Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Some aspects of dominance behavior in the song sparrow, (Melospiza melodia) Arcese, Peter 1984

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1984_A6_7 A74.pdf [ 3.85MB ]
Metadata
JSON: 831-1.0096044.json
JSON-LD: 831-1.0096044-ld.json
RDF/XML (Pretty): 831-1.0096044-rdf.xml
RDF/JSON: 831-1.0096044-rdf.json
Turtle: 831-1.0096044-turtle.txt
N-Triples: 831-1.0096044-rdf-ntriples.txt
Original Record: 831-1.0096044-source.json
Full Text
831-1.0096044-fulltext.txt
Citation
831-1.0096044.ris

Full Text

SOME ASPECTS OF DOMINANCE BEHAVIOR IN THE SONG SPARROW, (MELOSPIZA MELODIA) by PETER ARCESE B.A., University Of Washington, Seattle, Washington, 1981 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Zoology) We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA July 1984 fe) Peter Arcese, 1984 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r a n a d v a n c e d d e g r e e a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e a n d s t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may b e g r a n t e d b y t h e h e a d o f my d e p a r t m e n t o r b y h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t b e a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . D e p a r t m e n t o f T h e U n i v e r s i t y o f B r i t i s h C o l u m b i a 1956 Main M a l l V a n c o u v e r , C a n a d a V6T 1Y3 / D E - 6 (3/81) i i ABSTRACT I studied the effect of behavioral dominance on survival and recruitment in the song sparrow (Melospiza melodia). The aims of t h i s study were: 1) to determine i f superior dominance status allows high survival and p r i o r i t y of access to breeding t e r r i t o r y ; 2) to measure the influence of phenotypic characters on dominance status; 3) to test the hypothesis that f a m i l i a r i t y with a s i t e i s a prerequisite to achieving dominant status. Dominance was estimated by observing the agonistic encounters of young sparrows at feeders, and t h i s estimate was correlated with the subsequent survival and settlement of yearlings. Correlations were also sought between dominance and several c h a r a c t e r i s t i c s of in d i v i d u a l s . To study s i t e attachment, I temporarily confined 24 early-hatched birds in two groups. This allowed them experience in agonistic encounters, but prevented them from gaining s i t e attachments u n t i l a l l young birds had become independent. The dominance of these captive birds was estimated after their release and compared to that of control birds. In both sexes, higher proportions of dominants survived and settled than subordinates in each year of study. Only age and sex were consistently correlated with dominance; young hatched early were dominant to those hatched l a t e r , and males were more dominant than females. Overall, age accounted for 59% of the variation in dominance. Captive males and females were as dominant as control young of equal age, and were dominant to birds hatched l a t e r . These results support the hypotheses that dominant status in song sparrows allows high survival and p r i o r i t y of access to a breeding t e r r i t o r y . Natural selection should favor parents that raise many early offspring. F a m i l i a r i t y with a loc a l area was not a prerequisite to achieving dominant status. The assumption that large size is advantageous in agonistic encounters was not supported by this study, and a review of the l i t e r a t u r e suggests that many studies that support this assumption are based on inadequate analyses. TABLE OF CONTENTS ABSTRACT i i LIST OF TABLES v ACKNOWLEDGMENTS v i GENERAL INTRODUCTION 1 CHAPTER ONE: Correlates and Consequences of Behavioral Dominance in Wild Song Sparrows 3 Introduction 3 Methods 5 Study Area and Study Population 5 Morphology 6 Estimating Dominance 7 Temporary Removal Experiment 9 Results 10 Dominance, Survival, and Settlement 10 Eff e c t s of Sex, Size, and Age on Dominance 12 Effe c t s of Natal Care and Natal Environment on Dominance 20 Effec t of Captivity on Dominance 26 Discussion 29 Measures of Dominance 29 Consequences of Dominance 30 The Determinants of Dominance 32 Dominance and Natural Selection 38 CHAPTER TWO: Dominance, Body Size, and Habitat D i s t r i b u t i o n V 41 Introduction 41 Measures of Size and Determinants of Dominance 43 Problems in measuring body size 43 Sex, Age, and Size 46 Examples From the Literature 48 Habitat D i s t r i b u t i o n and Dominance in the Song Sparrow ..51 The Advantage of Size in Wild Song Sparrows 51 Results 53 Dominance and Habitat D i s t r i b u t i o n 56 Discussion 60 GENERAL SUMMARY 67 LITERATURE CITED 70 APPENDIX 82 v i LIST OF TABLES Table 1. Dominance, su r v i v a l , and settlement 13 Table 2. Sex and dominance 14 Table 3. Dominance and morphology 16 Table 4. Hatch date and morphology 17 Table 5. Ef f e c t of morphology c o n t r o l l i n g for hatch date ...18 Table 6. Dominance and hatch date 19 Table 7. Dominance and parental age 20 Table 8. Dominance and brood size 22 Table 9. Order of young within a brood 23 Table 10. The number of nestmates of the same sex 24 Table 11. Nestling morphology and hatch date 25 Table 12. Survival and settlement of captives 28 Table 13. Wing length and age 46 Table 14. Sex and dominance in juncos 49 Table 15. Morphology and the outcome of encounters . ., 55 v i i ACKNOWLEDGMENTS Several people contributed to the production of this thesis. Anne Helbieg, Doug Ried, Pieter Bets, and Michaela Waterhouse a l l helped to c o l l e c t data on Mandarte Island. Don Ludwig was indispensible in his advice on data analyses, and he suggested the weighting scheme used to estimate dominance more accurately. Dolph Schluter suggested the model on habitat d i s t r i b u t i o n , and he provided valuable c r i t i c i s m s that greatly improved Chapter Two. J i r o Kikkawa, Lisa Rotterman, Chuck Monet, F.R. Gehlbach, Arne Lundberg, Trevor Price, Jamie Smith, and Pieter Bets a l l were generous with unpublished data and manuscripts. Sievert Rohwer planted the seed of my interest in dominance behavior and in the l i v e s of individual animals, and he provided good c r i t i c i s m and much encouragement to t h i s project. Charley Krebs, Kim Cheng, Robin L i l e y , and Jamie Smith a l l read the entire thesis, and made valuable comments on i t . The l a t e r three individuals comprised my excellent research committee. Jamie Smith taught me the s k i l l s that enabled me to study song sparrows e f f e c t i v e l y , and he t i r e l e s s l y edited early drafts of this thesis. The staff of the biosciences data center provided essential technical support, and the Tswaout and Tseycum Indian Bands kindly allowed me to work on Mandarte Island. This thesis is dedicated to the memory of Anne Vallee, whose s p i r i t and love of l i f e is a constant i n s p i r a t i o n . $SIGNOFF 1 GENERAL INTRODUCTION Dominance behavior is prevalent among so c i a l vertebrates as diverse as f i s h and primates, and high dominance status often a f f e c t s an individual's access to resources c r u c i a l to survival and reproduction (Wilson 1975). However, l i t t l e i s known about the q u a l i t i e s of individuals that determine dominance status (Wilson 1975). This i s because researchers have generally concentrated their e f f o r t on only a small portion of the l i f e h istory of a species, and have had l i t t l e information about the prior h i s t o r i e s or the fates of the study subjects. The ideal species for a study of behavioral dominance is one that can be marked for individual i d e n t i f i c a t i o n , i s eas i l y observed, and i s r e l a t i v e l y s hort-lived. Song sparrows (Melospiza melodia) are such a species, and the population on Mandarte Island, B.C., Canada, is resident, and individuals can be color-marked as nestlings soon after they hatch. Yearling song sparrows roam in lose flocks before s e t t l i n g on t e r r i t o r i e s in late f a l l or early spring, and some aspects dominance behavior have been investigated within these flocks (Knapton 1973, Smith et a l . 1980). Thus, the Mandarte song sparrow population provided me with an excellent opportunity to investigate dominance behavior in d e t a i l . This thesis contains two chapters and an appendix. In the f i r s t chapter, I explore the advantages of dominant status among 2 yearlings, and assess the influence of several factors on dominance such as age, sex, morphology, s i t e f a m i l i a r i t y , and nestling size and condition. Chapter Two deals with one often-stated determinant of dominance, body si z e . The theory of sexual selection assumes that large size is advantageous in agonistic encounters to explain sexual size dimorphism in vertebrates (Selander 1972, Payne 1984). I review the evidence with regard to this assumption from bird studies, and provide e x p l i c i t tests with data from the song sparrow. In the Appendix, I describe a method to analyze win/loss data using standard s t a t i s t i c a l procedures that i s more accurate than methods previously used by behavioral ecologists. 3 CHAPTER ONE: CORRELATES AND CONSEQUENCES OF BEHAVIORAL DOMINANCE IN WILD SONG SPARROWS INTRODUCTION Characters that lead to high survival and reproductive success a,re favored by natural selection. Dominant status is t y p i c a l l y associated with superior survival and reproductive success among polygynous vertebrates (LeBoeuf and Peterson 1969, Wiley 1973, Hrdy 1977, Clutton-Brock et a l . 1982), and similar correlations have been documented for monogamous birds (Fretwell 1969, Knapton and Krebs 1974, Smith 1978, Kikkawa 1981). Since dominant status confers such advantages, an understanding of the determinants of dominance should provide insight into the operation of natural selection on individuals. Dominance behavior can c l e a r l y influence the reproductive success of individuals yet, few studies have investigated the causes and consequences of dominant and subordinate status in d e t a i l within one species. This i s because i t i s d i f f i c u l t to follow large numbers of animals from b i r t h to successful recruitment. Studies of dominance in birds have been largely r e s t r i c t e d to winter flocks which contained individuals of mixed sex and age (e.g. Sabine 1959, Fretwell 1969, Glase 1973, Ketterson 1979) or to the laboratory (e.g. C o l l i a s 1943, Baker and Fox 1978, Searcy 1979a). In only one bi r d study have the prior h i s t o r i e s and the fates of the individuals involved been 4 known (Kikkawa 1981). As a result, there is a suprising lack of understanding of the determinants of dominance status in wild birds. The song sparrow population on Mandarte Island, B.C., Canada, offers an unusual opportunity to study dominance behavior. Song sparrows there are resident, and the population has been completely color-marked since 1975. Nearly a l l young that r e c r u i t to the population are hatched on the island and have been previously banded as nestlings (Smith 1981). Thus, detailed observations on large numbers of birds of known parentage can be made. I studied dominance behavior among yearling song sparrows from 1982 to 1984 to assess the effect of dominance on survival and settlement. I also explored two classes of factors that might influence dominance. The f i r s t class involved aspects of the individual i t s e l f , such as sex, size, and age. The second class involved aspects of natal care and the natal environment, such as parental age, brood size, order within the brood, number of nestmates of the same sex, and nestling size and condition. In general, I expected dominance to depend upon sex (e.g. Glase 1973), and to increase with size (e.g. Searcy 1979a) and age (e.g. Smith et a l . 1980). If the condition of nestlings influences dominance, older parents might raise more dominant offspring because they are better at raising young (Smith 1981). S i m i l a r l y , dominance might decrease with increasing brood size i f larger broods are more d i f f i c u l t to provision, or dominance might increase with increasing brood 5 size i f early experience with nestmates p o s i t i v e l y a f f e c t s dominance. I tested whether dominance was p o s i t i v e l y related to the number of nestmates, as found by Boag and Alway (1980). I also examined several other factors that I thought might aff e c t dominance. In p a r t i c u l a r , I present the results of an experiment designed to determine i f prior residence in an area might affect yearling dominance. In this experiment, 24 birds were temporarily removed from the population to prevent them from acquiring s i t e attachments or establishing relations with f r e e - l i v i n g neighbors. METHODS Study Area and Study Population Mandarte Island i s a small (6 ha.), cigar-shaped island dominated by shrubs and grasses. It l i e s in the Haro S t r a i t approximately 20 km. N.N.E. of V i c t o r i a , B.C., Canada. A detailed description of the habitat c h a r a c t e r i s t i c s of the island, and the population biology of the song sparrows resident there, i s given by Tompa (1964). More recent accounts of the nesting phenology, heredity, and mating system of the sparrows can be found in Smith (1981), Smith and Dhondt (1980), and Smith et a l . (1982), who also provide d e t a i l s of the general methods employed on the island since 1974. In b r i e f , on Mandarte Island song sparrows t y p i c a l l y l i v e from one to four years, and are resident and monogamous. Males and females cooperate to raise 6 from one to four broods per year, but two is the modal number. The sexes are s l i g h t l y dimorphic in size, with males being larger, and both sexes mature at one year of age. I studied the population from the spring of 1982, when 25 females bred on the island, through the springs of 1983 and 1984, when 53 and 55 females bred. During t h i s period, a l l successful nests were found by observing females from a ladder or natural vantage point while they were nest-building, or when they returned to the nest after feeding themselves. Nestlings were f i t t e d with a numbered aluminum United States Fish and W i l d l i f e Service band, and from one to four colored p l a s t i c bands, at five or six days of age in most cases. This allowed me to follow individuals from their departure from the nest to their death, or u n t i l they dispersed from the island. Morphology I took three measurements from each nestling: weight to the nearest 0.5 gm; tarsus length to the nearest 0.1 mm, from the posterior end of the ti b i o t a r s u s to the anterior end of the lowest undivided scute; and length of the flattened wing to the nearest 0.5 mm. Age was known from the date of hatching, or was estimated by comparison with nestlings of known age. As an index of condition, I used the cube root of nestling weight divided by nestling tarsus length. Dividing weight by a skele t a l measure corrected for body size and gave a measure of 7 the 'fatness' of individuals (Slagsvold 1982, Boersma and Ryder 1984). I i d e n t i f i e d the number of young reaching independence by close observation of parents feeding fledged young, and by subsequent sightings or captures of young off their natal t e r r i t o r i e s . I sexed juveniles using a discriminant function which combined wing and tarsus length with body weight to successfully c l a s s i f y 114 of 116 (98.3 %) birds of known sex. The sex of these birds was known from their singing behavior, or by observing mated pairs. In order to pool data from nestlings of d i f f e r e n t age, birds of di f f e r e n t sex, and birds born in different years, I standardized variables using (x-xbar)/s.d. I captured over 85 percent of independent young produced each year using mist-nets. Upon capture, birds were weighed and measured as described above. Weights were corrected to 1200 h following Dhondt and Smith (1980). Only measurements from birds over 55 days of age were used to estimate size; by this time young have e s s e n t i a l l y completed growth (Smith and Zach 1979). Most birds were captured more than once. I therefore averaged measurements from successive captures of the same in d i v i d u a l . Estimating Dominance Dominance was estimated by observing aggressive interactions at feeders provisioned with m i l l e t , and calc u l a t i n g the proportion of t o t a l interactions won by each individual (Fretwell 1969, Baker and Fox 1978, Ketterson 1979, Rikkawa 8 1981, Watt et a l . 