Citation
Territory retention in red-winged blackbirds

Material Information

Title:
Territory retention in red-winged blackbirds the role of RHP, previous ownership and habitat quality
Creator:
Koshak, Dianne Catharine
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
vi, 37 leaves : ; 29 cm

Subjects

Subjects / Keywords:
Red-winged blackbird -- Habitat ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 33-37).
Thesis:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Department of Integrative Biology
Statement of Responsibility:
Dianne Catharine Koshak.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
22698257 ( OCLC )
ocm22698257
Classification:
LD1190.L45 1990m .K67 ( lcc )

Downloads

This item has the following downloads:


Full Text
TERRITORY RETENTION IN RED-WINGED BLACKBIRDS:
THE ROLE OF RHP, PREVIOUS OWNERSHIP AND HABITAT QUALITY
by
Dianne Catharine Koshak
B.A., University of Denver, 1982
M.A., University of Denver, 1983
B.A. University of Denver, 1984
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Arts
Department of Biology
1990


This thesis for the Master of Arts degree by
Dianne Catharine Koshak
has been approved for the
Department of Biology
by
Diana F. Tomback
Janis W. Driscoll
-Lri'nda k. Dixojr^
Date


Ill
Koshak, Dianne Catharine (M.A., Biology)
Territory Retention in Red-Winged Blackbirds: The Role
of RHP, Previous Ownership and Habitat Quality
Thesis directed by Associate Professor, Diana F. Tomback
In a removal/replacement experiment, the effects
of Resource-Holding Potential (RHP), habitat quality, and
previous ownership were measured concurrently to assess
their relative influence on territory ownership by marsh-
nesting Red-winged Blackbirds (Agelaius ohoeniceus).
According to the RHP hypothesis, original territorial
males should be larger and more aggressive than replace-
ment males. We found that original males had signifi-
cantly shorter wing and bill lengths but showed no dif-
ference in fighting abilities compared to replacement
males. However, males with higher capture weights
(whether original or replacement) retained significantly
more territories than lighter males. Neither previous
ownership nor habitat quality had measurable influence on
territory retention.
Signed __
Faculty member in charge of thesis


iv
ACKNOWLEDGMENTS
We thank Michael W. Monahan for permission to use
part of his study area, for access to his data, and for
suggestions throughout this study. Jan Driscoll, Linda
Dixon, and Marc Bekoff provided helpful comments on the
manuscript. Equipment and animal care facilities were
provided by the Department of Biology at the University
of Colorado at Denver. The study was funded by a Gradu-
ate Student Development Award to Dianne Catharine Koshak
from the University of Colorado at Denver.


V
CONTENTS
CHAPTER Page
I. INTRODUCTION ................................ 1
Review of proposed explanations ....... 1
II. METHODS ..................................... 5
III. RESULTS .................................... 11
Morphological measurements and weight .. 11
Aggression ............................... 20
Territory quality, size, and position .. 20
Previous ownership experience ....... 21
Seasonal effects ......................... 22
Multivariate Analysis .................... 23
IV. DISCUSSION ................................ 25
BIBLIOGRAPHY .................................... 33


vi
TABLES
Table Page
1. Morphological Means and Standard Deviations
for 1988 ................................. 12
2. Morphological Means and Standard Deviations
for 1987 ................................... 13
3. Kruskal-Wallis One-Way ANOVA Results: 1988
and 1987 ................................... 14
4. Rank Sum Results for Weight Characteristics:
1988 and 1987 .............................. 17
5. Wilcoxin Signed Rank Results for Weight Data:
1988 and 1987 ............................. 18
6. Compilation of Win/Loss Results for Removal
Males: 1988 and 1987
24


CHAPTER I
INTRODUCTION
Several hypotheses have been proposed to explain
why certain individuals of a species successfully obtain
and hold territories while others do not. These hypo-
theses are based on variables such as RHP (Resource-
Holding Potential), previous ownership, habitat quality
and chance events and circumstances. Since previous
studies have not measured the effects of these variables
concurrently, researchers have been unable to discern
their relative influence. Our study addresses this issue
by comparing the abilities of male Red-winged Blackbirds
(Agelaius phoeniceus) with various RHP levels, tenure
backgrounds, habitat quality, and uncorrelated asymme-
tries to regain territory ownership after they are re-
moved for a designated period.
Review of Proposed Explanations
Researchers have speculated that the ability to
obtain and hold a territory may be based on several
factors rather than any one (Hansen and Rohwer, 1986;
Maynard-Smith, 1976; Parker, 1974; Parker and Rubenstein,
1981). Resource-Holding Potential (RHP) is thought to be


2
the most important factor in determining territorial
ownership in Red-winged Blackbirds (Hansen and Rohwer,
1986). Parker (1974) defined the term Resource-Holding
Potential as a measure of male fighting ability. In a
territorial breeding system governed by RHP, characteris-
tics such as size, fighting experience, aggressiveness,
and weaponry such as bill length determine the outcome of
a territorial contest. RHP theorists presume that in-
creasing ownership time increases RHP (Parker, 1974).
Since only those males with superior RHP are supposedly
dominant enough to defend superior habitat in the RHP
model, the largest and most aggressive males become
residents of superior territories (Parker, 1974; Rohwer,
1982) .
If RHP is a determining factor in territory acguisi-
tion and maintenance, then certain predictions can be
made with respect to territorial male removal experi-
ments; (l) the RHP of original males and replacement
males should be similar because both individuals defend
the territory against encroaching neighbors with constant
RHP abilities (Eckert and Weatherhead, 1987) ; (2) since
the original owners have defeated challengers, they ought
to have higher RHP values than non-territorial males
(Parker, 1974) ; (3) victorious replacement males should
possess the highest RHP levels of the surrounding floater
population (Rohwer, 1982); (4) regardless of territorial


3
ownership status, victorious males should have superior
RHP attributes, and losing males should have inferior RHP
qualities.
Ownership experience on a territory presumably
generates advantages for territorial males since they
return to the same territory each year (Beletsky and
Orians, 1987; Pieman, 1987). Previous ownership produces
knowledge of the intrinsic qualities of a territory which
can be used to defend the territory from challengers
(Krebs, 1982; Wiens, 1973; Wittenberger, 1980). Since no
experimental evidence has been found to indicate that
increasing ownership time improves RHP (Rohwer, 1982),
previous ownership advantages are not directly linked to
RHP superiority.
Superior rather than inferior habitat seems to be
more contested in territorial disputes (Orians, 1961;
Stanton, 1986). If habitat quality is a determining
factor in territorial contests, then the duration or
intensity of a contest should increase with habitat
quality. Habitat quality itself might motivate males to
defend or infiltrate territories (Maynard-Smith, 1976;
Parker and Rubenstein, 1981). In fact, studies have
demonstrated that superior habitat is not necessarily
inhabited by superior RHP males (Eckert and Weatherhead,
1987a; Rohwer, 1982), and thus, RHP ability does not
entirely explain settlement patterns.


4
Uncorrelated asymmetries are also thought to
affect territorial disputes. Krebs (1982) and Maynard-
Smith (1976) defined uncorrelated asymmetries as influ-
ences which can only be explained by circumstantial
events. Random variables such as chance differences in
arrival time on a newly vacated territory may signifi-
cantly affect the outcomes of territorial disputes
(Eckert and Weatherhead, 1987b; Krebs, 1982; Pieman,
1987; Searcy, 1979). Perhaps, the non-territorial male
who is first to capitalize on the disappearance of a
territorial male obtains ownership regardless of his
fighting prowess of ability to assess habitat. Since
uncorrelated asymmetries are ambiguous, they are dif-
ficult or even impossible to measure experimentally.
Although the latest research efforts have eluci-
dated the independence of RHP, previous ownership, habi-
tat quality, and uncorrelated asymmetries (Hansen and
Rohwer, 1986; Rohwer, 1982; Eckert and Weatherhead,
1987a; Eckert and Weatherhead, 1987b), the relative
importance of these variables has not been clarified. By
analyzing these factors simultaneously, we established
the importance of each variable in influencing terri-
torial behavior and settlement patterns in Red-winged
Blackbirds.


CHAPTER II
METHODS
The study area was located 2 km east of
Lafayette, Colorado, and consisted of a 1 ha cattail
marsh (Typhus spp.). bordered by an alfalfa field (west),
pasture (east), abandoned homestead (north) and railroad
track (south). Red-winged Blackbird males appear on ter-
ritories in January, while females appear on territories
in April. First eggs are typically laid around 7-9 May
at the study site (M.W. Monahan, pers. comm.). Wooden
stakes (2 m) were placed in the study area at 20-m inter-
vals to provide a grid for mapping nests and territories.
Stakes were frequently used by males as display perches.
Eleven males were removed in 1987, while 21 males
were removed in 1988. Removed males were chosen randomly
from a pool of 33 (1987) and 32 (1988) banded or unbanded
males at the site. Removal experiments were conducted
three times during the 1987 breeding season, from 8-12
April (n = 4), 2-9 May (n = 3), and 25 May-1 June (n =
4). In 1988, three removal experiments were conducted,
all prior to the start of nesting: 26 March-4 April (n =
6), 12-24 April (n = 7), 25 April-8 May (n = 8).


6
Before removal, the fighting ability of each male
was assessed. A mounted male Red-winged Blackbird speci-
men in full songspread display was placed at the center
of each territory for 15 minutes. Using an aggression
test similar to that of Hansen and Rohwer (1986) and
Rohwer (1982), we scored each male for (1) number of
pecks on the mount, (2) number of times male landed on
\
mount, and (3) number of full songspreads. A weighted
point system similar to Rohwer's (1982) assigned 2 points
for each peck on the mount, 1 point for each landing on
the mount, and 0.5 points for each full songspread. A
quantitative aggression index for each territorial male
was tabulated which included pecks, landings and song-
spreads .
Males were captured with a mousetrap triggered
decoy trap in 1987 (Bray et al., 1975) and Glenhaven or
4-celled treadle traps baited with commercial birdseed
mix in 1988 (L.D. Beletsky, pers. comm.). After wing,
tail, metatarsus, and bill length were measured, males
were removed from the site and housed at the University
of Colorado at Denver campus in individual cages (91 cm x
61 cm x 51 cm). Each day during captivity, males were
given commercial bird seed mix, 20-40 mealworms, bird
gravel and fresh water. Captive birds were maintained on
10-12 hr. daylight periods to match natural photoperiods
during March through May. Males were returned 48-144


