THE
ROLE
OF PLUMAGE
DOMINANCE
POLYMORPHISM
RELATIONSHIPS
WHITE-THROATED
IN
OF THE
SPARROW
DORISJ. WATT,• C. JOHN RALPH,2 AND CARTERT. ATKINSON3
Departmentof Biology,Dickinson
College,Carlisle,Pennsylvania
17013USA
ABSTRACT.--We
studied dominancebehaviorsof captive winter flocksof White-throated
Sparrows(Zonotrichia
albicollis)
in centralPennsylvania.Preliminarytestswith a smallgroup
of birdsindicatedthat there might be differencesin behaviorsbetweencolormorphsif age
and sexualdifferenceswere controlled.Studiesof two larger groupsproducedconsiderable
variationin resultsbetweenthe groups.One group displayedstriking correlationsbetween
dominanceability and color. In this group white adult malesdominatedothersmore than
tan adult males did, whereas tan immature males, adult females, and immature females were
more often dominant than were white immatureand female birds. The secondgroup did
not displaysuchstriking differences.Tan adult and immaturefemaleswere more dominant
than white females,as in the first group, but not significantly.Immature malesdisplayed
the oppositetrend from the firstgroup--white morphswere moredominantthan tans.Adult
malesin the secondgroup showed no clear trend. We also found differencesin dominance
between tan and white femalesthat appearedto depend on season,white birds dominating
tan birds in a small group in the spring,a reversalof relationshipsdocumentedin the fall.
Within age-sexclasses,dominantfemalesin both large groupstendedto be duller in plumage brightness(scaledas an index) than were subordinatefemales.In one of the two large
groups,duller immature maleswere more dominant than brighter ones,while brighter adult
males tended to be more dominant than duller ones. We suggestthat the relationships
betweenimmaturemale plumageand dominancemay be influencedby the morph of adult
malesof specificdominancestatusin the flock. Spring plumage indiceswere more often
correlatedwith fall dominancedifferencesthan were fall plumage conditions,suggestinga
genetic or causalrelationshipbetween dominanceand plumage rather than a proximate
(statussignaling) function. Recordedmorph frequenciesfrom this study and the literature
supportthe hypothesisthat selectionis balancedbetweenthe two morphsand is dependent
on the advantagesor disadvantagesof aggressiveness
within a sex. Received9 March 1983,
accepted25 July 1983.
LOWTHER
(1961) described two color morphs
of the White-throated Sparrow (Zonotrichiaalbicollis),a white-striped and a tan-stripedform.
He showed that both morphs included males
and females.
In field observations
of 110 mated
pairs, he observednegative assortativemating:
white-striped males mated with tan-striped females,and tan-striped malesmated with whitestriped females(Lowther and Falls 1968). Var-
• Present address: Biology Department, Saint
Mary'sCollege,Notre Dame,Indiana46556USA.Requestsfor reprintsshouldbe sentto this address.
2 Present address: U.S.D.A. Forest Service, Red-
wood SciencesLaboratory,1700 Bayview Drive, Ar-
dy (1971)disagreedthat colorvariationwasdue
to two morphs;she found color typesto have
a continuous
distribution.
• PresentAddress:Departmentof PreventiveMedicine, Collegeof Veterinary Medicine, University of
Florida, Box J-136, Gainesville, Florida 32610 USA.
11o
recent studies
chromosomal dimorphism in the speciesthat
was always related to the plumage dimorphism; Atkinson and Ralph (1980) found birds
in breeding plumage to be bimodally distributed with regard to quantitative plumage measurements;and Rising and Shields (1980) described other morphological differences
between the color morphs.Severalof theseau-
thorssuggestedthat there might be behavioral
differences between the color morphs on the
wintering grounds.
We examined
cata, California 95521 USA.
More
havesupportedLowther'sinterpretation,however: Thorneycroft(1966, 1975) documenteda
differences
in dominance
be-
haviors between the two color morphs in winter flocks. Several attempts to find differences
in dominance behavior between the morphs
The Auk 101: 110-120. January 1984
January1984]
White-throated
SparrowDominance
during the nonbreedingseasonhave been made
(Harrington 1973, Hailman 1975, Ficken et al.