1984). This method gives a continuous measure of dominance that gives similar but less ambiguous results than h i e r a r c h i c a l ranking (Baker and Fox 1978). A l l observations were recorded between 15 July and 15 September of each year at one of four feeders placed in areas where juveniles were concentrated. The winner and loser of agonistic encounters were recorded only when the outcome was c e r t a i n , and when interactions were the result of a clear i n i t i a t i o n (gape, wings lowered, or head forward posture; see Kikkawa 1961, and Knapton 1973, p.76-77). In this way, I t r i e d to minimize observer s u b j e c t i v i t y . In 1982 and 1983, respectively, 1,808 and 3,670 agonistic encounters were recorded among f r e e - l i v i n g birds. In 1983, 642 encounters were also recorded among temporary captives. The accuracy of this dominance estimate depends upon the number of observations gathered per bird (Appendix). Therefore, when ca l c u l a t i n g single c o r r e l a t i o n s t a t i s t i c s each value in an X,Y pair was corrected for i t s mean value, and then multiplied by the number of observations that the dominance estimate was based upon. This weighting scheme is explained in d e t a i l in the Appendix. The angular transformation was not applied to the proportions of wins because they were approximately normally d i s t r i b u t e d . I used unweighted scores in a l l other s t a t i s t i c a l analyses, and only used those birds with over f i v e observations. While there i s reason to believe that a non-interacting class contains a large proportion of subordinates (Kikkawa 1981), I 9 d i d not f i n d a s i g n i f i c a n t c o r r e l a t i o n between the number of e n c o u n t e r s per i n d i v i d u a l and t h a t b i r d ' s dominance (r=0.146, n=175,p>0.05). Fo u r t e e n p e r c e n t (30 of 205) of y e a r l i n g s of known sex were observed i n fewer than f i v e i n t e r a c t i o n s , and these b i r d s were e x c l u d e d from most of the f o l l o w i n g a n a l y s e s . In no case d i d t h i s e x c l u s i o n s i g n i f i c a n t l y a l t e r the r e s u l t s of t h i s s t u d y . In a n a l y s e s of frequency d a t a , dominants are d e f i n e d as those b i r d s w i n n i n g g r e a t e r or e q u a l t o the median p r o p o r t i o n of t o t a l e n c o u n t e r s f o r t h e i r sex, and s u b o r d i n a t e s are those w i n n i n g fewer e n c o u n t e r s . An e x c e p t i o n t o t h i s scheme i s t a b l e 2, where b i r d s a r e d i v i d e d by the median s c o r e f o r the p o p u l a t i o n as a whole. I used G - t e s t s of independence, goodness of f i t , and h e t e r o g e n e i t y ( S o k a l and R o l h f 1969) t o t e s t f o r s i g n i f i c a n t a s s o c i a t i o n s between dominance c l a s s and v a r i o u s a t t r i b u t e s of i n d i v i d u a l s . Product-moment c o r r e l a t i o n s ( S o k a l and R o l h f 1969) were used t o t e s t f o r s i g n i f i c a n t r e l a t i o n s h i p s between c o n t i n u o u s c h a r a c t e r s . A l l s i g n i f i c a n c e l e v e l s r e p o r t e d i n t h i s t h e s i s are t w o - t a i l e d . Temporary Removal Experiment Between 6 and 18 June 1983, I removed 28 j u v e n i l e s from the p o p u l a t i o n , when a p p r o x i m a t e l y one h a l f of the j u v e n i l e p o p u l a t i o n had reached independence. Each b i r d was p l a c e d i n one of two a v i a r i e s measuring 2m x 3m x 2.3m i n h e i g h t . Four b i r d s d i e d w i t h i n 36 hours of c a p t u r e , presumably because they would not f e e d i n c a p t i v i t y . A l l o t h e r b i r d s m a i n t a i n e d t h e i r 10 weight well on a diet of l i v e mealworms (Tenebrio sp.), m i l l e t , and a mixture of crushed dog meal, chopped boiled egg, wheat germ, cod l i v e r o i l , crushed oyster s h e l l , and vitamin supplement. I provided water for drinking and bathing. One cage housed 10 females, and the second housed 12 males and 2 females. Dominance relations were allowed to equilibrate for one week after the la s t bird was added, before observations began. Dominance among caged birds was scored as described for f r e e - l i v i n g birds. The birds were released on 22 July 1983, when nearly a l l birds hatched that year had reached independence. RESULTS Dominance, Survival, and Settlement In 1982, 107 birds survived to independence, and in 1983, 128 birds did so. Thirty of these birds were not captured or sexed. Of these t h i r t y , only three v i s i t e d feeders regularly, and the remainder were seen only once or had disappeared from the island by early August. I therefore based my analyses upon the 205 birds of known sex. If dominant status i s advantageous to individuals, two l i k e l y benefits are increased survival and access to t e r r i t o r y . I considered birds to have 'survived' i f they remained on Mandarte Island on 30 A p r i l of the year following b i r t h . 11 'Settlers' are those birds known to have defended t e r r i t o r i e s for at least one month on or before 30 A p r i l . Some birds may have dispersed to survive or s e t t l e in other populations. Therefore, my estimates of survival and settlement are conservative. Two classes of male survivors existed on 30 A p r i l of each year; floaters (after S.M. Smith 1978) and s e t t l e r s . No female floa t e r s were known to exist beyond 15 A p r i l in either year of study-. Therefore, survival and settlement are equal for females. I used these d e f i n i t i o n s to test for relationships between dominance, settlement, and s u r v i v a l . Table 1 contains the numbers of individuals of different sex and s o c i a l status that survived as f l o a t e r s or s e t t l e r s , or had disappeared from Mandarte Island on 30 A p r i l of 1983 and 1984. Dominant males and dominant females survived and settled at s i g n i f i c a n t l y higher rates than subordinates of their respective sex in each year of the study. Thirty non-interacting birds survived (53.3%) and se t t l e d (36.7%) in proportions similar to subordinates (53.9% and 37.1%, respectively). The proportion of birds that survived in each year was dependent upon the sex considered. Female survival was s i g n i f i c a n t l y higher from 1982-1983 than from 1983-1984 (G=3.95, d.f.=1, p<0.05). Male sur v i v a l , however, was equal between years (G=0.025, d.f.=1, NS). This was because s i g n i f i c a n t l y more males survived as floa t e r s from 1983-1984 than from 1982-12 1983 (G=7.53, d.f.=1, p<O.Ol), whereas females that f a i l e d to s e t t l e in 1984 did not survive as f l o a t e r s . In 1983, males and females survived equally well (G=0.55, d.f.=l, NS), but males survived s i g n i f i c a n t l y better than females in 1984 (G=7.91, d.f.=1, p<0.005). The proportion of males and females that s e t t l e d in each year was similar (1983, G=0.60, d.f.=1, NS; 1984, G=0.96, d.f.=1, NS). But o v e r a l l , settlement was s i g n i f i c a n t l y higher in 1983 than in 1984 (G=6.63, d.f.=1, p<0.025). This was true for each sex (males, G=5.46, d.f.=1, p<0.025; females, G=3.95, d.f.=1, p<0.05). Therefore, while the number of birds that survived and that settled in each year was d i f f e r e n t , dominance consistently predicted the l i k e l i h o o d of individual survival and settlement in each year and within each sex. These results strongly suggest that dominance is a key determinant of which individuals of each sex survive and s e t t l e on Mandarte Island. Ef fects of Sex, Size, and Age on Dominance Among birds, dominance status has often been related to c h a r a c t e r i s t i c s of individuals such as sex, size, and age. In general, males dominate females, and dominance increases with size and age, though exceptions exist for each of these relationships (Chapter Two). In this section, I consider the effect of these three c h a r a c t e r i s t i c s on the dominance of yearling song sparrows. Table 1. The number of s e t t l e r s , f l o a t e r s , and birds absent from Mandarte Is. on A p r i l 30 of the year following hatch. Birds are divided by year, sex, and s o c i a l status. Gsur denotes the G - s t a t i s t i that results from comparing the status of a l l survivors to those absent, and Gset results from a comparison of s e t t l e r s to a l l others. In each case d.f.=l and sign i f i c a n c e i s indicated by: p<0.05 p<0.01**, p<0.001***. Year Group Class/Sex N Settled Floater Abs 1983 male dominant 20 16 3 1 subordinate 20 10 2 8 t o t a l 40 26 5 9 1983 female dominant 19 16 0 3 subordinate 19 10 0 9 t o t a l 38 26 0 12 1983 t o t a l dominant 39 32 3 4 subordinate 39 20 2 17 t o t a l 78 52 5 21 1984 male dominant 30 18 8 4 subordinate 31 7 13 11 t o t a l 61 25 21 15 1984 female dominant 17 11 0 6 subordinate 19 6 0 13 t o t a l 36 37 0 19 1984 t o t a l dominant 47 29 8 10 subordinate 50 13 13 24 t o t a l 97 42 21 34 Gsur=7.31** Gsur=4.33* Gset=4.33* Gsur=11.36*** Gsur=6.65** Gsur=3.86* 1 4 I found that dominance depended upon sex. In each year, about twice as many males were dominant, as compared to females (Table 2). In 1982, 38 females won, on average, 42.7 percent of their encounters, while 40 males won 57.5 percent of their encounters. In 1983, these figures were 41.4 percent for 36 females and 58.7 percent for 61 males. These data support the conclusions of others on song sparrows (Knapton 1976, Smith et a l . 1981). Table 2. The number of dominant and subordinate birds of each sex in each year of study. Significance levels are: p<0.025*, G-stat i st ic p<0.005** , p<0.00l * * * Year Sex N Dominant Subordinate 1 982 male 40 25 1 5 female 38 1 3 25 1 983 male 61 38 23 female 36 1 1 25 Total male 101 63 38 female 74 24 50 6.21 * 9.16** 15.44*** The theory of sexual selection assumes that large size is advantageous in aggressive competition to explain sexual size dimorphism in vertebrates (Darwin 1907, Selander 1972). I therefore expected to find positive correlations between dominance and the three morphological measures I recorded. Surprisingly, I instead found s i g n i f i c a n t negative correlations 1 5 between tarsus length and dominance within males and females in both years. Table 3 also shows that weight was p o s i t i v e l y related to dominance among females in 1982, and that wing length was p o s i t i v e l y related to dominance among females in 1983, but that no other correlation d i f f e r e d s i g n i f i c a n t l y from zero. However, morphology sometimes varied s i g n i f i c a n t l y with hatch date (Table 4), and, as I show below, dominance also depends upon hatch date. I therefore calculated p a r t i a l correlation c o e f f i c i e n t s for the effects of morphological characters on dominance, while c o n t r o l l i n g for the ef f e c t s of hatch date on morphology. Table 5 shows that dominance was independent of the morphological characters that I measured when the effect of hatch date on those characters was controlled s t a t i s t i c a l l y . S i m i l a r l y , among captive birds none of these three measures was .sig n i f i c a n t l y correlated with dominance. These results offer no support for the assumption that large size i s advantageous in agonistic encounters at feeders. Furthermore, they suggest that intrasexual competition can not explain sexual size dimorphism in this species. Several studies of birds have found that adults dominate juveniles (e.g. Rohwer et a l . 1981, Smith et a l . 1981), but the effect of age within a year-class has gone largely uninvestigated (but see Kikkawa 1981). I used hatch date as the independent variable to assess the eff e c t of age on dominance. The hatching period of young surviving to independence extended from 24 A p r i l to 4 July in 1982, and from 1 A p r i l to 8 July in T a b l e 3. C o r r e l a t i o n c o e f i c i e n t s f o r the r e l a t i o n s h i p between three measures of s i z e and dominance. S i g n i f i c a n c e l e v e l s are: p<0.05*, p<0.01**, p<0.001***. Year Sex N Weight Wing Length Tarsus Length male 36 0.280 0.285 -0.337* 1982 female 36 0.462** -0.044 -0.467** male 55 -0.173 0.255 -0.365** 1983 female 33 -0.29-5 0.469** -0.683*** T a b l e 4. C o r r e l a t i o n C o e f f i c i e n t s f o r the r e l a t i o n s h i p between three measures of s i z e and hatch date. S i g n i f i c a n c e l e v e l s a re: p<0.05*, p<0.01**. Year Sex N Weight Wing l e n g t h Tarsus l e n g t h male 36 -0.387* -0.163 -0.309 1982 female 36 -0.457** -0.088 0.054 male 55 0.041 0.323* -0.063 1983 female 33 -0.075 0.259 -0.286 T a b l e 5. C o r r e l a t i o n c o e f f i c i e n t s f o r the r e l a t i o n s h i p between three measures of s i z e and dominance, c o n t r o l l i n g f o r the e f f e c t s of hatch date on morphology. No c o r r e l a t i o n i s s i g n i f i c a n t l y d i f f e r e n t from z e r o . Year Sex N Weight Wing Length Tarsus Length male 36 -0.039 -0.130 -0.058 1982 female 36 -0.082 -0.145 -0.086 male 55 1983 female 33 0.202 0.118 0.181 -0.014 0.118 -0.115 19 1983. Thus, 70 and 98 days separated the e a r l i e s t and latest hatched young in 1982 and 1983 respectively. Table 6 shows that for both sexes, and in each year, dominance was strongly and negatively correlated with hatch date. Overall, hatch date accounted for 59% of the variation in dominance scores. Hatch date was even s i g n i f i c a n t l y related to dominance among captive birds (r=-.45, N=24, p=.026), where just 24 days separated the oldest bird from the youngest. Table 6. Correlation c o e f f i c i e n t s and c o e f f i c i e n t s of determination (r-squared) for the relationship between hatch date and dominance. A l l r-values are s i g n i f i c a n l t y d i f f e r e n t from zero (p<0.00l). Year Sex N Correlation r-squared 1982 male 40 -0.825 0.680 female 38 -0.560 0.314 1983 male 61 -0.760 0.578 female 36 -0.906 0.820 Total male 101 -0.755 0.570 female 74 -0.781 0.610 t o t a l 175 -0.768 0.590 Together, these results confirm that sex i s a key determinant of dominance within yearlings, but offer no support for the idea that morphology af f e c t s dominance. However, without detailed information on the effect of hatch date on d i f f e r e n t morphological characters, I might have concluded that dominance was p o s i t i v e l y , negatively, or not at a l l affected by size, depending upon the measure used. This finding underscores 20 the importance of knowledge of the prior h i s t o r i e s of individuals. Age was a powerful predictor of dominance, even among captives separated by l i t t l e more than three weeks. Hypotheses which could explain this relationship are discussed below (see Effect of Captivity on Dominance). Effects of Natal Care and Natal Environment on Dominance Dominance i s sometimes affected by factors e x t r i n s i c to the individual (e.g. Boag and Alway 1980, S a f r i e l 1981, Hrdy 1977). In this section, I consider the effects of several factors that could influence dominance. These factors include parental age, brood siz e , order within the brood, the number of nestmates of the same sex, and nestling size and condition. Table 7. The number of dominant and subordinate young produced by pairs of d i f f e r e n t age. Data from both years of study are combined. Class N Pair Type Year1ing Mixed Adult Dominant 88 22 30 36 Subordinate 87 20 29 38 To explore the effect of a parent's age and breeding experience on the dominance of young, I c l a s s i f i e d pairs into three groups based upon their previous experience at raising 21 young: inexperienced pairs ( f i r s t year birds), mixed pairs (one adult, one f i r s t year), and experienced (both adults). Table 7 shows that there was no tendency for more or less experienced pairs to have di f f e r e n t proportions of dominant and subordinate young. No study has investigated the effects of brood size on dominance, but there are reasons to expect that i t might have some influence (see Introduction). Table 8 shows the numbers of dominant and subordinate young reared in broods of one to four at the time of banding during 1982 and 1983. In each year, broods of three produced more dominant young than others, but t h i s difference was s i g n i f i c a n t only for both years combined (Table 8). Broods of one, two, and four produced similar proportions of dominant and subordinate young (G=0.14, d.f.