Ecology and Management of Breeding Waterfowl

Transcription

1 Ecology and Management of Breeding Waterfowl Brce D. ]. Batt, Alan D. Afton, Michael G. Anderson, C. Davison Ankney, Doglas H. Johnson, John A. Kadlec, and Gary L. Krap, editors 1/r;-- niversity of Minnesota Press Minneapolis and London

2 CHAPTER 8 Spacing Patterns Michael G. Anderson and Roger D. Titman. ntrodction The resorces all animals need for srvival and reprodction are distribted over environmental landscapes in patterns that vary in space and in time. ndividals shold adopt behavioral strategies that maximize their efficiency in acqiring resorces. For example, if individals are more sccessfl at acqiring resorces by contesting for them rather than scrambling for them, varios forms of direct competition, sch as territoriality, shold reslt. Spacing patterns and agonistic behavior appear to have evolved as compromises reslting from the action of many conflicting selective forces, most likely related to food, mates, risks of predation, and intraspecific and interspecific competition. Stdents of animal behavior crrently view social systems as the prodcts of conflicting "best soltions" for males and females (Davies and Krebs 1978). This perspective is important for nderstanding how selective forces mold social systems, and it has been implicit in models of social behavior for some time (Orians 1969; Emlen and Oring 1977). Discssions of conflicting male and female "interests" have led to new insights abot social strctre (Trivers 197; Jarman 197; Bradbry and Vehrencamp 1977; Daly and Wilson 1978; Rbenstein and Wrangham 1986). A tentative general conclsion from these stdies is that female spacing patterns probably respond mostly to ecological factors, sch as the distribtion of food or the natre of important predators, while male spacing patterns are related mostly to the distribtion and availability of females. This is becase, in the broadest sense, female reprodctive sccess is limited by access to energy and other resorces necessary for reprodction, while male reprodctive sccess is determined mostly by the nmber of females they can inseminate (Trivers 197). Active competition for resorces may take several forms. At one extreme, animals may simply scramble for available resorces, a strategy that has been described as exploitation, scramble competition, or diffse competition. At the opposite extreme is contest competition, which involves the establishment of exclsive territories or rigid dominance hierarchies that may be sed to acqire and maintain viral resorces (Miller 1967; Fretwell and Lcas 1970; Davies 1978, 1980; Davies and Hoston 198). The search for answers to why different species, poplations, or individals tilize different strategies is a central theme of modern behavioral ecology. We shall try to synthesize information concerning the diverse spacing patterns and agonistic behavior fond in waterfowl and then interpret this in fnctional and evoltionary terms. Additionally, we seek to identify important gaps or weaknesses in or nderstanding of waterfowl spacing behavior. We begin by reviewing the most freqently accepted hypotheses offered to explain waterfowl spacing behavior. We then discss some general ideas abot living in grops. We believe it is sefl to briefly consider waterfowl flocking, typical of a majority of species for most of the year, before developing ideas abot the costs and benefits to individals of varios forms of spacing ot. This is followed by a srvey of spacing patterns and associated agonistic behavior exhibited by waterfowl. There, we search for generalizations concerning differences in spacing, and for ecological correlations and phylogenetic trends associated with these differences. We conclde with a short discssion of general models of animal dispersion and animal contests that bear on spacing in waterfowl, a few recommendations for ftre research, and some comments on related isses in waterfowl management. t is difficlt to discss spacing behavior withot considering closely related isses sch as mating behavior or feeding ecology. We have tried to restrict orselves to the topic of spacing, bt we wander occasionally into other 1

3 MCHAEL G. ANDERSON AND ROGER D. TTMAN aspects of social behavior. Patterns of spacing, mating, and parental care are basic components of a species's social system and are tightly intertwined (see also chapters 6 and 7 of this volme). Becase of the relationship to ecology of fnctional and evoltionary qestions abot spacing, and becase of or personal interests, we concentrate on these isses, largely neglecting problems of proximate casation and ontogeny. A. Approaches to the Stdy of Spacing Behavior Adaptive aspects of spacing behavior have been stdied in three general ways, all of which have reslted in new ideas abot behavioral ecology and evoltion: (1) Comparative stdies of social behavior in closely related species have contribted mch to or nderstanding of the inflence of ecological factors on spacing systems (Crook 196, 196; Geist 197; Jarman 197; Wilson 197; Cltton-Brock and Harvey 1977; Rbenstein and Wrangham 1986). This approach involves contrasting behavior among closely related species, or among poplations of the same species occpying different habitats, and correlating these patterns with differences in ecological conditions (McKinney 1978; Cltton-Brock and Harvey 1979; Jarman 198; Endler 1986). Conversely, comparisons can be made between phylogenetically diverse animals in similar ecological settings. () Several theories have been developed regarding spacing behavior or patterns of dispersion in relation to fnctions sch as reglation of poplation density, optimality, foraging efficiency, or the fitness of competing individals (Hinde 196; Wynne-Edwards 196; Brown 196; Horn 1968; Brown and Orians 1970; Wiens 1976; Waser and Wiley 1979; Schoener 198). Brown's ( 196) concept of economic defendability was especially inflential as it emphasized for the first time the importance of viewing spacing systems as the prodct of cost/ benefit trade-offs made by individals. () A third approach, combining theory and empirical estimates, has been to measre differences in traits presmably associated with fitness, sch as mating sccess and food acqisition, among individals behaving in different ways (C. Smith 1968; Wolf and Hainsworth 1971; Wolf eta/. 197; Carpenter and Macmillen 1976; Mark! 1981; Harper 198). One conclsion is that optimal strategies are freqently conditional. For example, a tactic which may be profitable when a resorce is abndant may be replaced by another one when theresorce is scarce. Sch stdies have provided the first tests of economic models of spacing behavior and offer mch promise for frther advances sing species amenable to direct cost/benefit analyses.. Theories of Spacing Behavior A. Territory Althogh anecdotal acconts of what we now call territorial behavior date back to the Bible and early Greek writing, the first modern descriptions of territoriality come from Bernard Altm's 1868 book Der Vogel nd Sein Leben and Eliot Howard's Te"itory in Bird Life (190) (Welty 198). Howard's (190) thoghtflly written book made the stdy of territoriality poplar, and ornithologists qickly recognized that many breeding birds were dispersed, that food was associated with space, that song was a signal of territory ownership among passerine birds, that bright plmage and displays were often frther signals sed to advertise space occpation, and even that space ownership might serve to limit the prodction of yong. Many definitions of territory have been proposed. The simplest and most freqently sed one is that a territory is "any defended area" (Noble 199). Several others are worthy of note. Nice ( 191) arged that "territory in the male bird implies isolation, advertisement, fixation and intolerance." Tinbergen's (197) concise and descriptive definition is perhaps the best: territorial behavior consists of site attachment and hostility directed toward a certain category of other animal~. The term "territory" is generally applied to an area or space in which a particlar animal is aggressive and dominant with respect to certain intrders. Brown and Orians (1970) listed the essential characteristics of territory: (1) a fixed area; () defensive behavior is exhibited there; and () the area is exclsive with respect to rivals. Brown (1969) noted that "when activity spaces of different individals overlap broadly and the areas defended, if any, are not clearly demarcated, then se of the term 'territory' is not jstified." Behavior sed to defend an area can consist of overt defense (e.g., chases and attacks) or acts identifying the defender and making him conspicos (e.g., vocalizations and displays) (Brown and Orians 1970). Among birds a great diversity of territories exist. B. Economic Defendabiliry Early reviews of territorial behavior (Howard 190; Hinde 196; Tinbergen 197) cataloged the occrrence of territoriality in a great array of species, tried to classify the range of behavior observed, and strived to assign fnctions to varios territorial systems (e.g., mating territories, feeding territories, all-prpose territories). Defense was the diagnostic qality of a territory, and its fnction related to the resorce being defended. These smmaries were helpfl, bt sch approaches shed little light on qestions of when, or nder what circmstances, resorce defense was advantageos. Fr-

4 l 1 'b! ' SPACNG PATERNS thermore, rhe qestion of "advantageos to whom?... ro individals or to poplations?" was still largely nresolved. Wynne-Edwards (196) developed the hypothesis that territorial behavior evolved via poplation (grop) selection to reglate poplation density. Brown (1969) fond little evidence to spport this hypothesis and arged instead that territoriality evolved via individal selection reslting from aggressive competition. Lack (196, 1966, 1968), Williams (1966), Wiens (1966), Brown and Orians (1970), and others provide critiqes of Wynne-Edwards's hypothesis, the reslt being that today it has little spport. Along with a general shift to individal-selection thinking among behavioral ecologists (Williams 1966; Lack 1968), Brown's formal introdction of economic thinking into the analyses of animal spacing behavior inagrated a period of rapid advance (Brown 196; Brown and Orians 1970). Brown arged that individals shold defend territories when, and only when, it paid them to do so in terms of increased fitness. Brown conclded that an element of competition may be all that territorial systems in different species had in common, and that rather than search for fnctions it wold be more profitable to ask what factors selected for increased or decreased aggressiveness with respect to space. Therefore, the important determinant of territorial behavior is whether or not critical reqisites for reprodction or srvival are economically defendable. Of corse this idea can readily be extended to all forms of resorce defense, not jst territoriality. Brown and Davies and Hoston (198) envisioned thresholds for the expression of territorial behavior sch that when benefits of defense for individals exceeded the costs of defense, animals wold behave territorially. This simple graphical concept can be extended to offer predictions abot changes in behavior in response to territory size, competitor density, defense costs, and so on. Brown's idea nderlies virtally all recent empirical work on the economics of territory defense. The economics of territorial behavior has been careflly stdied in birds amenable to sch analyses, sch as Anna Hmmingbirds (Calypte anna), Golded-winged Snbirds (Nectarinia reichenowi), and a Hawaiian Honeycreeper (Vestiaria coccinea) (Stiles 1971; Gill and Wolf 197; Carpenter and Macmillen 1976). Forthese nectarfeeding birds that appear to optimize their behavior over a short time span, economic models seem to predict qite well when, and nder what circmstances, individals will be territorial. Similar stdies of nonbreeding Pied Wagtails (Motacilla alba) (Davies and Hoston 1981, 198) and Sanderlings (Calidris alba) (Myers et al. 1979, 1981) fond birds changing from territorial to flock behavior and back with flctations in food availability. To or knowledge, sch detailed energetic measrements have not been condcted on any species of waterfowl. Althogh accrate measrements of resorce availability are lacking, abandonment of breeding effort and dispersal have been recorded for Northern Pintails (Derksen and Eldredge 1980), Mallards (Krap et a/. 198), and Lesser Snow Geese (Ankney and Macinnes 1978) when breeding resorces are severely redced by droght or other harsh climatic conditions. No dobt measring the availability of key resorces for breeding waterfowl or condcting experimental maniplations of resorce availability wold be difficlt. However, sch energetics stdies wold be valable for nderstanding variations in spacing behavior among individals, poplations, or species. These stdies are most likely to be instrctive with species showing site-specific or temporal variation in spacing behavior where similar resorces can be compared. Detailed knowledge of at least two factors is reqired before field stdies of defense thresholds are possible. First the investigator mst know what resorce(s) is being defended. Defending animals may be optimizing their inclsive ficness (Hamilton 196), bt scientists almost never measre this. f food, mates, or another resorce is being measred instead, the investigator mst be satisfied that he is observing the same factor(s) that is inflencing fitness. The second important isse is the time scale for optimization (Davies and Hoston 198). s defense selected to assre access to a daily-varying food spply?... to a seasonal food spply?... to a territory that might provide critical resorces for yong 6 months from now?... or to real estate that might provide a breeding opportnity once every few years? Misinterpretation of either fnction or time scale cold lead to wholly inappropriate conclsions abot the economics of defense. C. Hypotheses abot Spacing in Waterfowl 1. History of deas abot Waterfowl Spacing and Chasing- Debates over Fnction and Motivation Brggeman (1876, in Palmer 1976, p. 89) recorded defensive behavior, site fixation, and bondary delimitation by Common Goldeneyes over 100 years ago. Bennett (198) saw a reglar pattern in the spacing of breeding Ble-winged Teal which Hochbam (19) considered territorial. Hochbam ( 19) emphasized the importance of territorial behavior in creating the dispersion of dcks in prairie marshes. His belief rhar territorial dcks established definite bondaries against intrders was dispted by Dzbin (19) and Sowls (19), who fond it difficlt to define strict bondaries and observed broad overlap of the home ranges of neighboring pairs in most Anas and Aythya species. For a while there was considerable debate concerning the applicability of the concept of territory to breeding

5 MCHAEL G. ANDERSON AND ROGER D. TTMAN t. i t f dcks, especially Mallards, Nonhern Shovelers, and Gadwall (Bezzel 199; Lehrer 1961; Gates 196; Hori 196; McKinney 196a). Later athors conclded that these species are indeed territorial (McKinney 1967; Dwyer 197; Seymor 197a, b; Titman 198). Of all behavior associated with spacing in dcks, aerial chases or prsit flights by paired male dabbling dcks (Anas spp.) have received the most stdy and stimlated the greatest controversy. The first extensive description of Mallard prsit flights was given by Heinroth (191 1), who believed that the chases reslted from the "sexal or raping rge of drakes." Geyr (19) sggested that certain prsit flights represented attempts by a paired male to keep a breeding area free from other pairs. He distingished between these prsit flights and others that occasionally ended in rape (forced coplation). These chases were later referred to as Vertreiben (driving away) and Hetzjagen (harassing chase), respectively (Geyr 199). Christoleit (199a, b) attribted conship fnction to prsit flights, considering them "mating play." He criticized Geyr's ideas, deeming them nreliable becase Geyr had observed tame or park Mallards. Wetmore (190), Hochbam (19), Lehrer (191, 19, 1961), Dzbin (197), Bezzel (199), and Wst (1960) described aerial chases associated with cortship for several species of srface-feeding dcks. They believed that prsit flights cold be divided into three categories: (1) grop flights associated with social display; () Vertreiben (Geyr 199), territorial defense flights (Hochbam 19), three-bird chases (Dzbin 197) or threebird flights (McKinney ); and () Hetzjagen (Geyr 199), grop chases (Dzbin 197), or attempted rape flights (McKinney ; Lebret 1961). Several athors have commented pon the difficlty of effectively separating the latter two categories (Dzbin 197; Weidmann 196; Wst 1960; McKinney 196a; R. Smith 1968; Titman 197). McKinney (196a) commented that "the prsit of the female of a strange pair by a paired male cold be motivated by either attack tendency, rape tendency, or a combination of both." Weidmann (196), observing tame Mallards, conclded that all prsit flights were motivated by the paired male's desire to coplate with other females. Lebret (1961) and Hori (196), also observing semi-tame Mallards, were similarly disenchanted with the view that a prsit flight might be territorially motivated. Lebret considered that explsion flights were de to the simple intolerance of breeding pairs toward each other, and Hori failed to see any male-to-male aggression, so he assmed that Mallards were not territorial. Geyr (19) appears to have been very astte in his interpretation that prsit flights fnction to keep an exclsive breeding area, especially since the territorial tendencies of semi-tame Mallards are mch less obvios than those of wild birds. Observing wild Mallards in Nonh America, Hochbam (19), Dzbin (19, 197, 1969), Sowls (19), McKinney (196a), and Titman (198) have associated prsit flights with spacing. Similarly, prsit flights in Gadwall, Ble-winged Teal, Northern Shoveler, Black Dck, and probably other species appear to fnction, at least in part, in spacing breeding pairs.. Sggested Costs and Benefits of Spacing Motivational considerations aside, different athors have qalitatively identified a variety of costs and benefits of spacing for waterfowl. The proximal costs of territoriality are chiefly the time and energy sed ro threaten and expel intrders and the loss of feeding time throgh vigilance (Patterson 198). The benefits of groping, considered below, are also lost to dispersed individals. The risk of injry dring aggressive enconters and exposre to predators dring agonistic displays mst also be considered (McKinney et al. 1978; Wishart 198; Nechterlein and Storer 198a). Waterfowl species appear to behave aggressively for a wide range of reasons. Athors have arged that defense serves many fnctions (Table 8-1 ), largely based on sbjective assessments of the place, context, and timing of aggressive behavior. Fnctions related to mates, access to food, and predation appear to be more niversally applicable to spaced-ot waterfowl than the other benefits identified. Most of these sggested benefits accre to individals or pairs; a few are dearly poplation-selection hypotheses (e.g., dispersion to redce predation or disease) and seem nlikely to be important selective forces.. Evoltion of Waterfowl Spacing Patterns Despite extensive stdy of behavior associated with spacing in waterfowl, relatively few attempts have been made to link variation in spacing patterns to possible evoltionary cases sch as differing ecological circmstances. nstead, most argments have centered on the presmed motivation or conflicting tendencies of birds behaving aggressively toward conspecifics (see above). Encoragingly, thogh, there seems to be growing interest in examining the ecological bases of waterfowl behavior (McKinney 197, 197, 1986; McKinney et al. 1978; Derrickson 1977; Afton 1980; Titman and Seymor 1981; Ndds and Ankney 198; Gathier 1987a, 1988). Considerable variation in spacing systems and agonistic behavior has been described for Anas, the beststdied gens (McKinney 197, 1986; Ndds and Ankney 198), and temporal changes in agonistic behavior associated with changing breeding stats have been described for a few species (e.g., Mallards [Rairaso 196; Titman 198], Nonhern Pintails [R. Smith 1968; Derrickson 1977), African Black Dcks [McKinney et

6 SPACNG PATTERNS Fnction Related to Mates Provide an opportnity to breed Provide an assred rendezvos site Redce interference with coplation Assre paternity Protect the female from harassment by conspecifics Related to Food Provide exclsive access to food Allow the female to feed ndistrbed (also see above) Provide food for yong Related to Nests Provide an assred nest site Defend nest from competitors, brood parasites, and predators Related to Predation Assre dispersion to redce predation Related to Disease and Parasitism Redce poplation density and ths ssceptibility to disease and parasites Table 8-1. Sggested benefits of spacing ot in varios species of waterfowl Species Northern Sheldck African Black Dck Northern Sheldck Northern Shoveler American Wigeon Mallard Anas & Aythya spp. Northern Sheldck American Black Dck Mallard Canada Goose Mallard Anas spp. & others American Wigeon Canvasback Lesser Snow Goose Common Eider Northern Sheldck Mallard American Wigeon Anas spp. Canvasback Canada Goose Ross Goose Bfflehead Anas spp. & others Northern Shoveler Bfflehead Pink-footed Goose Northern Sheldck Cape Shoveler Gadwall Mallard American Black Dck Canvasback Common Eider Barrow's Goldeneye Bfflehead Bfflehead Lesser Snow Goose Cygns spp. Canada Goose Barnacle Goose Anas spp. & others Anas spp. & others American Wigeon Sorce Yong 196 McKinney et al Hori 1969 Seymor 197b Wishart 198 Hmbrg et al. 1978, Titman 198 Hochbam 19 Hori 1969 Seymor and Titman 1978 Titman 198 Cooper 1978 Barash 1977, Titman 198 McKinney 197, 1986, McKinney eta/. 198 Wishart 198 Anderson 198Sa, b Minea and Cooke 1979 Ashcroft 1976 Patterson 198 Titman 198, Goodbrn 198 Wishart 198 McKinney 198 Anderson 198a, b Ewaschk and Boag 197, Cooper 1978 Ryder 197 Gathier 1987a McKinney 196Sa Seymor 197b Gathier 1987a, b nglis 1976 Williams 197, Patterson 198 Siegfried 196 Dwyer 197 Titman 198 Seymor and Titman 1978 Anderson 198a, b Ashcroft 1976 Savard 198 Gathier 1987b Gathier 1987b Minea and Cooke 1979 Kear 197 Cooper 1978 Owen and Wells 1979 McKinney 196a McKinney 196a Wishart 198 a/. 1978], Northern Shovelers [Poston 197; Seymor 197b,c], Canvasbacks [Anderson 198, 198a)). Only preliminary attempts have been made to test for habitatspecific differences in spacing behavior among dcks (Ndds and Ankney 198; Amat 198) despite indications from other grops that sch differences may be common (Dnbar 198; Oring 198). Rigoros tests of economic models of spacing have not been made with waterfowl Several writers, how- ever, have sed economic reasoning to try to explain patterns they have observed. Bilding on Brown's (196) concept of economic defendabiliry, McKinney (197) sggested that differences in territoriality among some Anas species, particlarly between territorial Northern Shovelers and nonterritorial Northern Pintails, might be explained by the relative defensibility of their contrasting foods and habitats. Derrickson (1978) also arged that Northern Pintails were not territorial

7 6 MCHAEL G. ANDERSON AND ROGER D. TTMAN becase the patchy and ephemeral natre of their main spring food spply (aqatic invertebrates in temporary water) made defense impractical. n addition to effects of resorce distribtion, factors sch as resorce predict ability and cost of defense may modify spacing behavior. For Canvasbacks, large-bodied diving dcks with high wing-loading that feed on patchily distribted and largely "hidden" benthic foods, Anderson (198a) arged that ncertainty of resorce availability and high costs of defense over large home ranges otweigh the possible benefits of long-term site defense. Conversely, McKinney et al. (1978), Eldridge (1986a, b) and others have arged that all reqisites needed for breeding by several river specialist dcks (e.g., African Black Dck, Torrent Dck, Ble Dck) are obtainable from welldefined segments of streams that pairs defend vigorosly year-rond. Comparative argments may be sefl to erect hypotheses for frther testing, bt very few waterfowl stdies have measred resorce distribtion. Descriptive data on resorce distribtions wold be sefl. Field ex periments that maniplate resorce characteristics and monitor behavioral responses wold be extremely val able. t has also been arged that mate garding is anticck oldry behavior by paired males that might select for spacing independent of any attribte of environmental resorces (Barash 1977; Wittenberger and Tilson 1980; McKinney et al. 198; Wishart 198; Goodbrn 198; McKinney 1986; Gathier 1988). ndependent of the work on dck spacing, stdents of goose behavior have developed several hypotheses to ex plain the pecliar natre of small territories in arctic breeding geese. Ryder (197) proposed that territory size in Ross' Geese, and perhaps in other colonial arctic geese, evolved as a balance between the need for spple mental food by territorial males and the costs of defending the female and nest site from conspecifics, which shold increase with increasing territory size. Territory size ths also may be affected by the size of prebreeding ntrient reserves of males. nglis (1976) rejected Ryder's hypothesis as an explanation for territorial behavior in celandic Pink-footed Geese becase: (a) nesting females were not harassed by conspecifics; and (b) territories broke down in early incbation, and ths were not defended when most feeding on the territory occrred. Rather, he arged that territories were most important in spplying food for the female prior to incbation. Minea and Cooke (1979) dismissed both of these hy potheses for Lesser Snow Geese and instead arged that male aggressiveness provides a "bffer zone" between the nest and potential nest parasites or rapists. They conclded that there was no good evidence to show that territorial ganders defended anything bt the female and the nest itself, althogh they conceded that all good nest sites were sed and competition for nest sites cold be an important isse. As evidence for their claims they noted that daily home ranges of breeding males and overlap between neighbors increased as the season progressed, and there was an inverse relationship between the freqency of forced coplation or parasitic laying and overlap in home ranges. nglis's (1976) hypothesis probably does not hold for Snow Geese becase they rely largely on stored ntrient reserves for reprodction. Simltaneosly, Owen and Wells (1979) presented a hypothesis for territoriality in Barnacle Geese based on stdies of both wild and captive birds. They arged that breeding grond food resorces are not so important for this species. Barnacle Geese typically nest where little food is available, and time spent in defense by males declines in late incbation. Also, Barnacle Geese defend water as well as land, and territory size varies with terrain. nstead, they reasoned that the main fnctions of territories are: (a) to maintain ownership of a nest site; and (b) to prevent egg dmping by intrders. Territory size is probably a fnction of the ease with which sites can be defended and perhaps the size of body reserves as Ryder (197) proposed (males lose weight dring incbation; the male sbjected to most intrsions lost most weight) (Owen and Wells 1979). They frther arged that in all species stdied to date, aggressiveness is most marked dring nest establishment becase all pairs are then intrders, bt in arctic species with short laying periods this shold decline qickly. Competition for nest sites and nonsexal harassment of females does seem to occr in some Canada Geese (Ewaschk and Boag 197; Cooper 1978). Of corse, there is no a priori reason to assme that territoriality in all these species has a single common fnction. Selection pressres will vary among species depending pon the extent to which birds rely on stored reserves for reprodction, the abndance of good nest sites, the prevalence of parasitic laying or forced coplations, and perhaps latitde or nesting habitat characteristics. More observational and experimental tests of specific predictions of varios hypotheses abot spacing in these and other colonial species are dearly needed.. Flocking and Spacing: The Costs and Benefits of Living in Grops Nearly all waterfowl are gregarios for most of the year. To nderstand the selective forces that lead to spacing ot, we think it important first to consider why waterfowl spend so mch of their lives in flocks. Life in flocks has rewards and liabilities that probably vary for each species. Seasonal changes in the balances of these factors case waterfowl to space ot. Most analyses of the costs and benefits of grop living