7
hours (1987) or 72 hours (1988) later to their terri-
tories. In 1987, removal intervals per bird varied with
respect to how fast replacement males arrived, whereas a
single removal interval was implemented in 1988 to mini-
mize captivity time.
During the interim between each male1s removal
and release, his territory was observed for presence and
behavior of replacement males. In most cases, several
males initially contested each territory, but only one
tended to dominate after 48 hours. If more than one
replacement male retained interim ownership, all males
were considered viable replacements. Once ownership
status was determined, fighting ability indexes and
morphological measurements for replacements were obtained
in the same manner as for the original males.
At time of release, removal males were weighed.
Percentage of body weight loss for various groups of
males or individual males was calculated as follows: (1)
mean weight loss in grams for group divided by mean
capture weight for group, and (2) weight loss in grams
for individual divided by capture weight for individual.
After release, both the ability of removal males to
regain ownership and the time necessary to become rein-
stated were monitored. In the second year, the time
required to recover a territory was classified into three
categories: (1) 1-5 days, (2) 6-10 days, (3) over 10


8
days. The following categories of territorial males were
designated after release: (1) Original males who were
territory owners prior to removal, (2) Replacement males
who became territory owners after the originals were
captured, (3) Victorious males who became the winners in
contests between original and replacement males, and (4)
Losing males who became the losers in contests between
original and replacement males.
Habitat quality of territories was assessed in
1988 only. Superior or inferior habitat was determined
by assigning high (10), moderate (5) or low (0-1) quality
points to the following vegetation variables: cattail
density, cattail height, and density of dead standing
cattail (estimated by density of cattail catkins). The
presence (10) or absence (0) of deep water ( >30 cm) was
also included. This weighted point system was based on
the findings that high cattail density (Lenington, 1980),
moderately high cattail height (Caccamise, 1977; Holm,
1973), high percentage of dead standing cattail (Goddard
and Board, 1979; Robertson, 1972), and deep water level
around cattails (Holm, 1973) may be indicative of super-
ior Red-winged Blackbird habitat.
A habitat quality score was calculated as fol-
lows: (1) each territory was divided into six equal
sectors; (2) each sector or proportions of sectors were
assigned a score using appropriate water and cattail


9
characteristics; (3) these scores were added to obtain a
composite height/water score for the territory; (4) the
entire territory (all six sectors) was given a high (10),
medium (5) or low (0) density score for both cattail and
cattail catkins; (5) the composite height/water score and
the cattail density and cattail catkin density scores
were combined to produce an overall habitat quality
index. As an example, cattail height on one male's
territory was estimated as 4.5 out of 6 sectors tall (45)
and 1.5 out of 6 sectors medium (7.5) and no deep water
(0). Overall cattail density was estimated as high (10)
with catkin density high (10). Therefore, this territory
obtained a habitat quality score of 72.5 points.
Territory sizes for each removal male were deter-
mined during both field seasons. Territory boundaries
from 10-15 sighting records were drawn on graph paper.
Each territory was cut out and weighed on a Fisher Scien-
tific Top-Loading Balance with 0.001 g precision.
Weights were converted to square meters. Territory
position was categorized as border or interior.
Since variations in habitat, ownership status and
trapping methods created unique circumstances for each
field season, the 1987 and 1988 data were analyzed separ-
ately. Nonparametric tests were mostly utilized for
comparisons because variances were heterogeneous (Bailey,
1981). However, discriminant analysis was applicable


10
since methods met basic requirements (Tobachnick, 1983).
We used Kruskal-Wallis One-Way ANOVAs, Rank Sum tests,
and Wilcoxin Signed Rank tests to analyze the importance
of male quality, habitat quality, and territory quality.
Chi-square and Fisher Exact tests were used to inves-
tigate the influence of previous ownership, seasonal
effects, and the influence of weight loss on re-estab-
lishment time (Bailey, 1981). With discriminant analy-
sis, we examined multivariate interrelationships among
the different factors (Tobachnick, 1983). Computations
were completed using both MICROSTAT and SPSS statistical
programs. Statistical results were considered signifi-
cant at P < 0.05.


CHAPTER III
RESULTS
During the 1987 field season, six males success-
fully regained their territories after being removed and
five did not. For 1988, there was a similar trend with
eleven males regaining territories after a 72 hour remo-
val while ten did not. This outcome approximates a 50%
success rate for both original owners and replacement
males at this study area. Morphological, behavioral, and
habitat comparisons became important components of this
study since actual win/loss contest results uncovered no
strict "residents always win" advantage for replacements
or previous ownership advantages for originals.
Morphological Measurements and Weight
Of the morphological measurements, only wing
length, bill length, capture weight, and release weight
seemed to be associated with territory ownership status.
Morphological means are reported in Tables 1 and 2.
All morphological measurements were compared
separately in 1988 and 1987 by ANOVA among original
victors, original losers, replacement victors, and re-
placement losers (Table 3). Capture and release weight


12
Table 1. Morphological means and standard deviations for
1988. Morphological measurements are in millimeters and
weights are in grams. Other abbreviations used: Orig =
original males, Rep = replacement males, OrigVic = ori-
ginal victors, OrigLos = Original losers, etc. See text
for details.
Categories of males n = Wing (ran) Tail (ran) Meta- tarsus (ran) Bill (ran) Capture weight (9) Release weight (9)
All
Orig 134.2 97.1 30.4 16.7 74.6 69.1
21 +2.1 +2.1 +1.0 +1.6 +3.8 +4.4

All
Rep 135.7 97.3 30.9 17.3 78.7 78.7
16 +1.4 +2.0 +1.0 +1.0 +4.0 +4.0

Orig
Viet 134.4 97.0 30.7 16.7 76.3 69.9
11 +1.6 +2.3 +0.8 +1.8 +3.7 +4.4

Orig
Lose 133.9 97.3 30.2 16.7 72.9 68.3
10 +2.6 +2.2 +1.1 +1.3 +3.1 +4.4

Rep
Viet 135.8 97.1 30.8 17.3 77.7 77.7
11 +1.2 +1.8 +1.0 +0.9 +4.5 +4.5

Rep
Lose 135.7 97.6 31.0 17.4 75.4 75.4
7 +1.8 +2.6 +0.9 +1.2 +3.3 +3.3
All
Viet 135.1 97.0 30.7 17.0 76.9 73.6
25 +1.6 +2.0 +0.9 +1.4 +4.0 +5.8

All
Lose 134.6 97.4 30.5 17.0 73.9 71.2
17 +2.4 +2.3 +1.1 +1.3 +3.3 +5.3


13
Table 2. Morphological means and standard deviations for
1987. Morphological measurements are in millimeters and
weights are in grams. Other abbreviations used: Orig =
original males, Rep = replacement males, OrigVic = origi-
nal victors, OrigLos = original losers, etc. See test
for details. **Insufficient Data (Only One Male in
Sample with Weight Data).
Categories Wing Tai l Meta- Bill Capture Release
of males tarsus weight weight
(mm) (run) (urn) (mn) (9) (9)
n =
All
Orig 134.3 98.4 30.7 15.2 72.9 67.8
11 +2.4 +2.5 +0.7 +3.4 +3.4 +29.4

All
Rep 133.6 97.4 30.4 16.6 73.5 73.5
8 +2.4 +2.8 +1.3 +2.8 +1.9 +1.9

Orig
Viet 134.5 98.8 30.7 15.0 74.0 68.8
6 +2.6 +2.7 +0.8 +3.0 +3.7 +3.2
Orig
Lose 134.0 98.0 30.7 15.4 71.6 66.8
5 +2.5 +2.3 +0.7 +4.1 +2.7 +2.5

Rep
Viet 134.2 97.2 30.8 16.5 74.0 74.0
5 +2.7 +2.4 +0.7 +3.3 +2.0 +2.0

Rep
Lose 132.7 97.7 29.7 16.9 72.0 72.0
3 +2.1 +4.0 +1.9 +2.4 + **

All
Viet 134.4 98.1 30.7 15.7 74.0 70.8
13 +2.5 +2.6 +0.7 +3.1 +3.1 +3.7

All
Lose 133.4 97.9 30.3 16.0 71.6 67.7
1 +2.3 +2.8 +1.2 +3.5 +2.4 +3.1


14
Table 3. Results of Kruskal-Wallis One Way ANOVA tests
for four categories of males including original victors,
original losers, replacement victors, and replacement
losers in 1988 and 1987. The first value reported is the
H statistic, and the second is the probability. Early
males were categorized as males removed before nesting
activity began. Abbreviations used: Capwt = capture
weight, Relwt = release weight, Met = metatarsus and Agg
= aggression.
Characteristic All males 1988 All males 1987 Early males 1988 (Round I/II) Early males 1987 (Round I)
Capwt 8.001 2.817 5.826 0.500
0.046 0.421 0.120 0.480
Relwt 19.186 6.198 14.554 1.125
<0.001 0.102 0.002 0.289
Uing 6.280 2.303 2.476 3.781
0.099 0.512 0.480 0.052
Tai l 0.615 1.458 2.961 1.125
0.893 0.692 0.400 0.289
Met 2.644 1.667 3.762 0.500
0.450 0.644 0.289 0.480
Bill 2.347 1.042 2.960 1.125
0.504 0.791 0.398 0.289
Agg 1.871 6.444 2.932 0.029
0.600 0.092 0.402 0.865


15
both differed significantly among the 1988 males (Table
3). To eliminate possible seasonal effects on morpho-
logical comparisons, males from prenesting removal rounds
were compared separately for 1988 and 1987. Release
weight differed significantly among the 1988 males (Table
3). Wing length was barely significant among the 1987
males (Table 3).
Rank Sum tests demonstrated that only original
losers consistently showed significantly smaller wing and
bill lengths than replacements. Originals (both victors
and losers) had significantly shorter wings than replace-
ments (both victors and losers) (z = -2.423, P = 0.01).
Both original victors (z = -1.871, P = 0.03) and losers
(z = 2.042, P = 0.02) had significantly shorter wings
than replacement victors but not replacement losers.
When removal/replacement pairs were examined with Wil-
coxin Signed Rank tests, original males (both victors and
losers) had significantly shorter wings than their re-
spective replacements (z = -2.286, P = 0.01). While
original losers also had significantly shorter wings than
their replacements (z = -1.750, P = 0.04), original
victors did not. The bills of original males (both
victors and losers) were significantly shorter than their
replacements (z = -1.982, P = 0.02). Original losers
also showed significantly shorter bills than their re-
placements (z = -2.100, P = 0.02), while original victors


16
did not. No similar wing or bill length correlations
were found in 1987.
Rank Sum tests indicated that original losers
were significantly lighter in weight at capture than
either original victors or other replacements during the
study. In 1988 but not 1987, all original males showed a
strong tendency to be lighter than all replacements but
only original losers were, in fact, significantly lighter
than all other territorial males (Table 4). In original/
replacement pair comparisons using Wilcoxin Signed Rank
tests, original losers had a strong tendency to be
lighter at capture than their particular replacements in
1987 but no other significant paired analysis trends were
seen in either field season (Table 5).
The results of the Rank Sum tests also show that
original males were significantly lighter at the time of
release than replacement males because they lost weight
while in captivity. For release weight comparisons, the
field weight of replacement males was compared to the
release weight of removal males to measure the effect of
weight loss. In 1988, at time of release, original males
(both victors and losers) were significantly lighter than
replacement males (victors and losers), and original
victors and losers were significantly lighter than both
replacement victors and losers (Table 4). In 1987,


