Testosterone stimulates reproductive behavior during autumn in mockingbirds (Mimus polyglottos)

HORMONES

AND

BEHAVIOR

25,

229-241 (1991)

Testosterone Stimulates Reproductive Behavior during Autumn in Mockingbirds (A4fnus polyglottos) CHERYL

A. LOGAN AND CHERYL ANN CARLIN

Department of Psychology, University of North Carolina at Greensboro, Greensboro. North Carolina 27412-5001 Mockingbirds normally secrete little or no testosterone during the period of autumnal territoriality. To determine the behavioral effects of exogenously administered testosterone, 20-mm lengths of Silastic tubing filled with crystalline testosterone were implanted into free-living resident mockingbirds during the autumn. Control residents were given sealed empty implants. Focal animal sampling showed that T-implanted males sang significantly more than controls. Perhaps as a consequence, a significantly greater percentage of the T-implanted males acquired mates. Though nest building does not naturally occur in autumn, Timplanted males also showed significantly more nest building than control males. However, T-implanted males only built if there was a female in the territory, suggesting a synergy between the presence of testosterone and social cues provided by the female. Examination of the effects of testosterone on territorial aggression showed that despite the high levels of territorial activity common in this species in autumn, territorial fights were unaffected by the presence of testosterone. One aggressive call, known to function in fall territorial defense, was significantly decreased in T-implanted versus control males. The presence of fall testosterone appears to stimulate a number of reproductive activities in mockingbirds, leaving autumnal aggressive interactions either unchanged or decreased. We discuss the application of these data to the effects of testosterone on the mockingbird’s reproductive behavior during the breeding season. Q IWI Academic PWS, IX

The activational effects of hormones on reproductive behavior and aggression ‘are known to vary widely across species (Crews and Moore, 1986; Moore and Marler, 1988). Associated patterns are characterized by a close causal connection between hormones and behavior (Crews, 1984), and the field endocrinology of songbirds has focused on the associated patterns common in migratory north temperate species. However, proximate mechanisms linking hormones and behavior may vary even among species showing associated patterns. For example, although sex steroids are widely known to affect both reproductive and territorial behavior in birds (Balthazart, 1983), the copulatory behavior of male white-crowned sparrows occurs in response to female solicitation, irrespective of the 229 0018-506x/91 $1.50 Copyright Q 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

230

LOGAN AND CARLIN

male’s hormonal state (Moore and Krantz, 1983). On the other hand, the male’s territorial aggression depends on hormonal cues (Moore, 1984). The northern mockingbird (Mimu.s polyglottos) differs in several ways from many of the species in which the relationship between hormones and behavior has been studied. For example, many pairs exhibit perennial monogamy with the duration of the pairbond lasting as long as 8 years (Breitwisch, personal communication). In the southeastern United States most birds do not migrate; rather, individuals defend their territories throughout the year. Males sing during much of the year, stopping only during the postnuptial molt and in early winter, and song appears to function in mate attraction in autumn as well as in the spring (Breitwisch and Whitesides, 1987; Logan and Hyatt, 1991; Merritt, 1985). Several unusual features also distinguish the mockingbird’s reproductive behavior. Males almost always initiate nest building, and although females often help line the nest, some males complete the nest with no help from the female. The species is very multibrooded, attempting as many as seven nests in a single breeding season (Logan, 1983), and males show a high degree of parental investment (Breitwisch, 1989). Despite intense territorial activity during the months of September and October, males secrete little or no testosterone (T) at this time of year. In addition, differences in autumnal territorial aggression and in singing and the apparent use of fall song in mate attraction occur in the absence of measurable levels of autumnal T (Logan and Wingfield, 1990). We took advantage of the lack of fall testosterone to examine the behavioral effects of exogenously administered testosterone at a time when it is normally absent. Long-acting testosterone capsules were implanted in freeliving mockingbirds (1) to assesswhether exposure to high levels of T can alter behavior in autumn, and (2) if effective, to determine whether T affects fall territorial aggression, fall reproductive behavior, or both. MATERIALS AND METHODS The study was conducted on a population of wild mockingbirds inhabiting the residential campus of the University of North Carolina at Greensboro. The majority of males holding spring breeding territories in this area over winter, defending the space they occupy throughout autumn and winter as well. Of the 25 males studied, 14 had held territory in the population for at least one season prior to implantation. Eight of these began the study unmated, and 6 were mated. Eleven birds established their territories during the season in which the implants were inserted; all of these were unmated at the start of the study. We began preexperimental mapping of residents’ territories in early September to determine the stability of residency and confirm mating status. Beginning on 24 September, 1988 and 27 September, 1989, 21 birds were captured in baited potters traps and implanted with 20-mm