=2, NS), but combined they produced fewer dominants than expected compared to a 50:50 r a t i o (G=4.10, d.f.=1, p<.05). Broods of three produced s i g n i f i c a n t l y more dominants than expected (G=4.44, d.f.=1, p<.05). Thus, there was no clear trend across brood siz e s . This result might be expected i f broods of three were more common early in the season, as early-hatched young were more dominant than those hatched later (table 6). To explore t h i s p o s s i b i l i t y , I classed a l l birds as 'early' or 'late' based upon the median hatch date for each year, and compared the frequency of early and late young from broods of three to those of a l l other broods combined. As expected, more early-hatched young came from broods of three than from broods 22 Table 8. The number of dominant and subordinate birds from broods of d i f f e r e n t size. Broods of one and two were pooled G - s t a t i s t i c to calculate the G-- s t a t i s t i c for 1982 and 1983. Year Class N Brood Size One Two Three Four 1982 Dominant 39 3 2 23 1 1 Subordinate 39 0 6 1 4 1 9 1983 Dominant 49 1 1 0 30 8 Subordinate 48 5 1 3 19 1 1 Total Dominant 88 4 1 2 52 1 9 Subordinate 87 5 19 33 1 0 of one, two, and four (G=7.05, d.f.=1 , P<0 .01 ) . 4.31, d.f.=2 NS 4.55, d.f.=2 NS 8.64, d.f.=3 p<0.05 This suggests that brood size influenced dominance through the intervening variable of age. S a f r i e l (1981) found that age determined dominance within broods of oysterchatchers (Heamatopus ostralegus). Dominance within a brood might therefore affect dominance after independence. I investigated t h i s p o s s i b i l i t y i n d i r e c t l y by exploring the effect of an individual's r e l a t i v e size within a brood on i t s eventual dominance status, assuming that r e l a t i v e size i s a measure of dominance within broods because i t r e f l e c t s the age of young. Eggs usually hatch asynchronously in the nests of song sparrows, p a r t i c u l a r l y in four-egg clutches, where the oldest young are t y p i c a l l y one day older than the youngest (unpubl. Observations). The r e s u l t i n g difference in the size 23 of young can be pronounced (up to 20% by weight). Table 9 l i s t s Table 9. The number of dominant and subordinate young from d i f f e r e n t sized broods in relation to their r e l a t i v e size within the brood. Data from both years of study are combined. Brood Size Class N Order Within Brood F i r s t S econd Third Fourth One Dominant 4 4 - -Subordinate 5 5 - - -Two Domi nant 1 2 9 3 -Subordinate 19 1 2 7 -Three Domi nant 53 19 13 21 -Subordinate 33 1 1 15 7 -Four Dominant 19 5 9 4 1 Subordinate 30 10 5 6 9 Total Dominant 88 37 25 25 1 Subordinate 87 38 27 1 3 9 the number of dominant and subordinate young from broods of one to four in r e l a t i o n to their r e l a t i v e size in the nest at banding. Surprisingly, I found no interaction between r e l a t i v e nestling size and later dominance class; once independent, the smallest young in nests were as l i k e l y to become dominant as the largest. Indeed, of the 64 largest young from broods of two, three, and four, 32 became dominant while 32 became subordinate. Si m i l a r l y , of the 49 smallest young in broods, 25 and 24 became dominant and subordinate. These results do not support the idea that dominance within a brood af f e c t s the dominance of independent birds. 24 Boag and Alway (1980) found that dominance could be increased in two species of Galliformes by experimentally increasing the number of nestmates of the same sex that an individual was raised with. To explore t h i s p o s s i b i l i t y in song sparrows, the sex of a l l young in a nest needed to be known. In many cases I did not have this information because some nestlings either f a i l e d to reach independence, or were not captured thereafter. I therefore pooled a l l available broods from 1982 and 1983 where the sexes of young were f u l l y known to increase my samples. Table 10 shows that females without s i s t e r s became dominant more often than those with one, or especially two s i s t e r s . However, males without brothers were dominant as often as those raised with one or two brothers. Table 10. The number of nestmates of the same sex in r e l a t i o n to dominance status. Data from both years of study are combined. Nestmates of Same Sex Sex Status N zero one two male dominant 45 21 1 7 7 subordinate 27 8 1 6 3 female dominant 32 1 5 16 1 subordinate 32 9 16 7 G-stat i st ic 3.04,d.f.=2 NS 6.29,d.f.=2 p<0.05 These results are inconsistent with those found by Boag and Alway (1980). I used p a r t i a l c o rrelation analysis to investigate the effects of nestling weight, tarsus and wing length, condition, 25 and hatching date on dominance. Only nestlings five or six days of age were used because samples of others were small. I pooled data for fiv e and six-day-olds after standardizing for their differences in size. Table 11 shows that in both 1982 and 1983 Table 11. Correlation c o e f i c i e n t s for three measures of nestling size and nestling condition in re l a t i o n to hatch date. Significance levels are: p<0.05*, p<0.0l**, p<0.00l***. Year N Weight Wing length Tarsus Length Condition 1982 76 0.434*** 0.540*** 0.377*** -0.234* 1983 175 0.358*** 0.206** 0.244** 0.035 nestling size increased s i g n i f i c a n t l y as the season progressed. However, condition declined s i g n i f i c a n t l y with advancing date in 1982, and was constant throughout 1983. Using data from 66 birds banded as f i v e or six-day-olds and later scored for dominance, I found no s i g n i f i c a n t correlations between dominance and nestling morphology after c o n t r o l l i n g for the effect of hatch date (weight,r=-0.074; wing,r=-0.012; tarsus,r=-0.084; condition,r=0.062; n=66 in a l l cases). Thus, while nestling morphology varied s i g n i f i c a n t l y throughout the 1982 and 1983 breeding seasons, these differences did not influence dominance. In t h i s section, I considered the e f f e c t s of several factors on the dominance status of yearling song sparrows. The previous breeding experience of parents was not associated with the eventual status of young. Nestling size r e l a t i v e to other 26 nestmates, or to those hatched at other times in the year, also did not predict dominance status. Although condition declined with hatch date in 1983, i t was not related to dominance among independent yearlings. Brood size was associated with dominance, but only through the effect of age. The dominance of females was negatively related to the number of s i s t e r s they were raised with, but there was no p a r a l l e l trend among males. Overall, the factors considered here had only minor effects on dominance compared to the strong effects of sex and age considered e a r l i e r . Effect of Captivity on Dominance The dominance of yearling song sparrows was c l o s e l y correlated with date of hatch (Table 6). Two hypotheses might explain that r e s u l t . F i r s t , dominance may develop with increasing experience in agonistic encounters. Individuals with more experience could be more s k i l l e d in f i g h t i n g and assessing the a b i l i t y of others. Second, dominance may be a function of an individual's f a m i l i a r i t y with a given s i t e (Brian 1949, Brown 1963). I tested the second hypothesis by temporarily holding 24 birds that hatched in the f i r s t half of the 1983 breeding period, u n t i l e s s e n t i a l l y a l l young hatched that year had reached independence. By this time, most of the early-hatched birds had r e s t r i c t e d d i s t r i b u t i o n s , and some males sang sub-song 27 within these l o c a l areas. Temporary captives were held in groups as described above (see Methods). Therefore, while captives were as old and experienced in agonistic ecounters as early-hatched free birds, they could not have become familiar with p a r t i c u l a r s i t e s or have established relations with free-l i v i n g birds. Three groups were thus created: early-hatched controls, early-hatched captives, and late-hatched controls. If s i t e dominance explains the negative c o r r e l a t i o n between age and dominance, I would expect captives to resemble late-hatched controls in dominance more clos e l y than they resemble early-hatched controls. To test this prediction, I compared the dominance of captives, estimated after release, to that of a group of control birds, chosen such that the mean hatch date of each group was approximately equal. Males and females were compared to controls separately to account for differences in dominance due to sex. Contrary to the prediction of the s i t e dominance hypothesis, the mean dominance of captive males (0.697, N=12) was very similar to that of the e a r l i e s t hatched free males (0.724, N=19, p>0.5, Mann-Whitney U-test). Captive females were also not s i g n i f i c a n t l y d i f f e r e n t from the e a r l i e s t hatched free females (0.512 vs. 0.451, means for 11 captive and 9 free females respectively; p>0.2, Mann-Whitney U-test). However, male and female captives were s i g n i f i c a n t l y more dominant than late-hatched free birds (p<0.01, Mann-Whitney U-tests). Furthermore, male and female captives survived at rates similar 28 to those of free birds (Table 12). A larger proportion of captive males and females settled than free birds, but this result only approaches significance when the data for both sexes are combined (G=3.63, d.f.=1, p>0.05). F a m i l i a r i t y with a par t i c u l a r s i t e was therefore not essential to dominant status for either males or females. These results are consistent with the hypothesis that dominance develops with experience in aggressive encounters- independently of the location of the in d i v i d u a l . However, dominance could develop independently of experience of any kind. I am currently conducting an experiment to determine whether experience in agonistic encounters is necessary to achieve dominant status. Table 12. The numbers of s e t t l e r s , f l o a t e r s , and birds absent from Mandarte Is. on 30 A p r i l 1984. Captive and free birds are divided by year, sex, and s o c i a l status. Group Class/Sex N Settled Floater Absent male 55 1 8 20 1 7 Control female 33 1 4 0 19 pooled 88 32 20 36 male 1 2 7 4 1 Captive female 1 2 7 0 7 pooled 24 1 4 4 6 29 DISCUSSION Physical attributes, amount of experience, and motivational state probably a l l contribute to success or f a i l u r e in agonistic encounters (Parker 1974, Brown 1975). When these factors are r e l a t i v e l y stable through time and variable among individuals, some individuals w i l l consistently win more encounters than others. Dominant status w i l l be especially advantageous when i t results in p r i o r i t y of access to resources c r u c i a l to survival and reproduction, and when i t does, the determinants of dominance w i l l be subject to selection. In the following discussion, I consider my method of estimating dominance, the determinants of dominance i d e n t i f i e d by t h i s study, and some evolutionary consequences of dominance behavior. Measures of Dominance In this study, dominance was estimated by observing agonistic encounters at feeders, but some birds did not v i s i t feeders. These few non-interactors survived and se t t l e d in proportions similar to subordinates (see Results). Kikkawa (1981) also believed that non-interacting silvereyes were usually subordinate. Dominance is t y p i c a l l y defined by an individual's a b i l i t y to gain access to resources, although dominance orders sometimes vary with the resource contested (Richards 1974, Syme 1974). My 30 measure of dominance estimated the p r i o r i t y of access to a food resource ( m i l l e t ) . I also determined which individuals gained access to t e r r i t o r y and which did not. This constitutes a second measure of dominance. I showed above that for males and females in each year of thi s study, these two measures of dominance were s i g n i f i c a n t l y correlated. This result supports the e a r l i e r findings of Glase (1973) for black-capped chickadees (Parus a t r i c a p i l l u s ) , of Knapton (1976) for song sparrows, and of Smith (1978) for rufous-collared sparrows (Zonotr ichia  capensis). These studies, however, involved many fewer individuals. Not only did dominance estimated at feeders successfully predict survival and settlement in thi s study, but i t also predicted which males survived as n o n - t e r r i t o r i a l floaters in 1984. There were few floaters in 1983 (Table 1). These results confirm that an empirical measure of success at feeders is a valuable tool for assessing the success of individuals within natural populations. Consequences of Dominance Previous studies have shown that dominant birds enjoy high survival rates compared to subordinates (Fretwell 1968, Glase 1973, Kikkawa 1981), though pronounced advantages were only observed by Baker et a l . (1981) when food was scarce. My results support this general view, but they also present an intriguing inconsistency. Males tended to dominate females in this study (Table 2), and I therefore expected males to survive 31 better than females. This expectation was supported in 1983-4, but males and females survived equally well in 1982-3 (see Results). One explanation for this could be that males only experience a s i g n i f i c a n t survival advantage over females when survival is generally low, since survival in 1983-4 was s i g n i f i c a n t l y lower than in 1984-3. However, when survival was equally high for each sex in 1982-3, dominants s t i l l survived s i g n i f i c a n t l y better than did subordinates within each sex (see Results). Thus, this explanation was not completely supported. An additional explanation could be that competition i s more intense within each sex than between the sexes. For example, survival, as measured here, might result from a) l i f e or death, or b) dispersal or non-dispersal. Dispersers could be • those birds which are not successful in competition for t e r r i t o r y (Gauthreaux 1978). Competition for t e r r i t o r y should be stronger within each sex than between them. Thus, even when survival was r e l a t i v e l y high, dominance could s t i l l have affected the number of birds that remained on the island through di s p e r s a l . Together, these explanations suggest that dominance influences dispersal within each sex in years even when survival is high, but that survival must be r e l a t i v e l y low before s i g n i f i c a n t between sex effects are observed. Further study is needed to test these explanations more thoroughly. The 32 important point here i s that, while dominance successfully predicted which birds survived in each year and sex, the more dominant sex (males) only survived better when recruitment was low. The relationship found between dominance and settlement was similar to the relationship between dominance and s u r v i v a l . These results support e a r l i e r work based on many fewer individuals (Odum 1942, Glase 1973, Smith 1976, 1978, Knapton 1976). The Determinants of Dominance I now consider the attributes of individuals that were associated with dominance. Sex i s the most frequently observed determinant of dominance (Brown 1975, Wilson 1975). Among birds, males t y p i c a l l y dominate females, though exceptions exist (Chapter Two). The results of this study further support t h i s general pattern. An additional point is that o v e r a l l , males survived s i g n i f i c a n t l y better than females (G=6.14, d.f.