17
Table 4. The results of Rank Sum tests comparing weights
between different categories of males in both 1988 and
1987. The first value for each comparison is the z value
and the second is the associated probability. Other
abbreviations used: Orig = original males, Rep =
replacement males original losers, , OrigVict etc. See = original victors, text for details. OrigLos =
Categor- ies of males compared Capture Weight 1988 (9) Release Weight 1988 (9) Categor- Capture ies of males Weight compared 1987 (9) Release Weight 1987 (9)
AllOrig AllOrig
VS. -1.548 -4.299 VS- -0.588 -2.687
AllRep 0.061 <0.001 AllRep 0.278 0.004
OrigVic OrigVic
VS. 2.007 0.756 VS- 1.095 0.940
OrigLos 0.022 0.225 OrigLos 0.137 0.174
OrigVic OrigVic
VS. -0.722 -3.062 VS- 0.000 -2.087
RepVic 0.235 0.001 RepVic 0.500 0.018
OrigLos OrigLos
VS. -1.610 -3.123 VS. -0.586 -1.464
RepLos 0.054 0.001 RepLos 0.279 0.072
RepVic RepVic
VS. 1.217 1.217 VS. 0.894 0.894
RepLos 0.112 0.112 RepLos 0.185 0.185
OrigLos OrigLos
VS. -2.531 -3.225 VS. -1.342 -2.236
RepVic 0.006 0.001 RepVic 0.090 0.013
OrigVic OrigVic
VS. 0.589 -2.537 VS. 0.500 -1.171
RepLos 0.278 0.006 RepLos 0.309 0.121
AllVic AllVic
VS. 2.362 1.109 VS. 1.473 1.420
AllLos 0.009 0.134 AllLos 0.070 0.078


18
Table 5. The results of Wilcoxin Signed Rank tests
comparing weights in grams between each original male and
his replacement male for both 1988 and 1987. The first
value for each comparison is the z value and the second
is the associated probability. Other abbreviations used:
Orig = original males, Rep = replacement males, OrigVic =
original victors, OrigLos = original losers, etc. See
text for details.
Categor- ies of males compared Capture Ueight 1988 (9) Release Ueight 1988 (9) Categor- ies of males compared Capture Ueight 1987 (9) Release Ueight 1987 (9)
AllOrig AllOrig
VS- -0.724 -3.209 VS. -1.095 -1.826
Their 0.235 0.001 Their 0.137 0.034
Rep Rep
OrigVic OrigVic
VS. -0.051 -2.310 VS- 1.000 -1.000
Thei r 0.480 0.010 Thei r 0.159 0.159
Rep Rep
OrigLos OrigLos
VS- -1.153 -2.201 VS. -1.604 -1.604
Thei r 0.124 0.014 Thei r 0.054 0.054
Rep Rep


19
original males (both victors and losers) were signifi-
cantly lighter than replacement victors but not replace-
ment losers (Table 4). Original males were significantly
lighter than their respective replacements at release
except for original victors in 1987 (Table 5).
In 1988, victors lost an average of 6.4g (7.3%)
of their body weight, and losers lost an average of 4.5g
(6.2%) of their body weight while in captivity. In 1987,
victors lost an average of 4.6g (6.2%) of their body
weight and losers lost an~average of 4.8g (6.7%) of their
body weight. While there was no significant tendency for
high weight loss to be correlated with territory forfei-
ture in 1988 (Rank Sum test, z = 1.209, P = 0.1132), the
small sample size in 1987 precluded this comparison.
There was no correlation using a Chi-Square test between
high weight loss and longer re-establishment time for
original victors in 1988. Lack of data on re-establish-
ment time prevented similar comparisons in 1987.
Ignoring prior residency status (original versus
replacement), the effect of body size on territory reten-
tion was analyzed by comparing weights between victors
and losers. In 1988, victors were significantly heavier
at time of capture than losers (Table 4). There was a
similar tendency, although not significant, for victors
to outweigh losers in 1987 (Table 5). These victor/
loser comparisons along with previous results suggest


20
that capture weight rather than weight loss may be as-
sociated with successful territory domination.
Aggression
Aggression scores did not appear to affect the
ability of males to compete for territories. Aggression
scores for 1987 ranged from 2.0 to 36.2 with a mean of
12.1. Aggression scores for 1988 ranged from 0.0 to 81.5
with a mean of 14.6. Although original loses in 1987
displayed aggression scores that were 50.6% lower than
original victors (Rank Sum test, z = 1.826, P = 0.03), no
other significant differences were found in either 1987
or 1988.
Territory Quality. Size and Position
Habitat quality, territory size, and territory
position (interior vs. exterior) did not seem to affect
the duration or final outcome of territorial disputes.
The habitat quality scores assessed in 1988 ranged from
21.0 to 72.5, with a mean of 54.2. Since nine high
quality territories were regained but seven were for-
feited, there was no evidence that superior quality
habitat was regained more often than inferior quality
habitat.
Territory sizes in 1987 ranged from 22 m2 to 169
m2 with a mean of 80 m2, while in 1988 they ranged from
17 m2 to 115 m2, with a mean of 54 m2. Territories re-


21
gained by the original owners were significantly larger
than those that were successfully defended by replacement
males in 1987 (Rank Sum test, z = 2.008, P = 0.02) but
not in 1988. Fisher Exact Probability tests indicated
there was also no significant tendency for territory
position to affect re-establishment. There were six
males on border and five males on interior territories in
1988, while four victors were on border and two victors
were on interior territories in 1987. Chi-square tests
determined that territory position had no significant
effect on time needed to regain ownership in 1988, and
time to regain ownership was not measured in 1987.
Previous Ownership Experience
Prior residency did not improve the likelihood of
reclaiming a territory. Originals were placed into two
categories: (l) older originals who occupied the same
territory during the previous season, and (2) younger
originals who were holding territories for the first
time. If the residence status of any male was unknown,
it was eliminated from resident effect analysis. There
were 8 older originals who won and 8 older originals who
lost in 1988. In 1987, there were 5 older originals who
won and 4 older originals who lost. Since there were
very few confirmed younger originals in either field
season, sample sizes were not large enough to compare
retention rates of younger originals.


22
There was no tendency for neighbors (1) to be
more common replacements than new males, or (2) to have
higher success in removal/replacement contests. New
replacements were defined as males who had never pre-
viously been seen at the study area, while neighbor re-
placements were males whose territories adjoined the
original males'. In 1988, there were 11 new replacements
who took over vacant territories and 10 neighbors who
expanded onto empty territories. In 1987, four new
replacement males took over vacant territories while
eleven neighbor replacement males expanded onto adjacent
territories, but these results were not significant
(Chi-square tests). Although seven out of ten neighbors
and three out of eleven new males successfully retained
territories in 1988, and two of four new males and four
of eleven neighbors successfully retained territories in
1987, new or neighbor status was not significantly corre-
lated with more successful territory retention (Fisher
Exact Probability tests).
Seasonal Effects
Date of removal may have affected the ability of
original males to regain territories. In 1987, the first
group of males was removed between 8-11 April, the second
group between 2-9 May (which corresponded to start of egg
laying), and the third group between 25 May-1 June (which
corresponded to peak nesting season). In 1987, all males


23
(4 out of 4) regained territories during the pre-nesting
interval, while only two out of five regained nesting and
post-nesting interval territories (Table 6). The small
sample size in 1987 precluded statistical analysis, but
breeding cycle effects could have been a confounding
variable. Although not statistically significant (Fisher
Exact Probability test), four out of six males regained
territories in the pre-nesting period but only seven out
of fifteen regained ownership after nesting activities
were underway in 1988 (Table 6).
Multivariate Analyses
Multivariate analyses of 1988 data determined
that male status was significantly correlated with wing
length and release weight. Discriminant analyses com-
pared males as originals versus replacements or victors
versus losers. Capture or release weight was analyzed
separately to ascertain any debilitating effects of
captivity. In a statistically significant comparison
between originals and replacements (p = .004,n = 31),
wing length was the first predictor variable (cc = 0.97)
and release weight was the second predictor variable (cc
= 0.74). No other significant correlations were found.


24
Table 6. Win/loss results for removal males in 1987 and
1988. Abbreviations used: Orig = original males, Orig-
Vict = original victors, OrigLos = original losers.
Year Interval Removal Period Round Weight Loss Means (9) Orig Males Orig Viet Orig Losers
1987 2-4 day 4-8/4-12 I 4.7+2.4 4 4 0
1987 7 day 5-2/5-9 II 5.0+2.0 3 0 3
1987 7 day 5-25/6-1 III 4.3+1.3 4 2 2
1987 total 4.7+1.8 11 6 5
1988 3 day 3-26/4-8 I 6.8+2.3 6 4 2
1988 3 day 4-12/4-24 11 6.6+3.3 7 4 3
1988 3 day 4-25/5-8 III 3.6+2.7 8 3 5
1988 total 5.4+3.1 21 11 10


CHAPTER IV
DISCUSSION
This investigation uncovered little evidence that
the previously predicted factors of RHP, previous owner-
ship, and habitat quality actually created asymmetrical
advantages that determined contest outcomes. Contrary to
the RHP hypothesis, (1) original males, particularly
original losers, tended to have significantly shorter
wing and bill lengths than replacement males when they
should have been larger or similar; (2) original males
and their respective replacements had significant wing
and bill length disparities when they should have been
similar; (3) morphological characteristics of replacement
victors and losers were not significantly different; and
(4) victors and losers showed no significant difference
in their aggression indexes. The RHP hypothesis was
confirmed only in capture weight analysis since males
(whether original or replacement) with low capture
weights tended to lose territories. Although previous
ownership and habitat quality did not seem to affect
outcomes of contests, seasonal effects and weight loss in
captivity could have obscured the impact of these


26
factors.
The RHP hypothesis predicts that body size is
positively correlated with the ability of males to defend
and maintain territories (Parker, 1974; Rohwer, 1982),
but we did not confirm this assumption. While wing
length is believed to be the most reliable measure of
both body size (Hamilton, 1961; James, 1970; Searcy,
1979) and dominance status (Eckert and Weatherhead,
1987a; Searcy, 1979; and Yasukawa, 1981a), capture weight
more accurately predicted dominance status in this study.
Even if our original males were heavy at the beginning of
the breeding season, the energetic demands of territory
owners (Beletsky et al., 1989) probably resulted in
weight loss. Searcy (1979) found a negative correlation
between increasing age and body weight. This could imply
that some younger males outcompeted older individuals in
our study, if we can assume that younger males had higher
capture weights. Even though capture weight is not a
highly reliable size measurement, it could still rep-
resent an important component of RHP in Red-winged Black-
birds that gives younger and/or non-territorial males an
asymmetrical advantage over older males as the breeding
season progresses.
Other research has also revealed that the mor-
phological characteristics of territorial males cannot be
predicted by the RHP hypothesis. Successful territorial