TESTOSTERONE

AND REPRODUCTIVE

BEHAVIOR

231

lengths of sealed medical grade Silastic tubes (o.d. 0.077 mm, i.d. 0.050 mm). For a given mating status (mated or unmated), males were randomly assigned to T-implant or empty implant (control) conditions. When birds had to be reassigned due to our inability to capture some individuals, reassignment was made via coin toss. Eleven birds received implants packed with crystalline testosterone (Sigma Chemical) and 10 received empty implants. All implants were inserted in the field with the unanesthetized bird restrained on a laparotomy board. Implants were placed subcutaneously in the birds’ abdomens, lying vertically along the flank. Unbanded birds were marked with colored leg bands and immediately released into their territories. Twelve of the 21 implanted birds were recaptured after behavioral sampling ceased. The presence of the implant was checked, and it was removed. Following recapture, but before removal of the implant, we took approximately 300 ~1 of whole blood from the wing vein using heparinized microhematocrit capillary tubes. No bird remained in a trap for more than 15 min before being removed, and all bleeding times ranged from 3 to 9 min. Whole blood was kept cool prior to centrifugation, and plasma samples were stored at - 20°C. Testosterone titers were measured at the University of Washington using radioimmunoassay (RIA) following separation and purification on celite : glycol microcolumns according to the methods of Wingfield and Farner (1975). Steroid fractions were assayed using standard curves calculated over the range of 2 to 500 pg and corrected for recovery percentages. Intraassay coefficients of variation are less than 12% as determined by Wingfield, Newman, Hunt, and Farner (1982). Interassay variation for samples measured in two different assays equaled 14%. Additional details on the assay procedure can be found in Wingfield and Farner (1975) and Hegner and Wingfield (1986). Focal animal sampling (Altmann, 1974) of behavior began a minimum of 3 days after the implants were inserted. Each sample consisted of 30 min of continuous observation of the resident’s activities. We collected from 8 to 11 30-min samples per bird distributed across approximately 14 days. Weather permitting, we approximated equal numbers of samples collected during morning (0700-1100) and afternoon (1500-1800) hr. Samples were initiated only after a bird had been sighted, and no bird was sampled more than twice per day. Three different observers monitored each bird, and two of the three were blind to the implant condition. Observers also monitored time-out periods, defined by the amount of time per sample during which visual and acoustical contact with the bird was lost. Samples were aborted if the observer lost contact with the bird for more than 10 min. To minimize the possibility that varying activity levels affect the timing of sampling, on arrival observers searched the bird’s territory for a fixed period of 10 min. If the bird could not be found, the observer moved to a different territory, returning to the original space

232

LOGAN AND CARLIN TABLE 1 Summary of Observations on T Implant and Control (Empty Implant) Males”

Observation/bird (min) Time-out/sample (min) Perches per sample

T implant

Control

Mann-Whitney U

300.0 3.0 15.8

300.0 3.1 10.3

n.s. n.s. P < 0.02

’ Values are group medians.