1978), but none of these investigatorsconsidered the sexor age of the birds. We studiedthe
interrelationshipsamong sex, age, and color
morph and found that a combination of characterswas a goodpredictor of dominancerank.
The purposeof this paper is to demonstratethe
relationships of these variables with social
dominance behavior of White-throated Sparrows in captive, nonbreeding flocks.
METHODS
Birdswere caughtin mist netsthroughoutOctober
1975 at Carlisle, Pennsylvania. Birds were individually marked at capturewith colored leg bandsand
weighed. We determinedage by the amount of skull
ossificationand sexby the unflattenedwing chord at
capture (Atkinson and Ralph 1980) or by laparotomy
the following spring.Birdswere kept in three indoor
aviaries (2 x 2 x 2 m) under controlled artificial
lights, initially on a winter light schedule. The daylength was increasedon 1 April to 12 hours light and
12 hours dark to stimulate prenuptial molt. By 16
April the birds were coming into breeding condition,
as indicated by molt and singing. We investigated
social relationshipswithin four groups of birds: a
smallgroupof 16 malesand femaleson 28 November
1975;two largegroupsof 51 and 54 birds in November and December1975;and a small group of 17 females on 16 April 1976.
Determinationof colormorph.--Even though Whitethroated Sparrows in breeding plumage can be divided into two distinct color morphs (Thorneycroft
1975, Atkinson and Ralph 1980), considerable individual variation exists in plumage characteristics
(Vardy 1971).Atkinson and Ralph (1980) demonstrated that variation in plumage in the fall ranged from
a bright to a dull extreme and that combinationsof
various plumage characteristicsinto an index resulted in a normal distribution. In addition, plumage
characteristicswere correlated with sex,age, and sea-
birds arbitrarily into "white" or "tan" groups: birds
with index valuesgreaterthan the medianwere called
"white"; those with less were "tan." Because such
plumage indices produced a relative comparisonof
brightnessamong birds and becausewe did not prepare karyotypes,our designationsof morph may not
coincidewith Thorneycroft'sgeneticallydetermined
"color morph." Atkinson and Ralph (1980), as well
as Thorneycroft (1966, 1975), however, found that,
although the two morphsare difficult to distinguish
in autumn,they are much more distinctin the spring.
Therefore,we used spring-plumageindex valuesto
approximate the morph class most similar to Thorneycroft'sgeneticallydetermined "color morph," and,
in fact, only 9 of the 105 birds would have differed
in their morph classification
betweenfall and spring
plumages.
Behavioral
observations.--Behavioral
interactions
were observedthrough a one-way window overlook-
ing a feeding platform in eachof the three aviaries.
During observationperiods, food was restricted to a
small dish on the platform. We recordedwinners and
losersof dominanceencountersat the feeding dish.
Winning birds supplanted,pecked, or chasedthe losing individuals. Using the methods of Sabine (1959)
and Brown (1975: 86), we determined dominance
hierarchiesfor the two smallgroupsof birds by constructingdominancematricesof encounters.Large
groups did not form clear linear hierarchies, precluding ranking of individual birds. Instead,we tallied dominancedyads for each morph-age-sexclass.
A dominance dyad between two birds [i.e. A > B: the
"pair relations" of Sabine (1959)] was designated
when
one bird was dominant
in more than one-half
of the encountersbetweenthe pair members,regardlessof the numberof interactions.The percentageof
dyads in which a bird was dominant was defined as
its "percentage dominance," or the number of individuals it consistently dominated in known relations. In the two large groups,the numbersof dominant dyadswere combinedfor membersof eachage-
sex-morphclassto give the percentagedominanceof
each class.
Statistics.--Because
the percentagedata were tested
son: males were generally brighter than females, and found to be normal, they were not transformed.
adults brighter than immature birds, and all birds Statistical tests included a test for differences bebecamebrighter in the spring (Atkinson and Ralph tween two percentages,product-momentcorrelation,
and Mann-Whitney U-Tests(Sokal and Rohlf 1969).
1980).