=1, p<0.0l). This result may explain the consistent skew in favor of males in the adult sex-ratio in this population (Smith et a l . 1 982) . Body size is generally thought to strongly influence dominance status (e.g. Searcy 1979a,b,d, Dawkins and Krebs 1978, Payne 1984). However, few authors consider the problems 33 of interpretation that arise because of the close correlations between size, sex, and age that are present in many species (Chapter Two). Most studies of i n t r a s p e c i f i c dominance in birds that have controlled for these factors have found no relationship between dominance and morphology (Chapter Two). This study provides another example of a species in which morphology does not have a strong influence on dominance. I used three measures of size, and I also knew the influence of hatch date on morphology and on dominance. Without this information, I might have concluded that dominance was p o s i t i v e l y , negatively, or not at a l l related to 'body size', depending upon the measure chosen. No other study of dominance in birds has studied the eff e c t of morphology on dominance in such d e t a i l . For this reason, the results of previous studies that have not considered relationships between age and size, and between age and dominance, should be c a r e f u l l y re-examined. Many authors assume that large size i s advantageous in agonistic encounters over mates and resources to explain sexual size dimorphism (e.g. Selander 1972, Searcy and Yasukawa 1983, Payne 1984). The results of thi s study offer no support for this assumption in a monogamous species. Similar studies should be carried out in polygamous species to test t h i s assumption further. The relationship between size and dominance i s given a more extensive consideration in Chapter Two. Dominance was strongly related to the age of yearlings for males and females in each year of this study. Overall, age 34 accounted for a s t r i k i n g 59% of the variation in dominance scores, and was the only s i g n i f i c a n t predictor of dominance among captive sparrows, where only 24 days separated the oldest bird from the youngest. This i s the most fascinating finding of this study. Among silvereyes, Kikkawa (1981) found a weak but not s i g n i f i c a n t tendency for older yearlings to dominate younger ones. However, Kikkawa's result has recently been confirmed after weighting dominance estimates by their errors (see Appendix; Kikkawa, personal communication). Glase (1973) used s k u l l o s s i f i c a t i o n to age yearling chickadees, and found a trend for earlier-hatched young to dominate those hatched l a t e r . Among great t i t s , late-hatched young disperse further (Dhondt and Huble 1968, Dhondt 1979), and survive less well (Perrins 1963, Dhondt 1979) than those hatched e a r l i e r . These observations support Dhondt and Huble's (1968) hypothesis that age-related dominance i s a driving force behind dispersal and survival among great t i t s . I suggest that these authors observed a widespread phenomenon among yearling birds that engage in frequent agonistic encounters, and that age-related dominance may be an important mechanism underlying juvenile dispersal in many t e r r i t o r i a l birds. I considered two hypotheses that could explain the cor r e l a t i o n between age and dominance. The experience hypothesis, proposed here, asserts that individuals acquire their f i g h t i n g s k i l l s through p a r t i c i p a t i o n in aggressive encounters, and that the le v e l of these s k i l l s i s the main 35 determinant of dominance. The s i t e dominance hypothesis (e.g. Brown 1969) states that f a m i l i a r i t y with an area increases the l i k e l i h o o d of winning encounters because l o c a l birds are more acquainted with strategic perches, cover, and food resources than are intruders. As an additional mechanism, Krebs (1983) proposed that t e r r i t o r y owners often win encounters because they are more l i k e l y to.escalate fights than are intruders. Owners might do this because they have invested energy establishing t e r r i t o r i a l boundaries with neighbors, and thus have more to lose than do intruders. I tested the s i t e dominance hypothesis by experimentally preventing older yearlings from gaining f a m i l i a r i t y with p a r t i c u l a r s i t e s or potential neighbors, while allowing them to gain experience in aggressive interactions. I released the captives when the youngest yearlings had become independent. Captives and late-hatched birds therefore had equal opportunity to develop s i t e f a m i l i a r i t y and relations with neigbors. The results of this experiment firmly rejected s i t e f a m i l i a r i t y as a prerequisite of dominant status among yearling song sparrows. Both male and female captives were as dominant at feeders after release as comparably-aged birds, and both groups of captives were s i g n i f i c a n t l y more dominant than birds hatched l a t e r . These results are consistent with the experience hypothesis. However, they are also consistent with hypotheses that explain dominance independently of experience of any p a r t i c u l a r kind. A test of the experience hypothesis is underway. 36 Of the several factors considered under the t i t l e 'Effects of Natal Care and Natal Environment', two were associated with dominant status: the number of nestmates of the same sex, and brood si z e . Brood size was a s i g n i f i c a n t predictor of dominance, but this was explained by the relationship between hatch date and dominance. Broods of three were more common e a r l i e r in the season, and thus produced more of the older yearlings. Boag and Alway (1980) found that among two species of Galliformes, dominance was p o s i t i v e l y correlated with the number of nestmates of the same sex in the natal brood. I found the opposite to be true in females, and found no effect in males. However, the differences in the dominance of females with varying numbers of s i s t e r s are small. Song sparrow young are a l t r i c i a l , but Galliformes have precocial young. Further data are needed to see i f these d i f f e r e n t results are general for birds with a l t r i c i a l and precocial young. Parental age had no e f f e c t on the proportion of dominants produced. S i m i l a r l y , neither the r e l a t i v e size of a young within a brood, nor the absolute size or condition of nestlings, was s i g n i f i c a n t l y associated with dominance after independence. Contrary to the pattern for two other small passerines (Perrins 1963, Howard 1980), the size of nestlings at a given age increased with hatch date in both years of this study. Perrins (1969) suggested that late-hatched great t i t s suffered high mortality after fledging, because they fledged in 37 poorer condition than young hatched e a r l i e r . Late-hatched song sparrows suffered poor survival compared to those hatched e a r l i e r in both years of this study, but they fledged in poorer condition in only one year. Garnett (1981) found that tarsus length in wild juvenile great t i t s was negatively related to date of hatch. He therefore proposed that the relationship observed by Perrins (1969) in the great t i t , was due to a dominance advantage gained by larger, early-hatched birds. To test this idea, he studied dominance among nine captive juvenile great t i t s and found that in the f i r s t week of study, age was s i g n i f i c a n t l y related to dominance. In the second week tarsus length was correlated with dominance, but age was not. Unfortunately, he did not account for his previous observation that hatch date and tarsus length were negatively related. Nevertheless, he suggested that size was the most important determinant of dominance, and that age influenced dominance only in the early development of young. My results show that the s t r i k i n g effect of age on dominance in the song sparrow per s i s t s well beyond the period postulated by Garnett (1981) for the great t i t , and that correlations between morphology and dominance may weaken when the relationships between date of hatch, morphology, and dominance are taken into account. I suggest that the relationship between hatch date and juvenile survival in the great t i t (Perrins 1969, Dhondt 1979), is at least partly due to an effect of hatch date on dominance similar to the one observed 38 in t h i s study. In summary, the c h a r a c t e r i s t i c s of individuals i d e n t i f i e d by this study that exhibited strong correlations with dominance were sex and hatch date. No other c h a r a c t e r i s t i c investigated was found to be strongly related to the dominance of yearling song sparrows. Dominance and Natural Select ion Dominant status was c l e a r l y advantageous to yearling song sparrows during the period of t h i s study. The determinants of dominance i d e n t i f i e d by th i s study suggest that selection for dominant young w i l l exert pressure on females to lay early in the season. Females that lay early may also raise more broods than those laying later (Smith 1981). Also, parents that raise many males may produce more recruit s than those raising many females, since males tended to survive better than females. However, selection for dominant young w i l l be opposed by countervailing selection pressures. For example, young hatched early are smaller at a given age than those hatched later (Table 11), they remain longer in the nest before fledging, and they are also more l i k e l y to die of exposure from unfavorable early season weather (Unpublished data). Furthermore, experimental evidence suggests that i t is d i f f i c u l t for females to lay early (Smith et a l . 1980), and these effects may balance those of low dominance among their o f fspring. 39 The present analyses only allow speculation about how these opposing selection pressures have affected the evolution of nesting phenology in the Mandarte song sparrow population, and further study is needed to determine their r e l a t i v e contributions. However, i f we are interested in an individual yearling, which controls neither i t s hatch date or i t s sex, my findings suggest that dominance status is largely the result of chance. Most pairs in t h i s study nested more than once per year, and they were therefore committed to raise both early and late-hatched young. Dominance has been shown to be heritable in several species of Galliformes (Craig et a l . 1965, Boag and Alway 1981, Boag 1982, Moss et a l . 1982). However, among the song sparrows studied here, P. Bets (personal communication) found no evidence for h e r i t a b i l i t y of dominance. Together, the above considerations suggest that selection for hatch date, and thus the dominance status of yearlings, w i l l result from a suite of balancing selection pressures. The dominance status achieved by a yearling sparrow probably depends largely on events beyond i t s control. Given this conclusion, i t would be interesting to know if individuals adopt d i f f e r e n t behavioral stratagies depending upon their status. In conclusion, dominant status as measured at feeders was a good predictor of survival and settlement among yearling song sparrows. This pattern was observed in two consecutive years, 40 even though population parameters d i f f e r e d markedly between years. Of the several phenotypic c h a r a c t e r i s t i c s investigated, only sex and age exercised strong effects on dominance. Twice as many males were dominant as were females, and age accounted for over 5 9 % of the t o t a l v a r i a t i o n in dominance among individuals summed over both years. Site f a m i l i a r i t y was not a prerequisite to achieving dominant status, and my results were consistent with the hypothesis that the amount of experience that an individual has had in agonistic encounters i s a key determinant of dominance. The dominance of yearling song sparrows is probably ultimately determined through the interaction of the selection pressures that affect laying date. 41 CHAPTER TWO: DOMINANCE, BODY SIZE, AND HABITAT DISTRIBUTION INTRODUCTION Darwin (1907, p.59) believed that sexual size dimorphism in birds " i s the result of the advantage gained by larger and stronger males over their r i v a l s during many generations", and this view persists (e.g. Searcy and Yasukawa 1983, Payne 1984). Even the smallest advantage to large individuals could act as a selective mechanism in competition for mates or resources, but some authors assume that the e f f e c t of size on dominance is large, and rely on i t to explain ecological processes such as the winter and breeding season d i s t r i b u t i o n s of birds (Fretwell 1969, 1972, Gauthreaux 1978, Dhondt et a l . 1979, Ketterson 1979, Lundberg et a l . 1981, Ulfstrand et a l . 1981). In support of this assumption, several studies have claimed that body size i s a key determinant of i n t r a s p e c i f i c dominance in birds (Fretwell 1968, Baker and Fox 1978, Ketterson 1979, Searcy 1979a, Garnett 1981). Though less often c i t e d , many studies of birds f a i l to support the idea that large size i s an advantage in agonistic encounters (Murchison et a l . 1935, Shoemaker 1939, C o l l i a s 1943, Tordoff 1954, Moore 1972, Glase 1973, Knapton 1973, Rohwer 1975, Schneider 1979, Kikkawa 1981, Price 1984, Watt et a l . 1984, D.A. Boag pers. com., t h i s study), or in obtaining mates ( L i l l 1974, Searcy 1979b, S t i l e s and Wolf 1979), breeding 42 t e r r i t o r i e s ( Knapton 1973, Yasukawa 1979, Hannon and Roland 1984, J.N.M. Smith and D. Schluter in prep.), or breeding positions within cooperative groups (Brown et a l . 1982). There are also exceptions of a d i f f e r e n t kind. For example, among wintering sparrows, smaller adult females often dominate larger juvenile males (Knapton 1973, Schneider 1979, Parsons and Baptista 1980, Rohwer et a l . 1981). F.R. Gelbach (personal communication) found that t e r r i t o r i a l male screech owls (Otus  asio) were s i g n i f i c a n t l y smaller than males without t e r r i t o r i e s . S i m i l a r l y , Jehl (1970) observed in two species of monogamous sandpipers, that small males were the f i r s t to attra c t mates. These findings a l l contradict the assumption that large individuals have an advantage in contests for resources, and they suggest that size does not determine dominance. Why, then, i s this assumption so popular ? One reason is that sexual selection, which assumes an advantage of large size in intrasexual competition (Darwin 1907), has so often provided an explanation for size differences between males and females (e.g. LeBoeuf and Peterson 1969, Hrdy 1977, Clutton-Brock et a l . 1982, Payne 1984). In addition, this assumption has much i n t u i t i v e appeal, and this may also explain i t s u n c r i t i c a l acceptance. In this paper, I offer an assessment of the assumption that large size is advantageous in agonistic ecounters, and is thereby a key determinant of i n t r a s p e c i f i c dominance in birds. 