only after at least 30 min had elapsed. Unless otherwise indicated, in all independent statistical comparisons we used the two-tailed Mann-Whitney U test (Siegel, 1956). Values for individual birds are means calculated across successive time samples. Group values are presented as medians. See Table 1 for a summary of observation conditions. Behaviors monitored included the locations of all perches and all foraging activities, the occurrence of aggressive interactions with other mockingbirds, including territory fights and chases and the ritualized dance defining territory boundaries (Hailman, 1960), and all vocalizations. The latter include song and two calls, the hew and chatburst, both commonly produced during fall aggressive encounters (Logan, Budman, and Fulk, 1983). Amount of song was measured by the number of 15set intervals per sample in which singing occurred. Though this is a gross measure of song, previous research has shown it an effective index of differences in amount of singing (Logan, 1983). The timing of each behavior or position change was recorded to the nearest 15 set; and observers recorded the location of each behavior on a map drawn to scale of the resident’s territory. Territory size was measured to within 10% of the asymptote of total space used by cumulating space measured across successive samples (Odum and Kuenzler, 1955). RESULTS Reproductive

Behavior

The behavior of T-implanted males differed from that of controls in several respects. Song increased considerably in T-implanted males (17 = 1.0, P < 0.01; Table 2). Perhaps because of a threefold increase in amount of singing, seven of the eight T-implanted males that were unmated when implanted acquired mates during sampling. However, not all females remained with their T-implanted males. In one territory, the female left while the male was still T-implanted; in two, the females left after the implant was removed. In four, females remained in the territories, and the pairs successfully reared young the following spring. None of the eight unmated control males obtained mates while sampling was underway, although two had mates the following spring. A x2 test on the frequency

TESTOSTERONE

AND REPRODUCTIVE

233

BEHAVIOR

TABLE 2 Summary of Behavioral Differences between T Implant and Control Males”

Song per sample Total nest building per bird Percentage samples with nest building Fights/sample with mockingbirds Percentage samples with mockingbird fights Chatbursts/sample Percentage samples with chatbursts Territory size (ha)b

T implant

Control

42.3 3.0 25.0

17.4 0.0 0.0

P < 0.02 P < 0.02

0.8

0.5

ns.

42.0

35.5

n.s.

1.3 25.0

5.3 47.0

P < 0.02

0.75

0.41

Mann-Whitney

U

P < 0.01

P < 0.02 P = 0.02

a Values are group medians. b In hectares (ha); conversion to ha was performed following statistical analyses.

with which T-implanted versus control males acquired mates during sampling was statistically significant (x*(l) = 5.14, P < 0.05). The effect of T implants on fall mate attraction is underscored by the fact that four males each attracted three mates while implanted with testosterone. One male had two females simultaneously residing in his territory, although none kept multiple females during the following spring. A comparison of mated and unmated T-implanted males was impossible because so many unmated males quickly obtained mates. Three had acquired mates by the time behavioral sampling began, within 3 days of implant, We arbitrarily divided the T-implanted males into two groups: those with females present on more than half of the samples were considered mated (n = 7); those with females present on fewer than half of the samples were considered unmated (n = 4). The median amount of song produced by the two groups did not differ. Although the number of birds is small, this comparison suggests that T eliminates the natural difference in song production distinguishing mated and unmated males in autumn. We were unable to capture and implant four additional males (two mated, two unmated) on whom we collected complete behavioral data. To increase the number of subjects, these birds are included as unimplanted controls in the following analyses of the effects of T on differences in song between mated and unmated males. Comparison of mated T-implanted males (n = 7) with unmated control males (n = 10) indicated that the amount of song produced by T-implanted mated males exceeded the normally high levels of song produced by unmated males (42.3 versus 19.2, respectively; U = 5, P < 0.01). T-implanted mated males also sang significantly more than control mated males (n = 4) who