In this study, we calculatedan index of plumage
brightness similar to Atkinson and Ralph's (1980),
using four of their five plumagecharacteristics:
percentageblack in lateral crown stripes,throat pattern
(asclassifiedby Lowther 1961),median crown-stripe
color, and superciliary stripe color. We did not include chest streaking becauseof its low variability
between seasons.The four plumage characteristics
were measuredat capture in the fall and again in
May after prenuptial molt.
We used the median
value of the indices to divide
RESULTS
SMALLGROUP(NOVEMBER)
We conducteda preliminary investigationof
a small group (16 birds) in November. The
group exhibited an essentiallylinear and stable
hierarchy (Fig. 1). Inspectionof physical characteristics
correlated
with
dominance
ranks of
112
WATT,RALPH,
ANDATKINSON
[Auk,Vol. 101
LOSER
BIRD
RANK
2
•,l
4 i
li113
!
z 9
12
•
I
I 4
•2 • • • • • • •2
4 3
2
3
3
•or 16 •hRe4hroa•ed •parrows •n •ove•ber. The h•erarchy
•or •ve •ns•ances(b•rd ranked •8 was do•nan•
b•rd a•acked
i 2
/
I•
a subordinate
iI
4 ! :5 :5i I t 2 I : 2 I :5
I
•g. 1. Dominance•a•r•
21
a do•nan•
•o b•rd ranked •2• e•c.). Two reversals (R) occurred where
b•rd.
these individuals (Table 1) indicated that sex
class with another
class would
be a convenient
was the best single predictor of dominance. meansof analysis.Beforewe couldproceedwith
this analysis,however, it was necessaryto deOther measures, such as morph, age, wing
length, or weight, did not explain dominance termine how the proportionof birds in eachof
rank simply. Among females, however, tan the two colormorphsin a classmight affectany
birds were more dominant than white birds.
measurementof aggression,as Hailman (1975,
That is, if birds of
The relationshipamong maleswith respectto pers.comm.)hassuggested.
morphwasnot clear.The resultsof thisprelim- one class encountered birds of another class at
inary studyled us to believethat differencesin random, then it would be possibleto use endominance behaviors between white and tan
counter frequenciesas a measure of domimorphsmight be found in largergroupswithin nance. If encounter frequencieswere not random,then comparisons
of morphssummedover
sex and/or age classes.
the variousage-sexclasseswould not be valid.
LARGE GROUPS
Appropriateness
of encounter-frequency
analysis.--We wished to investigate the role that
morph, age, and sex played in determining
dominance in the two large groups. Measures
involving the percentageof encountersof one
Therefore, we calculated the number of ex-
pectedencountersfrom the numbersof individualsin eachmorph-age-sexclassin the two
large groups(Table 2) and comparedtheseto
observed values (Tables 3 and 4). None of the
eight classes,in either of the two groups,encountered the other classesrandomly, that is,
January1984]
White-throated
Sparrow
Dominance
113
TABLE1. Relationshipsbetween dominancerank and physicalcharacteristics
of individual White-throated
Sparrowsrepresentedin Fig. 1.
Dominance
rank
Sex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Morph
M
M
M
M
F
F
F
F
F
F
M
F
F
F
F
F
Age
White
Tan
White
White
Tan
Tan
Tan
Tan
Tan
White
Tan
White
White
White
White
White
Adult
Adult
Adult
Immature
Immature
Immature
Immature
Adult
Immature
Adult
Adult
Adult
Adult
Adult
Adult
Immature
Wing length
Weight•
(mm)
(g)
Weightb
(g)
74
72
72
70
67
65
65
69
68
65
72
67
70
68
70
69
26.8
23.0
25.0
24.4
23.5
23.6
22.1
25.6
23.5
24.0
23.8
23.2
23.8
22.8
26.6
22.9
30.4
26.2
27.0
23.8
22.6
24.5
20.5
28.5
21.7
24.8
29.1
25.3
25.1
23.9
26.3
24.6
Weight at capture.
