly or indirectly. I am grateful to Professor D. Jenkins for helpful discussion over many years and for critical comments on this manuscript. REFERENCES Angelstam. P.. Lindstrom. E.C. &Widen, P. 1984. Role of predation in short-term population fluctuations of some birds and mammals in Fennoscandia. Oecologia 62: 199-208. Atkinson. I.A.E. 1985. The spread of commensal species of Rattus to oceanic islands and their effects on island avifaunas. ICBP Tech. Bull. 3 : 35-81. Ralser. D.J.. Dill. H.H. & Nelson, H.K. 1968. Effect of predator reduction on waterfowl nesting success. J. Wildl. Mgmt 32: 669682. Reasom, S.L. 1974. Intensive short-term predator removals as a game managemcnt tool. Trans. North Am. Wildl. Nat. Resour. Conf. 39: 230-240. Beissinger. S.R. & Bucher, E.H. 1992. Sustainable harvesting of parrots for conservation. In Beissinger. S.R. & Snyder, N.F.R. (eds) New World Parrots in Crisis: 73-115. Washington, D.C.: Smithsoriian Institution Press. Berndt. R. & Frantzen. M. 1964. Vom Einfluss des strengen Winters 1962/63 auf den Bruthestand der Kohlenbriiter bei Braunschweig. Om. Mitt. 16: 126-130. Blankinship. D.R. 1966. The relationship of White-winged Dove production to control of Great-tailed Grackles in the lower Rio Grande Valley of Texas. Trans. North Am. Wildl. Nat. Resour. Conf. 31: 45-58. 1994 LIMITATION O F B I R D B R E E D I N G D E N S I T I E S Boyd, A.W. 1932. Notes on the Tree Sparrow. Br. Buds 25: 278285. Brawn, J.D. & Balda, R.P. 1988. Population biology of cavitynesters in northern Arizona: Do nest sites limit breeding densities? Condor 9 0 61-71. Brown, C.R. & Brown, M.R. 1986. Ectoparasitism as a cost of coloniality in CliffSwallows (Hirundopyrrhonata).Ecology 67: 12061218. Brown, J.L. 1969. Territorial behaviour and population regulation in birds. Wilson Bull. 81: 293-329. Brush, T. 1983. Cavity use by secondary cavity-nesting birds and response to manipulations. Condor 85: 461466. Bump, G.. Darrow. R.W.. Edminster. F.G. & Crissey. W.F. 1947. The Ruffed Grouse. New York: Authority of the New York State Legislature. Cave. A.J. 1968. The breeding of the Kestrel, Falco tinnunculus L., in the reclaimed area Oostelijk Flevoland. Neth. J. Zool. 18: 313407. Chapman, B.R. & George, J.E. 1991. The effects of ectoparasites on Cliff swallow growth and survival. In Loye, J.E. & Zuk, M. (eds) Bird-parasite Interactions: 69-92. Oxford Oxford University Press. Chesness, R.A.. Nelson, M.M. & Longley. W.H. 1968. The effect of predator control on Pheasant reproductive success. J. Wildl. Mgmt. 32: 683-697. Coleman, J.D. 1974. The use of artificial nest sites erected for Starlings in Canterbury. New Zealand. N.Z. J. Zool. 1: 349-354. Cooper, R.J.. Dodge, K.M., Martinat, P.J.. Donahoe. S.B. & Whitmore, R.C. 1990. Effect of diflubenzuron application on eastern deciduous forest birds. J. Wildl. Mgmt 54: 486-493. Crawford, R.J.M. & Shelton, P.A. 1978. Pelagic fish and seabird inter-relationships off the coasts of South West and South Africa. Biol. Conserv. 14: 85-109. Creutz. G. 1949. Untersuchungen zur Brutbiologie des Feldsperlings (Passer m. rnontanus L.). 2001. Jahrb. 78: 133-172. Crissey. W.F. & Darrow, R.W. 1949. A study of predator control on Valcour Island. Res. Ser. No. 1.Albany, N.Y.: New York State Conservation Department, Division of Fish and Game. Currie. F.A. & Bamford. R. 1982. Songbird nest-box studies in forests in north Wales. Q. J. For. 76: 250-255. Cutforth. B. 1968. Eighth annual report of the Brandon Juniors nest box project, 1968. Blue Jay 2 0 188. Dahlsten, D.C. & Copper. W.A. 1979. The use of nesting boxes to study the biology of the Mountain Chickadee (Parus gambeli) and the impact on selected forest insects. In Dickson, J.G.. Conner. R.N., Fleet, R.R., Kroll. J.C. & Jackson, ].A. (eds) The Role of Insectivorous Birds in Forest Ecosystems: 217-260. London: Academic Press. Davies. N.B. 1978. Ecological questions about territorial behaviour. In Krebs, J.R.