234

LOGAN

AND

CARLIN

normally sing least in the fall (42.3 versus 18.6, respectively; U (onetailed) = 3, P = 0.02). Other reproductive behaviors previously seen only in the spring were also observed in T-implanted males. The clearest of these was male nest building. Nine of 11 T-implanted males began nest building; only one of 10 control males started to build (x*(l) = 3.84, P = 0.055). The total number of nest building responses per bird summed across all samples was significantly greater for T-implanted males (U = 11, P < 0.01; Table 2), as was the percentage of samples in which male twig selection or twig placement was seen (U = 11.5, P < 0.01; Table 2). However, testosterone was effective in stimulating male nest building only if a female was present in the territory. Females were present on a significantly greater percentage of the samples in which nest building occurred in T-implanted males (94% versus 6%, x’(l) = 24.5, P < 0.001). Only two of a total of 56 nest placement responses occurred in the absence of a female. We collected samples on five previously unmated T-implanted males both before and after a female arrived in the territory. All nest building occurred after the female arrived, although because the number of birds is small, the result only approaches significance (Wilcoxon’s matched-pairs signed rank test: T = 0, P = 0.06). Although they sang more than unmated controls, these males also showed a drop in song production after the female arrived (59.5 versus 34.7 intervals with song per sample; Wilcoxon’s matchedpairs signed rank test: T = 1, P = 0.06.) Mate acquisition in mockingbirds is often associated with a tandem flight display in which the male and a newly arrived female fly in tandem throughout the territory repeatedly perching together. Although the display is usually seen in the spring, four T-implanted males engaged in fall tandem flights with their new partners. In the most pronounced instance, the display lasted for the full 30-min observation period. During the sample, the male’s song production dropped (22 intervals with song/30 min compared to a mean of 57.5 t 24.13). Mated birds often exchange a low growling call, the hew, when near each other, and hewing increased during the display (31 as compared to a mean of 4.9 & 9.7 hews/30 min). The male and female appeared to continuously follow one another throughout the territory. The male made 37 position changes, and the female made 33. This is an increase for the male over his mean of 22.5 + 5.2. Most perches were of short duration (mean = 46 set). In 72.97% (27 of 37) of the perch changes, the second bird perched within 3 m of the other within 15 set of the arrival of the first, indicating that following was occurring. The male often left the perch within 60 set of the female’s arrival, as if to lead her around the territory. Forty-two percent of the joint perches occurred in prospective nest sites, characterized by dense broadleaf foliage and branching twig understructure (Joern and Jackson, 1983). Several of these locations also bear abundant winter fruit. The

TESTOSTERONE

AND REPRODUCTIVE

BEHAVIOR

235

focus on prospective nest sites and fruiting resources raises the possibility that in tandem flight the female is inventorying critical resources that will be available in the territory if she remains. Territorial Aggression We also examined the impact of testosterone implants on the expression of fall aggressive behavior. T-implant versus control males did not differ in the median number of territorial fights per sample (U (one-tailed) = 40, P > 0.20), the percentage of samples with territorial fights (U (onetailed) = 38, P > 0.20), or the median number of fights with other species (U (one-tailed) = 48, P > 0.20; Table 2). The chatburst, a call commonly associated with fall territorial encounters (Logan et al., 1983), was significantly decreased in T-implanted males. Control males produced chatbursts on 47.0% of the samples, while T-implanted males produced chatbursts on only 25.0% of the samples (U = 11.5, P < 0.02; Table 2); control males produced 5.3 chatbursts per sample, while T-implanted males produced only 1.3 calls per sample (U = 15, P < 0.02; Table 2). However, median territory size was significantly greater in T-implanted males (U = 22, P = 0.02; Table 2). Several of the control males shared boundaries with T-implanted neighbors; others did not. Including the data from four males that we were unable to capture, but for which we had complete behavioral observations, we analyzed differences in aggressive interactions in control males with at least one T-implanted neighbor (n = 7) and those with no T-implanted neighbors(n = 7). There was no evidence of increased aggression in males with T-implanted neighbors; neither song, nor any measure of aggressive behavior distinguished the two groups (territorial fights/30 min: 0.9 versus 0.4, U (one-tailed) = 14.5, P = 0.12; percentage of samples with territorial fights: 22.0% versus 41.0%, U (one-tailed) = 15, P = 0.13.) T-implanted and control groups did not differ in the median observation time per bird or in the median time-out period per sample (see Table 1). Therefore, differences cannot be accounted for on the basis of biased time sampling. However, the median level of activity (number of position changes per sample) was significantly greater in T-implanted males (see Table 1). Examination of the implants and hormone assaysperformed on plasma samples taken at recapture confirm that the above differences were in fact due to the testosterone implants. Implants were found intact in 100% (n = 12) of the recaptured birds. Plasma T levels sampled in 6 T-implanted birds recaptured within 1 month of implant averaged 10.3 + 4.9 rig/ml. DISCUSSION Sex steroids are thought to affect both territorial and reproductive behavior in birds (e.g., Sosskina, Prove, and Immelmann, 1980). How-