Weight at testing,28 November.
in proportion to class size. In general, males
were observedin aggressiveencountersmore
frequently, and females less frequently, than
expected(row totals in Tables3 and 4, observed
vs. expected). Therefore, encounters were not
occurring randomly, and comparisons of
morphs summed over classeswould not be an
appropriate analysis.
Similarly, an analysisof the number of wins
(the "attacks" of Hailman 1975) would not be
percentage dominance values of white adult
males with those of tan adult males, the num-
ber of dyads they form with other morph-agesex classesshould be equal. We tested these
values (Tables 5 and 6) and found that for all
eight comparisonsof age-sexclassesthere was
no significantdifferencebetween white and tan
morphs in the probability of forming dyadic
relationshipswith the other classes,with a single exception(significantat slightlygreaterthan
the P < 0.05 level). Becausethe assumptionof
equalprobability was met, we summarizedpercentagedominancevaluesfor eachmorph-agesex classin each group (Table 7) for testing
differencesbetweenthe morphswithin a given
age-sexclass.
Dominanceas measuredby dyad formation.Striking differencesin dominance between the
appropriate,becausethey would be a function
of encounter frequency. Also, an analysis of
percentage wins or losses,based on total encounters, would be inappropriate, becausea
particularly aggressivebird might inflate the
percentagewins for its class.
Appropriateness
of analysisof dyadformation.Another method of comparing dominance relationshipsof the two morphs within age-sex
classesinvolves the useof percentage
dominance,
the percentageof dyads in which a given class TABLE2. Numbers of birds in each morph-age-sex
was dominant
over
another
class. The
use of
percentagesalleviates the problem of nonrandom encounter frequencies,becauseonly one
dyadic relationship is recorded, regardlessof
the
number
birds.
Given
of interactions
that
a bird
between
is involved
the
two
in a set of
dyadic relationships,the percentage of those
relationships in which it is dominant is independent of encounter frequency. The only remaining problem is the distribution of dyadformation frequencies: in order to compare
classfor two large groups.
Group Group
Class
Tan adult males (TAM)
White adult males (WAM)
Tan immature males (TIM)
White immature males (WIM)
Tan adult females (TAF)
White adult females (WAF)
Tan immature females (TIF)
White immature females (WIF)
1
2
5
7
5
9
12
3
6
7
3
8
4
6
11
4
7
8
114
watt, RALPH,AND ATKINSON
;>
'-o ;>
co ;>
co,.o
co,.o
[Auk, Vol. 101
January1984]
White-throated
SparrowDominance
115
116
WATT,RALPH,
ANDATKINSON
[Auk,Vol. 101
TABLE7. Percentagedominancein two groupsof each morph-age-sexclassover classesother than its own.
Group 1
na
Group 2
%
tb
Tan
adult
males
White
immature
males
Tan
immature
males
224
74.6 }
162 79.6
264
53.0 }
151 39.1
1.16ns
Tan
adultfemales
White adult females
288
37
2.74**
1.39ns
Whiteadultmales
35.4
24.3
142
35.2 }
tan
immature
females 190
38.4
Whiteimmature
females
na
%
tb
85 68.9
51.8
} 2.71'*
187
36.4
148
69.6
} 6.16'*
196
69
4.23**
34.8
22.9 } 4.29**
176 43.75
0.60ns
192
• n - numberof dominancerelationships(dyads)consideredin the computationof percentagedominance.
bStatisticalcomparisons
were madewithin age-sexclassbetweenwhite and tan birds.** = a < 0.01;ns = not significant.
two morphs, within each age-sexclass,were
found in Group 2 (Table 7). In encountersbetween adult males, white morphs were more
often
dominant.
In
encounters
between
im-
the expressionof color morph in the spring.
We could not distinguishbetween causeand
effectunder this last hypothesis.We did assess
relationships between fall dominance and de-
mature males, adult females, or immature fe-
gree of changein plumagefrom fall to spring,
males, however, tan morphs were more often
dominant. Group 1 showedlessstriking differ-
however.
ences. In encounters
between
adult
females
or
immature females, tan morphs were also more
dominant than white morphs, but not significantly. In encountersbetween adult males or
immature males, tan morphs were more dom-
inant, a trend oppositeto that Shownby Group
2 birds. Only the relationship between immature males was significant, however.