8 SPACNG PATTERNS 7 have been done on grops of nonbreeding animals. The profitability of different social strategies is mch easier to estimate for animals not immediately concerned with reprodction (e.g., isses like mate defense do not complicate matters). Althogh little of this work has been condcted on waterfowl, general patterns have emerged that are of interest to stdents of waterfowl ecology. A. Advantages of Groping-Understanding the "Baplan" of Waterfowl 1. Decreased Ssceptibility to Predation Flocks provide protection from predators in several ways. Grops provide "cover," becase each individal's probability of being killed declines as grop size increases (Krebs and Davies 1987; Bertram 1978; Plliam and Millikan 198). For attacks occrring from the periphery, a position near the geometric center of a flock provides individal prey with the greatest protection (Krk 196; Patterson 196; Hamilton 1971). Dominant individals or family grops shold seek central positions in flocks, as long as there are no other penalties for doing so (e.g., poorer access to food). Predators may also be confsed by explosive escape behavior of flocked prey (Bertram 1978). ndividals in flocks benefit from the grop's greater total vigilance and ths earlier detection of predators, while simltaneosly spending less personal time alert (Powell 197; Siegfried and Underhill 197; Kenward 1978; Dimond and Lazars 197; Bertram 1978; Krebs and Davies 1987; Plliam 197; Plliam and Millikan 198). n nonbreeding Barnacle Geese and Pink-footed Geese, individals spent less time alert as flock size increased, bt the total nmber of alert birds in the flock at anv one time still increased (Drent and Sweirstra 1977; Lazars and nglis 1978). Also, birds resting farther from conspecifics (greater "domain of danger" sens Hamilton 1971) were alert more often. Lazars ( 1978) reported similar reslts for wintering Whitefronted Geese. n breeding Canvasbacks, however, mated pairs were generally no more alert when alone than when associating with grops of Canvasbacks (Anderson 198a). This hints that intraspecific competition (e.g., for mates or food) might override benefits of increased flock vigilance and case alert behavior to remain freqent in flocked breeding birds. Brood amalgamation or creching (Mnro and Bedard 1977a, b; Kehoe 1986; Eadie et a/. 1988) may redce the probability of mortality for individals in sch grops. n these cases breeding waterfowl appear to grop in response to predation pressre.. Enhanced Foraging Abilities Groping can improve an individal's foraging efficiency in several ways: (a) enhanced ability to locate food (variosly called social facilitation or locale enhancement); (b) cooperative foraging; (c) enhanced competitive ability; and (d) risk aversion. Birds appear to exploit each other for information concerning good food patches, either directly by keying on sccessfl foragers, or indirectly by parasitism of information from individals at roosts (Krebs and Davies 1987). The hypothesis that bird roosts or colonies might form largely to act as "information centers" (Ward and Zahavi 197) has been hotly debated (Krebs 197; Plliam and Millikan 198; Wittenberger and Hnt 198). Regardless of what selective forces shaped sch aggregations initially, there is clearly the potential for information transfer in sch flocks, whether exploitative or altristic. Bailey (1981) arged that postbreeding Redheads, which typically occr in flocks of several thosand individals, exploit each other for information on good foraging parches. Ydenberg and Prins (198) showed that for 0 anatid genera and Anas species, commnal roosting is most common among seed and plant eaters and least common among carnivores, a pattern similar to that fond in corvids and fringillids and possibly related to the relative profitability of grop foraging on different types of foods. Y den berg et a/. ( 198) presented evidence that winrering Barnacle Geese spend more time in flocks following days of poor foraging sccess. They sggested that this occrred becase more individals had to consider moving to nfamiliar and npredictable feeding areas and ths needed more time to gather relevant information from conspecifics. n the only experimental stdy of dck foraging grops that we are aware of, Harper ( 198) fed bread crmbs to park Mallards and fond that birds chose feeding sires that approximated an "ideal free distribtion" (Fretwell and Lcas 1970) in response to variation in food spplies in patches. This was complicated, however, by a few dominant individals that managed co partially monopolize resorces. Other dcks showed some tendency to avoid sites occpied by those individals. Similar dominance relationships seem certain to affect payoffs associated with analogos choices in breeding waterfowl. Cooperative foraging can enhance the feeding efficiency of individals, sch as by herding prey, by enabling grops to prevail in competition with other grops for patchy resorces, and perhaps by allowing animals to crop renewable resorces efficiently (Krk 197; Bertram 1978; Plliam and Millikan 198; Mc Naghton 198; Krebs and Davies 1987). For waterfowl, cooperative foraging has not been demonstrated, bt flocks of mergansers (Mergs spp.) feeding on schools of fish, or perhaps other dcks preying pon invertebrates (e.g., nonbreeding Northern Shovelers), may

9 8 MCHAEL C. ANDERSON AND ROGER D. TTMAN : benefit by herding or gathering prey when they forage rogcrhcr. Competition for limited resorces might be enhanced by grop formation. Family grops in flocks of nonbreeding breeding geese and swans arc dominant over pairs or single birds, and large families rend to dominate small families (Boyd 19; Raveling 1970; Scott 1980; Gregoire 198), rhs providing famil y members with better access to choice feeding and resting sires. n a similar manner, pairs of some dcks are able ro dominate single birds in winter and presmably gain better access to an important resorce sch as food (Pals 198; Hepp and Hair 198). More ephemeral alliances of conspecifics, sch as those between npaired fema les (e.g., wintering Canvasbacks, Anderson npbl.) or possibly brood hens (Anderson pers. obs.), may also reslt in individals gaining competitive advantages by joining wirh others. ~tcnaghton (198, 1986) proposed that, for some social grazers, gregariosness may be advantageos in generating optimal food yield from pastres as well a ~ providing protection from pred:ors. This was also pre dieted by Cody ( 197). Prins et a/. ( 1980) and Ydcnbcrg and Prins (1981) showed rhat for Brant social grazing reslts in maximm individal gains per nit rime from grass swards. Risk a\ ersion may also favor rhc formation of social grops in birds that face food shorr:tges (Thompson et nl. 197; Caraco 1980, 1981; Caraco et a/. 1980; Plliam and l\ lillikan 198). These models predict rh:tt, for patchy foods, feeding in flocks redces chc variance of food intake for individal birds and may even in crease mean rates of food inta ke from ephemeral parches. We know of no pblished data concermng p.ltrerns of variance in feeding rates for waterfowl rha r mighr address this. Orhcr possible advantages of groping have been proposed (e.g., thermal advantages of clmping in harsh weather, energetic benefits of formation flight, learning migratory pathways), br these arc nlikely to be important for breeding waterfowl. B. Costs of Groping-Good Reasons ro "Space Or" Conterbalancing the advantages of groping arc cosrs associated with flocking. Among these arc: increased conspicosness to predators; increased exposre to disease or parasites; increased competition for mates, food, and other resorces; and increased risk of exploitation by conspecifics (e.g., brood parasitism, cannibalism) (Alexander 197; Hoogland and Sherman 1976; \Xrangham and Rbenstein 1986). ]\lost of these costs of locking pertain to waterfowl, br their importance may vary between males and females, and nder differem ecological circmstances. For most species, for mosr of the year, these costs mse be more than offset by benefits of groping. Spacing behavior in breeding waterfowl is likely driven by competition for one or more limited resorces (inclding mates) that cannot be efficiently acqired or held by individals in flocks. C. Conflicts over Groping Con(Jicrs over flocking strategies may be common in waterfowl, particlarly between males and females dring the breeding season. Sexal selection pressres may favor different choices of grop size or associates by males and females, and rhcse may conflict with optimal choices for, say, efficient foraging. t wold be insrrcrive ro model these tradc-offs in some well-stdied species, particlarly a nonrerrirorial dck sch as Lesser Scap, Canvasback, or Northern Pintail, and then look for behavioral evidence of specific hypothesized confliers. Laying or incbating females and their mares, who wait for rhem alone or in flocks of variable composition, might provide a sefl system for sch stdies..\!ales waiting for nesting mares seem to be balancing a hosr of rrade-offs concerning costs and benefits of associating with varios ocher birds. lv. Spacing Patterns and Agonistic Behavio r in Breeding Waterfowl A. A Srvey of Diversity among and within Species Breeding waterfowl space themselves according to a variery of patterns and exhibit strikingly different inrenstrics of aggression in the process. Nice ( 191) proposed a classification of rcrrirorialiry based principally pon how the defended area is sed. Many waterfowl occpy Type A, Type B, and Type C territories according to her classification. Many others, however, do nor fir any co n ventional category well. For convenience of analysis and discssion, we have tried ro classify each species nto one of 6 newly defined "Types" with respect to spacing behavior. \Y/e realize char any classification rends ro oversimplify rhe diversity fond in natre. While we find rhese gropings sef l and necessary for rhe discssions that follow, we do not advocate debate over the derails of sch a system or strict adherence to irs defin itions. The spacing patterns observed in breeding w:erfowl appear ro inclde rhe following: Type 1-A rypical Type A rcrrirory within which mating, nesting, and foraging by adlts and rong occr. Sch territories are defended year-rond by Mte Swans, steamers (Tach) crini), and ri\ er dcks ( ~ lerganertini and spp. of Anarini). The defended areas arc exclsive, aggression is intense, and the occpants arc typically perennially monogamos. Except for the Ble Dck (M. Williams pcrs. cornm. ), corrship and pair

10 SPACNG PATTERNS 9 formation often occr elsewhere before the territory is established. Type -A Type A territory that is exclsive and defended vigorosly for the breeding season only. t is sed for coplation, nesting, brood rearing, and foraging by adlts and the yong. Sch territories are defended by the migrant northern swans. Type -A Type B territory defended for the breeding season. Aggressive behavior is sally intense and the area is sed for coplation and foraging by adlts. Nests may be within the bondaries or nearby. There is little or no overlap in territory bondaries. Among the Anarini, the Northern Shoveler and the Ble-winged Teal defend sch territories. Type -A Type B territory defended for the breeding season where bondaries freqently overlap br areas occpied are temporally exclsive. These areas are sed for mating and some foraging by the adlts, and nests are bilt within rhe territory or dose by. Prairie Mallards defend this rype. Type -Breeding pairs are dispersed and a moving area arond the female and the nest site (sometimes) are defended, or there may be little evidence of aggressive behavior. This pattern is characteristic of Ayrhyini, Melanitta spp., and Northern Pintails. Type 6-A Type C colonial breeding sitation (sens Wittenberger and Hnt 198) where the female and a restricted area arond rhe nest are defended. Several species of geese, inclding the Snow Goose, provide examples of this. Apart from the spacing patterns classified above, there are brood territories defended by members of the gens Bllcephala, where the female aggressively defends an area sed by the brood. n the following section we describe general spacing patterns and agonistic behavior for each tribe, and focs pon specific traits exhibited by individal species (see also Appendix 8-l). 1. Anseranatini The Magpie Goose, the only member of this tribe, is highl> gregarios and breeds colonially (Type 6 spacing) (Frith and Davies 1961). As many as 1 nests have occrred over 100 acres Uohnsgard 1978). Within breeding aggregations, individals gather in family grops. t appears that a male and female will mate for life, bt instances of bigamy (1 male- females) have been reported both in the wild and in captivity (Frith 1967; Johnsgard 1961). Agonistic behavior by the Magpie Goose has been described only in the context of nest and brood defense where both males and females may threaten conspecifics. No dominance hierarchy is apparent (Frith and Davies 1961).. Dendrocygnini Spacing behavior of the nine species of whistling dcks belonging to this tribe is poorly nderstood. The Flvos Whistling Dck sometimes breeds colonially (Type 6 spacing) (Dickey and Van Rossem 19), bt has also been recorded as having dispersed nests (Palmer 1976, : 18). There are hints that Cban Whistling Dcks and Black-bellied Whistling Dcks may sometimes breed colonially, bt there is even less information abot spacing of the remaining species. General descriptions of grond-nesting White-faced Whistling Dcks in Nigeria sggest dispersed and well-concealed nests (Serle 19). Lavery (1970) described dispersed nests for Plmed Whistling Dcks in pland sires and for Wandering Whistling Dcks near or over water. Perennial monogamy has been docmented in Blackbellied Whistling Dcks (Bolen 1971), althogh occasional mate changes occr (Delnicki 198). Long-term pair bonds are sspected in other species. Males incbate and assist with brood care in several species (D'ombrain 19a, b;johnstone 197; Flickinger 197; Bolen and Smith 1979; Chronister 198). Whistling dcks occpy stable terrestrial habitats and are primarily vegetarian. Most species are sexally monomorphic Uohnsgard 1978). The context and timing of agonistic behavior by Whistling Dcks have not been adeqately described, nor have details of male and female participation in threats and chases. However, there is little evidence of aggression among breeding birds. Pairs of breeding Black-bellied Whistling Dcks are often fond in loose flocks. Aggression between pairs may occr when they move close together at feeding or resting sites (Type spacing), or sometimes ar nest sites (C. D. Chronister pers. comm.). Nest defense is ephemeral, thogh, and consists mainly of males defending nest cavities that prelaying females are exploring. The White-backed Dck is more aqatic than others in this tribe. After pair formation, agonistic behavior is associated with defense of a mare (Clark 1969). Both males and females may threaten conspecifics bt this is not clear from pblished descriptions.. Anserini This tribe incldes eight species of swans and forteen species of geese. The behavior of this grop is generally well described (Kear 197). The white swans are strongly territorial, and the geese are either colonial or defend their mates and an area arond the nest. Spacing information is lacking for the Coscoroba Swan; for Black-necked Swans, Bewick's Swans, Swan Geese, Bean Geese, and Hawaiian Geese, only sparse descriptions appear in the literatre. The Mte Swan may pgnaciosly defend Type 1 territories year-rond, or in climates where the water freezes dring winter, they defend Type territories for a long period arond the breeding season (Kear 197;

11 60 MCHAEL G. ANDERSON AND ROGER D. TTMAN. Cramp and Simmons 1977; Birkhead and Perrins 1986). n rare instances in Erope some breed in colonies (Cramp and Simmons 1977; Birkhead and Perrins 1986). Type territories are defended by the northern swans (Banko 1960; Hansen et al. 1971; Kear 197; Palmer 1976; Cramp and Simmons 1977). n contrast, the Black Swan of Astralia and New Zealand breeds in large colonies (Kear 197; Johnsgard 1978). There are sggestions that Black-necked Swans and Coscoroba Swans have colonial tendencies, althogh solitary nesting appears to be normal (Kear 197). When not breeding or defending territories, swans flock together and with other waterfowl. Family bonds last almost a fll year, ntil the pair next breeds (Johnsgard 196; Scott 1980). All are sexally monomorphic and appear to mate for life, althogh data on marked birds are scarce. n some Mte Swan poplations bigamy and mate changes have been docmented at low freqencies (%-9%) (Kear 197; Palmer 1976; Birkhead and Perrins 1986). Territorial defense by Mte Swans is normally evoked by other adlt swans, and is most vigoros jst before and dring nesting (Kear 197; Birkhead and Perrins 1986). Mte Swans have killed conspecifics and ocher water birds in their territories (Palmer 1976; Birkhead and Perrins 1986). Their conspicos white plmage serves as a continos advertisement of their presence (Wynne-Edwards 196). Threats, Rotation Displays (Lind 198), and mild attacks and withdrawals at a territorial bondary, primarily by males, sally serve to repel interlopers. Fights occasionally occr, thogh, where individals cross necks, grab each other, and beat each other with their wings. The territory owner most often wins these engagements. The migrant northern swans occpy large territories. Areas defended by Trmpeter and Tndra Swans range from 0 to 60 ha and 10 to 0 ha per pair respectively (Kear 197). Territory acqisition is necessary before they will nest. Territories are defended before nest site selection and ntil cygnets are at least half-grown (Banko 1960). Trmpeter, Whooper, Tndra, and Bewick's Swans are all alike in their agonistic behavior (Johnsgard 196; Kear 197) and are very vocal. Territorial defense begins with a warning display where birds exhibit half-opened qivering wings and stretched necks, accompanied by calling. Before oven attack, the wings are spread flly. f an intrder persists, the defending male will chase it well beyond the territory bondary before retrning to his mate to perform the Trimph Ceremony (Heinroth 1911; this book, chapter 7). n aerial chases the prser holds its head high with neck arched, and will occasionally nip feathers from the tail of the prsed bird (Palmer 1976). Fights can occr in which opponents swim or stand breast-to-breast, bffeting each other with their wings. Territorial interactions are most freqent dring nest bilding and laying, when other swans are searching for territories. Aggressive behavior of Black Swans, Black-necked Swans, and Coscoroba Swans has not been stdied in detail, bt their displays seem somewhat similar to those of Mte Swans (Johnsgard 196). Colonial nesting (Type 6 spacing) is characteristic of breeding Graylag, Bar-headed, Lesser Snow, Ross', Barnacle, and Red-breasted Geese, and Brant. Aggregations of nests in loose colonies are reported for Pink-footed Geese (Owen 1980), Greater Snow Geese (Palmer 1976), and Emperor Geese (Eisenhaer and Kirkpatrick 1977). Graylag Geese may breed with nests dispersed as well as in scattered loose colonies (Cramp and Simmons 1977). The Canada Goose, a species of many races, shows spacing behavior ranging from territoriality (Type ) throgh nest and mate defense {Type ) to colonialiry (Type 6). Bean, White-fronted, and Lesser White-fronted Geese exhibit Type spacing behavior. Minimm nest separation within dense goose colonies has been recorded as follows: 1m, Graylag Geese (Newron and Kerbes 197); m on cliff ledges, Barnacle Geese (Owen 1980);. m, Ross' Geese {Ryder 1967); and m, Lesser Snow Geese (Kerbes 197). The densest nesting concentration has been recorded for the Barheaded Goose, where 100 nestslha occrred on an island at 000 m altitde {Owen 1980). Geese form long-standing pair bonds that may endre their entire lives (Johnsgard 196; Palmer 1976; Ogilvie 1978; Owen 1980). Yearlings do not leave their parents ntil they are forced away by the adlts dring or shortly after their retrn to the breeding gronds in spring (Ogilvie 1978). n contrast, Brant families may break p in late fall when parents no longer defend the grop (Jones and Jones 1966). Geese are gregarios otside the breeding season as they gather in large flocks. Within large overwintering flocks, freqent agonistic behavior is associated with the establishment of dominance relationships. n Canada Geese, size of family grops determines position in a social hierarchy (Raveling 1970). Paired males are the principal aggressors in interactions between families relating to food and favorite resting locations. Aggressive behavior becomes more apparent among jvenile birds as pairs form 7-18 months before their first nesting attempts (Owen 1980). Yong males defend preferred females and try to gide them away from competitors. Prsit flights occr in exceptional cases of intense competition for mates where there are extra npaired matre males (Owen 1980). Dring the breeding season all geese exhibit aggressiveness, whether they defend a mate and small territory in a colony, a moving area arond a mate, or a larger territory. Within the defended areas pairs display, coplate, nest, and do some feeding. n general, males are the

12 SPACNG PATERNS 61 primary defenders of these areas. For Barnacle Geese, territory size remains constant from nest site selection throgh incbation, after which broods leave the territory (Owen and Wells 1979). Size of the area defended declines dring incbation for Lesser Snow Geese (Minea and Cooke 1979) and for Cackling Canada Geese (Mickelson 197; Eisenhaer and Kirkpatrick 1977), bt not for other Canada Geese (Ewaschk and Boag 197; Cooper 1978). Ross' Geese contine to defend territories well into incbation (Ryder 197), while in Pink-footed Geese defense declines early in incbation (nglis 1976). n Giant Canada Geese, moving areas centered arond the female and nest are defended from arrival throgh incbation. Vigoros male/male fighting is restricted to prelaying and laying periods and dring the first week of incbation (Balham 19; Cooper 1978). Both males and females participate in aggressive displays, bt only males commonly chase and fight intrders. Sch chases are sally directed at intrding females. n captive Western Canada Geese, male aggression peaked from jst before nest initiation throgh incbation, and males and females were very aggressive early in the brood-rearing period (Akesson and Raveling 198). Dring the laying period, females as well as males were aggressive toward intrding pairs. Aggregation of other sbspecies of Canada Geese in loose colonies appears to reslt from habitat satration nder conditions of high poplation density (D. Raveling pers. comm.). n these circmstances geese may fight vigorosly, and occasionally deserr nests.. Cereopsini The Cape Barren Goose is the only member of this tribe. t appears to be territorial, perhaps most closely fitting or Type category. Giler (1967) recorded a maximm breeding density of one pair per 1. acres on sabella sland soth of Astralia. The closest nests were 1 m apart. Territories were strongly defended from late Janary throgh nesting in May and Jne. Adlts and yong remained on their territories for approximately 6 weeks after hatching. Paired birds were fond throghot the year, so bonds may be long-lasting. This sedentary terrestrial species is very aggressive in captivity. Males may rn or fly at opponents to strike them with wings that have bony knobs at the wrist, and pairs perform gooselike Trimph Ceremonies Uohnsgard 196).. Stictonettini One species, the Freckled Dck, fond in soth central Astralia, is classified in this tribe. Few observations of breeding behavior have been recorded, so the spacing behavior of this species is poorly nderstood. On several lakes, Braithwaite (1976) fond pairs and nests to be well dispersed. He considered pair bonds to be weak becase individals that were incbating were fond alone near nest sites. 6. Tadornini nformation abot spacing exists for only 8 of the 1 species in this tribe of sheldgeese and sheldcks. The Ble-winged Goose fond in Ethiopia is a bird of highlands and shows little aggressive behavior. Males in captivity may threaten smaller waterfowl and rn back to their mates in behavior similar to a Trimph Ceremony Uohnsgard 196). The Falkland Upland Goose defends Type territories that inclde feeding and nesting areas and access to water for breeding pairs (Smmers 198). Yong are often, bt not always, reared in the same territory. Males are primarily responsible for territory defense, and also evict yong from their natal home to which families retrn in spring prior to the next breeding attempt. Some adlts, especially failed breeders, leave their territories to molt. Most (76%) marked pairs retained the same mate between years and generally reoccpied the same breeding territory. The Kelp Goose defends feeding areas along marine shorelines (Pettingill196), bt generally little is known abot the behavior of the other for Soth American sheldgeese of the gens Chloephaga. They are terrestrial grazers and seem highly aggressive. The sheldgeese have bony protberances at the wrists that cold act as effective weapons in fights. The Orinoco Goose appears to occpy a Type 1 territory year-rond in the tropical Orinoco and Amazon basins. Pairs were observed spaced 1.6 km apart along river areas in Venezela (Delacor 19-6). Pairs of Egyptian Geese are very aggressive in their discrete Type territories, which are defended from well before egg laying throgh the rearing of yong (Cramp and Simmons 1977). Birds are gregarios except when nesting, bt a small part of the poplation seems to be dispersed in pairs throghot the year, each on a small permanent water area. Territory size averages abot 1 ha, and fighting is freqent. Long-term pair bonds seem likely bt are nconfirmed. Both parents tend yong. The Radjah Sheldck of Astralia is highly territorial and very aggressive Uohnsgard 1978). f conditions are favorable, a pair or family may remain within its territory all year. f there is droght, the birds move elsewhere. Ths this spacing system cold be classified as Type 1 or Type. Pairs appear to be spaced every.8 km (1.7 mi) along rivers, or they may se pools or lagoons. Nests typically are located in tree cavities within the territory or close by (Frith 1967). Northern, Astralian, New Zealand, and Rddy Sheldcks gather in large flocks otside the breeding season (Frith 1967; Yong 1970; Riggert 1977; Cramp and