236

LOGAN

AND

CARLIN

ever, there is increasing evidence that the two may be independent of one another (Moore, 1984; Moore and Marler, 1988). In many passerines, the pattern of male testosterone secretion during the breeding season reflects species-specific patterns of male aggressive challenge (Wingfield and Moore, 1986; Wingfield, Ball, Duffy, Hegner, and Ramenofsky, 1987), suggesting that one primary behavioral effect of testosterone in free-living passerines is in the control of aggression. Outside of the breeding season, this does not appear to be the case in the mockingbirds. Although they exhibit considerable territorial aggression in autumn (Laskey, 1936; Logan, 1987), T levels are basal in mockingbirds in autumn, and variations in plasma T are unrelated to differences in territorial aggression (Logan and Wingfield, 1990). The results reported above confirm the independence of testosterone and autumnal territorial aggression in this species. Measures of aggressive behavior were unchanged by the presence of T, and the production of the chatburst, a call that functions in fall territorial defense (Logan et al., 1983), was inhibited in T-implanted males. The absence of neighbor effects produced by T implants also reflects the independence of testosterone and autumnal territoriality in mockingbirds. Testosterone implants during the breeding season increased territorial aggression in both implanted male song sparrows and in their unimplanted neighbors (Wingfield, 1985). In mockingbirds, however, autumnal territorial aggression in unimplanted males with T-implanted neighbors did not differ from that seen in control birds with no T-implanted neighbors. Other behavioral effects of testosterone (see below) indicate that the absence of a change in aggressive behavior was not due to the insensitivity of the mockingbird’s brain to the presence of T in autumn. Interestingly, territory size increased in T-implanted males. Winter territory size may shrink in this species, possibly as a consequence of the energetic demands of colder winters. When shrinkage occurs, the vacated areas between occupied space are not contested, and the shrink is followed by a late winter size increase in preparation for breeding (Logan, 1987). We believe that the size increase seen in T-implanted males may be interpreted as an early adjustment for breeding. Males may expand territory size to amass breeding resources, rather than in response to territorial challenge. In other species, aggressive behaviors controlled by testosterone during the breeding season may be unaffected by changes in testosterone outside the breeding season (Hahn and Wingfield, 1988; Moore and Marler, 1987). The above results suggest that this is the case in mockingbirds as well. Testosterone implants were effective in altering several other activities usually associated with breeding. Song, mate acquisition, and nest building each increased in T-implanted males. The effects of T on mate attraction were very likely indirect; we propose that they depended upon the T-