& Davies. N.B. (eds) Ekhavioural Ecology. An evolutionary approach: 317-350. Oxford: Blackwell Scientific Publications. Dennis. R.H. & Dow. H. 1984. The establishment of a population of Goldeneyes (Bucephala clangula) breeding in Scotland. Bud Study 31: 217-222. DeReede, R.H. 1982. A field study on the possible impact of the insecticide diflubenzuron on insectivorous buds. Agro-Ecosystems 7: 327-342. Dhondt. A.A. & Eyckerman. R. 1980. Competition between the Great Tit and the Blue Tit outside the breeding season in field experiments. Ecology 61: 1291-1296. Dornbusch. M. 1973. Zur Siedlungsdichte und Ernarung des Feldsperlings in Kiefern-Dickungen. Der Falke 20: 193-195. 409 Duebbert, H.F. & Kantrud, H.A. 1974. Upland duck nesting related to land use and predator reduction. J. Wildl. Mgmt 38: 257-265. Duebbert. H.F. & Lokemoen. J.T. 1980. High duck nesting success in a predator-reduced environment. J. Wildl. Mgmt 44: 428-437. East, M.L. & Perrins. C.M. 1988. The effect ofnest-boxes on breeding populations of birds in broadleaved temperate woodlands. Ibis 130: 393-401. Enemar. A.B. & Sjostrand, B. 1972. Effects of the introduction of Pied Flycatchers Ficedula hypoleuca on the composition of a passerine bird community. Ornis Scand. 3: 79-87. Enoksson. B. & Nilsson, S.G. 1983. Territory size and population density in relation to food supply in the Nuthatch Sitta europaea. J. Anim. Ecol. 52: 927-935. Eriksson, M.O.G. 1982. Differences between old and newly established Goldeneye (Bucephala clangula) populations. Ornis Fenn. 59: 13-19. Errington, P.L. 1946. Predation and vertebrate populations. Quart. Rev. Biol. 21: 144-177, 221-245. Franzblau. M.A. & Collins, J.P. 1980. Test of a hypothesis of territory regulation in an insectivorous bird by experimentally increasing prey abundance. Oecologia 46: 164-1 70. Gauhl, F. 1984. Ein Beitrage zur Bruthiologie des Feldsperlings (Passer montanus). Vogelwelt 105: 176-187. Gauthier. G. & Smith, J.N.M. 1987. Territorial behaviour. nest-site availability, and breeding density in Buffleheads. J. Anim. Ecol. 56: 171-184. Greenwood. R.]. 1986. Influence of Striped Skunk removal on upland duck nest success in North Dakota. Wildl. SOC.Bull. 14: 6-11. Gustafsson. L. 1988. Inter- and intraspecific competition for nestholes in a population of the Collared Flycatcher Ficedula albicollis. Ibis 1 3 0 11-16. Haartman, L. von. 1971. Population dynamics. In Farner. D.S. & King, J.R. (eds) Avian Biology, Vol. 1:391-459. London: Academic Press. Haartman. L. von. 1973. Talgmespopulationen Lemsjoholm. Lintumies 8: 7-9. Hamerstrom, F.. Hamerstrom. F.N. & Hart, J. 1973. Nest boxes: An effective management tool for Kestrels. J. Wildl. Mgmt 37: 400-403. Haramis. G.M. & Thompson, D.Q. 1985. Density-production characteristics of box-nesting Wood Ducks in a northern greentree impoundment. J. Wildl. Mgmt 49: 429-436. Higuchi, H. 1978. Use of nest-boxes by birds according to forest types and the breeding density in forests with and without nestboxes. J. Jpn. For. SOC.6 0 255-261. Holroyd. G.L. 1975. Nest-site availability as a factor limiting population size of swallows. Can. Field-Nat. 89: 60-64. Hudson, P.J. 1986. The effect of a parasitic nematode on the breeding production of Red Grouse. J. Anim. Ecol. 55: 85-92. Hudson, P.J. & Dobson, A.P. 1990. The control of parasites in natural animal populations: Nematode and virus infections of Red Grouse. In Perrins. C.M.. Lebreton, J-D. & Hirons, G.J.M. (eds) Bird Population Studies: Their relevance to conservation and management: 413-432. Oxford: Oxford University Press. Jansson. C.. Ekman. J. & Briimssen. A. von. 1981. Winter mortality and food supply in tits (Parus spp.). Oikos 37: 313-322. Jarvinen. A. 1978. Nest-box studies in mountain birch forest at Kilpisjarvi. Finnish Lapland. Anser (Suppl. 