Failure to Detect Seasonal Changes in the Song System Nuclei of the Black-Capped Chickadee (Poecile atricapillus)

Failure to Detect Seasonal Changes in the Song System Nuclei of the Black-Capped Chickadee (Poecile atricapillus) T. V. Smulders, 1,2 M. D. Lisi, 1 E. Tricomi, 1 K. A. Otter, 3,4 B. Chruszcz, 3 L. M. Ratcliffe, 3 T. J. DeVoogd 1 1 Department of Psychology, Cornell University, Ithaca, New York 14853 2 School of Biology and Psychology, Newcastle University, Newcastle upon Tyne NE2 4HH, United Kingdom 3 Department of Biology, Queen s University, Kingston, Ontario K7L 3N6, Canada 4 Ecosystem Science & Management Program, University of Northern British Columbia, Prince George, British Columbia V2N 4Z9, Canada Received 18 November 2005; accepted 15 February 2006 ABSTRACT: Most temperate songbird species sing seasonally, and the brain areas involved in producing song (the song system) vary in size alongside the changes in behavior. Black-capped chickadees (Poecile atricapillus) also sing seasonally, and we find that there are changes in the stereotypy and the length of the fee-bee song from the nonbreeding to the breeding season. Yet despite these changes, we fail to find any evidence of seasonal changes in the song system. The song system of males is larger than that of females, as is typical in songbirds, but the ratio between the sexes is small compared to other species. We suggest three hypotheses to explain our failure to find seasonal variation in the chickadee song system. ' 2006 Wiley Periodicals, Inc. J Neurobiol 66: 000 000, 2006 Keywords: sexual dimorphism; social organization; vocal communication; Paridae; neuroethology INTRODUCTION Bird song is controlled by a set of interconnected nuclei in the brain, commonly known as the song system (reviewed by Brenowitz et al., 1997; Margoliash, 1997; Wild, 1997b; DeVoogd and Szekely, 1998). Correspondence to: T. V. Smulders (tom.smulders@ncl.ac.uk). Contract grant sponsor: Belgian American Educational Foundation Philips Fellowship. Contract grant sponsor: the Departments of Biology and Psychology at Cornell University. Contract grant sponsor: NIMH; contract grant number: MH-56093. ' 2006 Wiley Periodicals, Inc. Published online in Wiley InterScience (www.interscience.wiley. com). DOI 10.1002/neu.20281 These structures are involved in the acquisition, production and maintenance of the behavior (Nordeen and Nordeen, 1997; Wild, 1997a; Benton et al., 1998). Because song is mostly used in a reproductive context, the song system is sensitive to circulating steroid hormones (Schlinger, 1997). In most species studied, song is chiefly produced by the male. In those species, the size of the nuclei in the song system is strongly sexually dimorphic. In some duetting species, or species in which the females sing as well, the sexual dimorphism is less pronounced, but usually still there (Brenowitz and Arnold, 1986; DeVoogd et al., 1995; Gahr et al., 1998; MacDougall-Shackleton and Ball, 1999). In most temperate zone species, reproduction is a seasonal phenomenon, and the 1

2 Smulders et al. behaviors associated with it, such as song, are performed mainly during the breeding season. Associated with that, many species have been found to show a seasonal variation in the volume of some or all of the song nuclei studied (e.g., Nottebohm, 1981; Arai et al., 1989; Kirn et al., 1989; Brenowitz et al., 1991; Rucker and Cassone, 1991; Bernard and Ball, 1995; Smith et al., 1995; Li et al., 1996; Smith, 1996; Bernard et al., 1997; Brenowitz et al., 1998; Dloniak and Deviche, 2001; Caro et al., 2005). Seasonal plasticity has also been found in finer anatomical aspects of the song nuclei (DeVoogd et al., 1985). The functional role of seasonal changes in the song system is still under debate, with some hypothesizing a role of (re) learning (new) song elements (Nottebohm et al., 1986), others hypothesizing about a change in song stereotypy (Smith et al., 1997), and others focusing on changes in the amount of singing (Sartor et al., 2005). The black-capped chickadee (Poecile atricapillus) is an oscine song bird which produces a number of distinct vocalizations. The song of the black-capped chickadee is the fee-bee vocalization. It is delivered primarily (but not exclusively) by the male, often from a high perch during dawn chorus (Ficken et al., 1978). It is a long distance vocalization that seems mainly involved in challenging neighboring males in territorial disputes. Males countersing with each other and match