43 The f i r s t section examines some measures of body size, and i l l u s t r a t e s , with examples from the l i t e r a t u r e , how their use may be confounded by a dependence upon sex and age, or an independence from the mass of individuals. In the second section, I present data from a study of dominance among wild song sparrows (Melospiza melodia), and determine the slope of the relationship between the difference in the size of contestants and the probability of winning an agonistic encounter. Then, using a simple model, I estimate the slope of this relationship necessary to account for differences in the size of individuals between habitats observed by Lundberg et a l . (1981) and Ulfstrand et a l . (1981). MEASURES OF SIZE AND DETERMINANTS OF DOMINANCE Problems in measuring body size Body size i s a familiar term, but there i s no consensus on how to measure i t (see review by Clark 1979). Clark (1979) defined body size as the mass of the in d i v i d u a l , and I follow this d e f i n i t i o n here. Amadon (1943) recommended that the cube root of l i v e weight be used as the standard for i n t e r s p e c i f i c comparisons of body si z e . However, using weights for studies of dominance within species creates two problems. Most obvious, dominance is t y p i c a l l y determined by observing agonistic encounters at a food resource; individuals which are consistently successful are dominant to those that are less 44 successful (e.g. Brown 1963, Searcy 1979a). We might therefore expect dominants to outweigh subordinates. . Weight could also affect dominance; heavier birds might have an advantage, or l i g h t , hungry birds might fight harder than satiated ones. The result i s that correlations between l i v e weight and dominance can be variously interpreted. A second problem i s that weights change d a i l y and seasonally, especially in small birds (Clark 1979). As a result, researchers have sought indices of body size that are less variable and are not affected by dominance. Wing length, and to a lesser extent tarsus, b i l l , and keel length, have become popular indices of body size (e.g. Slagsvold 1982, Hannon and Roland 1984, Payne 1984). However, linear dimensions often f a i l to correlate well with either fat-free weight or l i v e weight (e.g. Power 1969, Bailey 1979, Halse and Skead 1983). Among male juncos (Junco hyemalis) , Helms et a l . (1967) found that wing length did not predict fat-free dry weight. The c o r r e l a t i o n was weak for resident females, and was s i g n i f i c a n t only for migrant females. They concluded that "extreme care must be used by authors using wing length as an indicator of mass" (p.571, 1967). Many authors report allometric relationships between wing length and body weight among populations d i s t r i b u t e d along l a t i t u d i n a l c l i n e s (e.g. Power 1969, James 1970 and references therein). Searcy (I979d) and Yasukawa (1979) reasoned that a positive relationship between mean wing length and mean body 45 weight between populations of red-winged blackbirds (Agelaius  phoenicius) (Power 1969), j u s t i f i e d their assumptions that the relationship also held within populations. In fact, Power (1969) found that within 10 male and 7 female populations of red-wings, there were 7 negative and 10 positive correlations between wing length and the cube root of body weight, just one being s t a t i s t i c a l l y s i g n i f i c a n t . Using wing length as an index of body size i s further complicated by the fact that i t increases with age (Table 13), and i s subject to seasonal variation due to wear (e.g. Brown et a l . 1983). In contrast, tarsus length and b i l l dimensions do not appear to change with age, at least in some passerines (P.T. Boag 1983 and references therein, Alatalo et a l . 1984), and tarsus length does not vary due to wear. The value of these or other characters as indices of body size w i l l depend upon their a b i l i t y to predict weight and other linear dimensions accurately. When linear dimensions are not strongly correlated, they may indicate shape more than si z e . P.T. Boag (1983) used p r i n c i p a l component analysis to id e n t i f y variables that described overa l l increases in the size of Galapagos finches (Geospiza spp.) better than weight or linear dimensions alone. This i s a promising technique to estimate the body size of birds when individual morphological characters are not closely correlated. 46 Table 13. Some species in which wing length increases with age. Study Mueller et a l . 1976 Mueller et a l . 1979 Mueller et a l . 1981 Newton et a l . 1983 L. Rotterman and C. Monet personal communication Gatti 1983 Slagsvold 1980 A. Lundberg pers. comm. Blake 1962, Ketterson 1979 Moore 1972 Wishart 1981 J.N.M. Smith Ulfstrand et Selander and Greenberg et N.B. Davies Searcy 1979c Rohwer et a l . Kikkawa 1981 in prep, a l . 1981 Johnston a l . 1972 pers. comm 1967 1981 Spec ies Accipiter g e n t i l l i s A c cipiter s t r i a t u s Accipiter cooperii Accipiter nisus Agelaius phoenicius Anas platyrhynchos Corvus corone cornix Ficedula hypoleuca Junco hyemalis Junco phaeonotus Mareca americana Melospiza melodia Parus major Passer domesticus Phasianus colchicus Prunella modularis Xanthocephalus xanthocephalus Zonotrichia querula Zosterops l a t e r a l i s Sex, Age, and Size Most populations include individuals of dif f e r e n t sex and age. The majority of bird species exhibit sexual size dimorphism (Amadon 1959), and males are usually the larger sex. In turn, dominance depends upon sex, and males t y p i c a l l y dominate females ( Shoemaker 1939, C o l l i a s 1943, Brian 1949, C o l l i a s and Taber 1951, Tordoff 1954, Marler 1955, Hinde 1956, Kikkawa 1961, Dixon 1965, Glase 1973, Knapton and Krebs 1974, Smith 1976, 1984, Baker and Fox 1978, Schneider 1979, Rohwer et a l . 1981, Smith et a l . 1981, Brawn and Sampson 1983, Peters and Grubb 1983 and references therein, Watt et a l . 1984, but 47 see Thompson 1960 and Coutee 1967). Wing length, a popular index of body size, increases with age in many species (Table 13; but see Ewald and Rohwer 1980), and age i s a primary determinant of dominance (Brown 1963, Moore 1972, Knapton 1973, 1976, Smith 1978, 1984, Searcy 1979a, Schneider 1979, Ewald and Rohwer 1980, Rohwer et a l . 1981, Kikkawa 1981, Smith et a l . 1980, De Vos 1983, Watt et a l . 1984, Chapter One). We should therefore expect positive correlations between wing length and dominance when the ef f e c t s of sex or age are not controlled for. Two main points were made in this section. F i r s t , body size must be determined accurately. If a single morphological character predicts dominance, but is not correlated with other measures of size, i t i s incorrect to argue that 'body size' determines dominance. A focus on body size should not obscure the importance of single characters either. Morphological characters may have d i f f e r e n t h e r i t a b i l i t i e s , or may be subject to d i f f e r e n t selection pressures (P.T. Boag 1983). Their d i f f e r e n t effects on-dominance should be as interesting as any shared e f f e c t . Second, characters that influence dominance, and covary with body size, must be i d e n t i f i e d and controlled for i f we wish to test the hypothesis that size determines dominance. Both age and sex may aff e c t size, but may independently influence dominance. These considerations provide the backround for a review of studies that have shown posit i v e correlations between size and dominance in birds. 48 Examples From the Literature The following examples i l l u s t r a t e the problems mentioned above. Each study claims to have found positive correlations between 'body size' and dominance. Fretwell (1969) found that wing length could predict dominance in the junco, and this result i s repeatedly c i t e d as evidence that body size is a key determinant of dominance (e.g. Fretwell 1972, Wilson 1975, Ketterson 1979, Morse 1980). However, Fretwell included birds of both sex in his analysis, and in juncos, males are larger than females (Balph 1975) and dominant to them with few exceptions (Moore 1972, Baker and Fox 1978, Balph 1975, Ketterson 1979). I partitioned his data (figure 7, p.17, 1969) into dominant and subordinate individuals (greater or less than 50% wins) of each sex using the c r i t e r i a given in Balph (1975) for sexing juncos by wing length. Because the sexes overlap in wing length (77-79mm), 19 birds could not be r e l i a b l y sexed, and I excluded these birds from my analysis. Table 14 shows that dominance was dependent upon sex, and th i s result i s independent of the sexing c r i t e r i a ; the data may be divided equally among the sexes over the zone of overlap without a l t e r i n g the r e s u l t . In addition, there was no correlation between wing length and dominance within males or females (r=-.02, n=l5, NS and r=-.l8, n=6, NS, res p e c t i v e l y ) . This shows that Fretwell's finding resulted from comparing males with females. Furthermore, wing length does not r e l i a b l y predict the fat-free dry weight of juncos (Helms et a l . 1967), and increases with age (Table 14). 49 Table 14. The effect of sex in Fretwell's figure 7 (see t e x t ) . It is l i k e l y that dominance depends on sex (Fisher's Exact Probability Test, two-tailed p=.01 1 , n = 21). Male Female Dominant 1 3 Subordinate 2 5 Other studies suffer similar problems. Baker and Fox (1978) found no relationship between dominance and wing length among juncos in the f i e l d , but did find a posit i v e correlation among captive birds. Discounting the f i e l d result, they conclude that "to predict dominance the single best source of information is wing length" (p.708, 1978). While th i s may be true, their procedure suffers from the same setbacks as Fretwell's. Ketterson (1979) recognized the problem of sex and age differences in size and dominance among juncos. She found a positive effect of wing length on dominance even when corrected for sex and age. These studies show that wing length can predict dominance in juncos. However, because wing length in juncos i s correlated with age and sex, and i s not correlated with fat-fr e e dry weight, speculation about the influence of body size on dominance i s risky at best. In the yellow-eyed junco (Junco phaeonotus), Moore (1972) found no correlation 50 between wing length, or t a i l length, and dominance among males or females in either of two years. Searcy (1979a) controlled for sex differences in size and dominance among captive red-winged blackbirds by studying only males. He further i d e n t i f i e d birds as either yearlings or adults (over one year). Wing length was p o s i t i v e l y correlated with dominance among adults, but not among juveniles. There are two problems with the interpretation of these r e s u l t s . F i r s t , wing length in male red-wings increases s i g n i f i c a n t l y , each year, up to six years of age (L. Rotterman and C. Monet personal communication). However, the age of the adults in Searcy's study and the effect of age past one year on dominance are unknown (Searcy personal communication). Maturation is delayed in red-wings, thus the effect of age past one year may be large. Second, wing length varies independently of body weight in red-wings (Power 1969), and may not indicate body si z e . Garnett (1981) studied dominance among captive great t i t s , and his conclusions, which depended on which week's data were considered, were discussed in Chapter One. To summarize this section, sex and age are key determinants of i n t r a s p e c i f i c dominance in birds, and these variables are often t i g h t l y linked to body s i z e . Authors have in general not accounted for these confounding factors when choosing indices of body si z e . Each study that has claimed support for the hypothesis that body size determines dominance in birds suffers 51 drawbacks which seriously l i m i t i t s interpretation. We c l e a r l y to need account for these factors when studying the influence of morphology on i n t r a s p e c i f i c dominance in birds. HABITAT DISTRIBUTION AND DOMINANCE IN THE SONG SPARROW In t h i s section, I present data from a study of dominance among wild song sparrows to determine the relationship between three morphological measures and the prob a b i l i t y of winning an agonistic encounter. Lundberg et a l . (1981) and Ulfstrand et a l . (1981) observed that male pied flycatchers and great t i t s , respectively, were larger in high quality breeding habitat than in lower quality areas. They attributed their observations to the e f f e c t of size related dominance in competition for breeding t e r r i t o r y . I used a simple model to estimate the slope of the rela t i o n s h i p between the difference in contestant size and the prob a b i l i t y of winning an encounter that would be necessary to account for the differences that they observed. By comparing the empirically derived slopes for song sparrows to those required to segregate birds into separate habitats on the basis of size, I provide a test of the hypothesis that size-related dominance i s a mechanism behind habitat segregation in birds. The Advantage of Size in WiId Song Sparrows In Chapter One, I discussed the determinants of dominance in the song sparrow. In thi s section, I re-examine the data on 52 the morphology of interacting sparrows to determine how the difference in the size of interactors affected the outcome of individual ecounters. I previously showed that larger and smaller birds in each sex won approximately the same proportion of encounters, but this analysis did not consider the r e l a t i v e size of competitors. The main purpose of this analysis was to provide a comparison for the results from a model presented below. The general methods employed here follow those described in Chapter One. For this analysis, I only included encounters between birds of the same sex in order to control for differences in size and in dominance between males and females (Chapter One), and only yearlings were considered to prevent the introduction of any bias in favor of larger, more dominant adults (Knapton 1976, Smith et a l . 