27
ownership does not seem to imply size similarities bet-
ween either neighboring males (Eckert and Weatherhead,
1987b; Yasukawa, 1981a) or original males and their
respective replacements (Eckert and Weatherhead, 1987b).
There is no evidence that non-territorial males are
smaller than territorial males (Eckert and Weatherhead,
1987c). In this study, significantly heavier and longer-
winged non-territorial males did not successfully oust
smaller original males until they were removed from the
study site. Even if the smaller territory holders were
the largest males at the time they established terri-
tories, the RHP hypothesis does not explain how smaller
males sustained ownership with larger males nearby in the
floater population. The morphological characteristics
associated with RHP apparently do not dictate territory
settlement and maintenance.
As in our study, Other research has indicated
that the high aggressiveness intrinsic to the RHP hypo-
thesis was not characteristic of many successful ter-
ritorial males. Yasukawa (1981b) determined that more
experienced males actually used fewer aggressive be-
haviors than new males except when they initially es-
tablished territory ownership. Dominance was shown to be
a measure-dependent quality by Tomback et al. (1989)
because different measures of behavior, including aggres-
sion, produced different hierarchies in Mule Deer (Odoco-


28
ileus hemionus hemionus). Since dominance status is
known to be energetically expensive (Hogstad, 1987),
Jackson (1988) proposed that extreme overt aggressive
behavior and/or increasing body size become disadvan-
tageous to dominance rank, which is contrary to the RHP
hypothesis. Since successful territory owners were not
always more aggressive or larger than replacement males,
yet weighed more at capture, this study could neither
clearly reject nor confirm the RHP hypothesis.
Our data did not support the hypothesis that
previous ownership of a territory guarantees territory
retention. In our study, eleven out of twenty-one origi-
nal males regained territories in 1987. Jakobsson (1988)
found even less convincing results in a study using
Willow Warblers (Phvlloscopus trochilus) where only four
of eleven retained ownership. In contrast, Beletsky and
Orians (1987) found that fifty of fifty-five original
Red-winged Blackbird males re-established territories.
Since it has been assumed that payoff asymmetries
(Parker, 1974) create large advantages for original
males, as observed by Beletsky and Orians (1987), perhaps
other factors created by the methods utilized in our
experiment produced unforeseen disadvantages for original
males.
Weight loss in captivity was a confounding vari-
able during this study and has also affected other remo-


val experiment results. In our study, the highest mean
body weight loss (7.3%) occurred in 1988 and resulted in
only 52% of males regaining territories. In a study by
Jakobsson (1988), a mean body weight loss of 5.2% cor-
responded to 36% retention. Duration of captivity also
seems to affect weight loss and successful territory
maintenance. While we found a 52% retention using a 72
hour removal period, Beletsky and Orians (1987) found a
90.9% retention rate for Red-winged Blackbirds with a 48
hour removal interval. When longer captivity periods
were used (144-168 hrs.), Beletsky and Orians (1989)
found an 8% retention rate and a 4.9% weight loss. They
determined that captivity periods over 120 hours severely
reduced territory retention ability. During our study,
substantial weight loss and/or excessive time in cap-
tivity did not preclude some males from regaining their
territories, but we cannot rule out the possibility that
captivity contributed to the loss of territories by other
original males.
Since retention abilities of original males may
have decreased as the breeding season continued, it is
possible that motivational changes occurred over time.
Higher risks than benefits may prompt males to avoid re-
establishing territories during the later periods of the
breeding season. Peek (1971) noted that although neigh-
boring males quickly incorporate empty territories early


30
in the breeding season, they often show no replacement
activity at all after peak nesting activity ceases.
Yasukawa (1981b) found that in later stages of the nest-
ing season, yearlings become actively territorial, where-
as territorial males tend to leave the area after their
young have fledged or nests have failed. Although this
study did not produce concrete evidence of altered ter-
ritorial motivation, the effect of time of year cannot be
ruled out.
The contrived nature of the removal experiment
may also have obscured the previous ownership hypothe-
sized for original males. Neighboring males seem to
become more aggressive after individuals are removed
(Krebs, 1982), and when there is a high natural influx of
new males (Pieman, 1987). Krebs (1982) speculated that
it is easier for a male who knows adjoining males to re-
establish ownership since males can predict aggressive
tendencies in their neighbors. A newcomer has to evalu-
ate each neighbor to find out if he is a threat (Krebs,
1982). If replacements dominate territories long enough
to assess their neighbors' combative ability, replacement
males may have acquired the asymmetrical advantage of
original owners (Krebs, 1982). Since these circumstances
would rarely occur in an unmanipulated system, the re-
sults of the removal study may not accurately reflect how
actual takeovers and turnovers occur (Beletsky and


31
Orians, 1987).
The study did not show that superior habitat was
regained more often or contested for longer periods.
There was also no conclusive evidence that high RHP males
occupied high quality territories in this study. Eckert
and Weatherhead (1987a) found that when males were placed
in captivity, the larger and most aggressive individuals
from inferior upland areas had higher dominance ranks
than males from superior marsh locations. Since this
study was the first removal experiment to attempt habitat
assessment, perhaps more emphasis on habitat quality in
future removal studies may help elucidate any effects
that habitat quality has on territorial disputes.
If habitat quality, previous ownership, RHP,
seasonal effects, weight loss, and replacement male
asymmetrical advantage were not determining factors in
the outcome of territorial contests, then other intan-
gible influences presumably had considerable impact.
Other researchers have speculated that site fidelity
might influence males to stay in inferior habitats, even
if they are capable of controlling superior territories
(Eckert and Weatherhead, 1987a; Pieman, 1987; Searcy,
1979). Although site fidelity did not influence re-
establishment in our study, it may enhance re-establish-
ment during the early stages of the breeding season
(Jakobsson, 1988; Pieman, 1987). Robertson (1971) sug-


32
gested that male Red-winged Blackbirds often will utilize
inferior habitat when superior vegetation is already
occupied. If stochastic circumstances create the oppor-
tunity for a male to dominate a territory and decrease
his risks of becoming injured, then chance events may
play a greater role in territory settlement patterns than
previously imagined (Eckert and Weatherhead, 1987a;
Eckert and Weatherhead, 1987b; Pieman, 1987; Searcy,
1979).


33
BIBLIOGRAPHY
Bailey, N.T.J. 1986. Statistical methods in biolocrv.
2nd ed. New York: John Wiley and Sons.
Beletsky, L.D. and Orians, G.H. 1987. Territoriality
among male red-winged blackbirds I. Removal experi-
ments and site dominance. Behav. Ecol. Sociobiol.
20. 339-349.
Beletsky, L.D. and Orians, G.H. 1989. Territoriality
among male red-winged blackbirds II. Testing hypo-
theses of territorial dominance. Behav. Ecol. Socio-
biol. 24. 333-339.
Beletsky, L.D. et al. 1989. Relationships of steroid
hormones and polygyny to territorial status, breeding
experience and reproductive success in red-winged
blackbirds. Auk. 106. 107-117.
Bray, O.L., et al. 1975. A trap for capturing terri-
torial male Red-winged Blackbirds. Western Bird
Bander. 50, 4-7.
Caccamisi, C.G. 1977. Breeding success and nest site
characteristics of the red-winged blackbirds. Wilson
Bull.. 89. 396-403.
Eckert, C.G. and Weatherhead, P.J. 1987a. Ideal
dominance distributions: a test using red-winged


34
blackbirds (Agelaius phoeniceusl. Behav. Ecol.
Sociobiol.. 20, 43-52.
Eckert, C.G. and Weatherhead, P.J. 1987b. Competition
for territories in red-winged blackbirds: is
resource-holding potential realized? Behav. Ecol.
Siociobiol.. 20. 369-375.
Eckert, C.G. and Weatherhead, P.J. 1987c. Owners,
floaters and competitive asymmetries among terri-
torial red-winged blackbirds. Anim. Behav.. 35,
1317-1323.
Goddard, S.V. and Board, V.V. 1979. Reproductive
success of red-winged blackbirds in North Central
Oklahoma. Wilson Bull.. 79, 283-289.
Hamilton, T.H. 1961. The adaptive significance of
intra-specific trends of variation in wing length and
body size among bird species. Evolution. 15. ISO-
195.
Hansen, A.J. and Rohwer, S. 1986. Coverable badges and
resource defence in birds. Anim. Behav.. 34. 69-76.
Hogstad, 0. 1987. It is expensive to be dominant. Auk.
104. 333-334.
Holm, C.H. 1973. Breeding sex ratios, territoriality
and reproductive success in the red-winged blackbird.
Ecology. 54. 356-365.


35
Jackson, W.M. 1988. Can individual differences in
history of dominance explain the development of
linear dominance hierarchies? Ethology. 79, 71-77.
Jakobsson, S. 1988. Territorial fidelity of willow
warbler (Phvlloscopus trochilus) males and success in
competition over territories. Behav. Ecol. Socio-
biol., 22, 79-84.
James, F.C. 1970. Geographic size variation in birds
and its relationship to climate. Ecology. 51, 365-
390.
Krebs, J.R. 1982. Territorial defence in the great tit
(Parus manor). Do residents always win? Behav.
Ecol. Sociobiol.. 11. 185-194.
Lenington, S. 1980. Female choice and polygyny in red-
winiged blackbirds. Anim. Behav.. 28. 347-361.
Maynard-Smith, J. 1976. Evolution and the theory of
games. Amer. Sci.. 64. 41-45.
Orians, G.H. 1961. The ecology of blackbird (Agelaius)
social systems. Ecol. Monoor.. 31. 85-312.
Parker, G.A. 1974. Assessment strategy and the evolu-
tion of fighting behavior. J. Theor. Biol.. 47, 23-
243.
Parker, G.A. and Rubenstein, D.I. 1981. Role assess-
ment, reserve strategy, and acquisition of informa-
tion in asymmetric animal contents. Anim. Behav..
29, 221-240.