DIFFERENCES WITHIN
Because we found
dominance behaviors than did fall values (Table 8). To demonstrate the nature of the cor-
relations,we plotted the spring-plumageindex
of individual birds in each of the two groups
againsttheir measuredpercentagedominance
(Fig. 2). A significant positivecorrelation was
found for adult malesin Group 2 (whiter birds
were more dominant). Significant negative relationships were found for immature males in
AGE-SEx CLASS
instances
We found that spring values of plumage
brightnesscorrelated more often with winter
of differences
in
Group 2, adult females in Group 1, and total
females(immature femalesvaried in the appropriate direction but were not quite significant
dominance between morphs within each agesex class,we investigated possible differences
involving more continuous plumage variation
within Groups 1 and 2. Specifically, we exam-
birds increasein plumage-indexvalue from fall
to spring (Atkinson and Ralph 1980),somebirds
ined
do so more than others. We found
whether
or not whiter
birds
were
more
dominantwithin eachage-sexclass(e.g.in adult
males)by analyzing correlationsof percentage
dominance with spring-plumage index values
and with fall-plumage index values. If birds
were using proximatecuesto evaluateprobable
dominancerelationshipsbefore interacting,i.e.
the statussignaling of Rohwer (1975), fall values should have the highest correlations with
dominance. If genetic differences in dominance ability were correlated with color morph
(apparentonly in spring),however,then spring
valueswould be mosthighly correlated.A third
and nonexclusive possibility was that winter
dominance behaviors in some way influence
when considered alone) (Table 8). While most
that the de-
gree of plumage brightening was negatively
correlated
with
fall
dominance
behavior
of
adult females in Group 2 and of all females
combined (Table 8). Females that were more
dominant in fall had lessplumagebrightening
in spring.
FEMALE MORPH DIFFERENCES IN TWO SEASONS
Thorneycroft (1975) noted that white femalesappearedto be more aggressivethan tan
females on the breeding grounds. We found
that tan females usually dominated white females and were
more
dominant
to other
birds
January1984]
White-throated
Sparrow
Dominance
117
TABLE
8. Correlationsof percentagedominancemeasureswith plumagemeasurestakenduring the fall and
spring and degree of plumage brightening from fall to spring. Correlationsbetween plumage and dominancewere more often significantin spring than in fall.
Correlation (r)
Samples
n
Fall plumage•
Spring plumage•
Group 1
Group 2
12
11
-0.142
0.490
-0.064
0.606
Total
23
0.142
0.201
Group 1
Group 2
14
10
0.410
-0.801'*
Total
24
- 0.193
47
0.067
-0.018
Group 1
Group 2
15
15
0.531'
0.101
-0.539*
-0.490
-0.146
-0.617'
Total
30
-0.242
- 0.376'
- 0.260
13
15
0.062
-0.202
0.210
-0.347
-0.281
-0.368
28
- 0.143
- 0.286
- 0.305
58
-0.209
-0.358**
-0.307*
Adult
Plumagebrightening•
males
Immature
0.144
males
Total males
Adult
0.181
0.172
0.243
0.781'*
- 0.254
-0.159
-0.327
- 0.201
0.170
females
Immature
females
Group 1
Group 2
Total
Total females
• ** = significant at a < 0.01; * - • < 0.05.
than
were
white
females.
Because our results
contradictedprevioussuggestions,we hypothesized that female dominance relationships
might change with season.Tan females could
be more dominant
than white
females
in fall
but becomemore subordinatein spring. To test
this hypothesis,we assembleda small group of
17 femalesfrom the larger hierarchiesand observed their dominance behavior in April. Following prenuptial molt, white females were
significantly more dominant than tan females
(Table 9; Mann-Whitney U-Test; U = 57, P <
0.05). Moreover, a significant positive correlation was found between individual springplumage indicesand percentagedominancein
April (r = 0.58, P < 0.05). Thus, it appearsthat
behavioral relationships between morphs, at
leastfor females,may changewith season.Becausemale birds were not present in this last
test, we cannot rule out the influence
males
might have had on female dominance.