1981). The t o t a l sample was further reduced because many encounters included birds for which I had incomplete morphological information. Wing and tarsus length and body weight are independent measures of size in yearling song sparrows (wing vs. weight, r=.116, NS; wing vs. tarsus, r=.048, NS; tarsus vs. weight, r=.293, p<0.0l; N=116). I used each of them to derive the r e l a t i o n s h i p between the difference in contestant size and the p r o b a b i l i t y of winning an agonistic encounter. I estimated the slope of each relationship using a least-53 squares multiple regression model (Sokal and Rohlf 1969). A value of one or zero was assigned to the outcome of encounters (y-coordinate) and I used the size difference of contestants as the x-coordinate. I also included the difference in the age of contestants as an x-variable, because I have previously shown that age is a powerful correlate of dominance, and is in turn s i g n i f i c a n t l y correlated with some morphological characters (Chapter One). By t h i s method, each encounter i s counted twice; as a win and as a loss. For tests of significance the degrees of freedom are therefore equal to half the sample si z e . Results Table 15 gives' the p a r t i a l regression s t a t i s t i c s for each character within each sex for both years. As expected from a similar analysis presented in Chapter One, the effect of the di f f e r e n t morphological characters was not the same between the sexes or the years of study. In 1982, only age played a s i g n i f i c a n t role in the outcome of encounters among males, and the trend for a l l morphological measures was negative. For females in 1982, both hatch date and tarsus length were highly s i g n i f i c a n t l y related with the probability of winning encounters. However, tarsus length had a negative effect on winning encounters, rather than a positive effect as would be predicted given the assumption that large size i s advantageous in agonistic encounters. However in 1983, wing length was s i g n i f i c a n t l y p o s i t i v e l y related to winning in males, and 54 contrary to results for 1982, no morphological variable exhibited negative e f f e c t s . Hatch date was highly s i g n i f i c a n t l y related to winning. For females in 1983, only hatch date was s i g n i f i c a n t l y related to the probability of winning encounters. Overall, for both males and females in each year, hatch date was by far the most powerful predictor of the outcome of encounters. These results demonstrate two important points: l) morphology can be p o s i t i v e l y or negatively correlated with the probability of winning agonistic encounters; 2) correlations between morphology and winning encounters may be s i g n i f i c a n t l y d i f f e r e n t between years and between males and females. The significance levels of these regressions rest on the assumption that interactions represent independent observations. I used each interaction because the outcome of encounters between two birds was not always the same. I therefore had no objective way by which to exclude individual interactions. The significance levels should therefore be viewed with some caut ion. I also determined the r e l a t i v e importance of size and sex to the outcome of encounters between males and females by comparing the size of winners and losers when differences in age were ten days or less. Males and females overlap very l i t t l e in wing length, so there were no cases where females were larger than males and nearly equal to them in age. Males and females 55 Table 15. Results of multiple regression of three morphological characters and hatch date on the probability of winning agonistic encounters. Degrees of freedom and standard errors are adjusted to one half the sample size (see text). Significance levels are: P<0.01**, p<0.00l***. P a r t i a l Regression Year Sex/variable C o e f f i c i e n t S.E. 1982 male n=760 weight -0.017 0.022 wing -0.005 0.021 tarsus 0.043 0.062 hatch date -0.014*** 0.002 1982 female n=490 weight -0.015 0.026 wing -0.015 0.026 tarsus -0.137** 0.062 hatch date -0.007*** 0.002 1983 male weight wing tarsus hatch date n=1930 0.014 0.030** 0.017 •0.008*** 0.014 0.011 0.025 0.0006 1983 female weight wing tarsus hatch date n=31 4 0.047 0.023 0.072 -0.006*** 0.041 0.043 0.079 0.002 often overlap in tarsus, however, and I therefore asked the question of how often females defeated males when they were of similar age. For both year's data combined, females won 132 of 880 encounters when smaller than the male contestant, 2 of 14 when equal in size, and 11 of 63 when larger than the male. Thus, the frequency with which females defeated males was independent of size as measured by tarsus length (G=0.89, d.f.=2, NS). Similar results were obtained using weight. 56 Dominance and Habitat D i s t r i b u t i o n Lundberg et a l . (1981) and Ulfstrand et a l . (1981) observed that male pied flycatchers and great t i t s , respectively, were larger in high quality plots of deciduous forest than in lower quality pine forest p l o t s . They interpreted these differences as the result of a dominance advantage to large males in agonistic encounters over t e r r i t o r y . Pied flycatchers and great t i t s are similar to song sparrows in mass and linear dimensions, and a l l three species share some important aspects of their natural history. In each species, males compete for t e r r i t o r i e s , and song sparrows and great t i t s form dominance hierarchies in the non-breeding season (Knapton and Krebs 1974, Brian 1949, Hinde 1952). We might therefore expect these species to be similar in the effect of morphology on dominance. A simple linear model may be used to determine how large the advantage of size must be to account for the size differences observed in pied flycatchers and great t i t s between habitats. The model assumes a large population of size N, in which the outcome of randomly occurring agonistic encounters depends upon the size of contestants. The winners of encounters go to habitat A, while the losers go to habitat B. The p r o b a b i l i t y of going to habitat A i s given by; P ± = a(X ±-X) +.0.5 (1) 57 where X is the mean size of males in the o r i g i n a l population, a is the slope of the relationship between size differences and the probability of winning an encounter, and x is the size of individual i . The mean size of birds in habitat A i s ; I X.P. x = 1 1 XA E P. . (2) l Which s i m p l i f i e s to; _ _ 2 X = X + 2aa . (3) where sigma-squared i s the variance with respect to size of the i n i t i a l population. Therefore, the mean size of birds in habitat A, the good habitat, depends upon the i n i t i a l population mean and variance, and the ef f e c t of size on winning (slope). Solving for a , a = 2a 2 . (4) With th i s model, we can estimate a for pied flycatchers 58 using the information given in Lundberg et a l . (1980, table 4). The mean size of males in deciduous and coniferous plots, is 50.16 mm2-gm and 48.95 mm2-gm, respectively. (units derive from a discriminant score based on wing and tarsus length, and weight). Since this model depends on the actual data, the observed mean size of birds in each habitat is assumed to equal the parametric mean, and the mean size of the i n i t i a l population is taken as the unweighted average of the two observed means. The variance of the i n i t i a l population i s estimated as the sums of squares within habitats plus the sums of squares between group means, divided by the t o t a l sample (Sokal and Rohlf 1969). Using these figures, a i s 0.24. For great t i t s , using the same procedure, and with figures from Ulfstrand et a l . (table 1, 1980), a is 0.33. These figures estimate the size advantage needed to produce the mean size differences between habitats observed by Lundberg et a l . (1980) and by Ulfstrand et a l . (1980). The estimates rely on two assumptions about the character of agonistic encounters: f i r s t , that encounters are random with respect to size, and second, that the i n i t i a l population i s large. Lundberg et a l . (1980) and Alatalo et a l . (1984) provide information which suggests that for pied flycatchers, each assumption • i s incorrect. Lundberg et a l . show that the deciduous plots are occupied e a r l i e r than are the coniferous pl o t s . Alatalo et a l . found that in the same population, large 59 males arr i v e e a r l i e r than smaller males. Therefore, the population of competing males i s smaller early in the season. Furthermore, the mean size of competing birds w i l l change as the season progresses, since large birds are s e t t l i n g as smaller birds a r r i v e . Competitors w i l l also be more a l i k e in size than expected i f birds met at random. A small population w i l l make chance differences in the size of birds between habitats a more common occurrence. Non-random contests between birds of similar size w i l l reduce the realized variance of the competitor population. Thus, the slope calculated above w i l l be too small (see equation 4). P a r a l l e l considerations apply to great t i t s , but I can not evaluate my assumptions for that species. Given that these slopes are r e a l i s t i c , or are under-estimated for pied flycatchers, they represent much stronger effects of size on the outcome of encounters than I found among song sparrows. In song sparrows, the largest p o s i t i v e slope observed was 0.031 (Table 15), an order of magnitude less than that calculated for great t i t s . There were no consistent effects of size on the outcome of encounters among song sparrows, and for birds in 1982, these effects were negative (Table 15). These results suggest that some other factor than dominance led to the differences in the size of males between habitats observed by Lundberg et a l . (1980) and Ulfstrand et a l . (1980). For pied flycatchers, the r e l a t i o n s h i p between 60 a r r i v a l time and body size observed by Alatalo et a l . (1984) could be s u f f i c i e n t to explain the difference between habitats. DISCUSSION Four points are made in the preceding sections. F i r s t , sex and age are t i g h t l y linked to body size in many species, and they may independently influence dominance. Ketterson (1979) found that wing length was a better predictor of dominance in juncos than was sex. My results were opposite to this finding. Female song sparrows won no more encounters when they were larger than males than when they were smaller. These opposite results may be due to the fact that Ketterson's conclusion was based on the results of a stepwise multiple regression in which sex was the la s t of four variables entered, and wing length was the f i r s t . She did not control for the strong correlation that exists between wing length and sex in juncos (Balph 1975), and thus much of the effect of sex on dominance would have appeared to have been 'explained' by wing length. We know that sex and age are key determinants of dominance, and that size covaries with these factors. If size and dominance are linked via an extra 'X' chromosome in birds, size is important, but not per se. Experimental evidence suggests that males may dominate females largely because of a hormonal influence on behavior. Male birds t y p i c a l l y have higher c i r c u l a t i n g levels of testosterone than females (Wingfield and 61 Farner 1978), and testosterone is known to d i r e c t l y influence aggressive behavior and the dominance status of birds (Rohwer and Rohwer 1978, Moss et a l . 1979, Searcy and Wingfield 1980, Watson and Parr 1981, Moore 1984). Second, studies that have found effects of size on dominance have either f a i l e d to control for factors known to affec t both size and dominance, or have used morphological measures of size, such as wing length, that are unrelated to the mass of individuals within populations. Thus, no strong empirical evidence exists to support the notion that large size is advantageous in agonistic encounters in birds. Third, t h i s conclusion is supported by the study of dominance among song sparrows presented here, in which age, sex, and size could a l l be accurately measured. In this study, I found both pos i t i v e and negative correlations between morphology and the pr o b a b i l i t y of winning agonistic encounters. No s i g n i f i c a n t e ffect of morphology on the outcome of encounters was consistent across years, or even among the sexes. These results show that large size confers no consistent advantage in agonistic encounters in the song sparrow, and that large size may sometimes be a disadvantage. Furthermore, they show that studies over a single year, or the observation of only one sex, may lead to spurious conclusions about the effect of size on dominance. 62 F i n a l l y , ecological processes that have been explained in terms of size-related dominance, such as the d i s t r i b u t i o n of t e r r i t o r i a l male birds observed by Lundberg et a l . (1980) and Ulfstrand et a l . (1980), require a much stronger effect of size on dominance than was found in the song sparrow. While t h i s does not exclude the p o s s i b i l i t y that size affects dominance more strongly in other species, i t does suggest that other processes may explain the observations. If dominance were unaffected by size, what factors could account for the habitat d i s t r i b u t i o n of male great t i t s and pied flycatchers? There are at least two p o s s i b i l i t i e s : 1) birds in better habitats may be older, and therefore larger (Table 13); 2) larger birds may be p h y s i o l o g i c a l l y more able to withstand poor weather that might be associated with early a r r i v a l . If the proportion of older, more experienced males was higher in deciduous plots, than in poorer coniferous plots, t h i s could explain the difference. However, for both species, males in deciduous plots also had s l i g h t l y longer b i l l s than those in coniferous p l o t s . In great t i t s at least, b i l l length does depend upon age (Ulfstrand et a l . 1981). Thus, some real size differences, uncorrelated with age, may e x i s t . A second p o s s i b i l i t y was suggested by work on pied flycatchers by Alatalo et a l . (1984). As discussed above, they found that tarsus length was negatively correlated with the a r r i v a l time of t e r r i t o r i a l males in spring. Lundberg et a l . (1981) found that male pied flycatchers occupied deciduous plots before coniferous 63 plo t s . This being so, we should expect birds in deciduous plots to be larger on average. Large males might survive better during periods of adverse weather than smaller birds (e.g. Ketterson and King 1975, Piersma 1984). Therefore, large males might arr i v e early to aquire superior t e r r i t o r i e s in the re l a t i v e absence of competitors. For smaller males, the benefits of early a r r i v a l may not outweigh the associated r i s k . In any case, t h i s difference in a r r i v a l time suggests that the observed habitat d i s t r i b u t i o n did not result from size-related dominance. If i t did, two questions need to be asked. F i r s t , where and when do competitive interactions occur? Second, why should the eff e c t of size be strong in these species, but weak in others of similar size and l i f e history (e.g. Moore 1972, Glase 1973, Schneider 1979, Kikkawa 1981)? Why does body size have l i t t l e influence on dominance in some birds? F i r s t , the character of aggressive encounters changes from species to species, and this may af f e c t the influence of large s i z e . Over 90% of the agonistic encounters observed between song sparrows in thi s study involved no bodily contact. When contact did occur, birds flew up against each other and used their feet, wings, and beaks in a b r i e f , acrobatic f l u r r y . I observed tumbling fights between birds on the ground only two or three times in over 5,000 encounters. Size could be more important in species that engage in shoving matches, where a g i l i t y , speed and fightin g s k i l l may y i e l d to bulk (e.g. Chum salmon, Schroder 1981; elephant seals, LeBoeuf 64 and Peterson 1969; red deer, Appleby 1982, Suttie 1983). Second, the narrow size range within most bird species may preclude large asymmetries in dominance based on siz e . Song sparrows interacted randomly with respect to size in thi s study (Arcese unpublished data). Thus for wing length, 95% of of a l l sex-specific agonistic ecounters involved birds that d i f f e r e d by less than five percent in siz e . In species other than birds, there is much evidence that demonstrates that differences in size of ten percent were large enough to s i g n i f i c a n t l y favor larger contestants in agonistic encounters (male chum salmon, Schroder 1981; mantis shrimp, Caldwell and Dingle 1979; the anemone Actina equina, Brace and Pavey 1978) . Intrasexual competition is frequently invoked to explain sexual size dimorphism in birds (e.g. Wiley 1973, Searcy I979d), es p e c i a l l y for polygamous species where variation in individual reproductive success may be much greater than in monogamous species (e.g. Searcy and Yasukawa 1983). This explanation i s strongly supported by correlations within taxonomic families that demonstrate that the degree of dimorphism between males and females is much larger in polygamous species than in monogamous species (e.g. Selander 1972, Payne 1984). Within species, however, I have shown that the assumption that large size confers an advantage in agonistic encounters is generally unsupported in studies of birds. Yet, intrasexual competition makes thi s assumption to explain sexual 65 size dimorphism. How can we resolve this paradox? I suggest three p o s s i b i l i t i e s . F i r s t , intrasexual competition should be less important in monogamous species, since the variance in the reproductive success of individuals is lower than in polygamous species. The two most detailed studies of the effect of morphology on dominance in birds are Kikkawa's study of silvereyes (1981) and this study. Neither study showed an advantage of large size, and both were conducted on monogamous species. The results of these studies strongly suggest that sexual size dimorphism in these species i s not the result of intrasexual competition. Price (1984a,b) showed that sexual selection favored large males in Geospiza f o r t i s , a monogamous Galapagos finch, but that this resulted from female choice of large males, rather than through male-male competition. He found no evidence for size-related dominance in this species (Price 1984a). Price (1984a,b) and Downhower (1978) each found that selection favored small females in Galapagos finches, because small females produce eggs more e f f i c i e n t l y and thus begin laying more quickly than larger ones. These studies, and others reviewed by Payne (1984), show that sexual size dimorphism in monogamous birds may be explained without invoking size-related dominance. Second, the studies of monogamous birds may not accurately model the relationship between size and dominance that might 66 exist in polygamous species. If polygamous species were morphologically more variable than the monogamous species studied thus far, larger asymmetries in the size of contestants could occur more often. This might enhance the advantage of large size in agonistic encounters. Or, the character of aggressive encounters may d i f f e r between monogamous and polygamous species (see above). Third, an advantage of large size in agonistic encounters may be r e a l , but be s l i g h t and d i f f i c u l t to measure. However, even the s l i g h t e s t advantage to large individuals could be selected for, and thus cause sexual size dimorphism through intrasexual competition. This argument requires only evidence from comparative studies (e.g. Payne 1984) to support or or reject the assumption that large size i s advantageous in agonistic encounters. 