13 6 MCHAEL G. ANDERSON AND ROGER D. TTMAN 'l ', J Simmons 1977; Williams 1979; Patterson 198). Within winrer flocks, agonistic behavior in the form of threats, attacks, avoidances, and fights is apparent as pair bonds are formed and tested. A dominance hierarchy exists among flocked Northern Sheldcks (Patterson 198) and is likely among New Zealand Sheldcks (Williams 1979). Pair bonds appear to be long-lasting, bt mates may be apart for some time each year. A majority of srviving Northern Sheldcks pair again with previos mates after migration each spring (Patterson 198). The dcks gradally become more territorial as the nesting period approaches, exhibiting site attachment and hostility toward conspecifics, while defending shoreline feeding areas. This can be classified as Type spacing, yet nests sitated in brrows are almost always located otside the territories. The pattern of spacing in Rddy Sheldcks appears to be the same (Cramp and Simmons 1977). Astralian Sheldck nests may be as far as.8 km away from feeding areas (Frith 1967). Northern Sheldck pairs are toleranr of one another away from their terrirories, as females nest close together (Parterson 198). Northern Sheldck territories, which average 1.7 ha, are not exclsive. There is overlap in areas sed by adjacent pairs (Yong 1970; Patterson 198). Srprisingly, territory size does not decline with increasing food density (Bxton 197). These feeding areas are defended by postring and calling, with only occasional attacks and rare fights. The Alert postre is often enogh ro indce retreat between well-established neighbors. New neighbors may fight, sally males with males and females with females (Yong 1970; Patterson 198). Males spend mch more time in aggressive interactions than do females, and are principally responsible for territory defense. For Northern Sheldck pairs, a territory serves as a meeting place and an exclsive ndistrbed feeding area from prelaying throgh incbation, bt the birds take their broods elsewhere to feed (Patterson 198). New Zealand Sheldck pairs most often rear their yong on ponds within their territories, and pairs are fond on their territories for considerable periods before and after nesting and brood rearing (Williams 1979). These characteristics, along with defense of an exclsive area dring the breeding period, lead s to classify their spacing as Type. For Astralian Sheldcks inhabiting a small island off western Astralia, the territory is an important feeding area for the brood (Riggert 1977). The average size of its defended area, which incldes a critical fresh water seepage, is 0. ha. nterestingly there is a female-biased sex ratio in the poplation Riggert (1977) examined. Perhaps conseqently, females of this species are the principal initiators of aggression, particlarly dring and jst after pair formation. Unpaired females spend mch time and effort seeking mates, and paired females are very aggressive when garding mates (Riggert 1977). After territories are established and nesting begins, males become the primary defenders, bt females also engage in threats and prsit flights at the onset. 7. Tachyerini All for steamer dcks of this tribe are highly aggressive and strongly territorial. These large diving dcks are perennially monogamos and occpy stable, strctrally simple water areas with predictable animal food resorces. Both sexes are involved in territorial disptes. Defended areas are contigos and reglarly spaced. The three flightless species (Magellanic, Falkland, and White-headed Flightless Steamer Dcks) are sedenrary and defend Type 1 territories year-rond in marine habitats (Pettingill 196; Weller 1976; Livezey and Hmphrey 198a). The Flying Steamer Dck reqires its Type territory to breed, bt birds may leave the area otside the breeding season. All steamers are interspecifically territorial, attacking congeners most intensely, bt other waterbirds as well (Livezey and Hmphrey 198a, b). Flying Steamer Dcks have killed other dcks (Red Shovelers, Yellow-billed Pintails) (Nechterlein and Storer 198a). Knobs on the birds' wings are sed in intense long-lasting (p to 0 min) fights (Pettingill196). Within the tribe, behavior patterns appear to be very similar. Males patrol their terricories vocalizing. Threat behaviors are performed by both males and females (Moynihan 198; Weller 1976; Livezey and Hmphrey 198; Nechterlein and Storer 198a). Attacks often lead to violent battles with wing thrashing dring which male Falkland Flightless Steamer Dcks have killed each other (Cawkell and Hamilton 1961 ). Often pairs engage opposing pairs with male fighting male and female fighting female. 8. Cairinini This extremely heterogeneos tribe may not be a valid taxonomic grop (Livezey 1986), and most of the species placed here have been little stdied. Spacing behavior is poorly known for 11 of 1 species. One grop of Cairinini are large in size, exhibit prononced sexal size dimorphism bt little plmage dimorphism, have simple vocalizations, weak pair bonds, and appear to be qite aggressive. The Mscovy typifies this grop. Males are aggressive to all other individals, inclding females, bt they do not seem to defend terricaries. Males may fight vigorosly with each other over positions in a social hierarchy, sing their sharp-clawed feet and their wings. Pair bonds do not appear to exist, and forced coplation (FC) is often observed. African Comb Dcks show extreme sexal size dimorphism and a complex mating system that incldes monogamy, harem polygyny, sccessive polygyny, and

14 SPACNG PATERNS 6 FC (Siegfried 1978). Males defend large (ca. 7 ha) Type or territories from which rival males are exclded. Conspicos Wing Flap displays seem ro serve as longdistance threat displays. Fights between rival males are common. Males that lose in competition with rival males generally lose both their territory and their harem. Female/female aggression is mostly associated with prospecting for a nest cavity. A dominance hierarchy may exist among females mated to the same polygynos male. Males freqently have to defend mates from FC attempts, so they are vigilant for other males as well as for predators. Males provide no parental care, bt hens may raise broods ndistrbed within a male's territory, especially if he is still defending a later-breeding female. Members of the genera Nettaps, Aix, Pteronetta, Callonetta, Chenonetta, and Amazonetta are small dcks that matre in their first year, have elaborate plmage, and may have relatively strong pair bonds. Most species are cavity nesters and have long incbation periods. When paired, captive males tend to chase other tpproaching males, bt little else is known abot their spacing behavior. Hartlab's Dck appears to be sedentary and territorial year-rond (Johnsgard 1978), inclding dring the brood-rearing period. The Green Pygmy Goose defends small feeding areas, occasionally fights, and seems ro maintain a strong pair bond year long (Frith 1967). The North American Wood Dck is not territorial, bt males defend their seasonal mare (Grice and Rogers 196). Males may chase each other and even beat each other with their wings, bt sally withot injry. The availability of nest cavities may determine spacing and density reslting in Type spacing. Mandarin Dcks appear to behave similarly (Brggers 1979). Maned Dcks are terrestrial grazers that associate in flocks for most of the year, bt nest in isolated tree cavities (Kingsford 1986). Pairs move among flocks, bt family grops persist for some time after the yong have fledged. Long-term pair bonds are probable (% of marked dcks remained together for > 1 year), and males take an active role in parental care (Kingsford 1986). Frith ( 198) believed that Maned Dck males defended brood territories from other males as soon as their mates commenced nesting. n a -year srdy of marked birds, Kingsford ( 1986) fond no evidence of territorial defense ocher than a single prsit flight. 9. Merganettini The Torrent Dck is the only species classified in this tribe. This small dck (90 g) lives in fast-flowing montain streams of western Soth America. Members of a pair cooperatively defend a Type 1 territory with well-defined bondaries that is occpied exclsively (Eldridge 1986b). Pairs appear to mate for life. Both sexes display at territory bondaries where most aggressive enconters occr. Territorial birds confront intrders of the same sex. nteractions involving pairs and lone males are long, and intense and elaborate displays are typically performed withot overt aggression (Eldridge 1979). Eldridge (1986b) observed only one fight lasting less than one second, involving a flrry of wing action between two females, bt Torrent Dcks show their wing sprs prominently in several display postres. 10. Anatini A diversity of spacing systems (Types 1,,,, and 6) has been recorded for this large cosmopolitan tribe known as the dabbling dcks. Panerns of spacing have been identified for of 9 species. No members of this tribe are known ro defend seasonal territories within which pairs also rear their yong (Type ); only rarely are any species colonial (Type 6). One species, the Speckled Teal, may sometimes nest in small colonies sing the nestholes of Monk Parakeets (Weller 1967), and for Anas species sometimes nest in dense concentrations on islands. Species defending Type 1 territories are fond in river habitats. These inclde the Ble Dck (tribal affiliation ncertain), the Salvadori Dck, and the African Black Dck (Kear and Steel1971; Kear 197; Siegfried 1968; Ballet al. 1978; Eldridge 1986a). Another species that might behave similarly is the Bronze-winged Dck, bt very little is known abot it (Johnsgard 1978). They are fond year-rond on territories that are aggressively defended. These species eat predominantly animal foods, lay small cltches ( to 6 eggs), and have very strong pair bonds. Territory ownership seems to be a prereqisite for sccessfl reprodction. The Salvadori Dck appears to be perennially monogamos. The African Black Dck and the Ble Dck have pair bonds lasting a year and longer, bt bonds may be broken following reprodctive failre, injry, or territorial displacement (Ball et a/. 1978; McKinney eta/. 1978; Eldridge 1986a). Extrapair FCs have not been recorded in any of these species. Territorial Ble Dcks and African Black Dcks react aggressively coward potential rivals whenever these are encontered in their territories. Each pair member typically confronts intrders of its own sex. Close cooperation of mates is an important factor in defense. n Ble Dcks, territorial adlt males interact with adlt males, which are the most freqent intrders, and females confront intrding females and jveniles of both sexes (Eldridge 1986a). A Ble Dck intrder that scceeds in evicting a resident bird of the same sex may also take over the loser's mate (M. Williams pers. comm.). For this species, territories are sally bordered by an ndefended area that is sed by npaired nonterritorial birds. For African Black Dcks, territorial bondaries are hotly dispted, and zones of overlap not exceeding 10% occr at the bondaries (Ball et al. 1978). For both spe-

15 . i l 11 : i i : '! 6 MCHAEL G. ANDERSON AND ROGER D. TTMAN cies, mates cooperate to evict intrders sing threats of escalating intensity and Swim-off or Walk-off mane vers. f threat displays are ineffective, they resort to aerial chases and fights. The three species of river dcks fight by grasping their opponents with their bills and pmmeling with their wings (Kear 197; McKinney et al. 1978; Eldridge 1986a). All have wing sprs that appear to be sed in the fights. Fighting by A. sparsa is considered by Ballet a/. (1978) as a major case of displacement and death of territorial pairs. Eleven Anas species defend Type territories. Strong territorial defense from before nest site selection ntil at least midincbation has been well established for Amer ican Wigeon (Wishart 198), Gadwall (Dwyer 197; Titman and Seymor 1981), American Black Dck (Seymor and Tieman 1978), Ble-winged Teal (Stewart and Titman 1980), and Northern Shoveler (McKinney 1967; Seymor 197a, b). The Cape Teal behaves in this fashion in captivity (Stolen and McKinney 198). Or interpretation of general descriptions of breeding behavior indicates that Brown Teal (Weller 197a; Dmbell 1986), Crested Dck (Johnsgard 1978), possibly Garganey (Cramp and Simmons 1977), Cinnamon Teal (McKinney 1970), and Cape Shoveler also belong in this category. Except for the Brown Teal, which resides yearrond where it breeds, these dcks arrive paired after spring migration and establish territories chat females choose and males defend. Males of Northern Hemisphere species sally depart before the yong hatch, and females lead their broods co find food where it is abndant, most freqently beyond the original territory. f males do remain with their broods, they rarely stay long, departing within two weeks, and the extent of their participation in parental care is problematic (McKinney 198, 1986). n contrast, males of tropical and Sothern Hemisphere species (e.g., Chiloe Wigeon, Cape Teal, Brown Teal, Chestnt Teal, and sometimes Speckled Teal) contribte parental care and do not leave their mates and broods ntil the latter part of the broodrearing period (Weller 1968a; Kear 1970; Siegfried 197; McKinney 198; Norman and McKinney 1987). Among those defending Type territories is a grop, the ble-winged dcks, whose members are recognized by behavioral and morphological similarities inclding a diagnostic ble patch on the pper wing (McKinney 1970). Members inclde the Northern, Cape, Red, and Astralasian Shovelers, and Ble-winged Teal, Cinnamon Teal, and Garganey. McKinney (1970) described the displays of for species of this grop in detail. Bondary disptes are intense and freqent (McKinney 1970). Pairs show strong site attachment to small, welldefined areas that are defended (Seymor 197a). Ag gressive interactions typically proceed from lowintensity threats to aerial chases and Circlar Fighting. Males sally direct their threats to the female of an intrding pair, bt there are freqent male-male interactions involving only two birds (McKinney 196a; Stewart and Titman 1980; Titman and Seymor 1981). Along with visal displays, wing noise by Cape and Northern Shovelers appears to advertise territorial occpation (McKinney 1970). The Gadwall arrives on breeding gronds in North America p to a month before laying eggs (Gates 196; Dwyer 197 ). Pairs show increasing intolerance of other pairs by exhibiting threat behavior and brief aerial chases (Dwyer 197). Meanwhile, pairs are more tolerant of npaired males. Jst prior to egg laying, territorial defense increases, and Three-bird Flights become the principal means of expelling intrders. A resident male chases the female of an intrding pair. Gadwall Threebird Flights are relatively short, and the resident male retrns to his mate on the territory (Dwyer 197). Progressing throgh the laying period into incbation, the resident male expands his range and spends more time away from his activity center. f his female leaves her nest while he is away, she may be sbjected to "har rying chases" (Gates 196), which on rare occasions reslt in extrapair FC. Occasionally Gadwalls nest on islands in concentrations as dense as 7 nestslha (Hammond and Mann 196; Newton and Campbell 197; Debbert eta/. 198). n sch cases, males may defend separate territories on mainland shores of the lake (lokemoen et al. 198). Agonistic behavior of the American Wigeon is very similar (Wishart 198). Here too, in contrast with the ble-winged dcks, territorial males are mch more likely to chase females of intrding pairs. n addition, Wishart (198) described Tandem-swimming where the territorial resident male swims behind a lone male that is swimming away in avoidance. Gadwall do this as well. Mean territory size is 7.8 ha. Wishart (198) observed American Wigeon in 666 interspecific interactions, most freqently involving Gadwall and Ble-winged Teal, both of which they dominated. The American Black Dck exhibits similar behavior, with males defending tidal pools in coastal areas, chosen by females, and ranging in size from 0.16 to.8 ha (Sey mor and Titman 1978). Territorial defense lasts 7- days, from abot days prior to laying ntil midincbation. Prsit flights directed by resident males toward intrding paired females are the principal means of defense. Althogh the chased female's mate may not join the flight, he leaves the territory after his mate is expelled. The majority of interactions occr between immediately adjacent pairs dring a -to- day period of territory establishment. After this there are typically only brief, ritalized enconters between neighbors at territory bondaries, and the most intense interactions involve strangers. On territories, resident males are

16 SPACNG PATERNS 6 mildly intolerant of npaired males in the presence of their mates (Seymor and Titman 1979). Spacing patterns have not been described for the astral teals other than the Brown Teal. For the flightless race of this species on Ackland sland, Weller (197a) recorded that males or pairs spaced along a coastal shoreline defended territories with aggressive displays. Chases occrred when females were present, and chestto-chest fighting was observed between territorial males (Weller 197a). Defense of Type territories occrs in Mallards (Titman 198) and Yellow-billed Dcks (Skead 1976). Areas defended are sally qite large, and they freqently overlap those of neighbors, althogh neighbors are not fond in regions of overlap at the same rime. Pairs that have migrated to the breeding gronds establish territories, coplate there, nest within the defended area or close by, and forage extensively there, althogh they may feed elsewhere, too. Males behave aggressively on their territories ntil midincbation, when they abandon the sires and their mares. Territorial behavior of Mallards in prairie pothole habitat is similar to that described for the Gadwall, American Wigeon, and American Black Dck above, except that areas occpied are larger, there is significant overlap of territory bondaries, the territorial period is shorter, and Mallards more freqently engage in extrapair FCs (Titman and Seymor 1981; Titman 198). n habitat with a high densiry of potholes, Mallards confine themselves to territories averaging 16 ha for a period of 1 to days from prenesting ntil early incbation (Titman 198). Becase Mallards occpy rather large home ranges, it is difficlt to exclde intrders from all parts of the territory at the same rime, which may explain why overlaps develop. Three-bird Flights are initiated by resident males as intrders are intercepted in flight or encontered on territories. Bondary disptes between neighbors involving a sccession of prsit flights also arise. A pair rypically dominates a pond in the absence of a neighboring co-owner. Territories are ths at least temporally exclsive. Tieman ( 198) never observed fighting among wild Mallards, bt fighting has been observed in contexts related to mate defense in artificial conditions sch as parks (Weidmann 196; Tieman and Lowther 197). Mallards are very adaptable birds char live in a wide variety of habitats, and their spacing is profondly inflenced by where they live. Stdies of spacing in semiarid pothole habitat and a forested environment illstrate this. Gilmer eta/. (197) fond that laying females occpied an average of 70 ha in a forested habitat, while Dwyer eta/. (1979) recorded se of ll ha dring rhe same breeding stage in semiarid pothole habitat. Dzbin (1969) noted that Mallard breeding reqirements are met in a wetland complex sally consisting of several ponds, particlarly if they are smaller chan acres. The stdy area of Dwyer et a/. ( 1979) had -1 wetland basins/km, which may partly explain the large difference in territory size (111 ha vs. 16 ha) between their stdy and Titman's (198) stdy where there were 7 wetlandslkm Some Mallards se islands for nesting, and these birds sally establish territories on lakeshores or wetlands as far as. km away from their nest sire (Newton and Campbell 197; Debbert eta/. 198; Lokemoen et al. 198). Mallard nests were fond as close as 0 em apart, averaging.7 m apart and reaching a maximm densiry of 89/ha on Miller Lake sland, North Dakota (Lokemoen et al. 198). Sparse descriptions of the spacing of Erasian Wigeon indicate that it probably defends Type territories (Hilden 196; Cramp and Simmons 1977). Typically pairs are well dispersed in semiwooded freshwater habitats dring the breeding season. Males defend their mares and a waiting area against conspecifics (Hilden 196). Newton and Campbell (197) recorded a dense breeding concentration on an island where -0 pairs nested 8 to m apart. Type spacing probably occrs in Green-winged Teal (McKinney 197), Hottentot Teal (Clark 1971), Pinkeared Dck (Frith 1967), and Marbled Teal (Cramp and Simmons 1977) (tribal affinities of the latter two species ncertain). Males defend their mares from conspecifics. A moving area arond the female is defended from rhe rime of pair formation ntil the female has finished laying eggs. nterestingly these are all small dcks, weighing between 0 and 00 g. Srprisingly little is known abot the breeding biology of wild Common Teal, Green-winged Teal, or any of the other species in this grop. Male Common Teal sally leave their mares late in the laying period prior ro incbation (Palmer 1976). Cramp and Simmons (1977) noted rhar in rhe Eropean race there is no indication of male defense of a territory or waiting area and that even dring laying males associate peaceflly. Hostile enconters do occr dring pair formation in captive birds, ranging from males threatening each other to fighting (McKinney 196b). Paired birds maintain strong bonds, br males spend very little rime chasing other males (McKinney 197). n captivity there is a high freqency of exrrapair FC (McKinney 197; McKinney eta/. 198). For Northern Pintails, breeding pairs and nests are well dispersed with little evidence of aggressive behavior (R. Smith 1968; McKinney 197, 197; Titman and Seymor 1981). Mates are nor closely defended, and intermale aggression is rarely observed (R. Smith 1968). t is possible that other pintails (e.g., Red-billed and Brown Pintails) behave similarly, bt these species re-

17 66 MCHAEL G. ANDERSON AND ROGER D. TTMAN i ' ''. :'. qire stdy. Weller (197b) observed some aggressive behavior among breeding Brown Pintails. Northern Pintails are seminomadic and se widely scattered, temporarily available shallow wetlands. Althogh winter social display is intense (R. Smith 1968), Northern Pintail pairs are not conspicosly spaced in migratory flocks (Palmer 1976). Weak pair bonds form bt dissolve early. Females chase other females away and appear to be more aggressive than males in defense of rheir mares. Males will threaten other males if their mares ncite (R. Smith 1968). On rhe breeding gronds, drakes are highly mobile and nor attached to specific sires (R. Smith 1968; McKinney 197; Derrickson 1978). There is a high freqency of prsit flights associated with social display ntil females are laying, and then associated wirh freqent FC (R. Smith 1968; Derrickson 1977; McKinney eta/. 198 ). Males do not consistently defend their mares dring FC attempts. n contrast, White-cheeked Pintail males in the Bahamas appear to defend variable territories from prelaying throgh incbation (some for as long as weeks into brood rearing) (L. Gminski-Sorenson 199). Males are very aggressive, particlarly dring their mares' laying period, and exclde all conspecifics from defended areas. Some males defend discrete areas even in the absence of their mares; others seem to defend a moving area arond the female. f pairs intrde, resident males sally direct chases at the females, as in typical Anas Three-bird Flights. Despite their commitment to territorial defense, most males freqently engage in extrapair FC (L. Gminski-Sorenson 199). 11. Aythyini Little qantitative information is available on spacing and chasing behavior in the pochards, bt something like Type spacing appears to be characteristic of most species. The pochards are a relatively homogeneos grop of medim-sized ( g) diving dcks that breed in freshwater habitats. There are 1 species in genera, distribted worldwide. The Canvasback of North America is the best-stdied member of the tribe. Breeding pairs occpy large and broadly overlapping home ranges (Erickson 198; Dzbin 19; Sgden 1978; Anderson 198a) that vary in size by consort stats and reprodctive stats. Breeding Canvasbacks commonly associate with conspecifics, most freqently jst after arrival on the breeding gronds and least freqently dring laying periods or renesting intervals. Males are primarily responsible for spacing, bt there is no evidence of anything more than ephemeral defense of any area other than the immediate space arond a pair. Agonistic behavior is highly variable (Anderson 198a). Low-intensity threats and short chases are typical within posrarrival and early prelaying flocks. Most of these reslt from contests over feeding or resting sires or are associated with cortship. As pairs near laying, males become mch more intolerant of most other dcks, especially male Canvasbacks. Chasing then is more common and less dependent on pair activity. Males remain vigilant, aggressive, and responsive to their mates ntil early in incbation, thogh the inten sity of this behavior declines rapidly as incbation proceeds. Dring renesting intervals and second laying periods, males are even more protective of their mates and more intolerant of conspecific males and other birds than dring the initial prelaying and laying periods (Anderson 198, 198a). Canvasbacks engage in relatively few interactions with other species of waterbirds, br win most sch enconters, presmably becase of their relatively large body size (cf. Bailey and Bart 197; Alexander 1980). Attacks on other species peak dring laying periods, and other diving dcks (Aythya spp., Rddy Dcks) are chased most freqently (Anderson 198a). This may be linked to greater overlap with other dcks in foods (pri marily benthic invertebrates) selected by Canvasbacks dring the laying period. Nest parasitism by Redheads has been shown to redce Canvasback reprodctive sccess (Erickson 198; Weller 199; Olson 196; Sgden 1980; Stodt 198; Boffard 198 ), and Redheads are threatened, chased, and avoided more freqently than any other species (Anderson 198a). The Erasian Pochard and the Ring-necked Dck appear to have spacing patterns very similar to that of Canvasbacks, involving broadly overlapping home ranges and consistent defense of, at most, a small area arond the pair (Bezzel 1969; Mendall 198; W. Hohman pers. comm.). Redheads show a pattern of agonisric behavior similar to Canvasbacks (Low 19; Weller 199; Sayler 198 and pers. comm.; Anderson pers. obs.), bt appear to move even more widely and spend considerable time targeting hosts for parasitic egg laying. Lesser Scap pairs normally confine their activity to a few small lakes, at least in central Manitoba (Hammell 197). No consistent defense of any territory is indicated for this species, intraspecific aggression appears to be infreqent (Hammell 197; A. Afton pers. comm.), and home ranges of individal pairs, thogh smaller than those of Canvasbacks, overlap to some extent. Tfted Dcks are highly gregarios and sometimes nest colonially as well as singly (Newton and Campbell 197; Cramp and Simmons 1977). Derails on movements of breeding pairs are lacking, bt in a London park poplation, males defended small moving areas arond the pair from both males and females (Gillham 1987). Red-Crested Pochards also may have a pattern of spacing similar to Canvasbacks (Lind 198, 196;