TESTOSTERONE

AND

REPRODUCTIVE

BEHAVIOR

237

induced increase in singing. This interpretation is supported by the decreased song production of T-implanted males that followed the arrival of females onto their territories. Mockingbirds normally sing throughout the fall, and previous research indicates that fall female removal increases song production (Logan and Hyatt, 1991). The fact that female arrival and female removal have inverse effects on song production (1) confirms the role of autumnal song in autumn mate attraction, and (2) implies that the effects of T implants on mate attraction resulted from increased singing produced by high levels of testosterone. Both song and mate acquisition normally occur in autumn (Logan and Hyatt, 1991), and the effects of testosterone appeared to enhance rather than activate these systems. However, in the presence of a female T implants also stimulated male nest construction, normally absent in autumn. In most passerine species in which hormone-behavior relationships have been studied, it is the female that builds the nest, and there are few data on the hormonal basis of male nest building in free-living birds (Collias and Collias, 1984). However, among the captive species studied, sex steroids are involved in the control of male nesting. For example, although they do not build, male ring doves, streptopelia risoria, gather nesting materials and bring them to the female. Control castrates do not gather, and gathering is restored by steroid replacement in castrated males (Martinez-Vargas, 1974; Erickson and Martinez-Vargas, 1975). In captive weaver birds, the administration of testosterone affects nest building in the presence of appropriate building materials (Crook and Butterfield, 1968). However, nest construction can occur in the absence of testosterone (Collias and Collias, 1984), and, because males engage in intense competition for nesting material, it is difficult to separate the effects of T on competition for a critical resource from its effects on nest construction per se (see Wingfield et al., 1987). Finally, male zebra finches show high levels of testosterone while nest construction is underway (Walters and Harding, 1988), and the presence of the female is critical. Castration plus steroid replacement do not affect the gathering of nesting material unless pairs have been together for some time (Harding and Sheridan, 1983). The results reported above show that testosterone may activate nest construction in free-living mockingbirds. As in zebra finches, the effectiveness of T in stimulating nest building depended on the presence of a female, suggesting a synergy between social signals and the impact of testosterone. However, mockingbirds that built were first seen building after the female had been in the territory for an average of only 1.9 ? 0.6 days. Four males that were unmated prior to implant were gathering material the day after the female was first seen in the territory. Though we have no data on whether female presence or some specific response of the female is critical to male nest building, the duration of the union

238

LOGAN

AND

CARLIN

seems less important than in zebra finches. However, in mockingbirds the amount of nest building seen in T-implanted males was less than that normally observed in the spring. All but two building males developed a clear nest structure, but only two nests appeared to be fully constructed, and only one female was seen carrying lining. The incompleteness of most of the autumn nests may reflect the absence of involvement by the female during a season in which nest building normally does not occur. During the breeding season, male mockingbirds initiate nest building and females assist in lining the nest. Mockingbirds are multibrooded, and the male constructs a new nest for each of as many as seven nesting attempts. Therefore, nests are built throughout the breeding season (Logan, 1983; Logan, Hyatt, and Gregorcyk, 1990; Zaias and Breitwisch, 1989). Plasma samples collected from free-living males engaged in normal spring nest building indicate that T levels are high, averaging 2.8 rig/ml (unpublished data). This suggests that testosterone may be necessary for the onset of male nest building during the spring as well. Moreover, male mockingbirds resume singing prior to and during the construction of each successive nest (Logan, 1983). Recent work indicates that the song produced during spring nest building may be one factor controlling the onset of nest building during successive nesting attempts (Logan et al., 1990). Little is known about the stimuli controlling the resumption of singing during spring renesting. However, it is possible that singing begins before T levels increase and that increased T levels reach values that, with a female present, stimulate spring nest construction. If the effectiveness of T in stimulating autumn nest building can be generalized to the spring, the results of T implants suggest an elaborate synergy between social, vocal, and endocrine events controlling the initiation of renesting in the spring. However, precise conclusions on the role of song and testosterone in the coordination of spring renesting require further research on natural cycles of endocrine secretion and behavior during the breeding season. Moore and Marler (1988) have proposed that the evolutionary flexibility now thought to exist between hormones and behavior may reflect phylogenetic conflicts in the use of hormonal signals for temporally incompatible behaviors. They contend that adaptively incompatible behaviors should not depend upon the same hormonal signal. Nothing is known about the control of aggression in mockingbirds during the breeding season. However, it is possible that the protracted demands of year-round territorial aggression and the precise sociotemporal coordination of male nest building in the spring are in conflict with one another in mockingbirds. Many different cues may stimulate aggression throughout the year in this perennially territorial species. But, nest building must occur in a welldefined period, under precise social and environmental cues. Control of both by the same endocrine cue might produce one under conditions more

TESTOSTERONE

AND REPRODUCTIVE

BEHAVIOR

239

appropriate for the other. The resulting conflict may have released territorial aggression from control by testosterone, leaving its primary activational effect on reproductive behavior. ACKNOWLEDGMENTS Thanks are due to Laura Hyatt and Jackie Spencer for their untiring help in the field, and to John Wingfield and Lynn Erckmann who performed the steroid assays. We are grateful for comments on an earlier draft of the manuscript provided by Reed Hunt and by two anonymous reviewers. This research was supported by the University of North Carolina Research Council.