36
Peek, F.W. 1971. Seasonal change in the breeding
behavior of the male red-winged blackbird. Will.
Bull.. 83, 383-395.
Pieman, J. 1987. Territory establishment, size, and
tenacity by male red-winged blackbirds. Auk. 104.
405-412.
Robertson, R. 1972. Optimal niche space of the red-
winged blackbird I. Nesting success in marsh and
upland habitat. Canadian J. Zool.. 50, 247-263.
Rohwer, S. 1982. The evolution of reliable and unreli-
able badges of fighting ability. Amer. Zool.. 22.,
531-546.
Searcy, W.A. 1979. Male characteristics and pairing
success in red-winged blackbirds. Auk. 96, 353-363.
Stanton, P.A. 1986. Comparison of avian community dyna-
mics of burned and unburned coastal sage scrub.
Condor. 88, 285-289.
Tobachnick, B.G. and Fidel1, L. 1983. Multivariate
Statistics. New York: Harper and Row.
Tomback, D.F. et al. 1989. Measuring dominance and
constructing hierarchies: An example using Mule
Deer. Ethology. 82, 275-286.
Wiens, J.A. 1973. Interterritorial habitat variation in
grasshopper and savannah sparrows. Ecology. 54, 87-
884


37
Wittenberger, J.F. 1980. Animal social behavior.
Boston: Duxbury Press.
Yasukawa, K. 1981a. Male quality and female choice of
mate in the red-winged blackbird (Agelaius phoe-
niceus). Ecology. 62, 922-929.
Yasukawa, K. 1981b. Territory establishment in red-
winged blackbirds: importance of aggressive behavior
and experience. Condor. 81. 258-264.


Full Text

PAGE 1

TERRITORY RETENTION IN RED-WINGED BLACKBIRDS: THE ROLE OF RHP, PREVIOUS OWNERSHIP AND HABITAT QUALITY by Dianne Catharine Koshak B.A., University of Denver, 1982 University of Denver, 1983 B.A. University of Denver, 1984 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Department of Biology 1990

PAGE 2

This thesis for the Master of Arts degree by Dianne Catharine Koshak has been approved for the Department of Biology by Date

PAGE 3

iii Koshak, Dianne Catharine (M.A., Biology) Territory Retention in Red-Winged Blackbirds: The Role of RHP, Previous ownership and Habitat Quality Thesis directed by Associate Professor, Diana F. Tomback In a removal/replacement experiment, the effects of Resource-Holding Potential (RHP), habitat quality, and previous ownership were measured concurrently to assess their relative influence on territory ownership by marshnesting Red-winged Blackbirds (Agelaius phoeniceus). According to the RHP hypothesis, original territorial males should be larger and more aggressive than replacement males. We found that original males had significantly shorter wing and bill lengths but showed no difference in fighting abilities compared to replacement males. However, males with higher capture weights (whether original or replacement) retained significantly more territories than lighter males. Neither previous ownership nor habitat quality had measurable influence on territory retention. Signed is

PAGE 4

iv ACKNOWLEDGMENTS We thank Michael w. Monahan for permission to use part of his study area, for access to his data, and for suggestions throughout this study. Jan Driscoll, Linda Dixon, and Marc Bekoff provided helpful comments on the manuscript. Equipment and animal care facilities were provided by the Department of Biology at the University of Colorado at Denver. The study was funded by a Graduate Student Development Award to Dianne Catharine Koshak from the University of Colorado at Denver.

PAGE 5

CONTENTS CHAPTER I. INTRODUCTION ....................... Page 1 Review of proposed explanations . 1 I I METHODS 5 III. IV. RESULTS 11 Morphological measurements and weight 11 . . . . . . . . 2 0 Territory quality, size, and position.. 20 Previous ownership experience ..... 21 Seasonal effects .................. 22 Multivariate Analysis ................ 23 DISCUSSION ................. 25 BIBLIOGRAPHY . . . . . . . . . . . . . 3 3 v

PAGE 6

TABLES Table 1. Morphological Means and Standard Deviations for 1988 . . . . . . . . . . . . . . . . 12 2. Morphological Means and Standard Deviations for 1987_ . . . . . . . . 13 3. Kruskal-Wallis One-Way ANOVA Results: 1988 and 19 8 7 . . . . . . . . . . . . . . . . . 14 4. Rank Sum Results for Weight Characteristics: 1988 and 1987 . . . . . . . . . . . . . . 17 5. Wilcoxin Signed Rank Results for Weight Data: 1988 and 1987 .. . . . . . . . . . . . . 18 6. Compilation of Win/Loss Results for Removal Males: 1988 and 1987 ................. 24 vi

PAGE 7

CHAPTER I INTRODUCTION several hypotheses have been proposed to explain why certain individuals of a species successfully obtain and hold territories.while others do not. These hypotheses are based on variables such as RHP (ResourceHolding Potential), previous ownership, habitat quality and chance events and circumstances. Since previous studies have not measured the effects of these variables concurrently, researchers have been unable to discern their relative influence. Our study addresses this issue by comparing the abilities of male Red-winged Blackbirds (Agelaius phoeniceus) with various RHP levels, tenure backgrounds, habitat quality, and uncorrelated asymmetries to regain territory ownership after they are removed for a designated period. Review of Proposed Explanations Researchers have speculated that the ability to obtain and hold a territory may be based on several factors rather than any one (Hansen and Rohwer, 1986; Maynard-Smith, 1976; Parker, 1974; Parker and Rubenstein, 1981). Resource-Holding Potential (RHP) is thought to be

PAGE 8

2 the most important factor in determining territorial ownership in Red-winged Blackbirds (Hansen and Rohwer, 1986). Parker (1974) defined the term Resource-Holding Potential as a measure of male fighting ability. In a territorial breeding system governed by RHP, characteristics such as size, fighting experience, aggressiveness, and weaponry such as bill length determine the outcome of a territorial contest. RHP theorists presume that increasing ownership time increases RHP (Parker, 1974). Since only those males with superior RHP are supposedly dominant enough to defend superior habitat in the RHP model, the largest and most aggressive males become residents of superior territories (Parker, 1974; Rohwer, 1982). If RHP is a determining factor in territory acquisition and maintenance, then certain predictions can be made with respect to territorial male removal experiments: (1) the RHP of original males and replacement males should be similar because both individuals defend the territory against encroaching neighbors with constant RHP abilities (Eckert and Weatherhead, 1987); (2) since the original owners have defeated challengers, they ought to have higher RHP values than non-territorial males (Parker, 1974); (3) victorious replacement males should possess the highest RHP levels of the surrounding floater population (Rohwer, 1982); (4) regardless of territorial

PAGE 9

3 ownership status, victorious males should have superior RHP attributes, and losing males Should have inferior RHP qualities. ownership experience on a territory presumably generates advantages for territorial males since they return to the same territory each year (Beletsky and Orians, 1987; Picman, 1987). Previous ownership produces knowledge of the intrinsic qualities of a territory which can be used to defend the territory from challengers (Krebs, 1982; Wiens, 1973; Wittenberger, 1980). Since no experimental evidence has been found to indicate that increasing ownership time improves RHP (Rohwer, 1982), previous ownership advantages are not directly linked to RHP superiority. superior rather than inferior habitat seems to be more contested in territorial disputes (Orians, 1961; Stanton, 1986). If habitat quality is a determining factor in territorial contests, then the duration or intensity of a contest should increase with habitat quality. Habitat quality itself might motivate males to defend or infiltrate territories (Maynard-Smith, 1976; Parker and Rubenstein, 1981). In fact, studies have demonstrated that superior habitat is not necessarily inhabited by superior RHP males (Eckert and Weatherhead, 1987a; Rohwer, 1982), and thus, RHP ability does not entirely explain settlement patterns.

PAGE 10

Uncorrelated asymmetries are also thought to affect territorial disputes. Krebs (1982) and MaynardSmith (1976) defined uncorrelated asymmetries as influences which can only be explained by circumstantial events. Random variables such as chance differences in arrival time on a newly vacated territory may significantly affect the outcomes of territorial disputes (Eckert and Weatherhead, 1987b; Krebs, 1982; Picman, 1987; Searcy, 1979). Perhaps, the non-territorial male who is first to capitalize on the disappearance of a territorial male obtains ownership regardless of his fighting prowess or ability to assess habitat. Since uncorrelated asymmetries are ambigu.ous, they are difficult or even impossible to measure experimentally. 4 Although the latest research efforts have elucidated the independence of RHP, previous ownership, habitat quality, and uncorrelated asymmetries (Hansen and Rohwer, 1986; Rohwer, 1982; Eckert and Weatherhead, 1987a; Eckert and Weatherhead, 1987b), the relative importance of these variables has not been clarified. By analyzing these factors simultaneously, we established the importance of each variable in influencing territorial behavior and settlement patterns in Red-winged Blackbirds.

PAGE 11

CHAPTER II METHODS The study area was located 2 km east of Lafayette, Colorado, and consisted of a 1 ha cattail marsh (Typhus RR.), bordered by an alfalfa field (west), pasture (east), abandoned homestead (north) and railroad track (south). Red-winged Blackbird males appear on territories in January, while females appear on territories in April. First eggs are typically laid around 7-9 May at the study site (M.W. Monahan, pers. comm.). Wooden stakes (2 m) were placed in the study area at 20-m intervals to provide a grid for mapping nests and territories. Stakes were frequently used by males as display perches. Eleven males were removed in 1987, while 21 males were removed in 1988. Removed males were chosen randomly from a pool of 33 (1987) and 32 (1988) banded or unbanded males at the site. Removal experiments were conducted three times during the 1987 breeding season, from 8-12 April (n = 4), 2-9 May (n = 3), and 25May-1 June (n = 4). In 1988, three removal experiments were conducted, all prior to the start of nesting: 26 March-4 April (n = 6), 12-24 April (n = 7), 25 April-8 May (n = 8).

PAGE 12

6 Before removal, the fighting ability of each male was assessed. A mounted male Red-winged Blackbird speci-men in full songspread display was placed at the center of each territory for 15 minutes. Using an aggression test similar to that of Hansen and Rohwer (1986) and Rohwer (1982), we scored each male for (1) number of pecks on the mount, {2) number of times male landed on \ mount, and (3) number of full songspreads. A weighted point system similar to Rohwer's (1982) assigned 2 points for each peck on the mount, 1 point for each landing on the mount, and 0.5 points for each full songspread. A quantitative aggression index for each territorial male was tabulated which included pecks, landings and song-spreads. Males were captured with a mousetrap triggered decoy trap in 198_7 (Bray et al., 1975) and Glenhaven or 4-celled treadle traps baited with commercial birdseed mix in 1988 (L.D. Beletsky, pers. comm.). After wing, tail, metatarsus, and bill length were measured, males were removed from the site and housed at the University of Colorado at Denver campus in individual cages (91 em x 61 em x 51 em). Each day during captivity, males were given commercial bird seed mix, 20-40 mealworms, bird gravel and fresh water. Captive birds were maintained on 10-12 hr. daylight periods to match natural photoperiods during March through May. Males were returned 48-144

PAGE 13

hours (1987) or 72 hours (1988) later to their territories. In 1987, removal intervals per bird varied with respect to how fast replacement males arrived, whereas a single removal interval was implemented in 1988 to minimize captivity time. 7 During the interim between each male's removal and release, his territory was observed for presence and behavior of replacement males. In most cases, several males initially contested each territory, but only one tended to dominate after 48 hours. If more than one replacement male retained interim ownership, all males were considered viable replacements. Once ownership status was determined, fighting ability indexes and morphological measurements for replacements were obtained in the same manner as for the original males. At time of release, removal males were weighed. Percentage of body weight loss for various groups of males or individual males was calculated as follows: (1) mean weight loss in grams for group divided by mean capture weight for group, and (2) weight loss in grams for individual divided by capture weight for individual. After release, both the ability of removal males to regain ownership and the time necessary to become reinstated were monitored. In the second year, the time required to recover a territory was classified into three categories: (1) 1-s-days, (2) 6-10 days, (3) over 10