18 SPACNG PATTERNS 67 Cramp and Simmons 1977; Gillham 1987), thogh there have been few observations of wild birds. There are insfficient data for any other Aythya species to provide sefl comparisons. Few species have been stdied dring the breeding season sing marked wild birds. However, we can find no pblished descriptions to sggest that any species shows a pattern of dispersion or agonistic behavior that is very different from those described above. No reports of strict territoriality appear to be reliable, and a few other species probably have overlapping pair home ranges dring the breeding season {Ferrginos White-eye, Dement'ev and Gladkov 19; Astralian White-eye, Frith 1967). n smmary, pochard spacing is characterized by a combination of more or less overlapping home ranges among breeding pairs, withot strict territorial defense. n most species, pairs, and especially males, maintain some space arond themselves when in the presence of conspecifics-a spacing pattern originally described as "mated-female distance" by Koskimies and Rotamo {19) (inclded in or Type spacing). To what extent sch defense in different species is the prodct of mate garding verss indirect parental investment by males is not clear. 1. Mergini Within this tribe, two general types of spacing behavior appear. The eiders, scorers, and mergansers tend either to nest colonially or to exhibit Type spacing involving nest and mate defense. The goldeneyes, in contrast, are strongly territorial (Type ), defending separate territories dring the early breeding season and the brood-rearing periods. Spacing behavior is not described for of the 0 species in this tribe. Colonial nesters (Type 6 spacing) typically fond on coastal islands inclde the Common Eider, the Red-breasted Merganser, and the Smew (johnsgard 1978; Yong and Titman 1986). Redbreasted Mergansers will, however, nest solitarily on riverbanks or lakeshores. Colonial nesting as well as solitary nesting has also been docmented for the White-winged Scorer, which breeds on inland lakes (Brown and Brown 1981), and for the Oldsqaw, which nests on the high arctic tndra (Alison 197). Nests of Black Scorer and Srf Scorer are dispersed adjacent to small boreal lakes (Palmer 1976; R. Titman pers. obs.), as are those of the Hooded Merganser, along streams or near woodland ponds farther soth {Palmer 1976; Johnsgard 1978). The Steller's Eider is the most solitary nester of the eiders. Spectacled Eiders are sally spaced as single pairs on a small pond, or several pairs on a large one. King Eiders sally have widely scattered nests, bt occasionally nest semicolonially on islets {Palmer 1976; Johnsgard 1978). Common Mergansers apparently defend preferred loafing sites along northern rivers (Palmer 1976). Harleqin Dcks breeding on streams in celand show variable spacing {Bengtson 1966, 197). Females as well as males are strongly philopatric and re-pairing of former mates is common. Hostile enconters between breeding birds of both sexes are freqent. On some streams, breeding pairs may be loosely territorial {Type ), bt others, apparently in high-density sitations, show Type spacing with freqent aggressive interactions. The Oldsqaw presents an interesting combination of colonial nesting with defense of separate territories {Alison 197). The females Alison {197) stdied nested within their territories (n = 6) or nearby, dose to other Oldsqaw nests (n = 9). However, females always fed on their mates' small (0. ha) territories, which were defended ntil the yong hatched. Broods most often went elsewhere. Only males defend territories, mostly sing prsit flights to chase females of intrding pairs, reminiscent of Three-bird Flights of the gens Anas. ntrders are greeted with Bill-toss, accompanied by vocalization and other threat postres, before being chased. Fights are rare, bt intrding males apparently try to defend their mates dring prsit flights. n Alison's {197) stdy, decoys of adlt conspecifics placed in territories elicited aggressive responses from resident males, decoys of sbadlts elicited only mild threats, and decoys of other dck species drew no response. When pairs or a member of a pair were removed from a territory, they were replaced by another pair. Common and Barrow's Goldeneyes and Bffleheads are strongly territorial, both intra- and interspecifically {Savard 198, 198; Savard and Smith 1987; Gathier 1987a, b). They defend discrete Type territories adjacent to nesting habitat from spring arrival ntil late in incbation. Neighbors are aggressively exclded, and bondaries are dearly defined. They also defend separate brood territories, which may incorporate a portion of the original territories, bt are often distinct. Territorial males of this gens exclde all conspecifics and most other waterfowl from the area they defend by sing threat displays, nderwater attacks, aerial prsits, and fights (Johnsgard 196; Palmer 1976; Savard 198, 1986; Savard and Smith 1987; Gathier 1987b). Most aggression is directed by males toward males of intrding pairs. Females rarely defend this spring territory, bt they will occasionally threaten other females. Fights occasionally occr between neighboring males as bondaries are established (Palmer 1976). nteractions between neighboring pairs are generally more ritalized, while those involving strangers are more overt and intense. Defended areas for the Barrow's Goldeneye average 0.7 ha (Savard 1986) and for the Bfflehead, 0.8 ha {Gathier 1987b). Males depart dring late incbation and do not attend broods. After hatch, females are aggressive in defense of

19 68 MCHAEL G. ANDERSON AND ROGER D. TTMAN ' l! l ~ brood territories against other females and yong, sing threat and agonistic displays similar to those of the males. The brood territories of Barrow's Goldeneye averaged 0.91 ha in Savard's (1986) stdy. nterspecific interactions between members of the Bcephala species are similar. The Bfflehead is dominated by Common or Barrow Goldeneye. Between the latter two species the otcome is npredictable (Savard 198). Barrow's Goldeneye males also defend winter feeding territories along rocky shorelines where they feed and may be rejoined by their previos mates (Savard 1986). 1. Oxyrini This grop incldes 8 species that show prononced sexal dimorphism in plmage or size and have elaborate displays. n most cases these displays are sed by males to attract females, bt male Maccoa Dcks and possibly the Astralian and Argentine Ble-billed Dcks may se some of the same displays in defense of Type territories where females nest and feed (Siegfried 1976a; Siegfried and van der Merwe 197). Siegfried (1976a) recorded p to 8 females nesting in the territory of one male Maccoa Dck. Male Msk Dcks appear to defend a territory consisting of shoreline and reedbeds for an nknown length of time. Pair bonds are not apparent, and males spend mch time trying to attract potential mates Uohnsgard 1966a; Braithwaite and Frith 1969). Aggression between displaying males may be intense, and vigoros fights have been observed Uohnsgard 1966a). Male Rddy Dcks and, likely, White headed Dcks defend a female and the moving area arond her (Type spacing) (Siegfried 1976a, b; H. Hays in Palmer 1976, vol. ). Little is known abot the spacing behavior of the Masked Dck or the Blackheaded Dck. B. nterspecific Territoriality Althogh interspecific aggression may be widespread among waterfowl, it appears to be infreqent, and there are relatively few reports of this in wild birds. The most dramatic examples come from reports of interspecific killing: e.g., by Mte Swans (Palmer 1976; Cramp and Simmons 1977); Tndra Swans (Ely eta/. 1987); and steamer dcks, Tachyeres spp. (Livezey and Hmphrey 198a; Nechrerlein and Storer 198a). Tndra Swans and Brant interact aggressively with a variety of waterbird species as well as other geese and dcks (Palmer 1976). Observations of aggressive prsits between waterfowl species have been recorded for three Bcepha/a spp. (Savard 198, 198; Savard and Smith 1987; Gathier 1987a, b), 1 Anas species (Skead 1977; Ball eta/. 1978; Titman and Seymor 1981; Wishart 198; Connelly and Ball 198), Canvasbacks (Anderson 198a), and the swans and steamer dcks indicated above. The contexts and probably the fnctions of sch interactions differ considerably from one grop ro another. An interesting debate has considered the proximate and ltimate cases of interspecific territoriality and killing in steamer dcks (Livezey and Hmphrey 198a, b; Nechrerlein and Storer 198a, b; Mrray 198). Livezey and Hmphrey (198a) arged that the large size and aggressive tendencies of steamer dcks evolved becase of the open, linear natre of their littoral habitats associated with predictable, defendable food resorces. They hypothesized that intrageneric competition for food predisposes these species to extreme aggressiveness. Both Livezey and Hmphrey (198a, b) and Nechterlein and Storer (198a, b) acknowledged the probable redction of competition for food reslting from interspecific territoriality, bt reasoned that this is nor a sfficient explanation for violent attacks on other species. Other explanations offered indde protection of broods, adaptive "play," and nonadaptive "inertial" aggression (Livezey and Hmphrey 198a, b; Mrray 1981, 198). Nechterlein and Storer (198a, b) sggested that the costs of extreme aggression, in wasted time and energy and in risks of injry and predation, are negligible for rhe very large steamer dcks and that the benefits are sfficient to release this behavior. They also speclated that sexal selection might favor males char exhibit aggressive belligerency and ths display their abilities to attendant females. Frther data on contexts of steamer dck aggressiveness, along with assessments of resorce characteristics, are needed. n the Bcephala species, intrageneric territoriality is very similar to intraspecific territoriality and may have evolved becase of competition for nest sites and food, and to ensre that females are ndistrbed dring their nesting efforts (Savard 198, 198; Gathier 1987a). Savard and Smith (1987) have recentjy provided the most detailed and objective assessment of interspecific aggression in any waterfowl species. Barrow's Goldeneye are interspecifically territorial dring both smmer (males in spring, brood hens in midsmmer) and winter (especially paired males). Their analysis of Barrow's Goldeneye attacks on different species sggested that diet overlap and competition for food might be the most important selective factor for interspecific aggression in those dcks. They also cold not reject the possibility that attacks on dabbling dcks (especially Ble-winged Teal) were simply "misdirected" aggression. nterspecific territoriality in zones of recent geographic overlap by ecologically similar Ble-winged and Cinnamon Teal (Connelly and Ball 198) and between Mallard and Black Dck (Titman pers. obs.) appears to be occrring, perhaps for the same reasons that have promoted the evoltion of intraspecific territoriality. n other cases, interspecific agonistic behavior appears co be indiscriminate and possibly a reslt of mis-

20 SPACNG PATERNS 69 taken identity in reaction to intrders (Palmer 1976; Titman and Seymor 1981). nterspecific prsits often are very short and rarely involve intense aggression. Ely et al. (1987) sggest that there is likely little selection against interspecific aggression in species sch as Tndra Swans. Skead (1977) has sggested that aggression toward other species can be associated with longer-lasting pair bonds. He cited as an example the Cape Teal, which showed a higher freqency of interspecific interactions than other Anatids with shorter bonds. Low freqencies of occrrence have been cited to spport the view that interspecific territoriality is incidental or nonadaptive (Mrray 1981). However, Wishart (198) noted that American Wigeon interacted more freqently than expected with Gadwall and Ble-winged Teal, implying prposefl explsion. Wishart (198) also detected dominance relationships in interspecific enconters related to body size and weapons (sch as sharp, pointed bills sed by grebes to displace dcks). n breeding Canvasbacks, interspecific aggression is directed primarily at brood parasites (especially Redheads) and other diving dcks that presmably are more likely to be resorce competitors (Anderson 198a and npbl.). nterspecific aggression, resorce overlap, and behavioral and morphological similarities in other territorial and nonterritorial species merit frther stdy (e.g., Astralian waterfowl where many sympatric species depend on tree cavities for nest sites). C. Patterns We searched for patterns of association between spacing behavior and several other behavioral, ecological, and morphological parameters sing Spearman rank correlation analysis. For each species of waterfowl that we cold assign to a spacing system category, we also assigned indices of pair-bond strength/permanence, habitat type/stability, diet, sexal dimorphism, and brood parasitism, as well as female body weight, modal cltch size, and home range or territory size (Appendix 8-1). Exploratory analyses were done on the entire data set, and then for the Anserini and Anatini separately, as these tribes are comparatively large and well stdied. 1. Spacing Systems and Pair Bonds One of only significant relationships that emerged was a correlation between spacing behavior and or index of pair bond length/strength. Strong, long-lasting spacing systems were associated with long-term pair bonds (for all species, r = 0.67, P < 0.1; for Anatini, r = 0.7, P < 0.0). Conversely, species with weak pair bonds or bonds of short dration are seldom territorial. Pair bonds were ranked from perennial monogamy to loose associations withot obvios bonds, and spacing was ranked from strong year-rond territoriality (Type 1) to dispersion with mate garding only (Type } (see Appendix 8-1 for details). We do not arge for any case/effect relationship in this association. Rather, we think that both spacing systems and mating systems (in part) respond to certain patterns of resorce distribtion. The common thread among the strongest territorial systems may be that limited, defensible resorces of the territories are critical for sccessfl reprodction, are permanently available, and these are defended, either continally or seasonally, by long-term mates who often mst cooperate in territory defense and whose individal interests in the territory are qite similar. This association is not absolte, thogh. Not all species showing perennial monogamy are territorial (e.g., Dendrocygna, Chenonetta), and some territorial species may have weak or ephemeral pair bonds (e.g., certain Oxyra). Others have pointed ot that territoriality and mating systems are closely related (McKinney 196 a, 197, 198; Titman and Seymor 1981; Gathier 1986) and that weaker bonds, sometimes associated with extra pair forced coplation, are fond in species of dcks that do not exhibit intense aggressive behavior with respect to space (e.g., Mallard, Northern Pintail, Lesser Scap) (McKinney et al. 198; Afton 198).. Resorce Predictability Unfortnately, we know little concerning the predictability or distribtion of critical environmental resorces or the costs associated with defense of these for most species. t appears that the most strongly territorial species occpy stable and predictable environments that have assred water levels, and presmably food spplies, while species sing ephemeral wetlands are less likely to defend specific sites. McKinney (197, 1986) sggested that these same factors cold explain variation in territorial behavior in Anas. Gathier (1986, 1988) proposed that habitat stability might also affect the risk of nest failre, and ths inflence the benefits of mate garding and strong territoriality. This argment hinges on the assmption that males behave territorially largely to protect their paternity and that there are significant trade-offs for males between territorial behavior and the prsit of secondary reprodctive opportnities. The latter trade-off may not be as stringent as formerly thoght (e.g., McKinney 197). For instance, some White-cheeked Pintail males defend territories and still engage in extrapair cortship and forced coplations (L. Gminski-Sorenson 199).. Sex Roles Male waterfowl are involved in defense of space for breeding pairs in all species where defense occrs. The only exception appears to be with brood territories defended solely by females in some species of Mergini. Male prominence in territory defense is common among

21 70 MCHAEL G. ANDERSON AND ROGER D. TTMAN birds (e.g., Lack 1968) bt nevertheless is notable in migratory waterfowl becase mate choice by females has nothing to do with territory qality (McKinney 1986; Anderson et al. 1987; Rohwer and Anderson 1988). This strong male involvement shold be expected, given the higher parental investment typical of female waterfowl {especially in dcks), resltant male-biased sex ratios, and generally stronger sexal selection operating on males. These ideas, together with observations of the widespread threats to paternity among waterfowl {e.g., McKinney et al. 198; McKinney 198, 1988), have cased some to arge that defense of paternity shold be viewed as nearly the only case for mare defense (as well as for monogamy and territoriality) in waterfowl (e.g., Emlen and Oring 1977; Wittenberger and Tilson 1980). McKinney (1986) and Rohwer and Anderson (1988) have arged that this wrongly oversimplifies the diversity of factors affecting waterfowl mating systems and mate defense. We agree, br it seems clear that protection of paternity is a priority for nearly all male waterfowl and probably, to varying degrees, nderlies all forms of mare defense and helps explain the male's niversal role in spacing behavior. Mate defense can be accomplished by means other than territorial behavior. t appears that threats to paternity from sperm competition may be inversely related, at least crdely, to degree of territoriality (McKinney 198, 1988). Frthermore, defense of space may vary greatly among species, over time, by area, and by opponent, in ways that cannot be explained simply by threats to paternity. Ths other factors also mst be important in defense and spacing, and these most likely relate to resorces in some way. Another qestion of interest regarding sex roles is, nder what circmstances do females become involved in defense of space? This happens almost exclsively in species with permanent monogamy or freqent remating by previos mates, and in circmstances of strong territorial defense {Type 1 or ). n species sch as Steamer Dcks, Ble Dcks, or African Black Dcks, mates may cooperate to evict intrders from their highly contested territories, or incoming females may challenge a resident female for both her mate and territory. Female defense seems to occr either where the efforts of both mates are needed occasionally to repel intrders or nder circmstances where females risk being displaced from critical and limited resorces that will affect their opportnity to breed (e.g., McKinney et al. 1978). Female hole-nesting dcks may compete directly with one another for access to a nest cavity. The few exceptions to these generalizations abot sex roles are interesting. With more females than males in his stdy poplation, Riggert (1977) fond females competing for mates and being active in mate defense. Other examples of species or poplations showing role reversal shold be soght and stdied.. Timing of Defense The timing of defense can provide additional cles concerning the fnction of aggresive behavior in different species. Althogh the riming of defense in relation to the stage of reprodction has not been adeqately described for most species, among those species stdied to dare where male defense is seen, defense of a mate or territory occrs at least from late prelaying throgh the laying period. Since this period corresponds with the period when females can be fertilized, it again sggests that protection of paternity by paired males is a common fnction of spacing behavior. Defense by many species lasting into or throgh the incbation period is probably related to competition for other resorces, sch as food or nest sites, or to assre access to critical resorces that will be needed later (e.g., for brood rearing or sbseqent breeding attempts). Defense of brood-rearing habitat is nsal in waterfowl. Most females or pairs roam rather widely dring brood rearing and exploit prodctive habitats providing temporally abndant resorces (e.g., prairie potholes, arctic goose pastres) that may not be economically defendable. Streams, coastlines, and some relatively resorce-poor environments are exceptions to this, and brood territories are common among river-dwelling dcks, some coastal species, and several species of Mergini that breed on small lakes, mostly in the sothern boreal forest zone.. Coloniality Coloniality in breeding waterfowl seems to be associated with either: {a) rare safe-nesting sites, or (b) habitats that offer exceptionally good brood-rearing conditions. Both of these sitations provide individals with advantages of groping against predators and access to a critical limited resorce. Examples of the former inclde islands in prairie wetlands, coastal islands with good nesting sbstrate, and abndant bt dmped nest cavities. Exceptional brood-rearing conditions may be largely responsible for dmping by many arctic-breeding geese, or by waterfowl exploiting highly dynamic environments (e.g., Astralian Black Swans or Pink-eared Dcks) where extensive feeding sires for pairs are not defended, and birds dmp in proximity to restricted brood habitat. Arctic geese are the most stdied of these waterfowl. Most are colonial, perhaps in response to selection for predator avoidance in open tndra habitats, and becase their clmped food resorces are only briefly available each smmer. Raveling (1989) has shown that groping confers advantages to Alaskan Black Brant. He reported higher nest srvival in larger colonies. We fond char body size (average female body mass) and

22 SPACNG PATERNS 71 spacing system were inversely correlated in the tribe Anserini (r = -0.88, P < 0.01). Larger-bodied swans generally defend Type 1 or territories, bt smaller-bodied geese typically exhibit mate defense or coloniality. Differences in spacing system and body size in geese may be related to breeding latitde. Smaller-bodied geese (and smaller swans) tend to breed at higher latitdes (Hanson 196; Ogilvie 1978; Dnn and Mac nnes 1987). The Greater Snow Goose appears to be an exception. Hanson ( 196: p. 77) arged that body size of breeding Canada Geese was inversely related to the severity of climate. Becase incbation period, dration of the molt, and fledging time are all greater for larger geese, and length of the breeding season is shorter in the far north, Hanson reasoned that arctic breeding selected for relatively smaller body size. The same may be tre in other geese (Ogilvie 1978). Conversely, body size in relation to temperatre stress may affect the distribtion of wintering geese (Hanson 196; Lefebvre and Raveling 1967; Lack 197; Ogilvie 1978). Becase they are at greater risk from predators, smaller geese may also benefit more from groping in open arctic habitats. nteresting combinations of largely separate pair territories with clmped nesting occr among species from several tribes, e.g., Northern Sheldck, Astralian Sheldck, Rddy Sheldck, Gadwall, Mallard, Red-breasted Merganser, and Long-tailed Dck. For the Mallard and Gadwall, this appears to be associated with rare safenesting islands that alone do not provide adeqate food resorces for breeding pairs. For the other species, sch patterns likely reslt from limited appropriate nesting habitat combined with separate defensible feeding and mating areas. Va:lable frther insights concerning the costs and benefits of coloniality might come from stdies of island-nesting prairie dcks or colonial waterfowl of the Sothern Hemisphere. 6. Problems Perhaps the clearest conclsion from or search for general patterns is that we still do not have sfficient information to allow many generalizations abot the evoltion of spacing systems. A major gap in or view is the pacity of data concerning the identity, distribtion, and predictability of critical resorces needed by breeding waterfowl (see section V-B below). Also, there are srprisingly few detailed data on the timing, context, and sbtle variations in defensive behavior that are needed to discriminate among hypotheses of adaptive fnction (e.g., Goodbrn 198; Anderson 198a). This points to the need for carefl design in ftre stdies of waterfowl spacing behavior to consider these aspects. V. Models of Animal Dispersion and Spacing Behavior A. Patterns of Dispersion Patterns of dispersion reslt from both short-term spacing and long-term dispersal. n migratory species, poplation density is the prodct of both poplation settling behavior, inflenced by habitat qality (Fretwell and Lcas 1970) and philopatry (Greenwood and Harvey 198), and the spacing of individals within available habitat. Factors affecting philopatry in breeding waterfowl are discssed elsewhere in this volme (chapters 11 and 1). Or discssion will be limited to spacing patterns within settled poplations. Using geometric models, Horn (1968) and C. Smith (1968) showed that for central-place foragers (e.g., animals with nest sites or dens), separate home ranges cold be exploited more effectively by individals than shared ranges in environments with dispersed, stable resorces (e.g., predictable foods evenly distribted at reglar intervals across the landscape). Opposite conditions favored central colonies in Horn's model (see also Wittenberger and Hnt 198). However, clmped resorces that vary in space and time are not necessarily more efficiently exploited by aggregated individals (Waser and Wiley 1979); it depends on specific assmptions. For instance. if the dimensions of available habitat exceed an individal's foraging range, sch conditions can favor widely overlapping bt dispersed home ranges (Waser and Wiley 1979), which are typical of several breeding dcks (Anderson 198a). Waser and Wiley extended Horn's model to accont for animals operating withot discrete central bases and showed that overlapping home ranges are favored when the availability of limiting resorces varies greatly across locations and over time (e.g., perhaps Northern Pintails). n contrast, exclsive se of space is favored in environments with evenly dispersed stable resorces, especially when site familiarity confers some advantage to individals or renewal rates of resorces are high. This might well describe the habitats of several territorial dabbling dcks on the prairies. Wiens (1976) arged that differences in environmental heterogeneity cold select for different patterns of dispersion by which individals wold maximize their efficiency in acqiring needed resorces. He sggested that territorial defense shold be most common where resorce defense is inexpensive, resorce clmping is low, and resorce predictability is high. n theory, the synchrony of settling may affect territory size and arrangement beyond any differences in environmental resorces (Maynard Smith 197). Knapton