67 GENERAL SUMMARY Dominance behavior is a widespread phenomenon both within and among species, and s o c i a l dominance in animals has intrigued b i o l o g i s t s for decades (e.g. Wilson 1975:279-297). Yet there is no consensus, and indeed there i s l i t t l e information, on the q u a l i t i e s of individuals that determine dominance • status. This study was undertaken to document the correlates of dominance status and consequences of dominance for the individual within one species. I f i r s t tested whether dominance could influence recruitment. If i t does, then natural selection w i l l operate on those characters which determine dominance. I therefore also sought correlations between several factors that had been found to, or that I thought could, influence dominance in the song sparrow. The results of t h i s investigation were s t r i k i n g l y c l ear. Individual song sparrows with high dominance status had superior survival and access to t e r r i t o r y , as compared to subordinates. I found that only two factors consistently predicted dominant status. These were sex and date of hatch. Males were more often dominant than females, and males survived better than females o v e r a l l . Thus, dominance could explain the skewed sex r a t i o favoring males on Mandarte Island (Smith et a l . 1982). 68 Variation in the dominance of yearlings was largely explained by their age; birds hatched e a r l i e r in the season dominated those hatched l a t e r . I showed by experiment that this was not because early-hatched birds had a greater f a m i l i a r i t y with l o c a l areas than others. Early-hatched birds that were deprived of this f a m i l i a r i t y by confinement in groups, but were thus allowed experience in agonistic encounters, were as dominant as early-hatched birds that were not confined. However, they were dominant to late-hatched birds that lacked comparable experience in agonistic encounters. I propose that the amount of experience an individual has had in agonistic encounters is a key determinant of dominance. Morphology only a weak influence on dominance among yearling sparrows. This r e s u l t , along with others reviewed in Chapter Two, represents a serious challenge to current theories about the evolution of sexual size dimorphism in birds (e.g. Searcy and Yasukawa 1983). However, morphological variation in the monogamous song sparrow could be too low to observe asymmetries in the size of competitors large enough to produce a strong effect of s i z e . More careful studies of the type described in this thesis are required in order to assess the importance of morphology in intrasexual competition. These are especially needed in species that are polygamous or are highly morphologically variable. Few studies have investigated the correlates and 69 consequences of dominance b e h a v i o r among so many i n d i v i d u a l s whose h i s t o r i e s were known. A major c o n t r i b u t i o n of t h i s t h e s i s was t o show the importance of such complete i n f o r m a t i o n t o the study of dominance. T h i s study was c a r r i e d out i n a p o p u l a t i o n of monogamous t e r r i t o r i a l b i r d s . However, dominance b e h a v i o r i s widespread among s o c i a l a n i m a l s , and the two f a c t o r s i d e n t i f i e d i n t h i s study t h a t i n f l u e n c e d dominance, sex and age, a r e common a t t r i b u t e s t o many of them. I b e l i e v e t h a t t h i s t h e s i s s u g g ests v a l u a b l e avenues f o r f u r t h e r r e s e a r c h i n t o the o r i g i n s and the s i g n i f i c a n c e of dominance b e h a v i o r . 70 LITERATURE CITED Alatalo, R.V. , Lundberg,A.,& Stahlbrandt,K. 1984. Female mate choice in the pied flycatcher Ficedula hypoleuca. Behav. Ecol. Sociobiol. in press. Amadon, D. 1943. Bird weights as an aid in taxonomy. Wilson B u l l . 55:164-177. Amadon, D. 1959. The significance of sexual differences in size among birds. Proc. Am. Phil.Soc. 103:531-536. Appleby, M.C. 1982. The consequences and causes of high s o c i a l rank in red deer stags. Behaviour 82:259-273. Appleby, M.C. 1 9 8 3 . The probability of l i n e a r i t y in hierarchies. Anim. Behav. 31:600-608. Bailey, R.O. 1979. Methods of estimating t o t a l l i p i d content in the redhead duck (Aythya americana) and an evaluation of condition indices. Can. J. Zool. 57: 1830-1833. Baker, M.C. and Fox, S.F. 1978. Dominance, sur v i v a l , and enzyme polymorphism in dark-eyed juncos, Junco hyemalis . Evolution 32:697-711. Baker, M.C, Belcher, C.S., Deutsch, L.C., Sherman, G.L., and Thompson, D.B. 1981. Foraging success in junco flocks and the e f f e c t s of s o c i a l hierarchy. Anim. Behav. 29:137-142. Balph, M.H. 1975. Wing length, hood coloration, and sex r a t i o in dark-eyed juncos wintering in Northern Utah. Bird-banding 46:126-130. Blake, CH. 1962. Wing length of the slate-colored junco. Bird-banding 33:97-99. Boag, D.A. 1982. How dominance status of adult Japanese quail influences v i a b i l i t y and dominance status of their offspring. Can. J. Zool. 60:1885-1891. 71 Boag, D.A. and Alway J.G. 1980. Effect of s o c i a l environment within the brood on dominance rank in gallinaceous birds (Tetraonidae and Phasianidae) Can. J. Zool. 58:44-49. Boag, D.A. and Alway J.G. 1981. H e r i t a b i l i t y of dominance status among Japanese q u a i l : A preliminary report. Can. J. Zool. 59:441-444. Boag, P. T. 1983. The h e r i t a b i l i t y of external morphology in Darwin's ground finches (Geospiza) on I s l a Daphne Major, Galapagos. Evolution 37:877-894. Boersma, D. and Ryder J.P. 1984. Reproductive performance and body condition in nesting r i n g - b i l l e d g u l l s . J. F i e l d Ornith. 54:374-380. Brace, R.C. & Pavey, J. 1978. Size-dependent dominance hierarchy in the anemone Actina equina . Nature, Lond. 247:289-290. Brawn,J.D. & Samson, F.B. 1983. Winter behavior of Tufted Titmice. Wilson B u l l . 95:222-232. Brian, A.D. 1949. Dominance in the great t i t . Scot. Nat. 61:144-1 55. Brown, J.L. 1963. Aggressiveness, dominance and s o c i a l organization in the S t e l l a r Jay. Condor 65:460-484. Brown, J.L. 1975. The evolution of behavior. W.W. Norton and Company Inc., New York. 761pp. Brown, J.L., Brown, E.R. & Brown, S.D. 1982. Morphological variation in a population of grey-crowned babblers: correlations with variables a f f e c t i n g s o c i a l behavior. Behav. Ecol. Sociobiol. 10:281-287. Caldwell, R.L. and Dingle, J. 1979. The influence of size d i f f e r e n t i a l on agonistic encounters in the mantis shrimp Gonodactylus v i r i d u s . Behavior 68:255-264. 72 Clark, G.A. 1979. Body weights of birds: a review. Condor 8J_ 193-202. Clutton-Brock T.H., Guinness, F.H., and Albon, S.D. 1982. Red deer: behavior and ecology of two sexes. The University of Chicago Press, Chicago. Collias,N.E. 1943. S t a t i s t i c a l analysis of factors which make for success in i n i t i a l encounters between hens. Am. Nat. 77:519-538. C o l l i a s , N.E. & Taber, R.D. 1951. A f i e l d study of some grouping and dominance relations in ring-necked pheasants. Condor 53:265-275. Connell, C.E., Odum, E.P. & Kale, H. 1960. Fat-free weights of birds. Auk 77:1-9. Coutee, E.L. 1967. Agonistic behavior in the American goldfinch. Wilson B u l l . 79:89-109. Craig, J.V., Ortmann, L.L., and Guhl, A.M. 1965. Genetic selection for soc i a l dominance a b i l i t y in chickens. Anim. Behav. 13:114-131. Darwin, CR. 1907. The descent of Man and selection in r e l a t i o n to sex. New York, New York:J.A.Hi 11 and Company. Dawkins, R. and Krebs, J.R. 1978. Animal signals: information or manipulation? In Behavioural Ecology. An evolutionary approach. Eds. Krebs, J.R. and Davies, N.B. Sinauer Associates, Inc. Sunderland, Massachusetts, pp 282-312. De Vos, G.J. 1983. Social behavior of black grouse: an observational and experimental f i e l d study. Ardea 71:1-1 03. Dhondt, A.A. 1979. Summer dispersal and survival of juvenile great t i t s in southern Sweden. Oecologia 42:139-157. . 73 Dhondt, A.A. and Huble, J. 1968. Fledging date and sex in re l a t i o n to dispersal in young t i t s . Bird Study 15:127-1 34. Dhondt,A.A., Eyckerman, R. & Huble, J. 1979. W i l l great t i t s become l i t t l e t i t s ? B i o l . J. Linn. Soc. 11:289-294. Dixon,K.L. 1965. Dominance-subordination relationships in the mountain chickadee (Parus gambeli) . Condor 67:291-299. Downhower, J.F. 1976. Darwin's finches and the evolution of sexual size dimorphism in body si z e . Nature 263:558-563. Ewald,P.W. & Rohwer, S.A. 1980. Age, coloration and dominance in non-breeding hummingbirds: a test of the asymmetry hypothesis. Behav. Ecol. Sociobiol. 7:273-279. Fretwell, S.D. 1969. Dominance behavior and winter habitat d i s t r i b u t i o n in juncos (Junco hyemalis) . Bird-banding 40: 1-25. Fretwell, S.D. 1972. Populations in a seasonal environment. Princeton,New Jersey: Princeton University Press. G a t t i , R.C. 1983. Spring and Summer age separation techniques for the mallard. J . Wildl. Manage. 47:1054-1062. Garnett, M.C. 1981. Body size, i t s h e r i t a b i l i t y and influence on juvenile survival among great t i t s , Parus major . Ibis 123:31-41. Gauthreaux, S.A. 1978. The ecological s i g n i f i c a n c e of behavioral dominance. In: Perspectives in Ethology vol.3 (eds. P.P.G. Bateson & P.H. Klopfer), pp.17-54. Gi l b e r t , N. 1973. Biometrical interpretation. Oxford University Press, London. 125pp. Glase, J. 1973. Ecology and s o c i a l organization of the black-capped chickadee. Living Bird 12:235-267. 74 Greenberg, R.E., Etter, S.L. & Anderson, W.L. 1972. Evaluation of proximal primary feather c r i t e r i a for aging wild pheasants. J. Wil d l . Manage. 36:700-705. Greenwood, P.J., Harvey, P.H., and Perrins, CM. 1979. The role of dispersal in the great t i t : the causes, concequences, and h e r i t a b i l i t y of d i s p e r s a l . Jour. Anim. Ecol. 48: 123-142. Halse, S.A. & Skead D.M. 1983. Wing moult, body measurements and condition indices of spur-winged geese. Wildfowl 34,108-114. Hamilton, T.H. 1961. The adaptive significance of i n t r a s p e c i f i c trends of variation in wing length and body size among bird species. Evolution 15:180-195. Hannon, S.J. & Roland, J. 1984. Physiological factors influencing t e r r i t o r y a c q u i s i t i o n and retention in the Willow Ptarmigan. Can. J. Zool. in press. Helms, CW., Aussiker, E.B., Bower, E.B. & Fretwell, S.D. 1967. A biometrical study of the major body components of the slate-colored junco, Junco hyemalis . Condor 69:560-578. Hinde, R.A. 1956. A comparative study of the courtship of certain finches (Fringi11idae) . Ibis 98:1-23. Howard, R.D. 1983. Sexual selection and v a r i a t i o n in reproductive success in a long-lived organism. Am. Nat. 122:301-325. Hrdy, S.B. 1977. The langurs of Abu. Harvard University Press. Cambridge Massachusetts. 360pp. James, F.C 1970. Geographic size variation in birds and i t s relationship to climate. Ecology 51:365-390. Jehl, J.R. 1970. Sexual selection for size differences in two species of sandpipers. Evolution 24:311-319. 75 Ketterson, E.D. 1979. Aggressive beahvior in wintering dark-eyed juncos: determinants of dominance and their possible r e l a t i o n to geographic v a r i a t i o n . Wilson B u l l . 91 : 371-383. Ketterson, E.D. and King, J.R. 1979. Metabolic and behavioral responses of fasting in the white-crowned sparrow (Zonotr ichi a luecophrys gambeli i ) . Physiol. Zool. 50:115-129. . Kikkawa, J. 1961. Social behavior of the white-eye Zosterops l a t e r a l i s in winter flocks. Ibis 103:428-442. Kikkawa, J. 1981. Winter survival in relation to dominance classes among silvereyes, Zosterops l a t e r a l i s  chlorocephala , of Heron Island, Great Barrier Reef. Ibis 122:437-446. Knapton, R.W. 1973. Some ecological aspects of s o c i a l beahvior in the song sparrow, Melospiza melodia ). M.Sc. thesis, University of B r i t i s h Columbia, Vancouver. Knapton, R.W. 1976. Dominance hierarchies in winter song sparrows. Condor 78:567-569. Knapton, R.W. & Krebs, J.R. 1974. Settlement patterns, t e r r i t o r y size, and breeding density in the song sparrow (Melospiza melodia) . Can. J. Zool. 52:1413-1420. Krebs, J.R. 1983. T e r r i t o r i a l defence in the great t i t (Parus  major): do residents always win? Behav. Ecol. Sociobiol. LeBoeuf, B.J. and Peterson, R.S. 1969. Social status and mating a c t i v i t y in elephant seals. Science 1633:91-93. L i l l , A. 1976. Lek behavior in the golden-headed manakin (Pipra  erythrocephala) in Trinidad (West Indies). F o r t s t s c h r i f t Verhaltenstorschung, heft 18. Lundberg, A., Alatalo, R.V., Carlson, S. & Ulfstrand, S. 1981. Biometry, habitat d i s t r i b u t i o n and breeding success in the Pied Flycatcher, Ficedula hypoleuca . Ornis Scand. 12:68-79. 76 Marler,P. 1955. Studies of fighti n g in chaffinches. 1. Behaviour in r e l a t i o n to the s o c i a l hierarchy. B r i t . J. Anim. Behav. 3:111-117. Mayr, E. 1972. Sexual selection and natural sel e c t i o n . In Sexual selection and the descent of man, ed. B.G. Campbell, pp.87-104. Chicago:Aldine. 378pp. Moore, M.C. 1984. Changes in t e r r i t o r i a l defense produced by changes in c i r c u l a t i n g levels of testosterone: a possible hormonal basis for mate-guarding behavior in white-crowned sparrows. Behavior 88:215-226. Moore, N.J. 1972. Ethology of the Mexican junco (Junco phaenotus  pallidus) . Ph.D. thesis, University of Arizona, Arizona. Moss, R., Watson, A., Rothery, P., and Glennie, W.W. 1982. Inheritance of dominance and aggressiveness i t captive red grouse Lagopus lagopus scoticus. Aggressive Behav. 8:1-18. Morse, D.H. 1980. Behavioral mechanisms in ecology. Harvard University Press, Cambridge, Mass. Mueller, H.C., Berger, D.D., and A l l e z , G. 1976. Age and sex variation in the size of goshawks. Bird-Banding 47:310-318. Mueller, H.C., Berger, D.D., and A l l e z , G. 1979. Age and sex differences in the size of sharp-shinned hawks. Bird-Banding 50:34-44. Mueller, H.C., Berger, D.D., and A l l e z , G. 1981. Age, sex, and seasonal differences in size of Cooper's hawks. J . F i e l d Ornith. 