PAGE 14

8 days. The following categories of territorial males were designated after. release: (1) Original males who were territory owners prior to removal, (2) Replacement males who became territory owners after the originals were captured, .(3) Victorious males who became the winners in contests between original and replacement males, and (4) Losing males who became the losers in contests between original and replacement: males. Habitat quality of territories was assessed in 1988 only. Superior or inferior habitat was determined by assigning high (10), moderate (5) or low (0-1) quality points to the following vegetation variables: cattail density, cattail height, and density of dead standing cattail (estimated by density of cattail catkins). The presence (10) or absence (0) of deep water ( >30 em) was also included. This weighted point system was based on the findings that high cattail density (Lenington, 1980), moderately high cattail height (Caccamise, 1977; Holm, 1973), high percentage of dead standing cattail (Goddard and Board, 1979; Robertson, 1972), and deep water level around cattails (Holm, 1973) may be indicative of superior Red-winged Blackbird habitat. A habitat quality score was calculated as follows: (1) each territory was divided into six equal sectors; (2) each sector or proportions of sectors were assigned a score using appropriate water and cattail

PAGE 15

9 characteristics; (3) these scores were added to obtain a composite heightjwater score for the territory; (4) the entire territory (all six sectors) was given a high (10), medium (5) or low (0) density score for both cattail and cattail catkins; (5) the composite heightjwater score and the cattail density and cattail catkin density scores were combined to produce an overall habitat quality index. As an example, cattail height on one male's territory was estimated as 4.5 out of 6 sectors tall (45) and 1.5 out of 6 sectors medium (7.5) and no deep water (0). overall density was estimated as high (10) with catkin density high (10). Therefore, this territory obtained a habitat quality score of 72.5 points. Territory sizes for each removal male were determined during both field seasons. Territory boundaries from 10-15 sighting records were drawn on graph paper. Each territory was cut out and weighed on a Fisher Scientific Top-Loading Balance with 0.001 g precision. Weights were converted to square meters. Territory position was categorized as border or interior. Since variations in habitat, ownership status and trapping methods created unique circumstances for each field season, the 1987 and 1988 data were analyzed separately. Nonparametric tests were mostly utilized for comparisons because variances were heterogeneous (Bailey, 1981). However, discriminant analysis was applicable

PAGE 16

10 since methods met.basic requirements (Tobachnick, 1983). We used Kruskal-Wallis One-Way ANOVAs, Rank Sum tests, and Wilcoxin Signed Rank tests to analyze the importance of male quality, habitat quality, and territory quality. Chi-square and Fisher Exact tests were used to investigate the influence of previous ownership, seasonal effects, and the influence of weight loss on re-establishment time (Bailey,. 1981). With discriminant analysis, we examined multivariate interrelationships among the different factors (Tobachnick, 1983). Computations were completed using both MICROSTAT and SPSS statistical programs. Statistical results were considered significant at P < 0.05.

PAGE 17

CHAPTER III RESULTS During the 1987 field season, six males successfully regained their territories after being removed and five did not. For 1988, there was a similar trend with eleven males regaining territories after a 72 hour removal while ten did not. This outcome approximates a 50% success rate for both original owners and replacement males at this study area. Morphological, behavioral, and habitat comparisons became important components of this study since actual winjloss contest results uncovered no strict "residents always win" advantage for replacements or previous ownership advantages for originals. Morphological Measurements and Weight Of the morphological measurements, only wing length, bill length, capture weight, and release weight seemed to be associated with territory ownership status. Morphological means are reported in Tables 1 and 2. All morphological measurements were compared separately in 1988 and 1987 by ANOVA among original victors, original losers, replacementvictors, andreplacement losers (Table 3). Capture and release weight

PAGE 18

12 Table 1. Morphological means and standard deviations for 1988. Morphological measurements are in millimeters and weights are in grams. Other abbreviations used: Orig = original males, Rep = replacement males, OrigVic = original victors, OrigLos = Original losers, etc. See text for details. Categories of males n = ALL Orig 21 ALL Rep 16 Orig Viet 11 Orig Lose 10 Rep Viet 11 Rep Lose 7 ALL Viet 25 ALL Lose 17 Wing (nm) 134.2 :!:2.1 135.7 :!:1.4 134.4 :!:1.6 133.9 .:!:,2.6 135.8 .:!:,1.2 135.7 .:!:,1.8 135.1 :!:1.6 134.6 .:!:,2.4 Tail (nm) 97.1 :!:2.1 97.3 .:!:,2.0 97.0 .:!:,2.3 97.3 .:!:,2.2 97.1 .:!:,1.8 97.6 :!:2.6 97.0 :!:2.0 97.4 .:!:,2.3 Metatarsus (nm) 30.4 :!:1.0 30.9 .:!:,1.0 30.7 .:!:,0.8 30.2 .:!:,1.1 30.8 .:!:,1.0 31.0 :!:0.9 30.7 :!:0.9 30.5 .:!:,1.1 BiLL (nm) 16.7 :!:1.6 17.3 .:!:,1.0 16.7 .:!:,1.8 16.7 :!:1.3 17.3 .:!:,0.9 17.4 .:!:,1.2 17.0 .:!:,1.4 17.0 .:!:,1.3 Capture weight (g) 74.6 :!:3.8 78.7 :!:4.0 76.3 :!:3.7 72.9 :!:3.1 77.7 .:!:,4.5 75.4 .:!:,3.3 76.9 .:!:,4.0 73.9 .:!:,3.3 Release weight (g) 69.1 :!:,4.4 78.7 :!:4.0 69.9 :!:,4.4 68.3 :!:4.4 77.7 :!:,4.5 75.4 .:!:,3.3 73.6 .:!:,5.8 71.2 :!:5.3

PAGE 19

13 Table 2. Morphological means and standard deviations for 1987. Morphological measurements are in millimeters and weights are in grams. Other abbreviations used: Orig = original males, Rep = replacement males, OrigVic = original victors, origLos = original losers, etc. See test for details. **Insufficient Data (Only One Male in Sample with Weight Data). Categories of males n = All Orig 11 All Rep 8 Orig Viet 6 Orig Lose 5 Rep Viet 5 Rep Lose 3 All Viet 13 All Lose 1 Wing Cnm> 134.3 ;!:2.4 133.6 ;!:2.4 134.5 ;!:2.6 134.0 ;!:2.5 134.2 ;!:2.7 132.7 ;!:2.1 134.4 ;!:2.5 133.4 ;!:2.3 Tail Cnm) 98.4 ;!:2.5 97.4 ;!:2.8 98.8 ;!:2.7 98.0 ;!:2.3 97.2 ;!:2.4 97.7 ;!:4.0 98.1 ;!:2.6 97.9 ;!:2.8 Meta tarsus .Cnm) 30.7 ;!:0.7 30.4 ;!:1.3 30.7 ;!:0.8 30.7 +0.7 30.8 ;!:0.7 29.7 ;!:1.9 30.7 ;!:0.7 30.3 ;!:1.2 Bill Cnm) 15.2 ;!:3.4 16.6 ;!:2.8 15.0 ;!:3.0 15.4 +4.1 16.5 ;!:3.3 16.9 ;!:2.4 15.7 ;!:3.1 16.0 ;!:3.5 Capture weight (g) n.9 ;!:3.4 73.5 ;!:1.9 74.0 ;!:3.7 71.6 ;!:2.7 74.0 ;!:2.0 n.o + ** 74.0 ;!:3.1 71.6 ;!:2.4 Release weight (g) 67.8 :!:29.4 73.5 ;!:1.9 68.8 ;!:3.2 66.8 ;!:2.5 74.0 ;!:2.0 72.0 + ** 70.8 ;!:3.7 67.7 ;!:3.1

PAGE 20

14 Table 3. Results of Kruskal-Wallis One Way ANOVA tests for four categories of males including original victors, original losers, replacement victors, and replacement losers in 1988 and 1987. The first value reported is the H statistic, and the second is the probability. Early males were categorized asmales removed before nesting activity began. Abbreviations used: Capwt = capture weight, Relwt = release weight, Met = metatarsus and Agg = aggression. Characteristic All males All males Early males Early males 1988 1987 1988 1987 (Round I/II) (Round I) Capwt 8.001 2.817 5.826 0.500 0.046 0.421 0.120 0.480 Relwt 19.186 6.198 14.554 1.125 <0.001 0.102 0.002 0.289 Wing 6.280 2.303 2.476 3.781 0.099 0.512 0.480 0.052 Tail 0.615 1.458 2.961 1.125 0.893 0.692 0.400 0.289 Met 2.644 1.667 3.762 0.500 0.450 0.644 0.289 0.480 Bill 2.347 1.042 2.960 1.125 0.504 0.791 0.398 0.289 Agg 1.871 6.444 2.932 0.029 0.600 0.092 0.402 0.865

PAGE 21

15 both differed significantly among the 1988 males (Table 3). To eliminate possible seasonal effects on morphological comparisons, males from prenesting removal rounds were compared separately for 1988 and 1987. Release weight differed significantly among the 1988 males (Table 3). Wing length was barely significant among the 1987 males (Table 3). Rank Sum tests demonstrated that only original losers consistently showed significantly smaller wing and bill lengths than replacements. Originals (both victors and losers) had significantly shorter wings than replacements (both victors and losers) (z = -2.423, P = 0.01). Both original victors (z = -1.871, P = 0.03) and losers (z = 2.042, P = 0.02) had significantly shorter wings than replacement victors but not replacement losers. When removal/replacement pairs were examined with Wilcoxin Signed Rank tests, original males (both victors and losers) had significantly shorter wings than their respective replacements (z = -2.286, P = 0.01). While original losers also had significantly shorter wings than their replacements (z = -1.750, P = 0.04), original victors did not. The bills of original males (both victors and losers) were significantly shorter than their replacements (z = -1.982, P = 0.02). Original losers also showed significantly shorter bills than their replacements (z = -2.100, P = 0.02), while original victors

PAGE 22

did not. No similar wing or bill length correlations were found in 1987. 16 Rank Sum tests indicated that original losers were significantly lighter in weight at capture than either original victors or other replacements during the study. In 1988 but not 1987, all original males showed a strong tendency to be lighter than all replacements but only original losers were, in fact, significantly lighter than all other territorial males (Table 4). In original/ replacement pair comparisons using Wilqoxin Signed Rank tests, original losers had a strong tendency to be lighter at capture than their particular replacements in 1987 but no other significant paired analysis trends were seen in either field season (Table 5). The results of the Rank Sum tests also show that original males were significantly lighter at the time of release than replacement males because they lost weight while in captivity. For release weight comparisons, the field weight of replacement males was compared to the release weight of removal males to measure the effect of weight loss. In 1988, at time of release, original males (both victors and losers) were significantly lighter than replacement males (victors and losers), and original victors and losers were significantly lighter than both replacement victors and losers (Table 4). In 1987,