23 7 MCHAEL G. ANDERSON AND ROGER D. TTMAN. i ~~ i and Krebs (197) fond evidence in territorial Song Sparrows (Melospiza melodia) that synchronos serdemenr reslted in righter packing of territories than did asynchronos settlement. Given observed variability in homing by individal dcks (Blohm 1979; Afton 198; Anderson 198a), the possibiliry rhar riming of arrival may inflence patterns of dispersion deserves stdy. Unfortnately, roo few data are available on spatial and temporal variability of resorces needed by breeding waterfowl ro allow rests of rhese models. B. Territory Size Schoener (198) attempted to reconcile previos efforts ar modeling optimal territory size. He offered specific predictions concerning how territory size (as well as feeding time and defense time) shold change as environmental conditions change for: (a) time minimizers; (b) energy maximizers wirh rime constraints; and (c) energy maximizers with processing constraints (Schoener 1971). He modeled an increase in food density, increased intrder pressre, and simltaneos changes in these two factors, all modified by assmptions of how defense costs might change with increasing territory size. Unfortnately, for waterfowl we Jack the data on how costs and benefits change with territory size and intrder pressre, and so we cannot rest Schoener's (198) specific predictions. Lima (198) also modeled territory size, br in a variable environment where food density and intrder pressre varied in a stochastic fashion. The predictions of these models wold be worth testing with waterfowl, bt are complicated by the fact that breeding birds of different sex or reprodctive srage might be optimizing different things. The possibility of changing optimization criteria raises other interesting problems. For instance, many nonterritorial species of dcks show a decline in home range dring the breeding season-from a prelaying maximm ro smaller areas throgh egg laying and incbation (Table 8-). Even for some territorial species, pairs se fewer ponds as nesting proceeds (e.g., American Black Dcks, Ringelman et a/. 198). Traditional wisdom arges rhar large home ranges dring prelaying are associated with a search for sitable nest sites. At the same time, females srely gain knowledge abot the distribtion of resorces in their environment, bt that still does nor explain why average travel distances decline. A shift by nesting hens from energy maximization dring prelaying to time minimization dring incbation may explain sch changes. n many species (Table 8-), males apart from their nesting mates expand their home ranges as incbation proceeds. Mch of this movement is probably associated with seeking additional mating opportnities (McKinney 198), rather than with a shift in foraging strategy, bt we know little abot the nderlying cases of seasonal change in home range size. Declines in territory Species American Black Dck Mallard Nonhern Pintail Gadwall American Wigeon Ble-winged Teal Table 8-. Temporal changes in home range size for breeding dck pairs Reprodctive stage Prelaying & laying ncbation Total Laying Prenesting Laying ncbation Laying to incbation Prenest & Nest Nest only Arrival to incbation Early Arrival Preincbation Prelaying to third week of incbation Home range (hal ll a b c 8 d Sorce Ringelman et a/. 198 Dwyer eta/ Gilmer et a/. 197 Titman 198 Derrickson 1978 Gates 196 Wishan 198 Stewart and Titman 1980 Canvasback Postarrival 7 Anderson 198Sa Prelaying 17 Laying 0 ncbation 18 Renest interval 6 Males of marked pairs expanded their ranges significantly after the first week of incbation. bpis females showed hints of decline from laying to incbation. 'No data presented, bt stated home range declined throgh the sea son..jno change in territory size. size have been associated with increasing density of prey (reviewed by Davies and Hoston, 198). The sal explanation is that animals adjst their behavior to defend only sfficient resorces to meet their needs. For waterfowl, in one of years Gathier (1987b) fond a significant inverse correlation between territory size of breeding Bffleheads and food densiry. Patterson (198) reported conflicting evidence concerning territory size and food density in Northern Sheldcks. n one stdy Northern Sheldcks actally defended larger territories when food spplies were greater (Bxton 197). Smaller territories in better habitats cold instead be the reslt of increased competition rhar forces competitors into smaller areas (Schoener 198). Myers et a/. (1979) showed that increased intrder pressre was at least partly responsible for decreases in Sanderling (Calidris alba) territory size. Poplation densiry, presmably proportional to intrder pressre, can inflence waterfowl territorial behavior, too. With increased densiry nder the same habitat conditions, Titman (198) fond Mallard territory sizes that were 9% of those nder normal conditions. High densiries were - pairs/ km, de to the release of hatchery-reared birds that re-

24 SPACNG PATERNS 7 Table 8-. Correlations between body weights and breeding poplation densities for North American prairie dcks Stdy area/grop Years (n) Species (n) R l P-vale Sorce Redvers, Saskatchewan Dabblersb 1 Divers Total Losana, Alberta Dabblers 1 Divers Total Caron, Saskatchewan Dabblers 6 Divers Total Prairie Pothole Region Dabblers 1 Divers Total Body weights were taken from Appendix 8-1. banas spp. < Aythya spp. and Oxyra iamaicensis s Stodt Smith Leitch and Kaminski Batt et al O.ot 0.77 trned to breed verss -7. pairs/km normally. There was also a lower freqency of Three-bird Flights, and these prsits were shorter in dration nder conditions of increased density. Under extreme crowding, where intrder pressre greatly escalates the cost of defense, territorial behavior of Mallards breaks down (Titman and Lowther 197). Gathier (1987b) reported a negative correlation between territory size of Bfflehead and the density of all congeners on his stdy area. Variation in competition may also partly explain variation in waterfowl home ranges in different habitats (Ndds and Ankney 198). An individal's best choice of defense options may depend sbstantially on what other individals do. C. Home Range, Body Size, and Poplation Density There is a positive correlation between home range size and the body weight of a variety of animals (McNab 196; Schoener 1968). There is also some indication that diet frther inflences home range size. McNab (196) and Schoener (1968) sggested that carnivores tend to have relatively larger home ranges for their body size than herbivores or omnivores. Mace and Harvey (198) arged that resorce patchiness, in addition to habitat prodctivity and energetic needs, is important in determining home range size. Alernatively, Janes (1986) sggested that the positive relationship between body size and home range cold be sprios, becase both factors correlate inversely with poplation density. Large birds tend to have large home ranges, bt they also tend to exist at low densities. This relationship appears to be strong for raptors and other carnivores and weak for omnivores and herbivores. A negative correla- rion between poplation density and body size is not apparent for North American prairie dcks (Table 8- ). Ndds and Ankney (198) fond a significant positive relationship between body size and home range size of prairie dcks and sggested, based on slimmer evidence, that home range size may differ between large marsh and pothole habitats. Althogh sch correlations sggest a relationship between spacing and resorce distribtion, they do not distingish case from effect. We cannot be certain that body size and diet largely dictate home range size nless factors sch as defense of other resorces and variable resorce patchiness can be rled ot. D. A Better Analytical Description of Spacing Behavior? Detailed stdies of marked individals in wild poplations have demonstrated variability in how animals compete for resorces. Lmping complex and diverse patterns of behavior in only a few catchall categories, sch as "territorial," may not be very sefl. McKinney ( 196a) made essentially the same point in his review of waterfowl spacing many years ago. We agree with Waser and Wiley (1979, p.60) that agonistic behavior, activities, and isolation from neighbors vary in complex ways among individals and among species, and that this variation shold be recognized and stdied. "Territory" and like terms may mask important variation. Or categorization is also too simplistic, althogh we have attempted to describe more general variations on waterfowl spacing than any previos review. Descriptive data on most species crrently are not adeqate for more refined comparisons. Waser and Wiley (1979) introdced three descriptors

25 7 MCHAEL G. ANDERSON AND ROGER D. TTMAN \ i \ of dispersion and agonistic behavior that cold enhance qantitative comparisons of spacing attribtes berween species or poplations. They envisioned animals occpying a two-dimensional srface where any location can be identified by a pair of x, y coordinates. The measres rl:,. defined consist of: (1) Activity Fields, a smmary of \\. -:re an individal spends irs rime; () solation Fields, an index to the relative exclsiveness with which an animal ses any portion of its home range; and () Aggression Fields, at any location, the probability of an animal attacking or retreating from an opponent. These measres wold greatly enrich descriptions and comparisons of spacing for species where individals concentrate their activities in relatively few scattered locations (e.g., dcks in a complex of small prairie wetlands). Sch data cold aid descriptions of reactions between known specific opponents, and stdy of Activity and solation fields cold reveal patterns of avoidance as well as defense when copled with mathematical indices (correlation coefficients or the like) to overlap in fields among individals. A major advance may reslt from an objective means of discriminating among sbtly different social systems. We had considerable difficlty, for instance, categorizing Mallards, which show different behavior in different habitats, and acconting for the sbtle differ ences among races of Canada Geese. With the addition of data on resorce variability, Waser and Wiley's techniqes may also provide an objective measre of behavioral response to environmental variation, sch as prairie droght. E. Game Theory and the Evoltion of Spacing Behavior An important property of settling behavior mentioned above, particlarly related to the concept of ideal free distribtion (Fretwell and Lcas 1970; Fretwell 197), is that the optimal soltion for any individal is freqency dependent; i.e., it depends pon what other individals in the poplation do. This basic notion has been central co the development of game theory, which has provided new insights on many aspects of animal behavior (Maynard Smith 198; Parker 198). To or knowledge, no one has sed the insights offered by game theory to stdy spacing behavior in waterfowl. We believe this will be fritfl. The approach of game theory is to define a set of possible alternative phenotypic strategies that animals might play against one another and then solve a mathematical game for an eqilibrim (Maynard Smith and Price 197; Maynard Smith 197, 1976; Parker 198). The soltion, or evoltionarily stable strategy (ESS), is defined as a strategy that, if adopted by most members of a poplation, cannot be invaded by the spread of a rare alternative strategy. Lendrem (1986) offers a very readable elementary introdction to game theory for ecologists. This approach shold be especially helpfl for the stdy of behavior becase the fimess payoff to a behavioral strategy mst often depend strongly on the strategies being played by other individals (Maynard Smith 198; Parker 198; Parker and Hammerstein 198). The stdy of animal contests, clearly relevant to spacing behavior, may be the most promising area in biology for the application of game theory (Maynard Smith 198; Parker 198). The main vale of ESS models is heristic; they are designed co reveal the logical possibilities inherent in contest sitations (Maynard Smith 198). Under many circmstances there can be more than one stable soltion co a contest sitation, and ethologists freqently enconter high variability in the social strategies of wild animals. Until ESS models formalized thinking abot how differences in the vale of resorces, resorce holding potential (RHP) (sens Parker 198), or other asymmetries among individals shold affect the otcome of contests, they were nor examined. For example, the Parallel Walk and other displays between Red Deer (Cervs elaphs) stags competing for mares have been convincingly interpreted in terms of mtal assessment of RHP (Cleron-Brock and Albon 1979), and the behavior of competing males conforms qire well to expectations from ESS models. The properties and context of this display, in which competing males slowly walk parallel preceding withdrawal or fighting, are strikingly similar to the Sneak of breeding male Canvasbacks, and in many respects, to the Tandem Swim of American Wigeon, Slimoff manevers of African Black Dcks (McKinney eta/. 1978), and Head-back postres of the White-backed Dck (Clark 1969). We strongly sspect that these and other threat displays associated with spacing in waterfowl (e.g., bondary displays in strongly territorial species like Mte Swans [Lind 198)) might be better nderstood sing game theory analysis. Game theory analyses may provide enlightenment for many other problems in waterfowl spacing behavior. For example, in most prairie dcks, territory owners displace intrders in the vast majority of interactions. n Mallards, Titman (198) fond that resident chasing males retrned to the site where Three-Bird Flights originated > 8% of the time. ntrders left the area> 80% of the time. Established residents were almost never displaced by intrders at the same stage of their breeding cycle. The same was tre for territorial American Wigeon (Wishart 198). Similarly, male Canvasbacks nearly always won aggressive enconters that they initiated (Anderson 198a). Why shold this be? Srely all resident birds are not that sperior to intrders. Are the birds playing a simple strategy of "retreat if a pond is occpied,, or are asymmetries in resorce vale or RHP responsible for these lopsided otcomes? Shold this

26 SPACNG PATERNS 7 convention change predictably with time, resorce abndance, or poplation density? A related isse is the way that territory ownership is established. How do dispersion patterns develop each spring in migratory species? Female experience mst be a major factor (chapter 11), bt males do most of the chasing of other pairs. How do males choose where and when to chase conspecifics and how to respond to individal competitors? For most prairie dcks, settling is highly asynchronos (Hmbrg eta/. 1978; Ohde eta/. 198; Anderson 198a). s this simply a matter of inability on the part of some birds to arrive early or to compete sccessflly, or are some individals making what for them is the best of a difficlt sitation by setding later, avoiding mch competition, bt incrring the costs of delayed breeding in a seasonal environment? Prsit flights in the gens Anas have cased long debates over apparently conflicting motivational tendencies and probable mltiple fnctions. Tactics of interlopers, chasers, and defending mates shold be explored sing objective predictions of behavior from a priori ESS modeling in order to shed new light on this intriging behavior pattern. ESS thinking has led to many new ideas concerning the evoltion of signaling behavior. Mch early thinking abot displays focsed on selection for nambigos transfer of information between animals (Bastock 1967; Sebeok 1977; Smith 1977). Ethologists are now looking at displays with new qestions in mind-considering blffing verss honest signaling, salesmanship, maniplation, mind reading, exploitation, and deception (Trivers 1971; Dawkins and Krebs 1978; Rohwer 1977; Zahavi 1979; Dawkins 198; Krebs and Dawkins 198 ). These views are beginning to affect the analysis of waterfowl cortship (chapter 7) and shold be extended to threat displays and other aspects of spacing behavior. V. A Critiqe of Waterfowl Spacing Stdies and Recommendations for Ftre Research Dring or review of the literatre on spacing behavior, we were disappointed to note that there are virtally no references to waterfowl stdies in recent general reviews of spacing. Why shold this be, if waterfowl are as well stdied as we claim (see ntrodction)? We sspect there are at least two reasons. First, waterfowl stdies are sally logistically difficlt and expensive. This means that members of the family do not lend themselves well to qick experimental stdies of new concepts. On the other hand, the diverse radiation of waterfowl offers great opportnity for comparative analyses of social behavior (McKinney 1978, 1986), and it is in this way that waterfowl stdies now contribte most to general ideas abot social evoltion. The second possibility is that those of s interested in waterfowl spacing behavior have lost step with theoretical progress in this rapidly changing field and are rarely prsing qestions of general interest to behavioral ecologists. n the following section we try to identify the main gaps in or knowledge of waterfowl spacing behavior and some collective shortcomings in or approaches to stdying it. A. Empirical Needs and Comparative Stdies An obvios conclsion from this srvey is that nmeros gaps remain in or nderstanding of the basic life histories of many species. Simple descriptive observations concerning aggressive behavior and breeding dispersion wold be of vale, especially for varios whistling dcks (Dendrocygnini); several sheldcks and sheldgeese (Tadornini); 17 species of tropical, Asian, Astralian, and Soth American dabbling dcks (Anatini); the white-eyes and narrow-billed pochards (Aythyini); many Cairinini and Oxyrini; and several arctic sea dcks (Mergini). Even for the better-known species there are only meager data concerning derails of agonistic behavior. Carefl field stdies of individally marked birds of most species are still needed. t is important to learn exactly who chases whom, where, and when (Goodbrn 198; Anderson 198 a). We still know very little abot how spacing behavior changes with age, sex, and reprodctive stats, and how interactions vary in different contexts or between different individal competitors. Until sch data are available, qantitative comparative stdies will not be possible (cf. Cltton-Brock and Harvey 1977; Harvey and Mace 198), and these are needed if we are to develop and critically test general hypotheses concerning the evoltion of spacing behavior in waterfowl. Ftre descriptive stdies of waterfowl spacing behavior shold: (1) docment male and female roles in agonistic interactions with conspecifics in relation to individal characteristics (e.g., sex, pairing stats) and behavior of the participants, context, and the location of important resorces; {) report temporal changes in behavior, especially in relation to breeding stats; and () describe the reslting pattern of dispersion among birds, both spatially and temporally. These sggestions are similar to McKinney's (196a) of many years ago. B. Resorces There is a serios lack of information abot the distribtion and abndance of key resorces that waterfowl appear to be defending. Few stdies (Patterson 198; Gathier 1987b) have attempted to qantify the distribtion of important resorces, and no one has attempted to maniplate resorces and look for behavioral responses. Only preliminary attempts have been made to discover and nderstand variation in ag-

27 76 MCHAEL G. ANDERSON AND ROGER D. TTMAN. i,; gressive behavior and home ranges within species (Ndds and Ankney 198; Amat 198). Until measrements of resorce characteristics and associated behavior are made for a variety of species or poplations, we can do little else bt speclate abot the ecological bases of the patterns we observe. We regard this as perhaps the greatest single deficiency in waterfowl spacing stdies to date. C. Tests of General Models Althogh most waterfowl systems present formidable logistical difficlties for testing economic models of spacing behavior, a few sitations, sch as where species maintain permanent river territories, may be amenable to sch stdies. Regardless, cost/benefit thinking has been inflential and shold contine to gide the selection of specific research qestions on spacing behavior for the foreseeable ftre. Similarly, stdies of settling behavior in migratory breeding waterfowl and the dyn.amics of home range establishment will be difficlt, bt perhaps feasible, in relatively isolated stdy areas with small poplations of marked birds. These shold be prsed, becase many qestions remain nanswered concerning the biology of "floaters," the costs and benefits of breeding sire philoparry, and the cases and conseqences of "overflight" of traditional breeding gronds when environmental conditions change (see Johnson 1986; chapters 11 and 1). With the exception of prsit flights in the gens Anas, chasing behavior and threat displays in waterfowl have received little detailed stdy in the context of spacing. There is ndobtedly mch interesting, nexplored variation waiting to be discovered. nsights from game theory thinking and tests of specifically constrcted ESS models shold be employed in ftre stdies of agonistic behavior. Several specific sggestions along these lines were made above. Observations of captive breeding birds might be particlarly sefl for sch stdies, becase details of display exchanges, inclding sbtle differences in orientation or intensity, can be recorded more easily; long-term relationships berween known individals can be stdied more reliably; and experimental maniplations shold be more feasible than with wild birds. D. Long-term Field Stdies Long-term stdies of known individals are needed to assess changes in agonistic behavior with age or experience and to stdy the fimess payoffs of different spacing strategies nder variable environmental and social circmstances. Only long-term stdies in natral environments can provide sch data. Stdies by Cooke et a/. (198), Patterson (198), Newton (1986), Woolfenden and Fitzpatrick (1977, 1978, 198), Watson (1977), Watson and Moss (197), Watson eta/. (198a, b), Moss and Watson (198), Nolan (1978), Colson (1966, 197, 198), Cltton-Brock eta/. (198), and Goodall (1986) provide good models for long-term single-species research. E. Social Relationships Rbenstein and Wrangham (1986) have emphasized the need for stdies of long-term relationships among individals who know each other and have histories of interaction. Long-term relationships may inflence the costs and benefits of interactions in ways that make no sense to one observing interacting individals for the first time. For instance, a vigoros fight over some trivial item may be inexplicable withot knowing that a longterm dominance relationship berween the rwo contestants is at stake. We agree there is a need for sch analyses, particlarly among individals in sedentary poplations or among tightly philopatric individals in migratory poplations, inclding bt not limited to close kin, where the opportnity for development of long-term social relationships is greatest. F. Density-Dependent Changes in Behavior Few stdies have examined the effects of poplation density on the breeding behavior of waterfowl {Titman and Lowther 197; Titman 198), bt these sggest that sbstantial disrption of normal patterns may occr at high densities. Frther stdies are needed to measre the impact of changing poplation density on spacing behavior and dispersion, and ltimately on reprodctive sccess and srvival. Sch stdies shold be condcted over sfficient years to assess variability in natral poplations. The long-term stdies of Northern Sheldcks by Patterson and his colleages (Patterson 198; Patterson eta/. 198) are exemplary bt niqe. Field experiments might be sed to modify poplation density while also monitoring appropriate control poplations. Sch an approach has the additional advantage of removing possible confonding effects of year and density. The small-nit management stdy nder way in North Dakota and Minnesota by the United States Fish and Wildlife Service is sch an experiment, designed in pan to determine the effects of increased density on Mallard behavior and reprodction. f management of breeding waterfowl becomes more intensive in the ftre, it is critically important that we gain a clearer nderstanding of the long-term effects of poplation density on spacing and reprodctive sccess in breeding birds.

28 SPACNG PATTERNS 77 V. Discssion for Management A. Ethology and Waterfowl Management Unlike investigators working with many animals, waterfowl biologists freqently ask qestions for three fndamentally different reasons: (1) to solve specific wildlife management problems; () to answer qestions abot the basic biology of waterfowl; and () to test theoretical hypotheses. Unfortnately, investigators employing these different approaches freqently contribte little to solving each others' problems. This shold not be so with the stdy of spacing behavior. ntensive management of breeding waterfowl demands a clear nderstanding of spatial relationships among breeding birds and possible behavioral limits to habitat carrying capacity. We believe this contines to be a promising area for prodctive cooperation between ethologists and waterfowl managers. B. Spacing and Poplation "Reglation" The pblication of Wynne-Edwards's classic Animal Dispersion in Relation to Social Behavior ( 196) shered in a period when mch attention was given to territorial behavior as a factor in limiting animal nmbers. Literatre on this sbject is filled with argments, often vage, abot levels of selection and whether social be havior in general, and territoriality in particlar, may have evolved to limit poplation growth and prevent overexploitation of resorces. For the great majority of animal species, that view has been refted (.B. above). Nevertheless, the possibility of poplations being limited as a conseqence of territorial behavior or other forms of spacing behavior remains an interesting qestion for poplation biologists and wildlife managers. The scale at which sch limitation might occr is an important consideration. Local poplations might be limited by spacing behavior, br this does nor necessarily imply that whole species' poplations are limited in a similar manner. Several stdies with a variety of birds have shown that spacing behavior can indeed limit the nmber of individals breeding in a finite stdy area (Orians 1961; Krebs 1971; Warson 1977; see Davies 1978 for review). Limitation of a local waterfowl poplation by social behavior has been demonstrated only for Northern Sheldcks (Patterson 1980; Patterson eta/. 198) and perhaps Bffleheads (Gathier and Smith 1987) and Long-railed Dcks (Alison 197), althogh sch limitation seems likely for at least several other species with restricted habitats (e.g., African Black Dcks, Ball et al. 1978; McKinney et al. 1978) or for isolated poplations in limited habitat (e.g., Mallards, Hmbrg eta/. 1978; Ohde et al. 198). The distinction between local and overall poplation limitation by social behavior is important for poplation managers to nderstand. Speclations abot behavioral cases of poplation limitation in geographically widespread species (e.g., Mallards, Aldrich 197; Pospahala et al. 197) are likely to be misleading, althogh local limitation might occr in some years and in certain habitats. n contrast, geographically restricted species, with specialized habitat reqirements chat ltimately constrain poplation size, may well be candidates for proximal behavioral limitation, and chis possibility shold not be ignored by conservationists. The extent to which local poplations might be restricted by social behavior in a variety of species deserves frther stdy. For intensively managed poplations, local densities might well exceed the limits beyond which birds will not crowd frther (or some other density-dependent constraint may become effective). n sch cases, these limits shold be identified so that additional innovative techniqes to increase pair densities might be attempted, or conversely, so that seless efforts to frther increase poplations might be avoided. ntensively managed plots of dense nesting cover or island refges are examples of the kinds of management practices that might be so affected. n sch cases, it will also be important to nderstand longerterm effects of crowding, inclding impacts on resorces, brood srvival, philopatry, and reprodction in srronding areas. Acknowledgments For spport while writing this paper we thank the North American Wildlife Fondation throgh the Delta Waterfowl and Wetlands Research Station (Anderson), the nstitte for Wetland and Waterfowl Research (Anderson), and McGill University (Titman). We are gratefl to several persons cited in the text who provided access to npblished data and observations, and to Brce Batt, John Ryder, Frank McKinney, Dog johnson, and Gilles Gathier for helpfl comments on the entire manscript. Diane Chronister, Leigh Fredrickson, and Dennis Raveling offered sefl comments on specific tribal acconts. Rodger Titman is also gratefl to Harry Beach of Kochibogac National Park for providing office space dring the preparation of this paper.