52:112-126. Murchison, C., Pomerat, CM. & Zarrow, M.X. 1935. The experimental measurement of a so c i a l hierarchy in Gallus  domesticus . V. The post-mortem measurement of anatomical features. J. Soc. Psychol. 6:172-181. Newton, I., Marquiss, M. & V i l l a g e , A. 1983. Weights, breeding and survival in European sparrow hawks. Auk 100:344-354. 77 Odum, E.P. 1942. The annual cycle of the black-capped chickadee. 12:113-119. Parker, G.A. 1974. Assessment strategy and the evolution of figh t i n g behavior. J. Theor. B i o l . 47: 223-243. Perrins, C. 1963. Survival in the great t i t , Parus major. Proc. XIII Intern. Ornith. Congr.:717-728 Peters, Wm. D. & Grubb, T.C. 1983. An experimental analysis of sex-specific foraging in the downy woodpecker, Picoides  pubescens . Ecology 64:1437-1443. Piersma, T. 1984. Estimating energy reserves of great crested grebes Podiceps c r i s t a t u s on the basis of body dimensions. Ardea 72:119-126. Power, D.M. 1969. Evolutionary implications of wing and size v a r i a t i o n in the red-winged blackbird in relation to geographic and climatic factors: a multiple regression analysis. Syst. Zool. 18:363-373. Price, T.D. 1984a. Sexual selection on body size, t e r r i t o r y and plumage variables in a population of Darwin's Finches. Evolution, in press. Price, T.D. 1984b. The evolution of sexual size dimorphism in Darwin's finches. Am. Nat. 123, in press Rand, A.L. 1961. Some size gradients in North American birds. Wilson B u l l . 73:46-56. Richards, S.M. 1974. The concept of dominance and the methods of assessment. Anim. Behav. 22:914-930. Rohwer, S.A. 1975. The s o c i a l significance of avian winter plumage v a r i a b i l i t y . Evolution 29:593-610. Rohwer, S.A. and Rohwer, F.C. 1978. Status signaling in Harris' sparrows: experimental deceptions achieved. Anim. Behav. 26:1012-1022. 78 Rohwer, S.A., Ewald, P.W. and Rohwer, F.C. 1981. Variation in size, dominance and appearance within and between sex and age classes of Harris' sparrows. J. F i e l d Ornithol. 52: 291-303. Rolhf, F.J. and Sokal, R.R. 1969. S t a t i s t i c a l tables. W.H. Freeman and Company. San Francisco. Sabine, W.S. 1959. The winter society of the Oregon junco: intolerance, dominance, and the pecking order. Condor 61: 110-135. S a f r i e l , U.N. 1981 - Social hierarchy among s i b l i n g s in broods of oystercatchers, Haematopus ostralegus . Behav. Ecol. Sociobiol. 9:59-64. Schneider, K.C.J. 1979. Dominance hierarchies, foraging and predation in winter flocks of the white-throated sparrow (Zonotrichia a l b i c o l l i s ) . Ph.D. thesis Princeton University, New Jersey. Schroder, S.L. 1981. The role of sexual selection in determining o v e r a l l mating patterns and mate choice in chum salmon. Ph.D. thesis, University of Washington, Seattle, Washington. Searcy, W.A. 1979a. Morphological correlates of dominance in captive male red-winged blackbirds. Condor 81 : 417-420 . Searcy, W.A. 1979b. Male c h a r a c t e r i s t i c s and pairing success in red-winged blackbirds. Auk 96:353-363. Searcy, W.A. 1979c. Size and mortality in male yellow-headed blackbirds. Condor 81:304-305. Searcy, W.A. I979d. Sexual selection and body size in male red-winged blackbirds. Evolution 33:649-661. Searcy, W.A. 1982. The evolutionary e f f e c t s of mate selection. Ann. Rev. Ecol. Syst. 13:57-85. 79 Searcy, W.A. and Wingfield, J.C. 1980. The effects of androgen and antiandrogen on dominance and aggressiveness in male red-winged blackbirds. Horm. Behav., 14:126-135.. Selander, R.K. & Johnston, R.F. 1967. Evolution of the house sparrow. I. Intrapopulation variation in North America. Condor 69:217-258. Selander, R.K. 1972. Sexual selection and dimorphism in birds. In: Sexual selection and the descent of man 1871 -1971(ed. B.G. Campbell), pp.180-230. Chicago:Aldine Atherton. Shawcross, J.E. 1982. Agonistic behavior over food and perch space in the male Quelea quelea. Anim. Behav. 30:901-908. Shoemaker, H.H. 1939. Social hierarchy in flocks of the canary. Auk 56:381-406. Slagsvold, T. 1982. C r i t e r i a for estimating the condition of birds: relationship between fat content and body size dimensions in the hooded crow Corvus corone cornix. Ornis Scand. 13:141-144. Smith, J.N.M. 1981. Cowbird parasitism, host fi t n e s s , and age of the host female in an island song sparrow population. Condor 83: 152-161. Smith, J.N.M. & Zach, R. 1979. H e r i t a b i l i t y of some morphological characters in a song sparrow population. Evolution 33:460-467. Smith, J.N.M. and Dhondt, A.A. 1980. Experimental confirmation of heritable morphological variation in a natural population of song sparrows. Evolution 34:1155-1158. Smith, J.N.M. and Roff, D.A. 1980. Temporal spacing of broods, brood size, and parental care in song sparrows (Melospiza melodia). Can. J. Zool. 58:1007-1015. Smith, J.N.M., Montgomerie, R.D., Taitt,M.J. & Tom-Tov, Y. 1980. A winter feeding experiment on an island song sparrow population. Oecologia 47:164-170. 80 Smith, J.N.M., Yom-Tov, Y, and Moses, R. 1982. Polygyny, male parental care, and sex r a t i o in song sparrows: an experimental study. Auk 99:555-564. Smith, S.M. 1976. Ecological aspects of dominance hierarchies in black-capped chickadees. Auk 93:95-107. Smith, S.M. 1978. The "Underworld" in a t e r r i t o r i a l sparrow: adaptive stratagy for f l o a t e r s . Am. Nat. 112:571-582. Smith, S.M. 1984. Flock switching in chickadees: why be a winter floater? Am. Nat. 123:81-98. Sokal, R.R. & Rohlf, F.J. 1982. Biometry. San Francisco, Calif.,:W.H. Freeman and Company. S t i l e s , G. and Wolf, L. 1979. Ecology and evolution of lek mating behavior in the long-tailed hermit hummingbird. Ornith. Mongr. pp.27-88. Suttie, J.M. 1983. Body size: dominance relationships in red deer stag calves. Anim. Behav. 31:610-611. Syme, G.J. 1974. Competetive orders as measures of so c i a l dominance. Anim. Behav. 22:931-940. Syme, G.J., Pollard, J.S., Syme, L.A., and Reid, R.M. 1974. An analysis of the limited access measure of dominance in the laboratory rat. Amin. Behav. 22:486-500. Thompson, W.L. 1960. Agonistic behavior in the house finch part 2: factors in aggressiveness and s o c i a l i t y . Condor 62:378-402. Tompa, F.S. 1964. Factors determining the numbers of song sparrows (Melpspiza melodia, Wilson) on Mandarte Island, B.C., Canada. Acta. Zool. Fenn. 109:3-68. Tordoff, H.B. 1954. Social organization and behavior in a flock of captive non-breeding red c r o s s b i l l s . Condor 56:346-358. 81 Ulfstrand, S., Alatalo, R.V., Carlson, A. & Lundberg, A. 1981. Habitat d i s t r i b u t i o n and body size of the great t i t , Parus  major. Ibis 123:494-499. Watson, A,. & Parr, R. 1981. Hormone implants a f f e c t i n g t e r r i t o r y size and aggressive and sexual behavior in red grouse. Ornis Scand. 12:55-61. Watt, J.D., Ralph, C.J. & Atkinson, C T . 1984. The role of plumage polymorphism in dominance relationships of white-throated sparrows. Auk 101:110-120. Wiley, R.H. 1973. T e r r i t o r i a l i t y and non-random mating in sage grouse, Centrocercus urophasianus. Wilson, E.O. 1975. Sociobiology: the New Synthesis. Cambridge, Mass.,Harvard University Press. Wishart, R.A. 1981. Wing-feather c r i t e r i a for age separation of american wigeon. J.Wildl. Manage. 845:230-235. Yasukawa, K. 1979. T e r r i t o r y establishment in red-winged blackbirds: importance of aggressive behavior and experience. Condor 81:258-264. 82 APPENDIX Researchers have used several d i f f e r e n t methods to estimate dominance (reviewed by Syme 1974). Two methods, however, are most commonly employed by fieldworkers. Each of these uses the outcome of agonistic encounters over some resource, t y p i c a l l y food, to determine the dominance of individuals. The individual that achieves a supplant (the winner) i s said to dominate the one that is supplanted (the loser; Brown 1975). The Hierarchical Rank Method uses the outcome of paired encounters to construct dominance matrices (e.g. Sabine 1959, Brown 1963), within which individuals are ranked from one to n, where n equals group size. The Wins/Total Method simply divides the t o t a l number of wins per individual by the t o t a l number of encounters that the individual was observed in (e.g. Fretwell 1969, Kikkawa 1980). A variation on t h i s method i s the win/loss r a t i o employed by Searcy (1979). Each of these methods has drawbacks. The Hierarchical Rank Method assumes that dominance i s t r a n s i t i v e , and t h i s assumption i s not always j u s t i f i e d (Appleby 1983). In addition, when the number of individuals being sampled is large, dominance matrices are d i f f i c u l t to construct because the number of possible pair combinations increases as one half the square of the group si z e . This method also precludes the use of parametric s t a t i s t i c s . On the other hand, the Wins/Total method, as I s h a l l show here, is a very inaccurate measure of 83 dominance when the number of samples per individual is low. This method does, however, allow the use of parametric s t a t i s t i c s , and i t makes no assumption about the t r a n s i t i v i t y of dominance r e l a t i o n s . Baker and Fox (1978) compared these two methods using data from captive and f r e e - l i v i n g juncos (Junco  hyemalis) and found that the Wins/Total method provided unambiguous re s u l t s , whereas the rank method sometimes produced more than one equally v a l i d hierarchy. Baker and Fox (1978) did not, however, consider the effect that sample size has on estimation error. I studied dominance in a population of song sparrows (Melospiza melodia) where the number of interacting birds exceeded TOO. Thus, more than 5,000 pair combinations were possible in any one season. I therefore chose the Wins/Total method as the only feasible one for estimating dominance in the f i e l d among large groups of birds. After choosing this method, I had to devise a scheme to account for estimation error. Previous researchers have simply excluded from their analyses birds with less than five observations in order to reduce error in dominance estimates (e.g. Fretwell 1969, Kikkawa 1980). However, the 95% confidence l i m i t s for a value of 50% based even on ten observations are plus or minus 31.3%. Such a large error is unacceptable. Here, I account for the e f f e c t that estimation error has on least-squares type analyses, using standard s t a t i s t i c a l techniques. 84 The accuracy of any individual percentage increases with the number of samples upon which i t i s based. Thus, the confidence i n t e r v a l for a value of 50% is reduced from plus or minus 31.3% to plus or minus 10.1% as sample size increases from ten to 100 (Rohlf and Sokal 1969). Gilbert (1973) points out that the variance of a percentage, based upon N observations, is proportional to 1/N. He therefore recommended that percentage data based on r a d i c a l l y d i f f e r e n t samples be weighted by one over the variance of the estimate. In the case of percentages based on N successive binomial t r i a l s , the appropriate weight i s simply N (Gilbert 1973). Weighting has the effect of increasing the r e l a t i v e contribution of estimates based on larger, more accurate samples to the t o t a l (weighted) sums of squares. My data were dis t r i b u t e d approximately normally, and the angular transformation did not improve their f i t . I therefore wished to determine the appropriate weight empirically, rather than immediately accepting the theore t i c a l solution suggested above. I did t h i s by estimating the relationship between the variance of the estimate and the number of observations i t was based on. The data used for t h i s purpose were obtained by observing the outcomes of agonistic encounters between color-marked yearling song sparrows at feeding tables. Details of the study s i t e and methods employed are given -in Chapter One. To estimate the variance of the Wins/Total dominance measure, I used data from twenty-five birds which had been observed in over 85 100 encounters (range 112-200), and whose sequence of interactions in time was f u l l y known. Then, assuming that the variance is proportional to the mean-squared error of the estimates, I calculated the deviation of each successive estimate from the f i n a l estimate as indicated by: Deviation = ( wins/total) — (wins/total) N,bird N observation f i n a l Where N is the observation number from one to 100. This generated 2,500 deviations. I summed the squared deviations for a l l birds over each observation number, and then multiplied this sum of the squares by 1/N to obtain the mean-squared error of the estimate which corresponded to observation N. A curve f i t t e d to these points had the equation: -.93 Y = .24 * X , r-squared =0.98 Where Y i s the mean-squared deviation and X i s the observation number. The weight is simply one over the X-term, which is very close to the t h e o r e t i c a l l y predicted weight of N suggested by G i l b e r t (1973). I therefore accepted N as the appropriate weight. 86 I next wished to test the v a l i d i t y of t h i s weighting scheme using the actual data. To do t h i s , I f i r s t assumed that a sample dominance estimate more accurately reflected a bird's average chances of winning an encounter as the sample size of the estimate increased. I then chose a variable (hatch date) that was found to be highly correlated with the weighted dominance estimate. If the weighted correlation is correct, then similar results should be reached i f only the best estimated samples are used to approximate the c o r r e l a t i o n , as i f the weighted analysis were used. As expected, when the acceptable minimum number of samples per estimate is increased from zero to one hundred, the cor r e l a t i o n c o e f f i c i e n t between dominance and hatch date increases l i n e a r l y from 0.21 to 0.86 respectively (1983 data, n=112). The second figure i s very similar to those obtained from weighted analyses (Table 6, Chapter One). The same effect of sample size on the co r r e l a t i o n between two variables was also observed for correlations between morphology and dominance (Table 4, Chapter One). In cases where no s i g n i f i c a n t correlations were found between dominance and an independent variable in weighted analyses, they were also absent when only estimates based on large samples were used. These results confirm that this weighting scheme i s a valuable tool to account for errors in dominance estimates based on win/loss data. It i s important to note, however, that t h i s 87 procedure is no alternative to the rigorous c o l l e c t i o n of data. If a l l estimates are based on very small samples, errors in sample estimates may prevent a s t a t i s t i c a l c o rrelation from being found when a corr e l a t i o n of b i o l o g i c a l significance does indeed e x i s t . In summary, researchers that have used win/loss data to estimate the dominance of individuals have not seriously considered the effect that estimation error may have had on thei r analyses. The common procedure to deal with this problem has been to exclude from analyses a l l estimates based on fewer than five samples, and to treat a l l those based on more than f i v e as equally good. The analyses presented here show that t h i s is not a v a l i d assumption. This f a i l u r e to account for estimation error in th i s case means that b i o l o g i c a l relationships may be p a r t i a l l y or t o t a l l y obscured. The effect that error in estimates has can be accounted for using the standard s t a t i s t i c a l procedures reviewed here. I j u s t i f y t h i s procedure for the analyses presented in t h i s thesis. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0096044/manifest

Comment

Related Items