PAGE 23

17 Table 4. The results of Rank Sum tests comparing weights between different categories of males in both 1988 and 1987. The first value for each comparison is the z value and the second is the associated probability. Other abbreviations used: Orig = original males, Rep = replacement males, origVict = original victors, OrigLos = original losers, etc. See text for details. CategorCapture Release categorCapture Release ies of males Weight Weight ies of males Weight Weight compared 1988 1988 1987 1987 (g) (g) (g) (g) AllOrig AllOrig vs. -1.548 -4.299 vs. -0.588 -2.687 All Rep 0.061 <0.001 All Rep 0.278 0.004 OrigVic OrigVic vs. 2.007 0.756 vs. 1.095 0.940 OrigLos 0.022 0.225 OrigLos 0.137 0.174 OrigVic OrigVic vs. -0.722 -3.062 vs. 0.000 -2.087 RepVic 0.235 0.001 RepVic 0.500 0.018 OrigLos OrigLos vs. -1.610 -3.123 vs. -0.586 -1.464 RepLos 0.054 0.001 RepLos 0.279 0.072 RepVic RepVic vs. 1.217 1.217 vs. 0.894 0.894 RepLos 0.112 0.112 RepLos 0.185 0.185 OrigLos Origlos vs. -2.531 -3.225 vs. -1.342 -2.236 RepVic 0.006 0.001 RepVic 0.090 0.013 OrigVic OrigVic vs. 0.589 -2.537 vs. 0.500 -1.171 Replos 0.278 0.006 Replos 0.309 0.121 All Vic All Vic vs. 2.362 1.109 vs. 1.473 1.420 All los 0.009 0.134 All los 0.070 0.078

PAGE 24

18 Table 5. The results of Wilcoxin Signed Rank tests comparing weights in grams between each original male and his replacement male for both 1988 and 1987. The first value for each comparison is the z value and the second is the associated probability. Other abbreviations used: Orig = original males, Rep = replacement males, OrigVic = original victors, OrigLos = original losers, etc. See text for details. CategorCapture Release CategorCapture Release ies of males Weight Weight ies of males Weight Weight compared 1988 1988 1987 1987 (g) (g) (g) (g) AllOrig AllOrig vs. -0.724 -3.209 vs. -1.095 -1.826 Their 0.235 0.001 Their 0.137 0.034 Rep Rep OrigVic OrigVic vs. -0.051 -2.310 vs. 1.000 -1.000 Their 0.480 0.010 Their 0.159 0.159 Rep Rep Origlos OrigLos vs. -1.153 -2.201 vs. -1.604 -1.604 Their 0.124 0.014 Their 0.054 0.054 Rep Rep

PAGE 25

19 original males (both victors and losers) were significantly lighter than replacement victors but not replacement losers (Table 4). original males were significantly lighter than their respective replacements at release except for original victors in 1987 (Table 5). In 1988, victors lost an average of 6.4g (7.3%) of their body weight, and losers lost an average of 4.5g (6.2%) of their body weight while in captivity. In 1987, victors lost an average of 4.6g (6.2%) of their body weight and losers lost an-average of 4.8g (6.7%) of their body weight. While there was no significant tendency for high weight loss to be correlated with territory forfeiture in 1988 (Rank Sum test, z = 1 . 209, P = 0 . 1132), the small sample size in 1987 precluded this comparison. There was no correlation using a Chi-Square test between high weight loss and longer re-establishment time for original victors in 1988. Lack of data on re-establishment time prevented similar comparisons in 1987. Ignoring prior residency status (original versus replacement), the effect of body size on territory reten-. tion was analyzed by comparing weights between victors and losers. In 1988, victors were significantly heavier at time of capture than losers (Table 4). There was a similar tendency, although not significant, for victors to outweigh losers in 1987 (Table 5). These victor; loser comparisons along with previous results suggest

PAGE 26

that capture weight rather than weight loss may be associated with successful territory domination. Aggression 20 Aggression scores did not appear to affect the ability of males to compete for territories. Aggression scores for 1987 ranged from 2.0 to 36.2 with a mean of 12.1. Aggression scores for 1988 ranged from 0.0 to 81.5 with a mean of 14.6. Although original loses in 1987 displayed aggression scores that were 50.6% lower than original victors (Rank sum test, z = 1.826, P = 0.03), no other significant differences were found in either 1987 or 1988. Territory Quality, Size and Position Habitat quality, territory size, and territory position (interior vs. exterior) did not seem to affect the duration or final outcome of territorial disputes. The habitat quality s.cores assessed in 1988 ranged from 21.0 to 72.5, with a mean of 54.2. Since nine high quality territories were regained but seven were forfeited, there was no evidence that superior quality habitat was regained more often than inferior quality habitat. Territory sizes in 1987 ranged from 22 m2 to 169 m2 with a mean of 80 m2 while in 1988 they ranged from 17 m2 to 115 m2 with a mean of 54 m2 Territories re-

PAGE 27

21 gained by the original owners were significantly larger than those that were successtully defended by replacement males in 1987 (Rank Sum test, z = 2.008, P = 0.02) but not in 1988. Fisher Exact Probability tests indicated there was also no significant tendency for territory position to affect re-establishment. There were six males on border and five males on interior territories in 1988, while four victors were on border and two victors were on interior territories in 1987. Chi-square tests determined that territory position had no significant effect on time needed to regain ownership in 1988, and time to regain ownership was not measured in 1987. Previous ownership Experience Prior residency did not improve the likelihood of reclaiming a territory. Originals were placed into two categories: (1) older originals who occupied the same territory during the previous season, and (2) younger originals who were holding territories for the first time. If the residence status of any male was unknown, it was eliminated from resident effect analysis. There were 8 older originals who won and 8 older originals who lost in 1988. In 1987, there were 5 older originals who won and 4 older originals who los.t. Since there were very few confirmed younger originals in either field season, sample sizes were not large enough to compare retention rates of younger originals.

PAGE 28

22 There was no tendency for neighbors (1) to be more commonreplacements than new males, or (2) to have higher success in removal/replacement contests. New replacements were defined as males who had never previously been seen at the study area, while neighbor replacements were males whose territories adjoined the original males'. In 1988, there were 11 new replacements who took over vacant territories and 10 neighbors who expanded onto empty territories. In 1987, four new replacement males took over vacant territories while eleven neighbor replacement males expanded onto adjacent territories, but these results were not significant (Chi-square tests). Although seven out of ten neighbors and three out of eleven riew males successfully retained territories in 1988, and two of four new males and four of eleven neighbors successfully retained territories in 1987, new or neighbor status was not significantly correlated with more successful territory retention (Fisher Exact Probability tests). Seasonal Effects Date of removal may have affected the ability of original males to regain territories. In 1987, the first group of males was removed between 8-11 April, the second group between 2-9 May (which corresponded to start of egg laying), and the third group between 25 May-1 June (which corresponded to peak nesting season). In 1987, all males

PAGE 29

23 (4 out of 4) regained territories during the pre-nesting interval, while only two out of five regained nesting and post-nesting interval territories (Table 6). The small sample size in 1987 precluded statistical analysis, but breeding cycle effects could have been a confounding variable. Although not statistically significant (Fisher Exact Probability test), four out of six males regained territories in the pre-nesting period but only seven out of fifteen regained ownership afternesting activities were underway in 1988 (Table 6). Multivariate Analyses Multivariate analyses of 1988 data determined that male status was significantly correlated with wing length and release weight. Discriminant analyses compared males as originals versus replacements or victors versus losers. Capture or release weight was analyzed separately to ascertain any debilitating effects of captivity. In a statistically significant comparison between originals and replacements (p = .004,n = 31), wing length was the first predictor variable (cc = 0.97) and release weight was the second predictor variable (cc = 0.74). No other significant correlations were found.

PAGE 30

24 Table 6. Win/loss results for removal males in 1987 and 1988. Abbreviations used: Orig = original males, Orig-Viet = original victors, OrigLos = original losers. Year Interval Removal Round Weight Orig Orig Orig Period Loss Males Viet Losers Means (g) 1987 2-4 day 4-8/4-12 4.7:!:2.4 4 4 0 1987 7 day 5-2!5-9 II 5.0:!:2.0 3 0 3 1987 7 day 5/6-1 III 4.3:!:1.3 4 2 2 1987 total 4.7:!:1.8 11 6 5 1988 3 day 3/4-8 6.8:!:2.3 6 4 2 1988 3 day II 6.6;!:3.3 7 4 3 1988 3 day 4/5-8 III 3.6:!:2.7 8 3 5 1988 total 5.4:!:3.1 21 11 10

PAGE 31

CHAPTER IV DISCUSSION This investigation uncovered little evidence that the previously predicted factors of RHP, previous ownership, and habitat quality actually created asymmetrical advantages that determined contest outcomes. Contrary to the RHP hypothesis, {1) original males, particularly original losers, tended to have significantly shorter wing and bill lengths than replacement males when they should have been larger or similar; {2) original males and their respective replacements had significant wing and bill length disparities when they should have been similar; {3) morphological characteristics of replacement victors and losers were not significantly different; and {4) victors and losers showed no significant difference in their aggression indexes. The RHP hypothesis was confirmed only in capture weight analysis since males (whether original or replacement) with low capture weights tended to lose territories. Although previous ownership and habitat quality did not seem to affect outcomes of contests, seasonal effects and weight loss in captivity could have obscured the impact of these

PAGE 32

26 factors. The RHP hypothesis predicts that body size is positively correlated with the ability of males to defend and maintain territories (Parker, Ropwer, 1982), but we did not confirm this assumption. While wing length is believed to be the most reliable measure of both body size (Hamilton, James, Searcy, 1979) and dominance status (Eckert and Weatherhead, Searcy, and Yasukawa, 1981a), capture weight more accurately predicted dominance status in this study. Even if our original males were heavy at the beginning of the breeding season, the energetic demands of territory owners (Beletsky et al., 1989) probably resulted in weight loss. Searcy (1979) found a negative correlation between increasing age and body weight. This could imply that some younger maies outcompeted older individuals in our study, if we can assume that younger males had higher capture weights. Even though capture weight is not a highly reliable size measurement, it could still represent an important component of RHP in Red-winged Black birds that gives younger andjor non-territorial males an asymmetrical advantage over older males-as the breeding season progresses. Other research has also revealed that the morphological characteristics of territorial males cannot be predicted by the RHP hypothesis. successful territorial