DISCUSSION
Dominance
andplumagemorph.--Lowtherand
Falls (1968) and Falls (1969) found that white
White-throated Sparrows are more aggressive
and are more persistent and frequent singers
than are tan sparrows.Thorneycroft(1975) suggested that white females might be more aggressivethan their tan counterparts.Other investigators (Harrington 1973, Hailman 1975,
Ficken et al. 1978) have suggestedthe samefor
nonbreeding groups--white morphs are more
aggressivethan tan morphs. Hailman (1975)
reanalyzedmuch of Harrington's data and concluded that the data suggesting that white
morphs were more aggressivethan tan morphs
were inconclusive, becausethe relative proportions of morphs were not known. Taking into
accountthe proportions of the morphs present
during the encounters,Ficken et al. (1978) concluded that white morphs are more frequently
the aggressorsthan are tan morphs. They sug-
gestedthat there is a behavioral difference between the two morphs. Among the interpretations they discussedwas the possibility of a
differential sex or age bias in the morph pop-
ulations present. More important, they also
suggested that, like its coloration, the tan
morph's relative lack of aggressivenessrepresents a neotenic condition, in that in both be-
havior and morphology it retains characteris-
tics of young birds in adulthood: "tan head-
118
WATT,RALPH,
ANDATKINSON
[Auk, Vol. 101
io•7
ADULT
o
FEMALES
o
o •e
o
0 ß
o•
0
o
o
0
o
0
[]
o
0
25
ADULT
MALES
i.iJ
o
z
z
ioo
IMMATURE
IMMATURE
FEMALES
I--
[]
7•
MALES
[]
Z
•J
•.•
[]
[]
•J
•.
5o
ß O
O
O
O
ß
eO
o
25.
ß
,o
o
I0
20
[]
ß
oo
ß O0 ß
30
40
PLUMAGE
I
BRIGHTNESS
0
0
INDEX
Fig. 2. Plot of percentagedyads in which individuals were dominant in November againsttheir plumagebrightnessindex asmeasuredafter spring molt (seetext for explanation). Circles (O) and squares([•) represent
birds in Groups 1 and 2, respectively.
stripe and relative unaggressivenessare characteristics
of birds in their
first fall of life."
Our finding of relatively greater dominance
in tan females
in the fall is the reverse
of the
above reports for the spring and breeding seasons, and it disagrees with the neotenic hypothesis of Ficken et al. (1978). We feel that
among females, in particular, a reversal in relative
dominance
occurs
between
seasons.
It is
possiblethat harassmentof white females by
dominant
males
could
force
them
to lower
becamelessaggressivetoward potential mates.
It is also possible, however, that white female
birds actually becomemore aggressive,relative
to tan females, in the spring. This could be due
to hormonal changesaffecting each morph differently. Future studiesshould be able to determine whether the presence of males or a
change in physiology is responsible for these
changesin relative dominance levels.
At this time, we do not know if adult males
exhibit such a seasonalchange in dominance;
levels in the fall hierarchy, and, subsequently in one group we found white adult maleswere
in the spring, their levels might rise as males more dominant than tan adult males. Immature
January1984]
White-throated
Sparrow
Dominance
malesin one group were more dominant when
white, in the other group when tan. This obviously requires further testing, especially of
all-malegroupsin both springand fall. Sexual
differencesmay existbetweenthe colormorphs
in their relationship to dominance rank, and
further studiesrequire assessment
of sexualeffectsin the analysis.
Correlations
between
fall dominance
behav-
ior and the "spring" rather than the "fall" color morph support the hypothesis(Ficken et al.
1978) that behavioral differences between the
color morphs are genetic rather than being
basedon appearances
alone(i.e. statussignals).
We cannot rule out the possibility, however,
that behavioral
differences
in the fall affect the
expressionof color in spring plumages.
Sex differences
in morphratios.--Thorneycroft
(1975)found amongyoung femalesthe ratio of
tan to white was 1:1, while among malesit was
43:57. By breeding season, the ratios were
74:26 in females
and
32:68
in males.