their frequency to that of their opponent (Ratcliffe and Weisman, 1985; Hill and Lein, 1987; Horn et al., 1992; Shackleton and Ratcliffe, 1994; Otter et al., 2002; Mennill and Ratcliffe, 2004a,b), while females use it to evaluate male quality (Otter and Ratcliffe, 1993; Ratcliffe and Otter, 1996; Otter et al., 1997; Mennill et al., 2002). The song is heard mostly during the breeding season, but on rare occasions also during winter (Smith, 1991). Like all oscine songs, the fee-bee is a learned vocalization (Shackleton and Ratcliffe, 1993; Kroodsma et al., 1995). Surprisingly, it shows remarkable geographical stability, with only minor deviations in isolated populations (Smith, 1991; Gammon and Baker, 2004; Gammon et al., 2005). There has been little research to date on the song system of chickadees or other Paridae. Song nuclei HVC (used as a proper name) and RA (robust nucleus of the arcopallium) show an increase in volume from winter to the breeding season in blue tits (Parus caeruleus) on Corsica (Caro et al., 2005). The only published study to look at black-capped chickadees to date showed that in captivity, photostimulated birds had larger HVC and area X (and marginally larger RA) than either photorefractory or photosensitive birds (MacDougall-Shackleton et al., 2003). As discrepancies between captive and wild populations of birds in seasonal changes in the song system have been described before (Leitner et al., 2001b), it is important to verify laboratory studies with field studies. We therefore decided to investigate seasonal patterns in singing and the associated song system nuclei in wild black-capped chickadees. METHODS Behavior Song Recordings. Winter songs were recorded near Kingston, Ontario, during January and February 1997 with a Sony WM-D6C cassette recorder and a Sennheiser MKH816 microphone, using a Sennheiser MZA16-T-U preamplifier. Playback of less than 1 min, consisting of chick-a-dee calls and fee-bee songs from a local, unfamiliar male, was broadcast to elicit the birds to sing. Songs were obtained from a total of 23 birds, 12 of which produced more than five songs and were therefore included in the analysis. Of these 12 birds, 2 birds were in their first winter and were therefore inexperienced singers. Analyses were performed both including and excluding these two birds. The results are the same, so that the results reported include all 12 birds. Dawn chorus songs were recorded at the same location during April and May 1994 and 1995 using either a Sennheiser MKH816 or Audiotechnica AT815a microphone lined to a Sony WMD6C, Sony WMD3, or Panasonic RQ-L335 tape recorder. All 17 birds sang almost continuously during the recording session, such that all were included in the analysis. Most of the birds used for the analysis were individually color-marked; it is very unlikely that any bird recorded in spring was also recorded in winter. Song Analysis. All songs were digitized from a Marantz PMD221 tape player, through a MacRecorder into a Power Macintosh 7600/120, using Canary 2.1 (sample rate, 22.05 khz; sample size, 16 bits). Sonograms were made of each song and the following parameters were measured: length of fee, bee, and gap between fee and bee, as well as starting and ending frequencies of both fee and bee (Fig. 1). From these Figure 1 Sonogram of one fee-bee vocalization, with all the different measurements marked in the diagram.

Chickadee Song System Seasonal Patterns 3 original measures, other measures could be derived. Because we are especially interested in stereotypy, we calculated the coefficient of variation (CV ¼ 100% SD/mean) for each feature and each bird (Weisman et al., 1990). CVs of the most important song features (total song length, fee end frequency, bee end frequency, fee/bee ratio) were compared between the two groups of birds using independent sample t tests. Results were considered significantly different from each other when p < 0.05 and Bonferroni correction was applied to control for multiple testing. Table 1 Sample Sizes Feb Apr Jun Aug Oct Dec Male Adult 2 5 4 5 1 3 Juvenile 5 2 Female Adult 2 3 1 3 3 Juvenile 5 5 1 Anatomy Subjects. Subjects were black-capped chickadees, caught in 1992 1993 near Ithaca, New York, under state and federal permits. A total of 50 birds were caught at six different time points of the year [5 February (February), 30 March 9 April (April), 18 May 10 July (June), 13 18 August (August), 30 October 5 November (October), and 17 18 December (December)]. These were the same