3): 107-11 1. Johnson, L.L. 1967. The Common Goldeneye Duck and the role of nesting boxes in its management in north central Minnesota. J. Minnesota Acad. Sci. 34: 110-1 13. 4 10 I. NEWTON Jones. R.E. & Leopold. AS. 1967. Nesting interference in a dense population of Wood Ducks. J. Wildl. Mgmt 31: 221-228. Kallander, H. 1981. The effects of provision of food in winter on a population of the Great Tit Parus major and the Blue tit P. caerule11s. Ornis Scand. 12: 244-248. Keith. I,.B. 1963. Wildlife's Ten-year Cycle. Madison, Wis.: University of Wisconsin Press. Keith, L.R. & Rusch. D.H. 1988. Predation's role in the cyclic fluctuations of RuHed Grouse. Proc. Int. Om. Congr. XIX: 699732. Kissam. J.B.. Noblet, K.E. & Garris. G.1. 1975. Large scale aerial treatment of an endemic area with abate granular larvicide to control black flies (Diptera: Simulidae) and suppress Leucoc!jtozoon srnithi of turkeys. 1. Med. Entomol. 12: 359-362. Krcbs, J.K. 1971. Territory and breeding density in the Great Tit, l'urus major L. Ecology 52: 2-22. lack. D. 1954. 'The Natural Kegulation of Animal Numbers. Oxford: Oxlord University Press. Lynch. G.M. 1972. Effect of strychnine control on nest predators of dabbling ducks. J. Wildl. Mgmt 36: 436-440. Marcstriim. V., Kenward. R.E. & Engren, E. 1988. The impact of predation on boreal tetraonids during vole cycles: An experimenLa1 study. J. Anirn. Ecol. 57: 859-872. Ilarjakangas, A. 1987. Does the provision of extra food in winter affect the abundance of Finnish Black Grouse? In Hudson, P. & Lovel, T. (eds) Fourth International Grouse Symposium: Paper No. 20. West Germany: World Pheasant Association. Mayfield, H.F. 1983. Kirtland's Warbler, victim of its own rarity? Auk 100: 974-976. Mc1,aughliri. C.L. & Grice. D. 195.2. The effectiveness of large-scale erection of Wood Duck boxes as a management procedure. Trans. Worth Am. Wildl. Natur. Resour. Conf. 17: 242-259. Miller. G . R . , Watson. A. & Jenkins. D. 1970. Responses of Red Grouse populations to experimental improvement in their food. In Watson. A . (ed.) Animal Populations in Relation to Their Food Rcsourccs: 323-3 35. Oxford: Blackwell. M01ler, A.P. 1990. Effects of parasitism by a haematophagous mite on reproduction In the Rarn Swallow Hirundo rustica. Ecology 71: 2345--2%57. Ucwton, 1. 1979. Population Ecology of Raptors. Berkhamsted: Poyser. Sewton. I. 1980. The role of food in limiting bird numbers. Ardea 68: 1 1 - 3 0 . Newton. I. 1992. Experiments on the limitation of bird numbers by territorial behaviour. Biol. Rev. 67: 129-173. Newton. I. 1993. Predation and limitation of bird numbers. Current Ornith. 11: 143-198. Parker, H. 1984. Effect of corvid removal on reproduction of Willow Ptarmigan and Black Grouse. J. Wildl. Mgmt 48: 1197-1205. Prakall. D.R. & Bart. J.R. 1983. Impacts of aerial application of insecticides on forest birds. CBC Critical Reviews in Environmental Control 13: 117-165. t'earce. P.A. & Peakall, D.B. 1977. The impact of fenitrothion on bird populations in New Rrunswick. In Roberts, J.R.. Greenhalgh. R. & Marshall, W . K . (eds) Fenitrothion: The long-term effects of its use in forest ecosystems-Current status: 199-306. Publ. No. 1 5 389. Ottawa: National Research Council of Canada. I'rch, R.P., Sinclair. A.R.E.. Newsome. A.E. & Catling. P.C. 1992. Limits to predator regulation of rabbits in Australia: Evidence from predator-removal experiments. Oecologia 89: 102-112. lBIS 136 Potts. G.R. 1980. The effects of modern agriculture, nest predation and game management on the population ecology of partridges Prrdix prrdix and Alectoris ruJa. Adv. Ecol. Res. 11: 1-79. Probst, J.R. 1986. A review of factors limiting Kirtland's Warbler on its breeding grounds. Am. Nat. 117: 87-100. Raphael, M.G. & White, M. 1984. 