119
TABLE9. Relationships between dominance rank,
percentagedominance, and color characteristicsin
a group of 17 female White-throatedSparrowsin
April.
Dominance
Color
Springplumage
Percentage
rank
morph
index
dominance
White
White
White
Tan
White
Tan
Tan
White
White
Tan
White
White
Tan
Tan
Tan
Tan
Tan
36.84
26.15
35.51
19.80
37.42
24.15
23.13
36.83
34.33
24.96
35.66
35.52
13.63
8.78
21.79
11.26
16.12
93
69
67
76
75
63
69
69
44
40
29
40
19
29
20
15
13
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
In our
study,36%of the maleswere tan, and 64%were
white. Among females,62%were tan, and 38%
were white (Table 2). These values are intermediatebetween Thorneycroft'sratiosof young
morphs in the fall. [Wessel and Leigh (1941)
studied fall birds, but they sexed most of their
birds and thoseof his adult breeding birds and birds by plumage differences,a method now
probablyreflecta normal autumn ratio, at least known to be inappropriate.]
for central Pennsylvania.
Plumagebrightnesswithin age-sexclass.--We
Bodysizeand dominance.--Rising
and Shields found that, within age-sex classes,spring(1980) found that white males were slightly plumage brightnesswas a good indicator of
larger than tan males.Although they had small dominance in females, but not in males in two
samplesof females,they showed that in 8 of 9 of four groups. The pattern of significance
measurements,
white femaleswere larger than found in one group leads us to hypothesize
tan females. Based on body size alone, one that dominancein immature males might be
might have predicted that white birds should influenced by the relations of the adult males
dominate tan birds in both sexes. We found in
in their group. Similarly to a speculationmenthe analysis of the small hierarchy (Table 1), tionedaboveregardingfemalerank, adult males
however, that body size was not a good pre- may persecuteindividualsof their own morph
dictor of dominance rank, within either age- classselectively,so that in groupswhere white
sex classesor color morphs.
males are strongly dominant, immature white
Effectsof season.--Inthe studiesof Harring- males are strongly submissive.Studies of beton (1973) and Ficken et al. (1978), data were haviorsin groupswith different sex, age, and
collected from 22 April to 4 May, when the morph ratioscould clarify thesequestions.
majority of birds are in nuptial plumage
With regard to Rohwer's (1975) hypothesis
(Lowther and Falls 1968). Therefore, their re- that plumage variability in birds is used as a
suitscan bestbe compardto thoseof the breed- statussignal, we doubt that it is being usedas
ing studies(Lowtherand Falls1968,Falls1969). such by White-throated Sparrows.Plumage itOur results for a group of females in nuptial self is not a good predictor of dominance in a
plumage agree with these previous studies; mixed-sexgroup of this species.For example,a
white
birds were more dominant
than tan
tan female in the autumn could not predict the
relative status of a white bird without knowlbirds in late spring.
Our study is the first to investigateand doc- edge of its sex--if it were a male, it would be
ument
dominance
differences
between
color
dominant to her; if a female, subordinate. Ad-
120
WATTß
RALPH,
ANDATKINSON
[Auk,Vol. 101
ditionally, changesin signalswith seasonde- FALLS,J.B. 1969. Functionsof territorial song in the
White-throated Sparrow. Pp. 207-232 in Bird
creasethe predictivevalue of plumages.Plumvocalizations (R. A. Hinde, Ed.). Cambridgeß
age variation may be important in facilitating
Cambridge Univ. Press.
individual recognition,however (Shields1977).
FICKEN, R. W., M. $. FICKEN, & J.P. HAILMAN. 1978.
Evolution.--Linkagebetween color mutations
Differential aggressionin genetically different
and aggressivelevels has been documentedin
morphs of the White-throated Sparrow (Zonomammalsby Keeler et al. (1968, 1970).The intrichiaalbicollis).Z. Tierpsychol. 46: 43-57.
teractionbetweenplumage,sex,and hormonal HAILMANß
J.P. 1975. Analysisof aggressionin Whitecontrol of aggressionmay also be a considerthroated Sparrow types of different proportions.
able influence on White-throated Sparrows.