individuals as were used in a previous study (Smulders et al., 1995; Smulders et al., 2000; Shiflett et al., 2002). An additional two adult males were captured in April 2000. In the winter months, birds were caught with Potter s traps baited with food, and in the summer with song playback and mist nets. All birds were perfused within hours of being caught; age and sex were determined after perfusion. The birds were divided into two age classes, based on the separation of the two skull layers: juvenile (skull not completely pneumatized) and adult (pneumatized skull) birds (Smith, 1991). Juveniles could only be identified by these criteria in the June, August, and October samples. Details of sample sizes can be found in Table 1. Body weight and gonad weight were recorded for each bird. Histology. Birds were perfused transcardially with 0.8% saline and 10% formalin in 0.8% saline. The heads were then postfixed in 10% formalin/0.8% saline for at least 1 day, after which the brain was removed from the skull, weighed, and allowed to postfix for at least another day in formalin/saline. The brain was transferred to 10% formalin in 30% sucrose, until it sank (2 3 days). It was weighed again and embedded in 10% gelatin/30% sucrose, which was hardened in 10% formalin/30% sucrose. The brains were sliced on a freezing microtome at 40 m and transferred to microscope slides. Alternate sections were stained with Cresyl-violet stain and coverslipped with Permount 1 or Eukitt 1. Volume Measurements. We measured the volumes of five song control nuclei: HVC, RA, area X, LMAN (lateral magnocellular nucleus of the nidopallium) and nxiits (the tracheosyringeal portion of the nucleus of cranial nerve XII), as well as the control structure nxiil (the lingual portion of the nucleus of cranial nerve XII). Because not all the brains had been sectioned all the way into the brainstem, nxiits and nxiil could not be measured in all brains. In a previously published study on the same tissue, four more structures were measured: hippocampal formation (HF), entopallium [E; previously ectostriatum (Reiner et al., 2004)], nucleus rotundus (Rt), and the entire telencephalon (Tel) (Smulders et al., 1995). The different song nuclei were drawn using a camera lucida attached to the microscope. The drawings were then captured with a video camera (COHU) and digitized on a Macintosh IIci using NIH Image 1.54 to measure the surface areas. Volumes were calculated by multiplying the surface area with the distance between measurements (80 m for HVC, RA, area X, and nxiits; 40 m for LMAN), and adding those numbers. We measured left and right hemisphere of HVC, LMAN, nxiits, and nxiil. For RA and area X, only the left side was measured. For the analyses, we used the volume of the two hemispheres combined. For RA and area X, we did this by doubling the measured volumes. We only measured HVC proper; para-hvc (Kirn et al., 1989; Johnson and Bottjer, 1995; Brenowitz et al., 1998) could not be distinguished in these birds. The investigator drawing the sections and digitizing them was blind to the identity of the birds. Statistical Analysis. All statistical analyses were done using SPSS 11.0 for Windows. The main statistical technique used was the General Linear Model. Such a model tests for linear effects of each of the independent variables (which can be continuous or categorical) on a continuous dependent variable, while keeping the other independent variables constant. When all independent variables are categorical, it is identical to a multiway ANOVA. When we mention effects of several independent variables on a dependent variable, they are always the result of one such model, unless mentioned otherwise. Results are considered statistically significant for p < 0.05. Statistically controlling for total brain size can lead to spurious results, when total brain size itself varies seasonally (Smulders, 2002). Because telencephalon size in this data set is larger in the October sample than at other times of year (Smulders et al., 1995), we will do all analyses on the raw song system nucleus volumes. Only when a significant effect of season is found in which the nucleus is larger in October do we enter telencephalon into the model as a covariate, to test whether the difference in the nucleus is purely due to a difference in total brain size. When we do this, only the volume of the telencephalon exclusive of the telencephalic song system was used, in order to keep both measures independent of each other.