29 j:! : 78 MCHAEL G. ANDERSON AND ROGER D. TTMAN Appendix 8-1. Spacing patrems and other characteristics of breeding waterfowl Spacing Pair Habitat Home Female Plmage,. Tribe Sl!ecies s~stem bon db stability' Dietd ran8e (ha) weight (!1) dimorl!hism 1 Referencesd Anseranatini Ansanas ,9,61 semipa/mata Dendrocygnini Dendrocygna U gttata J : : i D. eytoni U-1 T ,61 D. bicolor, ,90,10,11 D. arcata 7 0,61 D. javanica U D. vidata D. arborea U-6 U ,61 D. atmna/is, ,61,10 Tha/assornis U ,61 leconots Anserini Cygns olor ,6,9 C. atrats 6 U , C. melancoryphs C. bccinator ,,90 C. cygns C. co/mbians ,6 C. bewickii ,6 Coscoroba U-1 U ,6 coscoroba Anser cygnoides U l 10 0,61,89 A. (abalis U-S ,7 A. albi(rons 7 0 8,10,1,90 A. erythrops ,61 A. anser, ,8,88,89,16 A. indics 6 U-1 U ,61,89 A. caerlescens ,10,9,6,90,1 0 ' A. rossii ,90,101 A. canagics ,61, Branta sandvicensis U- T ! B. canadensis ,0,0,,81 ' B. lecopsis ,61,88.!i B. bernie/a 6 1 O.QJ 1 0 1,81,88,90 ql B. rficollis 6 U ,61,89 A. brachyrhynchs ,66 Cereopsini Cereopsis U ,,61 novaehollandiae y Stictonetrini Stictonetta naevosa ,,61 Tadornini Cyanochen ! cyanopters Chloephaga U-1 U l melanoptera C. picta C. hybrida U- 01 0,61,9 C. poliocephala U-1 U C. rbidiceps Neochen jbata U-1 U Alopochen aegyptiacs Tadorna ferrginea U- U T. cana T. tadomoides 191,61,99 T. variegata ,18 T. cristata 61 T. tadorna ,8,9,96,1 T. radjah 80 0,61 Tach) erini Tachyeres ,71,86 patachonics T. pteneres U-1 1ll 61,11 T. brachypters 00 61,71,9,11

30 SPACNG PATTERNS 79 Appendix 8-1. (Contined) Spacing Pair Habitat Home Female Plmage Tribe Sj!ecies srstem bondb stability< Dietd range (ha) weight (G) dimorphism 1 References 11 Cairinini Plectropters gambensis Cairina moschata C. sctlata ,7 Sarkidiornis, ,10 melanotos Pteronetta hartlabi Nettaps plchells U ,61 N. coromandelians U ,61 N. arits U Callonetta U-1 1 U- 61 fecophrys Aix sponsa 0,10,1,90 A. galericfata 7,,,61 Chenonetta ;bata 1 T 1 800,61,19 Amazon etta U brasiliensis Merganerrini Merganetta armata 0 0,60,8 Anatini Hymenolaims ,61,6,6 malacorhychos Anas waigiensis ,6 A. sparsa ,79,61 A. penelope 60 9,1,1, A. americana ,1 A. sibifatrix ,1 A. falcata 8 1,61 A. strepera ,6,,90 A. formosa ] 0 61 A. crecca 0 1,,90 A. flavirostris U-6 l ,17 A. capensis ,119,1 A. bernieri 0 61,109 A. gibberifrons 7 1,61,67 A. castanea 0,61 A. a11ckfandica ,61,18,19 A. pfatyrhynchos ,1,7,9,90,10 A. laysanensis,6, A. rbripes ,90,111,11 A. melleri 0 61 A. rmdlata 817 0,61,100 A. poeciforhyncha ,,61 A. fzonica A. specfaris U A. specfaroides ,17 A. acta ,78,90,116,117 A. georgica 1 61,10 A. bahamensis 1 61 A. erythrorhyncha A. versicolor A. hottentota 0 7,61 A. qerqedla 00 1 A. discors ,1,90,118 A. cyanoptera 60 10,8,77,90 A. pfatalea 1 61 A. smithii ,77,11 A. rhynchotis 66,61 A. clypeata ,76,90,97,110 Malacorhynchs 0,61 membranaceos Marmaronetta ,1, angstirostris Aythyini Netta rfina ,,68,69

31 80 MCHAEL G. ANDERSON AND ROGER D. TTMAN i ' " ;:,. Tribe Mergini Oxyrini Species N. erythropthalma N. peposaca A. valisinerla A. ferina A. americana A. collaris A. astralis A. baeri A.nyroca A. innotata A. fligla A. novaeseelandiae A. marila A. affinis Somaterit~ mollissima S. spectabilis S. (1Scheri Polysticta stelleri Histrionics histrionics Clangla hyemalis Melanitta nigra M. perspicillata M. fsca Bcephala albeola B. islandica B. clangla Mergs ccllats M. a/bel/s M. octosetaces M. serrator M. sqamats M. merganser M. astralis Heteronetta atricapilla Oxyra dominica 0. jamaicensis 0. lecocephala 0. maccoa 0. vittata 0. astralis Bi:cira lobata Spacing system U-.s.s 6 s s S,6 s 6 S,6 s U S Pair bondb U- s U- U- s U- U- U- U- U- U- U- s U- 6 Appendix 8 1. (Contined) Habitat stabiliry' s 1 l. Dietl U- U- U- U- U S U-S s s s s s s s s Home range (ha) Female weight (g) SO 900 1SOO 0 78S 800 S S 100 6S SO 10 Plmage dimorphism References' 1,,,61,8 61,1,1,,10,6,16,1 10,90,107,1 10,7, , 61 1,8 61,87 10,1,90 1,10,,1 10,61,90 10,90,91 1,,90 1,,90 11,1,1,90,1,90 1,90 61,90,1,90 1,6, 7,90 90,10,10,106 90,10 10,8,90 9,1, 61,9,90,1,61 10,1,, ,1 61,90 10,90,11,11 1,6,77,61,7,11 61,61 19,,61 spacing system classified as detailed in the text: 1-year-rond strong territoriality; -fll breeding season territoriality; -early breeding season territorialiry; -territorialiry with overlap; -dispersion with mate garding; 6-colonialiry; U-nknown. bpair bonds ranked according to strength and stabiliry:!-perennial monogamy; -seasonal monogamy, strong bond of long tenre (brood rearing); forced coplation not recorded or very rare; -seasonal monogam)', bond of medim tenre (midincbat~on), low freqency of forced coplation; -seasonal monogamy, bond of shorr tenre, relatively freqent forced coplation; -monogamy with occasional bigamy; 6-no bond; U-nknown. <Habitat ranked according to stab1liry: 1 -tropical rivers, coastal waters; -temperate rivers, deep waters, inland coastal areas; -lakes, permanent waters, ditches, potholes; -very shallow prodctive ephemeral waters; -droght-vlnerable habitats, tndra ponds; T (not inclded in correlations) -terrestrial habitats. ddiet classified according to plant or animal content: 1-plant diet(> 90%); -7S% plant; -0% plant, SO% animal; -7% animal; S -animal diet(> 90%); U-nknown. Area occpied dring the breeding season in ha (territory or home range). 1 Pimage color dimorphism: 0-none; 1 -slight; -moderate; -strong or prononced. "References: 1) Afton 198S; ) Ali and Ripley 1968; ) Alison 197; ) Anderson 198; ) Anderson 198a; 6) Ballet a/. 1978; 7) Banko 1960; 8) Barry 1966; 9) Baer and Gltz von Blotzheim 1968; 10) Bellrose 1976; ) Bengtson 1966; 1) Bengtson 1970; 1) Bengtson 1971; 1) Bengtson 197; 1) Bennett 198; 16) Bezzel 1969; 17) Bolen 1967; 18) Braithwaite 1976; 19) Braithwaite and Frith 1969; 0) Brakhage 196; 1) Brown eta/. 198; ) Brown and Brown 1981; ) Brggers 1979; ) Clark 196; ) Clark 1966; 6) Clark 1969; 7) Clark 1971; 8) Connelly and Ball198; 9) Cooch 198; 0) Cooper 1978; 1) Cramp and Simmons 1977; ) Da and Kistchinski 1977; ) Dean and Skead

32 SPACNG PA'TERNS ; ) Dement'ev and Gladkov 19; ) Dmbell 1986; 6) Dwyer 197; 7) Dzbin 1969; 8) Eisenhaer and Kirkpatrick 1977; 9) Eldridge 1986a; 0) Eldridge 1986b; 1) Erskine 197a; ) Erskine 197b; ) Ewaschk and Boag 197; ) Frith 1967; ) Gates 196; 6) Gathier 1987a, b; 7) Gathier and Smith 1987; 8) Gillham 1987; 9) Gilmer et al. 197; SO) Gladstone and Manell 1968; 1) Grice and Rogers 196; ) Giler 1967; ) Hammell197; ) Hansen et al. 1971; ) Hilden 196; 6) Hochbam 19; 7) Hohman 198; 8) Hori 1969; 9) Johnsgard 1961; 60) Johnsgard 1966b; 61) Johnsgard 1978; 6) Kear 197; 6) Kear 197; 6) Kear and Steel1971; 6) Kerbes 197; 66) Kerbes et al. 1971; 67) Lavery 197; 68) Lind 198; 69) Lind 196; 71) Livezey and Hmphrey 198a; 7) MacKenzie and Kear 1976; 7) MacNae 199; 7) Mathiasson 196; 7) Matthews and Evans 197; 76) McKinney 1967; 77) McKinney 1970; 78) McKinney 197; 79) McKinney et al. 1978; 80) Mendaii19S8; 81) Mickelson 197; 8) Middlemiss 198; 8) Moffett 1970; 8) Morse et al. 1969; 8) Newton and Campbell197; 86) Nechterlein and Storer 198Sa; 87) Oliver 19; 88) Ogilvie 1978; 89) Owen 1980; 90) Palmer 1976; 91) Parmelee et al. 1967; 9) Panridge 196; 9) Patterson 198; 9) Perrins and Reynolds 1967; 9) Perringiii196S; 96) Pienkowski and Evans 198; 97) Poston 197; 98) Reid and Roderick 197; 99) Riggen 1977; 100) Rowan 196; 101) Ryder 1967; 10) Ryder 1971; 10) Rylander and Bolen 1970; 10) Savard 198; los) Savard 198; 106) Savard 1986; 107) Sayler 198; 108) Scott 198; 109) Scott and Lbbock 197; 110) Seymor 197b; 111) Seymor and Titman 1978; 11) Siegfried 196; 11) Siegfried 1968; 11) Siegfried 1976a; 11) Siegfried 1976b; 116) R. Smith 1968; 117) Sowls 19; 118) Stewan and Titman 1980; 119) Stolen and McKinney 198; 10) Titman 198; 11) Titman and Seymor 1981; 1) Trager 1971; 1) Weller 199; 1) Weller 1967; 1) Weller 1968a; 16) Weller 1968b; 17) Weller 197; 18) Weller 197; 19) Weller 197a; 10) Weller 197Sb; 11) Weller 1976; 1) Winterbottom 197; 1) Wishan 198; 1) Yong and Titman 1986; 1) Yong 1970; 16) Yong 197; 17) Molton and Weller 198; 18) Williams 1979; 19) Kingsford 1986; 10) Siegfried 1978; 11) Bolen and Rylander 198. References Afton, A. D. (1980). Factors affecting incbation rhythms of northern shovelers. Condor 8, Afton, A. D. (198). "Male and female strategies for reprodction in lesser scap." Ph.D. Dissenation. Univ. of Nonh Dakota, Grand Forks. Afton, A. D. (198). nflence of age and time on reprodctive performance of female lesser scap. Ak 101, -6. Afton, A. D. (198). Forced coplation as a reprodctive strategy of male lesser scap: a field test of some predictions. Behavior 9, Akesson, T. R., and Raveling, D. G. (198). Behaviors associated with seasonal reprodction and long-term monogamy in Canada Geese. Condor 8, Aldrich, J. W. (197). Disparate sex ratios in waterfowl. n "Breeding biology of birds." (D.S. Farner, ed.), pp Nat. Academy Sciences, Washington, D.C. Alexander, R. D. (197). The evoltion of social behavior. Ann. Rev. col. Syst., -8. Alexander, W. C. (1980). "The behavioral ecology and sociobiology of non-breeding diving dcks (Aythyini)." Ph.D. Dissertation. Clemson Univ., Clemson, Soth Carolina. Ali, S., and Ripley, S.D. (1968). "Handbook of birds of ndia and Pakistan." Vol.. Oxford Univ. Press, Oxford. Alison, R. M. (197). Breeding biology and behavior of the oldsqaw (Ciangla hyemalis L.). Ornithol. Monogr. 18, pp. Altm, B. (1868). "Der Vogel nd sein Leben." Mnster. Amat, J. A. ( 198). Prsit flights of Mallards and Gadwall nder different environmental conditions. Wildfow/, Anderson, M.G. (198). Parental investment and pair-bond behavior among canvasback dcks (Aythya valisineria, Anatidae). Behav. Ecol. Sociobiol. 1, Anderson, M.G. (198a). "Social behavior of breeding Canvasbacks: male and female strategies of reprodction." Ph.D. Dissenation, Univ. of Minnesota, Minneapolis. Anderson, M. G. ( 198b). Variations on monogamy in Canvasbacks (Aythya valisineria). Ornithol. Monogr. 7, Anderson, M.G., Hepp, G. R., McKinney, F., and Owen, M. (1987). Workshop Smmary: Cortship and pairing in winter. n "Waterfowl in winter." (M. W. Weller, ed.), pp Univ. Minnesota Press, Minneapolis. Ankney, C. D., and Macinnes, C. D. (1978). Ntrient reserves andreprodctive performance of female Lesser Snow Geese. Ak 9, Ashcroft, R. E. (1976). A fnction of the pair bond in the Common Eider. Wildfow/7, Bailey, R. 0. (1981). "The postbreeding ecology of the redhead dck (Aythya americana) on Long sland Bay, Lake Winnipegosis, Manitoba." Ph.D. Dissenation. McGill Univ., Montreal, Qebec. Bailey, R. 0., and Batt, B. D.J. (197). Hierarchy of waterfowl feeding with whistling swans. Ak 91, Balham, R. W. (19). "The behavior of the Canada Goose (Branta canadensis) in Manitoba." Ph.D. Dissenation. Univ. of Missori, Colmbia. Ball,. J., Frost, P. G. H., Siegfried, W. R., and McKinney, F. (1978). Territories and local movements of African Black Dcks. Wildfowl 9, Banko, W. E. (1960). "The Trmpeter Swan." Nonh American Fana Series No. 6, U.S. Fish Wild!. Serv., Washington, D.C. Barash, D. P. (1977). "Sociobiology and behavior." Elsevier Nonh Holland, nc., New York. Barry, T. W. (1966). "The geese of the Anderson River Delta, Northwest Territories." Ph.D. Dissertation. Univ. of Albena, Edmonton. Bastock, M. (1967). "Conship: an ethological stdy." Aldine Press, Chicago. Batt, B. D. J., Anderson, M. G., Anderson, C. D., and Caswell, F. D. (1989). The se of prairie potholes by Nonh American dcks. n "Northern prairie wetlands." (A. van der Valk, ed.), pp owa State Univ. Press, Ames. Baer, K. M., and Gltz von Blotzheim, U.N. (1968). "Handbck der Vogel Mitteleropas." Vol.. Akademische Verlagsgesellschaft, Frankfrt. Bellrose, F. C. ( 1976). "Dcks, geese and swans of Nonh America." Stackpole, Harrisbrg, Pennsylvania. pp. Bengtson, S.-A. (1966). Field stdies on the Harleqin Dck in celand. Wi/dfow/17, Bengtson, S.-A. ( 1970). Location of nest sites of dcks in Lake Myvatn area, Northeast celand. Oikos. 1, Bengtson, S.-A. (1971). Habitat selection of dck broods in Lake Myvatn area, Nonheast celand. Ornis. Scand., Bengtson, S.-A. (197). Breeding ecology of the harleqin dck Histrionics histrionics (L) in celand. Ornis Scand., Bennett, L. J. (198). "The Ble-winged Teal." Collegiate Press, Ames, owa. Bertram, B. C. R. (1978). Living in grops: predators and prey. n "Behavioral ecology-an evoltionary approach." (J. R. Krebs and N. B. Davies, eds.), pp Blackwell Sci. Pb!., Oxford. Bezzel, E. (199). Beitrage zr Biologie der Geschlechter bei Entenvogeln. Am:. Om. Ges. Bayern., 69-. Bezzel, E. (1969). "Die Tafelente." Nee Brehm Becherei. A. Ziemsem Verlag, Ltherstadt, Germany.

33 İ :,. ill f, 8 MCHAEL G. ANDERSON AND ROGER D. TTMAN Birkhead, M., and Perrins, C. M. {1986). "The Mte Swan." Croom Helm, London. Blohm, R. j. (1979). "Breeding ecology of the Gadwall in sothern Manitoba." Ph.D. Dissertation. Univ. of Wisconsin, Madison. Bolen, E. G. (1967). "The ecology of the Black-bellied Tree Dck in sothern Texas." Ph.D. Dissertation. Utah State Univ., Logan. Bolen, E. G. (1971). Pair bond-tenre in the black-bellied tree dck. f. Wild/. Manage., Bolen, E. G., and Smith, E. N. (1979). Notes on the incbation behavior of Black-bellied Whistling Dcks. Prairie Nat. 11, Bolen, E. G., and Rylander, M. K. (198). Whistling dcks: Zoogeography, ecology, anatomy. Texas Tech. Univ. Ms. Special Pb/. 0. Lbbock., Texas. Boffard, S. H. (198). Redhead parasitism of canvasback nests. J. Wild/. Manage. 7, Boyd, H. (19). On enconters between wild white-fronted Geese in winter flock~. BehaviorS, Bradbry, j. W., and Vehrencamp, S. L. (1977). Social organization and foraging in emballonrid bats. ll. Mating systems. Behav. Ecol. Sociobiol., Braithwaite, L. W., and Frith, H. j. (1969). Waterfowl in an inland swamp in New Soth Wales. CSRO Wild/. Res. 1, Braithwaite, L. W. (1976). Notes on the breeding of the Freckled Dck in the Lachlan River Valley. Em 76, Brakhage, G. K. (196). Biology and behavior of tb-nesting Canada geese. f. Wild/. Manage. 9, Brown, J. L. (196). The evoltion of diversity in avian territorial sys terns. Wilson Bll. 76, Brown, J. L. (1969). Territorial behavior and poplation reglation in birds: a review and re-evalation. Wilson Bll. 81,9-9. Brown, J. L. (197). "The evoltion of behavior." Norton and Co., New York. Brown,}. L., and Orians, G. H. (1970). Spacing patterns in mobile animals. Ann. Rev. Ecol. Syst. 1, 9-6. Brown, L. H., Urban, E. K., and Newman, K. (198). "The birds of Africa." Vol.. Academic Press, New York. Brown, P. W., and Brown, M.A. ( 1981 ). Nesting biology of the white winged scorer. J. Wild/. Manage., 8-. Brown, P. W., and Brown, M.A. ( 1981). Nesting biology of the white winged scorer. f. Wild/. Manage., 8-. Btggers, R. L. (1979). Nesting patterns of captive Mandarin Dcks. Wild(ow/0, -. Bxton, N. E. (197). "The feeding behavior and food spply of the Common Sheldck (Tadorna tadorna) on the Ythan Estary, Aberdeenshire." Ph.D. Dissertation. Aberdeen Univ., Aberdeen, U.K. Caraco, T. ( 1980). On foraging time allocation in a stochastic environment. Ecol. Monogr. 61, Caraco, T. (1981). Risk-sensitivity and foraging grops. Ecology 6, 7-1. Caraco, T., Martindale, S., and Whitten, T. S. {1980). An empirical demonstration of risk-sensitive foraging preferences. Anim. Behav. 8, Carpenter, F. L., and Macmillen, R. E. (1976). Energetic cost of feeding territories in an Hawaiian honeycreeper. Oecologia 6,1-. Cawkell, A.M., and Hamilton, J. E. (1961). The birds of the Falkland slands. bis 10, 1-7. Christoleit, E. (199a). Uber das Reihen der Enten. Beitr. Fortpfl. Biolog. Vogel jahrg. S, -6. Christoleit, E. (199b). Nochmals das Reihen der Enten. Beitr. Fortpfl. Biolog. Vogeljahrg. S, Chronister, C. D. (198). "Egg-laying and incbation behavior of Black-bellied Whistling Dcks." M.S. Thesis. Univ. Minnesota, Minneapolis. Clark, A. (196). The Maccoa Dck (Oxyra maccoa [Eyton]). Ostrich, Clark., A. (1966). The social behavior patterns of the Sothern Pochard Netta erythrophtha/ma brnnea. Ostrich 7, -6. Clark, A. (1969). The behavior of the White-backed Dck. Wildfowl 0,71-7. Clark, A. (1971). The behavior of the Hottentot teal. Ostrich, Cltton-Brock, T. H., and Albon, S.D. (1979). The roaring of red deer and the evoltion of honest advertisement. Behavior 69, Cltton-Brock, T. H., and Harvey, P. H. (1977). Primate ecology and social organization. f. Zoo/. London 18, 1-9. Cltton-Brock, T. H., and Harvey, P. H. (1979). Comparison and adaptation. Proc. Royal Soc. London, B. 0, 7-6. Cltton-Brock, T. H., Ginness, F. E., and Albon, S. D. (198). "Red deer: the behavior and ecology of rwo sexes." Univ. Chicago Press, Chicago. Cody, M. L. (197). Optimization in ecology. Science 18, Connelly,}. W., and Ball,.J. (198). Comparisons of aspects of breeding Ble-winged and Cinnamon Teal in eastern Washington. Wilson Bll. 96, Cooch, F. G. (198). "The breeding biology and management of the Ble Goose (Chen caerlescens)." Ph.D. Dissertation. Cornell Univ., thaca, New York. Cooke, F., Abraham, K. F., Davies, J. C., Findlay, C. S., Healey, R. F., Sadra, A., and Segin, R. j. (198). The La Perose Bay Snow Goose project-a 1-year report. Dept. Biolo~n, Qeen's Univ., Kingston, Ontario. Cooper, J. A. ( 1978). The history and breeding biology of the Canada Geese of Marshy Point, Manitoba. Wild/. Monogr. 611, 87 pp. Colson, J. C. ( 1966). The inflence of the pair-bond on the breeding biology of the Kittiwake gll, Rissa tridactyla. J. Anim. Ecol., Colson, J. C. (197). The significance of the pair-bond in the Kitti wake. Proc. nt. Ornithol. Congr. 1, -. Colson,}. C. (198). The poplation dynamics of the Eider Dck Somateria mollissima and evidence of extensive non-breeding by adlt dcks. bis 16, -. Cramp, S., and Simmons, K. E. L. (eds.) ( 1977). "Handbook of the birds of Erope, the Middle East, and North Africa: the birds of the western Palearctic." Vol. l. Ostrich to Dcks. Oxford Univ. Press, Oxford. Crook, J. H. ( 196). The evoltion of social organization and visal commnication in the weaver birds (Pioceinae). Behavior Sppl. 10, Crook, j. H. ( 196). The adaptive significance of avian social organization. Symp. Zoo/. Soc. Land. 1, Daly, M., and Wilson, M. (1978). "Sex, evoltion and behavior." Dxbry l>ress, North Scitate, Massachsetts. Da, C. P., and Kistchinski, S. A. ( 1977). Seasonal movements and distribtion of the Spectacled Eider. Wildfow/8, 6-7. Davies, N. B. (1978). Ecological qestions abot territorial behavior. n "Behavioral ecology-an evoltionary approach." (j. R. Krebs and N. B. Davies, eds.), pp Blackwell Sci. Pbl., Oxford. Davies, N. B. (1980). The economics of territorial behavior in birds. Ardea 68, 6-7. Davies, N. B., and Hoston, A.. (1981). Owners and satellites: the economics of territory defence in the pied wagtail, Motacilla alba. f. Anim. Ecol. SO, Davies, N. B., and Hoston, A.. (198). Time allocation between ter ritories and flocks and owner-satellite conflict in foraging pied wagtails, Motacilla alba. f. Anim. Ecol., Davies, N. B., and Hoston, A.. (198). Territory economics. n "Behavioral ecology-an evoltionary approach." nd ed. (j. R. Krebs and N. B. Davies, eds.), pp Blackwell Sci. Pbl., Oxford. Davies, N. B., and Krebs, j. R. (1978). ntrodction: ecology, natral selection and social behavior. n "Behavioral ecology-an evol-