As an evolutionary strategy for females,it
may be advantageousto be dominant to other
femalesin fall foraging flocksand then to become more subordinate in order to pair more
easilywith aggressivemales.This is the pattern
Bird-Banding46: 236-240.
HARRINGTON,
B.A. 1973. Aggressionin winter resident and spring migrant White-throated Sparrows in Massachusetts.
Bird-Banding44: 314-315.
KEELER,C., S. RIDGEWAYßL. LIPSCOMBß& E. FROMM.
1968. The geneticsof adrenal size and tameness
in colorphasefoxes.J. Hered. 59: 82-84.
seenin tan females.The sametrend (more agß T. MELLINGER, E. FROMM, & L. WADE. 1970.
gressivein fall but lessin the spring),however,
Melaninßadrenalin and the legacyof fear. J. Hewould be a disadvantagefor tan males when
red. 61: 81-88.
competingwith white malesfor territoriesin LOWTHER,J. K. 1961. Polymorphism in the Whitethroated Sparrow, Zonotrichiaalbicollis(Gmelin).
the spring. The apparent selectiveadvantages
Can. J. Zool. 39: 281-292.
of being tan, then, could be balancedthrough
ß & J. B. FALLS. 1968. White-throated Sparopposingpressureson malesand females.This
row. Pp. 1364-1392 in Life histories of North
hypothesisis supportedby the trend of deAmerican cardinalsßgrosbeaks,buntingsßtowcreasingrelative frequenciesof tan malesfrom
hees, finchesßsparrows and allies (A. C. Bent
autumn to spring and the increasing relative
and collaborators). U.S. Natl. Mus. Bull. 237.
frequency of tan females (Thorneycroft 1975). RISING, J. D., & G. F. SHIELDS. 1980. Chromosomal
White females would appear to be at a disadand morphological correlatesin two New World
vantagein both seasons(assumingsubordinasparrows(Emberizidae).Evolution 34: 654-662.
tion in winter and aggressivenessin spring
ROHWER,
S.A. 1975. The socialsignificanceof avian
winter plumage variability. Evolution 29: 593are selectively disadvantageousfor females),
610.
and, in fact, white femalesare proportionately
SABINE,
W. S. 1959. The winter society of the Orethe least numerousof any sex-morphclassby
gon
Junco: intolerance, dominance, and the
breedingseason(Thorneycroft1975).The white
pecking order. Condor 61: 110-135.
morph, then, could be maintainedby its gen-
eral advantage to adult males at all seasons.
SHIELDS,
W. M. 1977. The social significanceof avian winter plumage variability: a comment.Evo-
ACKNOWLEDGMENTS
SOKAL,R. R., & F. J. ROHLF. 1969. Biometry. San
lution
We thank J. B. Falls, J.P. Hailman, D. James, S.
31: 905-907.
Francisco, W. H. Freeman and Co.
THORNEYCROFT,
H. B. 1966. Chromosomalpolymorphism in the White-throatedSparrow,Zonotrich-
Rohwer, W. Shields,and R. Zink for help during the
courseof this studyand for helpful commentson the
ia albicollis(Gmelin). Science 154: 1571-1572.
manuscript. What forbearancewe received from the
1975. A cytogenetic study of the WhiteReineman Sanctuarywas appreciated.Bill Jefferies
throated Sparrow, Zonotrichiaalbicollis(Gmelin).
and Noel Potter provided helpful support for our
Evolution
29: 611-621.
researchat Dickinson College.
VARD¾, L. E. 1971. Color variation in the crown of
the White-throated Sparrow, Zonotrichiaalbicollis.
LITERATURE CITED
ATKINSON,C. T., & C. J. RALPH. 1980. Acquisition of
plumagepolymorphismin White-throatedSparrows. Auk
97: 245-252.
BROWNß
J.L. 1975. The evolution of behavior. New
Yorkß W. W. Norton and Co., Inc.
Condor
73: 401-414.
WESSEL,J.P., & W. H. LEIGH. 1941. Studies of the
flock organization of the White-throated Sparrow. Wilson
Bull. 53: 222-230.