PAGE 33

27 ownership does not seem to imply size similarities between either neighboring males (Eckert and Weatherhead, 1987b; Yasukawa, 1981a) or original males and their respective replacements (Eckert and Weatherhead, 1987b). There is no evidence that non-territorial males are smaller than territorial males (Eckert and Weatherhead, 1987c). In this study, significantly heavier and longerwinged non-territorial males did not successfully oust smaller original males until they were removed from the study site. Even if the smaller territory holders were the largest males at the time they established territories, the RHP hypothesis does not explain how smaller males sustained ownership with larger males nearby in the floater population. The morphological characteristics associated with RHP apparently do not dictate territory settlement and maintenance. As in our study, other research has indicated that the high aggressiveness intrinsic to the RHP hypothesis was not characteristic of many successful territorial males. Yasukawa (198lb) determined that more experienced males actually used fewer aggressive behaviors than new males except when they initially established territory ownership. Dominance was shown to be a measure-dependent quality by Tomback et al. (1989) because different measures of behavior, including aggression, produced different hierarchies in Mule Deer (Odoco-

PAGE 34

28 ileus hemionus hemionus). Since dominance status is known to be energetically expensive (Hogstad, 1987), Jackson (1988) proposed that extreme overt aggressive behavior andjor increasing body size become disadvantageous to dominance rank, which is contrary to the RHP hypothesis. Since successful territory owners were not always more aggressive or larger than replacement males, yet weighed more at capture, this study could neither clearly reject nor confirm the RHP hypothesis. our data did not support the hypothesis that previous ownership of a territory guarantees territory retention. In our study, eleven out of twenty-one original malasregained territories in 1987. Jakobsson (1988) found even less convincing results in a study using Willow Warblers (Phylloscopus trochilus) where only four of eleven retained ownership. In contrast, Beletsky and Orians (1987) found that fifty of fifty-five original Red-winged Blackbird males re-established territories. Since it has been assumed that payoff asymmetries (Parker, 1974) create large advantages for original males, as observed by Beletsky and Orians (1987), perhaps other factors created by the methods utilized in our experiment produced unforeseen disadvantages for original males. Weight loss in captivity was a confounding variable during this study and has also affected other remo-

PAGE 35

29 val experiment results. In our study, the highest mean body weight loss (7.3%) occurred in 1988 and resulted in only 52% of males regaining territories. In a study by Jakobsson (1988), a mean body weight loss of 5.2% correspondedto 36% retention. Duration of captivity also seems to affect weight loss and successful territory maintenance. While we found a 52% retention using a 72 hour removal period, Beletsky and Orians (l987) found a 90.9% retention rate for Red-winged Blackbirds with a 48 hour removal interval. When longer captivity periods were used (144-168 hrs.), Beletsky and Orians (1989) found an 8% retention rate and a 4.9% weight loss. They determined that captivity periods over 120 hours severely reduced territory retention ability. During our study, substantial weight loss and/or excessive time in captivity did not preclude some males from regaining their territories, but we cannot rule out the possibility that captivity contributed to the loss of territories by other original males. Since retention abilities of original males may have decreased as the breeding season continued, it is possible that motivational changes occurred over time. Higher risks than benefits may prompt males to avoid reestablishing territories during the later periods of the breeding season. Peek (1971) noted that although neighboring males quickly incorporate empty territories early

PAGE 36

30 in the breeding season, they often show no replacement activity at all after peak nesting activity ceases. I Yasukawa (1981b) found that in later stages of the nest-ing season, yearlings become actively territorial, where-as territorial males tend .to leave the area after their young have fledged or nests have failed. Although this study did not produce concrete evidence of altered territorial motivation, the effect of time of year cannot be ruled out. The contrived nature of the removal experiment may also have obscured the previous ownership hypothesized for original males. Neighboring.males seem to become more aggressive after individuals are removed (Krebs, 1982), and when there is a high natural influx of new males (Picman, 1987). Krebs (1982) speculated that it is easier for a male who knows adjoining males to re-establish ownership males can predict aggressive tendencies in their nei9hbors. A newcomer has to evaluate each neighbor to find out if he is a threat (Krebs, 1982). If replacements dominate territories long enough to assess their neighbors' combative ability, replacement males may have acquired the asymmetrical advantage of original owners (Krebs, 1982). Since these circumstances would rarely occur in an unmanipulated system, the re-sults of the removal study may not accurately reflect how actual takeovers and turnovers occur (Beletsky and

PAGE 37

31 Orians, 1987). The study did not show that superior habitat was regained more often or contested for longer periods. There was also no conclusive evidence that high RHP males occupied high quality territories in this study. Eckert and Weatherhead (1987a) found that when males were placed in captivity, the larger and most aggressive individuals from inferior upland areas had higher dominance ranks than males from superior marsh locations. Since this study was the first removal experiment to attempt habitat assessment, perhaps more emphasis on habitat quality in future removal studies may help elucidate any effects that habitat quality has on territorial disputes. If habitat quality, previous ownership, RHP, seasonal effects, weight. loss, and replacement male asymmetrical advantage were not determining factors in the outcome of territorial contests, then other intangible influences presumably had considerable impact. Other researchers have speculated that site fidelity might influence males to stay in inferior habitats, even if they are capable of controlling superior territories (Eckert and Weatherhead, 1987a; Picman, 1987; Searcy, 1979). Although site fidelity did not influence reestablishment in our study, it may enhance re-establishment during the early stages of the breeding season (Jakobsson, 1988; Picman, 1987). Robertson (197l) sug-

PAGE 38

32 gested that male Red-winged Blackbirds often will utilize inferior habitat when superior vegetation is already occupied. If stochastic circumstances create the opportunity for a male to dominate a territory and decrease his risks of becoming injured, then chance events may play a greater role in territory settlement patterns than previously imagined (Eckert and Weatherhead, 1987a; Eckert and Weatherhead, 1987b; Picman, 1987; Searcy, 1979)

PAGE 39

BIBLIOGRAPHY Bailey, N.T.J. 1986. statistical methods in biology. 2nd ed. New York: John Wiley and Sons. 33 Beletsky, L.D. and Orians, G.H. 1987. Territoriality among male blackbirds I. Removal experiments and site dominance. Behav. Ecol. Sociobiol. 2Q 1 3.39-349 Beletsky, L.D. and Orians, G.H. 1989. Territoriality among male red-winged blackbirds II. Testing hypotheses of territorial dominance. Behav. Ecol. Sociobiol., 24, 333-339. Beletsky, L.D. et al. 1989. Relationships of steroid hormones and polygyny to territorial status, breeding experience and reproductive success in red-winged blackbirds. Auk, 106, 107-117. Bray, O.L., et al. 1975. A trap for capturing territorial male Red-winged Blackbirds. Western Bird Bander, 50, 4-7. eaccamisi, e.G. 1977. Breeding success and nest site characteristics of the red-winged_blackbirds. Wilson Bull., 89, 396-403. Eckert, e.G. and Weatherhead, P.J. 1987a. Ideal dominance distributions: a test using red-winged

PAGE 40

blackbirds (Agelaius phoeniceus). Behav. Ecol. Sociobiol., 20, 43-52. Eckert, e.G. and Weatherhead, P.J. 1987b. Competition for territories in red-winged blackbirds: is resource-holding potential realized? Behav. Ecol. Siociobiol., 20, 369-375. Eckert, e.G. and Weatherhead, P.J. 1987c. Owners, floaters and competitive asymmetries among territorial red-winged blackbirds. Anim. Behav., 35, 1317-1323. Goddard, s.v. and Board, v.v. 1979. Reproductive success of red-winged blackbirds in North Central Oklahoma. Wilson Bull., 79, 283-289. 34 Hamilton, T.H. 1961. The adaptive significance of intra-specific trends of variation in wing length and body size among bird species. Evolution, 15, 180195. Hansen, A.J. and Rohwer, s. 1986. Coverable badges and resource defence_ in birds. Anim. Behav., d!, 69-76. Hogstad, 0. 1987. It is expensive tope dominant. Auk, 104, 333-334. Holm, C.H. 1973. Breeding sex ratios, territoriality and reproductive success in the red-winged blackbird. Ecology, 54, 356-365.

PAGE 41

Jackson, W.M. 1988. can individual differences in history of dominance explain the development of linear dominance hierarchies? Ethology, 79, 71-77. 35 Jakobsson, s. 1988. Territorial fidelity of willow warbler (Phylloscopus trochilus) males and success in competition over territories. Behav. Ecol. Socio biol., 22, 79-84. James, F.C. 1970. Geographic size variation in birds and its relationship to climate. Ecology, 51, 365-390. Krebs, J.R. 1982. Territorial defence in the great tit (Parus major). Do residents always win? Behav. Ecol. Sociobiol., 11, 185-194. Lenington, s. 1980. Female choice and polygyny in redwinged blackbirds. Anim. Behav., 28, 347-361. Maynard-Smith, J. 1976. Evolution and the theory of games. Amer. Sci., 64, 41-45. Orians, G.H. 1961. The ecology of blackbird (Agelaius) social systems. Ecol. Monogr., 31, 85-312. Parker, G.A. 1974. Assessment strategy and the evolution of fighting behavior. J. Theor. Biol., 47, 23-243. Parker, G.A. and Rubenstein, D.I. 1981. Role assessment, reserve strategy, and acquisition of information in asymmetric animal contents. Anim. Behav., 29, 221-240.

PAGE 42

Peek, F.W. 1971. seasonal change in the breeding behavior of the male red-winged blackbird. Will. Bull., 83, 383-395. Picman, J. 1987. Territory establishment, size, and tenacity by male red-winged blackbirds. Auk, 104, 405-412. Robertson, R. 1972. Optimal niche space of the redwinged blackbird I. Nesting success in marsh and upland habitat. Canadian J. Zool., 50, 247-263. 36 Rohwer, s. 1982. The evolution of reliable and unreliable badges of fighting ability. Amer. Zool., 22, 531-546. Searcy, W.A. 1979. Male characteristics and pairing success in red-winged blackbirds. Auk, 96, 353-363. Stanton, 1986. Comparison of avian community dynamics of burned and unburned coastal sage scrub. Condor, 88, 285-289. Tobachnick, B.G. and Fidell, L. 1983. Multivariate Statistics. New York: Harper and Row. Tomback, D.F. et al. 1989. Measuring dominance and constructing hierarchies: An example using Mule Deer. Ethology, 82, 275-286. Wiens, J.A. 1973. Interterritorial habitat variation in grasshopper and savannah sparrows. Ecology, 54, 87-884.

PAGE 43

Wittenberger, J.F. 1980. Animal social behavior. Boston: Duxbury Press. Yasukawa, K. 1981a. Male quality and female choice of mate in the red-winged blackbird (Agelaius phoeniceus). Ecology, 62, 922-929. 37 Yasukawa, K. 1981b. Territory establishment in redwinged blackbirds: importance of aggressive behavior and experience. Condor, 81, 258-264.