34 SPACNG PATERNS 8 rionary approach." (j. R. Krebs and N. B. Davies, eds.), pp Blackwell Sci. Pbl., Oxford. Dawkins, R. (198). "The exrended phenorype." W. H. Freeman, San Francisco, California. Dawkins, R., and Krebs, J.R. (1978). Animal signals: information or maniplation? n "Behavioral ecology-an evoltionary approach." (j. R. Krebs and N. B. Davies, eds.), pp Blackwell Sci. Pbl., Oxford. Dean, W. R. J., and Skead, D. M. (1977). The sex ratio in Yellowbilled Dck, Redbilled Teal and Sothern Pochard. Ostrich Spp/. 1, 8-8. Delacor, J. (19-6). "The waterfowl of the world." vols. Contry Life, London. Delnicki, D. (198). Mate changes by Black-bellied Whistling Dcks. Ak 100, Dement'ev, G. P., and Gladkov, N. A. (eds.). (19). "Birds of the Soviet Union." Vol.. (English translation, srael Program for Scientific Translation, Jersalem, 1967.) Derksen, D. V., and Eldredge, W. D. (1980). Droght displacement of pintails to the arctic coastal plain, Alaska.]. Wild/. Manage., - 9. Derrickson, S. R. (1977). "Aspects of breeding behavior in the Pintail (Anas acta)." Ph.D. Dissertation. Univ. of Minnesota, Minneapolis. Derrickson, S.R. (1978). The mobility of breeding pintails. Ak 9, Dickey, D. R., and Van Rossem,A.J. (19). The Flvos tree dcks of Bena Vista Lake. Condor, 8-0. Dimond, S., and Lazars, J. (197). The problem of vigilance in animal life. Brain Behav. and Evol. 9, D'ombrain, A. F. (19a). Notes on yong of plmed tree-dck. Em, -7. D'ombrain, A. F. (19b). Nesting notes on plmed tree-dcks. Em,80-8. Drent, R. H., and Sweirstra, P. (1977). Goose flocks and food: field experiments with Barnacle Geese (Branta lecopsis) in winter. Wildfow/8, 1-0. Debbert, H. F., Lokemoen, J. T., and Sharp, D. E. (198). Concentrated nesting of mallards and gadwall on Miller Lake sland, North Dakota.]. Wild/. Manage. 7, Dmbell, G. (1986). The New Zealand Brown Teal: Wildfow/7, Dnbar, R.. M. ( 198). ntra specific variations in mating strategy. n "Perspectives in ethology." Vol.. (P. P. G. Bateson and P. H. Klopfer, eds.), pp Plenm Press, New York. Dnn, E. H., and Macinnes, C. D. (1987). Geographic variation in dtch size and body size of Canada Geese. ]. Field Ornithol. 8, -71. Dwyer, T. J. ( 197). Social behavior of breeding gadwalls in North Dakota. Ak 91, Dwyer, T. j., Krap, G. L., and Janke, D. M. (1979). Use of prairie pothole habitat by breeding mallards. j. Wild/. Manage., 6-1. Dzbin, A. (19). Some evidence of home range in waterfowl. Trans. N. Amer. Wild/. Con(. 0, Dzbin, A. (197). Pairing display and spring and smmer flights of the Mallard. Ble jay 1, Dzbin, A. D. ( 1969). Comments on carrying capacity of small ponds for dcks and possible effects of density on mallard prodction. Can. Wild/. Serv. Rep. Ser. 6, Eadie, J. M., Kehoe, F. P., and Ndds, T. D. (1988). Pre-hatch and post-hatch brood amalgamation in North American Anatidae: a review of hypotheses. Can. ]. Zoo/. 66, Eisenhaer, D. 1., and Kirkpatrick, C. M. (1977). Ecology of the Emperor goose in Alaska. Wild/. Monogr. 7, 6 pp. Eldridge, J. L. (1979). Display inventory of the Torrent Dck. Wildfow/0, -1. Eldridge, J. L. (1986a). Territoriality in a river specialist: the Ble Dck. Wildfow/7, 1-1. Eldridge, J. L. (1986b). Observations on a pair of Torrent Dcks. Wi/dfow/7, Ely, C. R., Bdea, D. A., and Swain, U. G. (1987). Aggressive enconters between tndra swans and greater white-fronted geese dring brood rearing. Commentary. Condor 87, 67. Emlen, S. T., and Oring, L. W. (1977). Ecology, sexal selection and the evoltion of mating systems. Science 197, 1-. Endler, J. A. (1986). "Natral selection in the wild." Princeton Univ. Press, Princeton. Erickson, R. C. (198). "Life history and ecology of the canvasback, Nyroca valisineria (Wilson), in sotheastern Oregon." Ph.D. Dissertation. owa State College, Ames. Erskine, A. J. (197a). "Bffleheads." Can. Wild/. Serv. Monogr.. Erskine, A. J. (197b). Poplations, movements and seasonal distribtion of mergansers in northern Cape Breton sland. Can. Wild/. Serv. Rep. Ser. 17. Ewaschk, E., and Boag, D. A. ( 197). Factors affecting hatching sccess of densely nesting Canada geese. J. Wild/. Manage. 6, Flickinger, E.L. (197). ncbation by a male Flvos tree dck. Wilson Bll. 87, Fretwell, S. D. ( 197). "Poplations in a seasonal environment." Princeton Monogr. Pop. Bioi.. Princeton Univ. Press, Princeton, New jersey. Fretwell, S. D., and Lcas, H. L., Jr. (1970). On territorial behavior and other factors inflencing habitat distribtion in birds.. Theoretical development. Acta Biotheoretica 19, Frith, H. J. (1967). "Waterfowl in Astralia." Angs and Robertson, Sydney. Frith, H. J. (198). "Waterfowl in Astralia." nd edition. Angs and Robertson,Sydne~ Frith, H. J., and Davies, S. J. j. F. (1961). Ecology of the Magpie Goose, Anseranas semipalmata Latham (Anatidae). C.S..R.O. Wild/. Res. 6, Gates, J. M. ( 196). Breeding biology of the gadwall in northern Utah. Wilson Bll. 7, -67. Gathier, G. (1986). An evoltionary model of territorial behavior in dcks. Poster presentation at nternational Ornithological Congress, Ottawa, Canada. Jne, Gathier, G. (1987a). The adaptive significance of territorial behavior in breeding bffleheads: a test of three hypotheses. Anim. Behav., Gathier, G. (1987b). Brood territories in Bffleheads: determinanrs and correlates of territory size. Can.]. Zoo/. 6, Gathier, G. (1988). Territorial behavior, forced coplations and mixed reprodctive strategy in dcks. Wildfow/9, Gathier, G., and Smith, J. N. M. (1987). Territorial behavior, nestsite availability, and breeding density in Bffleheads. ]. Anim. Ecof. 6, Geist, V. ( 197). On the relationship of social evoltion and ecology in nglates. Am. Zoo{. 1, 0-0. Geyr von Schweppenbrg, H. (19). Zr Sexalethologie der Stockente. J. Ornithol. 7, Geyr von Schweppenbrg, H. (199). Das Reihen der Stockenten. Beitr. Fortpflanzngsbiol. Voge/, Gill, F. B., and Wolf, L. L. (197). Economics of feeding territoriality in the golden-winged snbird. Ecology 6, -. Gillham, E. (1987). "Tfted Dcks in a royal park." Hythe Printers, Ltd. Hythe, Kent, U.K.

35 . ~ MCHAEL G. ANDERSON AND ROGER D. TTMAN 'i Gilmer, D. S., Ball,. j., Cowardin, L. M., Riechmann, j. H., and Tester, j. R. ( 197). Habitat se and home range of mallards breeding in Minnesota.]. Wild/. Manage. 9, Gladstone, P., and Martell, C. (1968). Some field notes on the breeding of the Greater Kelp Goose. Wildfow/19, -1. Goodall, j. (1986). "The chimpanzees of Gombe." Harvard Univ. Press, Cambridge, MassachseTtS. Goodbm, S. F. (198). Mate garding in the mallard Anas platyrhynchos. Omis. Scand. 1,61-6. Greenwood, P. j., and Harvey, P. H. (198). The natal and breeding dispersal of birds. Ann. Rev. Ecol. Syst. 1, 1-1. Gregoire, P. E. J. (198). "Behavior of family and other social grops in wintering and migrating Lesser Snow Geese." M.S. Thesis. Univ. Western Ontario, London. Grice, D., and Rogers, j. P. (196). "The wood dck in Massachserts." Mass. Div. Fisheries & Game. Final Rept., Project No. W-19-R. Giler, E. R. (1967). The Cape Barren Goose, its environment, nmbers and breeding. Em 66, 11-. Gminski-Sorenson, L. (199). Variable mating system of a sedentary tropical dck: the white-cheeked Pintail (Anas bahamensis bahamensis). Ak 109, Hamilton, W. D. (196). The genetical evoltion of social behavior.,.]. Theor. Bioi. 7, 1-. Hamilton, W. D. (1971). Geometry for the sdfish herd. J. Theor. Bioi. 1,9-11. Hammell, G. S. (197). "The ecology of lesser scap (Aythya affinis Eyton) in sothwest Manitoba." M.S. Thesis. Univ. of Gelph, Gelph, Ontario. Hammond, M. C., and Mann, G. E. (196). Waterfowl nesting is lands.]. Wild/. Manage. 0, -. Hansen, H. A., Shepherd, P. E. K., King, ]. G., and Troyer, W. A. (1971). The trmpeter swan in Alaska. Wild/. Monogr. 6. Hanson, H. C. (196). "The Giant Canada Goose." Sothern lllinois Univ. Press, Carbondale, llinois. Harper, D. G. C. (198). Competitive foraging in mallards: ideal free dcks. Anim. Behav. 0, 7-8. Harvey, P. H., and Mace, G. M. (198). Comparisons berween taxa and adaptive trends: problems of methodology. n "Crrent problems in sociobiology." (King's College Sociobiology Grop, eds.), pp Cambridge Univ. Press, Cambridge. Heinroth, 0. (1911). Beitrage zr Biologic, namentlich Ethologie nd Psychologic: der Anatiden. Proc. nt. Omithol. Congr., Hepp, G. R., and Hair, J. D. ( 198). Dominance in wintering water fowl (Anatini): effects on distribtion of sexes. Condor 86, 1-7. Hilden, 0. (196). Ecology of dck poplations in the island grop of Valassaaret, Glf of Bothnia. Ann. Zool. Fenn., Hinde, R. A. (196). The biological significance of territories of birds. bis 98, Hochbam, H. A. (19). "The Canvasback on a prairie marsh." Amer. Wildt. lnst., Washington, D.C. Hohman, W. L. (198). "Aspects of the breeding biology of ring necked dcks (Aythya collaris)." Ph.D. Dissertation. Univ. of Minne sora, St. Pal. Hoogland, J. L., and Sherman, P. \'(~ (1976). Advamages and disadvan rages of bank swallow (Riparia riparia) coloniality. col. Monogr. 6,-8. Hori, j. (196). Three-bird flights in the mallard. Wildfowl1, 1 1. Hori, J. (1969). Social and poplation stdies in the Sheldck. Wild fowl 0, -. Horn, H. S. ( 1968 ). The adaptive significance of colonial nesting in the Brewer's blackbird Ephags cyanocephals. Ecology 9, Howard, H. E. (190). "Territory in bird life." (196 reprint.) Collins Fontana, London. Hmbrg, D. D., Prince, H. H., and Bishop, R. A. (1978). The social organization of a mallard poplation in Northern owa. ]. Wild/. Manage., nglis,. R. (1976). Agonistic behavior of breeding Pink-footed Geese with reference to Ryder's hypothesis. Wildfowl 7, nglis,. R., and Lazars, J. (1981). Vigilance and flock size in brent geese: the edge effect. z. Tierpsychol. 7, Jarman, P. j. (197). The social organization of antelope in relation to their ecology. Behavior 8,1-67. Jarman, P. j. (198). Prospects for interspecific comparison in sociobiology. n "Crrent problems in sociobiology." (King's College Sociobiology Grop, eds.), pp. -. Cambridge Univ. Press, Cambridge. johnsgard, P. A. (1961). Breeding biology of the Magpie Goose. Wildfow/1, Johnsgard, P. A. (196). "Handbook of waterfowl behavior." Cornell Univ. Press, thaca, New York. johnsgard, P. A. (1966a). Behavior of the Astralian msk dck and ble-billed dck. Ak 8, johnsgard, P. A. (1966b). The biology and relationships of the Torrent Dck. Wildfow/17, johnsgard, P. A. (1978). "Dcks, geese and swans of the world." Univ. of Nebraska Press, Lincoln. Johnson, D. H. (1986). "Empirical Bayes estimates of breeding pop lations of North American dcks." Ph.D. Dissertation. North Da kota State Univ., Fargo. Johnstone, S. T. ( 197). On breeding whistling dcks. Avic. Mag. 6,. jones, R. D., Jr., and jones, D. M. (1966). The process of family dis integration in Black Brant. Wildfotv/17, janes, F. (1986). Poplation density and body size in birds. Am. Nat. 18, Kear,]. ( 1970). The adaptive radiation of parental care in waterfowl. n "Social behavior in birds and mammals." U. H. Crook, ed.), pp Academic Press, London. Kear, J. (197). Reprodction and family life. n "The Swans," (P. Scott, ed.), pp Michael Joseph, London. Kear, J. (197). Salvadori's Dck of New Ginea. Wildfow/6, Kear,J., and Steel, T. H. (1971). Aspects of social behavior in the ble dck. Notomis 18, Kehoe, F. P. (1986). "The adaptive significance of creching behavior m the White winged Scorer (Melanitta fsca deglandi)." M.S. Thesis. Univ. of Gelph, Gelph, Ontario. Kenward, R. E. (1978). Hawks and doves: artack sccess and selection in goshawk flights at wood pigeons.). Anim. Ecol. 7, Kerbes, R. H. (197). The nesting poplation of Lesser Snow Geese in the eastern Canadian Arctic: a photographic inventory of jne 197. Can. Wild/. Serv. Rep. Ser.. Kerbes, R. H., Ogilvie, M. A., and Boyd, H. ( 1971 ), Pink-footed geese of celand and Greenland: a poplation review based on an aerial nesting srvey of Pjorsarver in Jne, Wildfow/, -17. Kingsford, R. (1986). "Reprodctive biology and habitat se of the Maned Dck Chenonetta j11bata (latham)." Ph.D. Dissertation. Univ. of Sydney, Sydney, Astralia. Knapton, R. W., and Krebs,]. R. (197). Settlement patterns, territory size, and breeding density in the Song Sparrow (Melospiza melodia). Can. J. Zoo/., Koskimies, j., and Rotamo, E. (19). Zr fortplanzngsbiologie der Samente Melanitta f. fsca (L.). 1: Allgemeine Nisttkologie. Finn. Pap. Game Res. 10, Krap, G. l., Klett, A. T., and Jorde, D. G. ( 198 ). The effect of vari able spring water conditions on Mallard reprodction. Ak 100,

36 SPACNG PATERNS 8 Krebs, J. R. (1971). Territory and breeding density in the great tit, Pars ma;or L. Ecology, -. Krebs, J. R. (197). Colonial nesting and social feeding as strategies for exploiting food resorces in the Great Ble Heron (Ardea herodias). Behavior 1, Krebs, J. R., and Davies, N. B. (1987). "An introdction to behavioral ecology." nd edition. Blackwell Sci. Pb., Oxford. Krebs, J. R., and Dawkins, R. (198). Animal signals: mind-reading and maniplation. n "Behavioral ecology-an evoltionary ap proach." nd edition U.R. Krebs and N.B. Davies, eds.). Blackwell Sci. Pbl., Oxford. Krk, H. (196). Predators and anti-predator behavior in the blackheaded gll (Lars ridibnds). Behavior Sppl. 11, Krk, H. (197). "The spotted hyena." Univ. of Chicago Press, Chicago. Lack, D. (196). Evoltionary ecology.]. Appl. Ecol. l, 7-. Lack, D. (1966). "Poplation stdies of birds." Clarendon Press, Ox ford. Lack, D. (1968). "Ecological adaptations for breeding in birds." Methen and Co., London. Lack, D. (197). "Evoltion illstrated by waterfowl." Harper and Row, New York. Lavery, H. J. (1970). The comparative ecology of waterfowl in north Qeensland. Wildfowlll, lavery, H. J. (197). The Grey Teal at saline droght-refges in north Qeensland. Wild(ow/1, 6-6. Lazars, J. (1978). Vigilance, flock size and domain of danger size in the White-fronted Goose. Wild(ow/9, 1 1. Lazars, J., and nglis,. R. (1978). The breeding behavior of the pink-footed goose: parental care and vigilant behavior dring the fledging period. Behavior 6, lebret, T. (191). The display flight of the srface feeding dcks in their winter qarters. British Birds,1-1. Lehrer, T. (19). Die Verfolgngsflge der Enten. J. Ornithol. 96, - 9. Lehrer, T. (1961). The pair formation in the annal cycle of the mal lard, Anas platyrhynchos L. Ardea 9, Lefebvre, E. A., and Ravc:ling, D. G. (1967). Distribtion of Canada Geese in winter as related to heat loss at varying environmental rem perarres. j. Wildl. Manage. 1, 8-6. Leitch, W. G., and Kaminski, R. M. (198). Long-term wetland-waterfowl trends in Saskatchewan prairies. J. Wildl. Manage. 9, 1-. Lendrem, D. (1986). "Modelling in behavioral ecology." Croom Helm, london. Lima, S. l. (198). Territoriality in variable environments: a simple model. Am. Nat. 1, Lind, H. (198). Eine Unterschng iiber das Balzverhalten der Kol benente (Netta rfina Pallas). Z. Tierpsychol. 1, Lind, H. (196). Zr Analyse des sexellen Verhaltens der Kolbc:nente, Netta r{ina (Pallas). Z. Tierpsychol. 19, Lind, H. (198). The rotation display of the Mte Swan Cygns olor: synchronised neighbor responses as instrment in the territorial defence strategy. Ornis Scand. 1, Livezey, B. C. (1986). A phylogenetic analysis of recent Anseriform genera sing morphological characters. Ak 10, Livezey, B. C., and Hmphrey, P. S. (198). Mechanics of steaming in steamer-dcks. Ak 100, Livezey, B. C., and Hmphrey, P. S. (198a). Territoriality and interspecific aggression in Steamer-Dcks. Condor 87, Livezey, B. C., and Hmphrey, P. S. (198b). nterspecific aggression in steamer-dcks. Commentary. Condor 87, 68. Lokemoen, J. T., Debbert, H. F., and Sharp, D. E. ( 198). Nest spacing, habitat selection and behavior of waterfowl on Miller Lake s land, North Dakota./. Wild/. Manage. 8, 09-. Low, j. B. (19). Ecology and management of the redhead, Nyroca americana, in owa. Ecol. Monogr. 1, -69. Mace, G. M., and Harvey, P. H. (198). Energetic constraints on home range size. Am. Nat. 11, McKelvey, S.D. (1977). The Meller's Dck on Maritis: irs stats in the wild and captive propagation. Game Bird Breeders', Avicltrists', Zoologists' and Conservationists' Gazette, May-Jne 1977, MacKenzie, M. J. S., and Kear, J. (1976). The White-winged Wood Dck. Wildfowll7, -17. McKinney, F. (197-6). Reports of annal seminars held at the Delta Waterfowl Research Station. (Mimeo). McKinney, F. (196a). Spacing and chasing in breeding dcks. Wildfow/16, McKinney, F. (196b). The displays of the American Green-winged Teal. Wilson Bll. 77, McKinney, F. (1967). Breeding behavior of captive shovelers. Wildfow/18, McKinney, F. (1970). Displays of for species of ble-winged dcks. Living Bird 9, 9-6. McKinney, F. (197). Ecoethological aspects of reprodction. n "Breeding biology of birds." (D.S. Farner, ed.), pp National Academy of Sciences, Washington, D.C. ~fckinney, F. (197). The evoltion of dck displays. n "Fnction Jnd evoltion in behavior." (G. Baerends, C. Beer, and A. Manning, eds.), pp Clarendon Press, Oxford. McKinney, F. ( 1978). Comparative approaches to social behavior in closely related species of birds. Adv. Stdy Behav. 8, 1-8. McKinney, F. (198). Primary and secondary male reprodctive strat egies of dabbling dcks. Ornithol. Monogr. 7, McKinney, F. (1986). Ecological factors inflencing the social systems of migratory dabbling dcks. n "Ecological aspects of social evol tion." (D.. Rbenstein and R. W. Wrangham, eds.), pp Princeton Univ. Press, Princeton, New Jersey. McKinney, F. (1988). Mate garding and the confidence of paternity in waterfowl. Acta Congr. nt. Ornithol. 19, McKinney, F., Derrickson, S. R., and Minea, P. (198). Forced coplation in waterfowl. Behavior 86, 0 9. McKinney, F., Siegfried, W. R., Ball,. J., and Frost, P. G. H. (1978). Behavioral specializations for river life in the African Black Dck (Anas sparsa Eyton). Z. Tierpsychol. 8, McNab, B. K. (196). Bioenergetics and the determination of homerange size. Am. Nat. 97, ~facnae, W. (199). Notes on the biology of the Maccoa dck. Bokmakierie 11, 9-. McNaghton, S. J. (198). Grazing lawns: animals in herds, plant form, and coevoltion. Am. Nat. 1, McNaghton, S. J. (1986). Grazing lawns: on domesticated and wild grazers. Am. Nat. 18, Markl, H. (1981). "Evoltion of social behavior: hypotheses and em pi rica Tests." Verlag Chemie, Basel. Mathiasson, S. (196). The Bean Goose in Skane, Sweden with re marks on occrrence and migration throgh Northern Erope. Acta. Vertebratica, 19-. Matthews, G. V. T., and Evans, M. E. (197). On the behavior of the White-headed Dck with especial reference to breeding. Wildfowl,6-66. Maynard Smith, J. (197). The theory of games and the evoltion of animal conflicts. j. Theor. Bioi. 7, Maynard Smith,J. (1976). Grop selection. Qart. Rev. Bioi. 1,77-8. Maynard Smith, J. (198). "Evoltion and the theory of games." Cambridge Univ. Press, Cambridge. Maynard Smith, J., and Price, G. R. (197). The logic of animal conflict. Natre, Land. 6, 1-18.