REFERENCES Altmann, J. (1974). Observational study of behaviour: Sampling methods. Behaviour 49, 227-267. Balthazart, J. (1983). Hormonal correlates of behavior. In D. Famer, J. King, and L. Parks (Eds.), Aviun Biology, Vol. VII, pp. 221-305. Academic Press, New York. Breitwisch, R. (1989). Mortality patterns, sex ratios, and parental investment in monogamous birds. In D. Power (Ed.), Current Ornithology, Vol. 6, Plenum, New York. Breitwisch, R., Merritt, P. G., and Whitesides, G. H. (1986a). Parental investment by the northern mockingbird: Male and female roles in feeding nestlings. Auk 103, 152-159. Breitwisch, R., and Whitesides, G. H. (1987). Directionality of singing and non-singing behaviour of mated and unmated northern mockingbirds, Mimes polyglottos. Anim. Behav. 35, 331-339. Cheng, M.-F. (1986). Female cooing promotes ovarian development in ring doves. Physiol. Behuv. 37, 371-374. Collias, N. E., and Collias, E. C. (1984). Nest Building and Bird Behavior, Princeton Univ. Press, Princeton, NJ. Crews, D. (1984). Gamete production, sex hormone secretion, and mating behavior uncoupled. Horm. Behuv. 18, 22-28. Crews, D., and Moore, M. C. (1986). Evolution of mechanisms controlling mating behavior. Science 231, 121-125. Crook, J. H., and Butterfield, P. A. (1968). Effects of testosterone propionate and luteinizing hormones on agonistic and nest building behaviour of Queleu quelea. Anim. Behuv.

16, 370-384. Erickson, C. J., and Martinez-Vargas, M. C. (1975). The hormonal basis of cooperative nest building. In P. Wright, P. Caryl, and D. Vowles (Eds.), Neural and Endocrine Aspects of Behaviour in Birds, pp. 91-109. Elsevier, Amsterdam. Hahn, T., and Wingfield, J. C. (1988). Testosterone and “Year-Round” Territoriality in the Rufous Song Sparrow. Paper presented at the Western Regional Conference of Comparative Endocrinology, Seattle, WA. Hailman, J. P. (1960). Hostile dancing and fall territory of a color-banded mockingbird. Condor 62, 464-468.

Harding, C. F., and Sheridan, K. (1983). Hormonal specificity and activation of sexual behavior in male zebra finches. Horm. Behav. 17, 111-133. Hegner, R. E., and Wingfield, J. C. (1986). Behavioral and endocrine correlates of multiple brooding in the semicolonial house sparrow Passer domesticus. 1. Males. Horm. Behav. 20, 294-312. Joern, W. T., and Jackson, J. F. (1983). Homogeneity of vegetational cover around the nest and avoidance of predation in mockingbirds. Auk 100, 497-499.

240

LOGAN AND CARLIN

Laskey, A. R. (1936). Fall and winter behavior of mockingbirds. Wilson Bull. 48, 241-255. Logan, C. A. (1983). Reproductively dependent song cyclicity in mated male mockingbirds. Auk 100, 404413. Logan, C. A. (1987). Fluctuations in fall and winter territory size in the Northern Mockingbird (Minus polyglottos). J. Field Ornithol. 58(3), 297-305. Logan, C. A., Budman, P. D., and Fulk, K. R. (1983). Role of Chatburst versus song in the defense of fall territory in mockingbirds (Mimus polyglottos). J. Camp. Psychol. 97, 292-301.