1Jse of snags by cavity-nesting birds in the Sierra Nevada. Wildl. Monogr. 86: 1-66. Richmond, M.L., Henny, C.J.. Floyd. R.L.. Mannan. R.W.. Finch, D.M. & DeWeese, L.R. 1979. Effects of sevin-4-oil. dimilin and orthene on forest birds in northeastern Oregon. U S . For. Serv. Res. Pap. PSCO-148. Ryel, LA. 1981. Population change in the Kirtland's Warbler. Jack-pine Warbler 59: 76-91. Samson. F.R. & Lewis, J.]. 1979. Experiments on population regulation in two North American parids. Wilson Bull. 91: 222-233. Savard, J.L. 1988. Use of nest boxes by Barrow's Goldeneyes: Nesting success and effect on the breeding population. Wildl. Soc. Bull. 16: 125-1 32. Schrank. B.W. 1972. Waterfowl nest cover and some predation relationships. J. Wildl. Mgmt 36: 182-186. Sih. A., Crowley. P.. McPeek, M.. Petranka. J. & Stohmeier. K. 1985. Predation, competition, and prey communities: A review of field experiments. Ann. Rev. Ecol. Syst. 16: 269-311. Siren. M. 1951. Increasing the Goldeneye population with nestboxes. Suomen Riista 6: 189-190. Smith, J.N.M., Montgomerie. R.D.. Taitt. M.J. & Yom-Tov. Y. 19F3. A winter feeding experiment on an island Song Sparrow population. Oecologia 47: 164-170. Spray, C.J., Crick, H.Q.P. & Hart, A.D.M. 1987. EHects of aerial applications of fenitrothion on bird populations of a Scottish Pine plantation. 1. Appl. Ecol. 2 4 29-47. Stribling, H.L. & Smith, H.R. 1987. Efkcts of dimilin on diversity and abundance of forest birds. North 1. Appl. For. 4: 37-38. Tapper, S.C., Rrockless, M. & Potts. C.R. 1991. The Salisbury Plain experiment: The conclusion. The Game Conservancy Review of 1990: 87-91. Tapper. S.C.. Potts. G.K.. Reynolds. J., Stoate, C. & Rrockless. M. 1990. Salisbury Plain experiment-Year six. The Game Conscrvancy Annual Review of 1989: 42-49. 'I'rautman, C.G.. Fredrtckson. L.F. & Carter. A.V. 1974. Kelationship of Red Foxes and other predators to populations of Ringnecked Pheasants and other prey, South Dakota. Trans. North Am. Wildl. Nat. Res. Conf. 39: 241-255. van Balen. J.H. 1980. Population fluctuations of the Great Tit and feeding conditions in winter. Ardea 68: 143-164. van Balen. J.H.. Booy. C.J.H.. van Franeker, J.A. & Osieck, E.R. 1982. Studies on hole nesting birds in natural nest sites. 1. Availability and occupation of natural nest sites. Ardea 70: 1-24. van Riper, C., van Riper, S.G.. Coff. M.G. & Laird, M. 1986. The epizootiology and ecological significance of malaria in Hawaiian land birds. Ecol. Monogr. 56: 327-344. Village, A. 1983. The role of nest-site availability and territorial behaviour in limiting the breeding density of Kestrels. J. Anim. Ecol. 52: 635-645. Village, A. 1990. The Kestrel. Calton: Poyser. Warner. R.E. 1968. The role of introduced diseases in the extinction of the endemic Hawaiian avifauna. Condor 70: 101-120. Waters, J.R.. Noon, B.R. & Verncr. J. 1990. Lack of nest site limitation in a cavity-nesting bird community. J . Wildl. Mgmt 54: 239-245. 1994 LIMITATION OF BIRD BREEDING DENSITIES Watson, A. &Moss, R. 1979. Population cycles in the Tetraonidae. Ornis Fenn. 56: 87-109. Weinrich, J.A. 1988. Status of Kirtlands Warbler 1987. Jack-pine Warbler 66: 155-158. Willebrand. T. 1988. Demography and ecology of a Black Grouse 41 1 (Tetrao tetrix L.) population. Published doctoral thesis, Uppsala University. Yom-Tov, Y. 1974. The effect of food and predation on breeding density and success, clutch size and laying date of the Crow (Corvus corone). J. Anim. Ecol. 43: 4 7 9 4 9 8 .