37 86 MCHAEL G. ANDERSON AND ROGER D. TTMAN. i i il~ ' Mayr, E. (19). Bernand Altm and the territory theory. Proc. Linnean Soc. New York and 6, -8. MendaU, H. (198). "The Ring-necked Dck in the Nonheast." Bll. Univ. of Maine, Orono. Mickelson, P. G. (197). Breeding biology of cackling geese and associated species on the Ykon-Kskokwim Delta, Alaska. Wild/. Monogr.. Middlemiss, E. (198). The Sothern Pochard Nena erythrophtha/ma brnnea. Ostrich Sppl., 1-. Miller, R. S. ( 1967). Pattern and process in competition. Adv. col. Res., 1 7. Minea, P., and Cooke, F. {1979). Territoriality in Snow Geese or the protecrion of parenthood-ryder's and nglis's hypotheses re assessed. Wi/dfow/0, Moffett, G. M. ( 1970). A stdy of nesting Torrent Dcks in the Andes. Living Bird 9, 7. Morse, T. E., Jakabosky,J. L., and McCrow, V. P. {1969). Some aspects of the breeding biology of the hooded merganser.}. Wild/. Manage., Moss, R., and Watson, A. {198). Adaptive vale of spacing behavior in poplation cycles of red grose and other animals. n "Behavioral ecology-ecological conseqences of adaptive behavior." (R. M. Sibly and R. H. Smith, eds.), pp Blackwell Sci. Pb!., Oxford. Molton, D. W., and Weller, M. W. (198). Biology and conservation of the Laysan Dck (Anas laysanensis). Condor 86, Moynihan, M. (198). Notes on the behavior of the Flying Steamer Dck. Ak 7, Mnro, J., and Bedard, J. {1977a). Creche formation in the Common Eider. Ak 9, Mnro, J., and Bedard, J. ( 1977b). Gll predation and creching behav ior in the Common Eider.}. Anim. col. 6, Mrray, B. G., Jr. (1981). The origins ofadaptive interspecific territorialism. Bioi. Rev. Camb. Philos. Soc. 6, 1-. Mrray, B. G., Jr. (198). nterspecific aggression in steamer-dcks. Commentary. Condor 87, 67. Myers,]. P., Connors, P. G., and Pitelka, F. A. (1979). Territory size in wintering sanderlings: the effecrs of prey abndance and intrder density. Ak 96, Myers, J.P., Connors, P. G., and Pitelka, F. A. (1981). Optimal terri tory size and the sanderling: compromise in a variable environment. n "foraging behavior: ecological, ethological and psychological aproaches." (A. C. Kamiland T. D. Sargent, eds.), pp Garland STPM Press, New York. Newton,. ( 1986). "The sparrowhawk." T. & A. D. Poyser, Ltd., Calton, Staffordshire, England. Newton,., and Campbell, C. R. G. (197). Breeding of dcks at Loch Leven, Kinross. Wi/dfow/6, Newton, 1., and Kerbes, R. H. (197). Breeding of Greylag Geese Anser anser on the Oter Hebrides, Scotland. ). Anim. col., Nice, M. M. (197). Stdies in the life history of the song sparrow,. Trans. Linnaean Soc. New York, 1-7. Nice, M. M. (191 ). The role of territory in bird life. Am. Mid/. Nat. 6,1-87. Noble, G. K. (199). The role of dominance in the social life of birds. Ak 6, 6-7. Nolan, V., Jr. (1978). The ecology and behavior of the Prairie Warbler Dendroica discolor. Omithol. Monogr. 6. Norman, F.., and McKinney, F. {1987). Cltches, broods and brood care behavior in Chesmt Teal. Wildfow/8, Ndds, T. D., and Ankney, C. D. (198). Ecological correlates of territory and home range size in Nonh American dabbling dcks. Wild fowl, 8 6. Nechterlein, G. L., and Storer, R. W. (198a). Aggressive behavior and interspecific killing by Flying Steamer-Dcks in Argentina. Con dor 87, Nechterlein, G. L., and Storer, R. W. {198b).lnterspecific aggression in steamer-dcks. Commentary. Condor 87,68. Ogilvie, M.A. (1978). "Wild geese." Bteo Books, Vermillion, Soth Dakota. Ohde, B. R., Bishop, R. A., and Dinsmore,]. J. (198). Mallard reprodction in relation to sex ratios.}. Wild/. Manage. 7, Oliver, W. R. B. (19). "New Zealand birds." nd edition. A. H. and A. W. Reed, Wellington. Olson, D. P. (196). "A stdy of canvasback and redhead breeding poplations, nesting habitats, and prodcrivity." Ph.D. Dissenation. Univ. of Minnesota, St. Pal. Orians, G. H. (1961). The ecology of blackbird (Age/ais) social systems. col. Monogr. 1, 8-1. Orians, G. H. ( 1969). On the evoltion of mating systems in birds and mammals. Am. Nat. 10, Oring, L. W. { 198). Avian mating systems. n "Avian biology." Vol. 6 (D. S. Farner, J. R. King, and K. C. Parkes, eds.), pp Academic Press, New York. Owen, M. ( 1980). "Wild geese of the world." Bats ford, London. Owen, M., and Wells, R. L. ( 1979). Territorial behavior in breeding geese-a re-examination of R) der's hypothesis. Wi/dfow/0, 0 6. Palmer, R. S. (1976). "Handbook ofnonh American birds." Vols. &. Yale Univ. Press, New Haven, Connecrict. Parker, G. A. (198). Evoltionarily stable strategies. n "Behavioral ecology-an evoltionary approach." nd edition. (}. R. Krebs and N. B. Davies, cds.), pp Blackwell Sci. Pbl., Oxford. Parker, G. A., and Hammerstein, P. {198). Game theory and animal behavior. n "Evoltion-essays in honor of John Maynard Smith." (P. J. Greenwood, P. H. Harvey, and M. Slatkin, eds.), pp Cambridge Univ. Press, Cambridge. Parmelee, D. F., Stephens, H. A., and Schmidt, R. H. ( 1967). The birds of sotheastern Vicroria sland and adjacent small islands. Nat/. Ms. C.m. Bll., 9 pp. Panridge, W. H. (196). Notes on the Brazilian merganser in Argen tina. Ak 7, Patterson, 1. J. (196). Timing and spacing of broods in the black headed gll Lars ridibnds. bis 107, -9. Patterson,. J. ( 1980). Territorial behavior and the limitation of poplation density. Ardea 68, 6. Patterson,. J. ( 198). "The Sheldck-a stdy in behavioral ecol ogy." Cambridge: Univ. Press, Cambridge. Patterson,. J., Makepeace, M., and Williams, M. (198). Limitation of local poplation size: in the Sheldck. Ardea 71, Pals, S. L (198). Dominance: relations, resorce se, and pairing chronology of gadwalls in winter. Ak 100, Perrins, C. M., and Reynolds, C. M. ( 1967). A preliminary stdy of the mte swan, Cygns olor. Wi/d(ow/18, 7-8. Pettingill, 0. S. ( 196). Kelp Geese and Flightless Steamer Dcks in the Falkland slands. Livit1g Bird, Pienkowski, M. W., and Evans, P. R. ( 198). Breeding behavior, prodctivity and srvival of colonial and non colonial Shc:ldcks, Tadoma tadoma (L.). Ornis Scand. 1, Pospahala, R. S., Anderson, D. R., and Hc:nny, C.}. (197). Poplation ecology of the: Mallard. : Breeding habitat conditions, size of the breeding poplations, and prodction indices. U.S. Fish and Wild/. Serv. Resor. Pb/. 11. Poston, H. J. (197). Home range and breeding biology of the sho\' eler. Can. Wild/. Serv. Rep. Ser.. Powell, G. V. N. (197). Expenmental analysis of the social vale of flocking by starlings in relation to predation and foraging. Anim. Behav., 01-0.

38 SPACNG PATERNS 87 Prins, H. H.Th., Ydenberg, R. C., and Drent, R. H. (1980). The inter action of Brent Geese Branta bernie/a and sea plantain Plantago maritima dring spring staging: field observations and experiments. Acta Bmo. Nederl. 9, Plliam, H. R. (197). On the advantages of flocking. f. Theor. Bioi. 8, 19-. Plliam, H. R., and Millikan, G. C. ( 198). Social organization in the non reprodctive season. n "Avian biology." Vol. 6 (D. S. Farner, J. R. King, and K. C. Parkes, eds.), pp Academic Press, New York. Raitaso, K. ( 196 ). Social behavior of the mallard, Anas platyrhyn chos, in the corse of the annal cycle. Finn. Pap. Game Res., 1-7. Raveling, D. G. (1970). Dominance relationships and agonistic behav ior of Canada Geese in winter. Behavior 7, Raveling, D. G. (1989). Nest-predation rates in relation to colony size of Black Brant. f. Wild/. Manage., Reid, B., and Roderick, C. (197). New Zealand Scap Aythya novaeseelandiae and Brown Teal Anas acklandica chlorotis in captivity. nt. Zoo Yearbook 1, 1 1 Riggen, T. L. (1977). The biology of the Montain Dck on Rottnest sland, Western Astralia. Wild/. Monogr.. Ringelman, J. K., Longcore, J. R., and Owen, R. B., Jr. (198). Breeding habitat selection and home range of radio-marked black dcks (Anas rbripes) in Maine. Can. f. Zoo/. 60, 1-8. Rohwer, S. (1977). Stats signalling in Harris's sparrows: some experiments in deception. Behavior 61, Rohwer, F. C., and Anderson, M.G. (1988). Female-biased philopatry, monogamy and the timing of pair formation in waterfowl. Crrent Ornithol., Rowan, M. K. (196). The Yellowbill Dck, Anas ndlata Dbois in sothern Africa. Ostrich Sppl., 1 6. Rbenstein, D. 1., and Wrangham, R. W. (1986). "Ecological aspects of social evoltion." Princeton Univ. Press, Princeton, New Jersey. Ryder, J. P. (1967). The breeding biology of Ross' goose in the Perry River region, Northwest Territories. Can. Wild/. Serv. Rep. Ser.. Ryder, J. P. (1971). Distribtion and breeding biology of the Lesser Snow Goose in Central Arctic Canada. Wildfow/, Ryder, J. P. (197). The significance of territory size in colonial nesting geese-an hypothesis. Wildfowl 6, Rylander, M. K., and Bolen, E. G. (1970). Ecological and anatomical adaptations of North American tree dcks. Ak 87: Savard, J.-P. L. (198). ntra- and interspecific competition between Barrow's Goldeneye (Bcephala islandica) and Bfflehead (Bcephala albeola). Can. f. Zoo/. 60, 9-6. Savard, J.-P. L. (198). Territorial behavior of Common Goldeneye, Barrow's Goldeneye and Bfflehead in areas of sympatry. Ornis Scand. 1, Savard, J.-P. L. (1986). "Territorial behavior, nesting sccess and brood srvival in Barrow's Goldeneye and its congeners." Ph.D. Dissertation. Univ. of British Colmbia, Vancover. Savard, J.-P. L., and Smith,J. N. M. (1987). nterspecific aggression by Barrow's Goldeneye: a descriptive and fnctional analysis. Behavior 10, Sayler, R. ( 198). "Brood parasitism and reprodction of canvasbacks and redheads on the Delta Marsh." Ph.D. Dissertation. Univ. of North Dakota, Grand Forks. Schoener, T. W. (1968). Sizes of feeding territories among birds. Ecology 9, Schoener, T. W. (1971). Theory of feeding strategies. Ann. Rev. Ecol. Syst., Schoener, T. W. (198). Simple models of optimal feeding-territory size: a reconciliation. Am. Nat. 11, Scott, D. K. (1980). Fnctional aspects of the pair bond in winter in Bewick's Swans (Cygns colmbians beweckii). Behav. Ecol. So ciobiol. 7, -7. Scott, D. K. (198). Wintering territoriality of mte swans, Cygns olor. bis 6, Scott, D. K., and Lbbock, j. (197). Preliminary observations on wa terfowl of Western Madagascar. Wildfow/, Sebeok, T. A. (ed.). (1977). "How animals commnicate." ndiana Univ. Press, Bloomington. Serle, W. (19). Frther field observations on Northern Nigerian Birds. bis 8, Seymor, N. R. (197a). Site attachment in the northern shoveler. Ak 91, -7. Seymor, N. R. (197b). Territorial behavior of wild shovelers at Delta, Manitoba. Wildfow/, 9-. Seymor, N. R. (197c). Aerial prsit flights in the shoveler. Can.}. Zoo/., Seymor, N. R., and Titman, R. D. (1978). Changes in activity patterns, agonistic behavior and territoriality of black dcks (Anas r bripes) dring the breeding season in a Nova Scotia tidal marsh. Can. f. Zoo/. 6, Seymor, N. R., and Titman, R. D. (1979). Behavior of npaired male black dcks (Anas rbripes) dring the breeding season in a Nova Scotia tidal marsh. Can. f. Zoo/. 7, 1 8. Siegfried, W. R. (196). The Cape Shoveller Anas smithii (Harten) in Sothern Africa. Ostrich 6, Siegfried, W. R. (1968). The black dck in the soth-western Cape. Ostrich 9, Siegfried, W. R. ( 197). Brood care, pair bonds and plmages in sothern African Anatini. Wildfo v/, -0. Siegfried, W. R. (1976a). Social organization in rddy and maccoa dcks. Ak 9, Siegfried, W. R. (1976b). Breeding biology and parasitism in the Rddy Dck. Wilson Bll. 88, Siegfried, W. R. (1978). Social behavior of the African Comb Dck. Living Bird 17, Siegfried, W. R., and Underhill, L. G. (197). Flocking as an anti-predator strategy in doves. Anim. Behav., Siegfried, W. R., and van der Merwe, F. (197). A description and inventory of the displays of the Maccoa Dck. Z. Tierpsychol. 7, 1-. Skead, D. M. ( 1976). "Social behavior of the yellow-billed dck and red-billed teal in relation to breeding." M.S. Thesis. Univ. of Natal, Soth Africa. Skead, D. M. (1977). nterspecific aggressive behavior of Anatini. Ostrich Sppl. 1, 1-1. Smith, A. G. ( 1971 ). Ecological factors affecting waterfowl prodction in the Alberta parklands. U.S. Br. Sport Fish. Wild/., Resorc. Pb/. 98. Smith, C. C. (1968). The adaptive natre of social organization in the gens of tree sqirrels Tamiascirs. Ecol. Monogr. 8,1-6. Smith, R.. (1968). The social aspects of reprodctive behavior in the pintail. Ak 8, Smith, W. J, (1977). "The behavior of commnicating: an ethological approach." Harvard Univ. Press, Cambridge, Massachsetts. Sowls, L. K. (19). Prairie dcks. Stackpole, Harrisbrg, Pennsylvania. Stewart, G. R., and Titman, R. D. (1980). Territorial behavior by prairie pothole Ble-winged Teal. Can. f. Zoo/. 8, Stiles, F. G. (1971). Time, energy and territoriality of Anna Hmmingbird. Science 17, Stolen, P., and McKinney, F. (198). Bigamos behavior of captive Cape Teal. Wildfow/, 10-1.

39 88 MCHAEL G. ANDERSON AND ROGER D. TTMAN ; r j i. Stodt, J. H. (1971). Ecological factors affecting waterfowl prodc rion in the Saskatchewan parklands. U.S. Br. Sport Fish Wild., Re sor. Pb/. 99. Stodt, J. H. (198). Habitat se and prodaivity of canvasbacks in soth western Manitoba, U.S. Fish Wild/. Serv. Spec. Sci. Rept. Wild/. 8. Sgden, L. G. (1978). Canvasback habitat se and prodaivity in Saskatchewan parklands. Can. Wild/. Serv. Occas. P1p.. Sgden, L. G. ( 1980). Parasitism of canvasback nesrs by redheads. f. Field Ornithol. 1, Smmers, R. W. ( 198 ). The life cycle of the Upland Goose Chloeph.Jga picta in the Falkland slands. bis 1, -. Thompson, W. D., Veninsky, 1., and Krebs, J. R. (197). The srvival vale of flocking in birds: a simlation model. f. Anim. Ecol., Tinbergen, N. (197). The fnctions of territory. Bird Stdy, 1-7. Titman, R. D. (197). "The role of the prsit flight in the breeding biology of the mallard." Ph.D. Dissenarion. Univ. of New Brns wick, Fredericton. Titman, R. D. (198). Spacing and three-bird flights of mallards breeding in pothole habitat. Can. f. Zoo/. 61, Titman, R. D., and Lowther,). K. (197). The breeding behavior of a crowded poplation of mallards. Can. f. Zoo/., Titman, R. D., and Seymor, N. R. (1981). A comparison of prsit flights by six Nonh American dcks of the gens Anas. Wildfowl, Trager, D. L. (1971). "Poplation ecology of Lesser Scap (Aythya affinis) in Sbaraic Taiga." Ph.D. Dissenation. owa State Univ., Ames. Trivers, R. L. (1971). The evoltion of reciprocal altrism. Qart. Rev. Bioi. 6, -7. Trivers, R. L. (197). Parental investment and sexal selection. n "Sexal seleaion and the descent of man, " (B. Campbell, ed.), pp Aldine Pb. Co., Chicago. Ward, P., and Zahavi, A. (197). The imponance of cenain assem blages of birds as "information centres" for food finding. bis 11, 17. Waser, P.M., and Wiley, R. H. (1979). Mechanisms and evoltion of spacing in animals. n "Handbook of nerobiology and behavior." Vol. (P. Marler and J. G. Vandenbergh, eds.), pp Plenm Press, New York. Warson, A. (1977). Poplation limitation and the adaptive vale of territorial behavior in Scottish red grose, Lagops L. scotics. n "Evoltionary ecology." (B. Stonehose and C. M. Perrins, eds.), pp Univ. Park Press, Baltimore, Maryland. Warson, A., and Moss, R. ( 197). A crrent model of poplation dy namics in red grose. Proc. nt. Ornithol. Congr. 1, Warson, A., Moss, R., and Parr, R. (198a). Effects of food enrichment on nmbers and spacing behavior of red grose. f. Anim. Ecol., Warson, A., Moss, R., Rothery, P., and Parr, R. (198b). Demographic cases and predictive models of poplation flctations in red grose. f. Anim. Ecol., Weidmann, U. ( 196). Verhalrenssrdien an der Srockenre (Anas platy rhynchos L.).. Das Aktionssysrem. Z. Tierpsychol. 1, \l\!ller, M. W. (199). Parasitic egg-laying in the redhead (Aythya americana) and other Nonh American Anatidae. Ecol. Monogr. 9, -6. ~ller, M. W. ( 1967). Notes on some marsh birds of Cape San Anto nio, Argentina. bis 10, ~ller,.m. W. (1968a). Notes on some Argentine anatids. Wilson Bll. so, ~ller,.m. W. (1968b). The breeding biology of the parasitic Black headed Dck. Living Bird 7, ~ller, M. W. (197). Ecological stdies of Falkland slands' water fowl. Wildfow/, -. Weller, M. W. ( 197 ). Habitat selection and feeding patterns of Brown Teal (Anas castanea chlorotis) on Great Barrier sland. Notornis 1,. 'X~er, M. W. (197a). Ecological stdies of the Ackland slands flighdess real. Ak 9, Weller, M. W. (197b). Ecology and behavior of the Soth Georgia Pintail Anas g. georgica. bis 117, Weller, M. W. (1976). Ecology and behavior of steamer dcks. Wild fowl 7,. Welty, J. C. (198). "The: life of birds." rd edition. W.B. Sanders, Philadelphia. Wetmore, A. ( 190). Observations on the habits of birds at Lake Br ford, New Mexico. Ak 7, 1 7. Wiens, J. A. ( 1966). Grop selection and Wynne-Edwards' hypothesis. Am. Sci., Wiens, J. A. (1976). Poplation responses to patchy environments. Ann. Rev. Ecol. Syst. 7, Williams, G. C. ( 1966). "Adaptation and natral selection." Princeton Univ. Press, Princeton, New Jersey. Williams, M. (1979). The social strctre, breeding and poplation dynamics of Paradise Sheldcks in the Gisborne East Coast District. Notornis 6, 1 7. Williams, M. J. (197). "Dispersionary behavior and breeding of Sheldck Tadorna tadorna on the River Ythan Estary." Ph.D. Dis senation. Aberdeen Univ., Aberdeen, U.K. Wilson, E. 0. (197). "Sociobiology: the new synthesis." Belknap Press, Cambridge, Massachsetts. Winterbottom, J. M. (197). The Cape Teal. Ostrich, Wishan, R. A. (198). "The behavioral ecology of the American wigeon (Anas americana) over its annal cycle." Ph.D. Dissenation. Univ. of Manitoba, Winnipeg. Winenberger, J. F., and Hnt, G. L., Jr. (198). The adaptive signifi cance of coloniality in birds. n "Avian biology." Vol. 8. (D. S. Farner, J. R. King, and K. C. Parkes, eds.), pp Academic Press, New York. Wittenberger, J. F., and Tilson, R. L. (1980). The evoltion of monogamy: hypotheses and evidence. Ann. Rev. col. Syst. 11, 197. Wolf, L. L., and Hainswonh, F. R. ( 1971). Time and energy bdgets of territorial hmmingbirds. Ecology, Wolf, L. L., Hainswonh, F. R., and Gill, F. B. (197). Foraging effi ciencies and time bdgets in nectar-feeding birds. Ecology 6, Woolfenden, G. E., and Fitzpatrick, J. W. (1977). Dominance in the Florida Scrb jay. Condor 79, 1 1. Woolfenden, G. E., and Fitzpatrick, J. W. (1978). The inheritance of territory in grop-breeding birds. Bioscience 8, Woolfenden, G. E., and Fitzpatrick, J. W. (198). "The Florida Scrb Jay: demography of a cooperative-breeding bird." Princeton Monogr. Pop. Bioi. 0. Princeton Univ. Press, Princeton, New Jersey. Wrangham, R. W., and Rbenstein, D.. (1986). Social evoltion in birds and mammals. n "Ecological aspects of social evoltion." (D. 1. Rbenstein and R. W. Wrangham, eds.), pp. 70. Princeton Univ. Press, Princeton, New Jersey. Wst, W. (1960). Das Problem des Reihens der Enren, besonders vom Anas strepera. Proc. nt. Ornithol. Congr. 1, Wynne-Edwards, V. C. (196). "Animal dispersion in relation to social behavior." Oliver and Boyd, Edinbrgh, U.K. Ydenberg, R. C., and Prins, H. H.Th. (1981). Spring grazing and the maniplation of food qality by barnacle geese. f. Applied col. 18, -. Ydenberg, R. C., and Prins, H. H.Th. (198). Why do birds roost com mnally? n "Coastal waders and wildfowl in winter." (P.R. Evans,