Logan, C. A., and Hyatt, L. E. (1991). Mate attraction by autumnal song in the northern mockingbird, (Mimus polygloztos). Auk, in press. Logan, C. A., Hyatt, L. E., and Gregorcyk, L. (1990). Song playback stimulates nest building during clutch overlap in mockingbirds (Mimns polyglottos). Anim. Behav. 39, 943-953.

Logan, C. A., and Wingfield, J. C. (1990). Autumnal territorial aggression is independent of plasma testosterone in mockingbirds. Horm. Behav. 24, 568-581. Martinez-Vargas, M. (1974). The induction of nest building in the ring dove (Streptopeliu risoriu). Hormonal and social factors. Behaviour 50, 123-151. Martinez-Vargas, M. C., and Erickson, C. J. (1973). Some social and hormonal determinants of nest building behaviour in the ring dove (Streptopelia risoria). Behaviour 45, 12-37. Merritt, P. (1985). Song Function and the Evolution of Song Repertoires in the Northern Mockingbird, Mimuspolyglorros. Ph.D. dissertation, University of Miami, Coral Gables, FL. Moore, M. C. (1984). Changes in territorial defense produced by changes in circulating levels of testosterone: A possible hormonal basis for mate-guarding behavior in whitecrowned sparrows. Behaviour 88, 215-226. Moore, M. C. (1987). Castration affects territorial and sexual behaviour of free-living male lizards, Sceloporus jarrovi. Anim. Behav. 35, 1193-1199. Moore, M. C., and Krantz, R. H. (1983). Evidence for androgen independence of male mounting behavior in white-crowned sparrows. Horm. Behav. 17, 414-423. Moore, M. C., and Marler, C. (1987). Effects of testosterone manipulation on nonbreeding season territorial aggression in free-living male lizards, Sceloporus jarrovi. Gen. Comp. Endocrinol.

65, 225-232.

Moore, M. C., and Marler, C. A. (1988). Hormones, behavior and the environment: An evolutionary prespective. In M. H. Stetson, (Ed.), Processing of Environmental Information in Vertebrates, pp. 71-84. Springer-Verlag, New York. Odum, E., and Kuenzler, E. (1955). Measurement of territory size and home range in birds. Auk 72, 128-137. Siegel, S. (1956). Non-parametric Stun&x McGraw-Hill, New York. Sosskina, R., Prove, E., and Immelmann, K. (1980). Hormonal mechanisms avian behavior. In A. Epple & M. Stetson (Eds.), Aviun Endocrinology, pp. 533-547. Academic Press, New York. Walters, Michael J., and Harding, Cheryl F. (1988). The effects of an aromatization inhibitor on the reproductive behavior of male zebra finches. Horm. Behav. 22, 207-218. Wingfield, J. C. (1985). Short-term changes in plasma levels of hormones during establishment and defense of a breeding territory in male song sparrows, Melospiza melodia. Horm. Behav. 19, 174-187. Wingfield, J. C., Ball, G. F., Dufty, A. M., Hegner, R. E., and Ramenofsky, M. (1987). Testosterone and aggression in birds. Amer. Sci. 75, 602-608. Wingfield, J. C., and Famer, D. S. (1975). The determination of five steroids in avian plasma by radioimmunoassay and competitive protein binding. Steroids 26, 311-327. Wingfield, J. C., and Moore, M. C. (1986). Hormonal, social and environmental factors in

TESTOSTERONE

AND REPRODUCTIVE

BEHAVIOR

241

the reproductive biology of free-living birds. In D. Crews (Ed.), The Psychobiology of Reproductive Behavior, pp. 148-175. Prentice-Hall, Englewood Cliffs, NJ. Wingfield, J. C., Newman, A. L., Hunt, G. L., and Farner, D. S. (1982). Endocrine aspects of female-female pairing in the western gull (Larus occident& wymani). Anim. Behav. 30, 9-22. Zaias, J., and Breitwisch, R. (1989). Intrapair cooperation, fledgling care, and renesting by northern mockingbirds (Mimes polyglottos). Ethology 80, 94-110.