4 Smulders et al. RESULTS Behavior The stereotypy in total song length was higher during the breeding season dawn chorus (CV ¼ 5.1% 6 2.4%) than during the winter [CV ¼ 9.5% 6 5.0%; t(27) ¼ 3.116, p ¼ 0.016 after Bonferroni correction]. The other three measures analyzed (stereotypy in the frequencies of fee and bee, as well as the stereotypy in fee/bee frequency ratio) did not differ between seasons. To further investigate the difference in stereotypy in total song size, we looked at the stereotypy of the three elements (fee, gap, and bee) separately, and found that only fee length varied less in the breeding season (CV ¼ 9.9% 6 5.0%) than in the winter [CV ¼ 17.6% 6 8.2%; t(27) ¼ 3.153, p ¼ 0.012 after Bonferroni correction]. After observing the differences in stereotypy in both total song length and fee length, we subsequently also investigated mean total song length and fee length, and found both to be shorter in winter (total, 916 6 78 ms; fee, 348 6 58 ms) than in the breeding season [total, 973 6 59 ms; fee, 390 6 43 ms; total, t(27) ¼ 2.217, p ¼ 0.035; fee, t(27) ¼ 2.209, p ¼ 0.036]. Anatomy Gonads. Testes of adult males differed significantly in size across the season [F(5,14) ¼ 4.150; p ¼ 0.016]. A Fisher LSD post-hoc test showed that testis mass in June birds was significantly higher than at any of the other times of year, except for April. There were no differences between any of the other time points (Fig. 2). The range of testes sizes in April varied from similar to the nonbreeding season, to close to the June sample, such that the average was between June and the other times of year, and the variability was large. Adult ovaries did not vary significantly across seasons. HVC. HVC could be measured in a total of 48 birds. In our sample, HVC volumes co-varied with telencephalon volume to such a degree that the October HVC volumes were slightly larger than at other times of the year [F(5,39) ¼ 2.523, p ¼ 0.045]. As explained in the methods, we therefore enter telencephalon volume into the analysis as a co-variate. Total telencephalon size significantly predicted HVC size [F(1,38) ¼ 5.746, p ¼ 0.022]. HVC was larger in males than in females [F(1,38) ¼ 15.85, p < 0.001; M/F ratio, 1.66] and there was a significant interaction between sex and age [F(1,38) ¼ 7.679, p ¼ 0.009; Fig. 3(A)]. There was no difference across seasons Figure 2 Seasonal pattern of the masses of gonads of adult birds. There is a clear seasonal pattern in the testes, but not in the ovaries of the birds in our samples. Closed symbols represent males, and open symbols females. [F(5,38) ¼ 1.537, p ¼ 0.203], nor was there a main effect of age [F(1,38) ¼ 0.607, p ¼ 0.441]. To investigate the sex/age interaction, we reran the analysis for adults and juveniles separately. We did not find any sex differences in HVC size in juveniles [F(1,12) ¼ 0.357, p ¼ 0.561], but they were very pronounced in adults [F(1,23) ¼ 21.492, p < 0.001]. Looking at it in each sex separately, there was a significant age effect in the males [adults > juveniles, F(1,15) ¼ 16.49, p ¼ 0.001], but not in the females [F(1,13) ¼ 2.98, p ¼ 0.108]. We also ran the seasonal analysis on adult birds only and found the same result. Because this decreased our sample size, we also compared clear breeding season adults (April and June) with clear nonbreeding season adults (October and December; both males and females), and again found no differences. Finally, we also compared just adult breeding males [with testes larger than 90 mg, in the April and June samples (Phillmore et al., 2006)] with nonbreeding adult males (testes smaller than 10 mg). The results were again no different from those of the main analysis. RA. RA could be measured in a total of 50 birds. RA volumes were larger in males than in females [F(1,42) ¼ 29.418, p < 0.001; M/F ratio, 1.39], but did not differ among age classes [F(1,42) ¼ 0.184, p ¼ 0.670] or seasons [F(5,42) ¼ 0.525; Fig. 3(B)]. Comparing adults in all three manners explained for HVC made no difference to the results. Area X. Area X could be measured in a total of 50 birds. Area X volumes were larger in males than in females [F(1,42) ¼ 38.446, p < 0.001; M/F ratio, 1.60], but did not differ among age classes [F(1,42) ¼

Chickadee Song System Seasonal Patterns 5 Figure 3 Seasonal patterns in the volumes of the different song control regions. There are no significant overall seasonal effects in any of the regions. There is an interaction between sex and season for nxiits, indicating a larger nxiits in April males. All song control regions were larger in males than in females. Closed symbols represent males, and open symbols females. Circles represent adult birds, and triangles juveniles (up to 6 months after hatching). 1.224, p ¼ 0.275] or seasons [F(5,42) ¼ 1.334, p ¼ 0.269; Fig. 3(C)]. Comparing adults in all three manners explained for HVC made no difference to the results. LMAN. LMAN could be measured in a total of 47 birds. LMAN was larger in males than in females [F(1,37) ¼ 27.729, p < 0.001; M/F ratio, 1.39] and there was a significant interaction between season

6 Smulders et al. and age [F(2,37) ¼ 9.476, p < 0.001]. There was no main difference across seasons [F(5,37) ¼ 0.990, p ¼ 0.437], nor was there a main effect of age [F(1,37) ¼ 2.176, p ¼ 0.149; Fig. 3(D)]. Looking at only those time points in which we had both juveniles and adult birds (June, August, and October), we see that juvenile LMAN volumes get smaller across this time period, whereas adult volumes increase. For neither age class by itself is this trend significant, but the interaction between the two trends is. Comparing adults in all three manners explained for HVC made no difference to the results. nxiits. nxiits could be measured in a total of 31 birds. It was larger in adults than in juveniles [F(1,18) ¼ 5.926, p ¼ 0.026] and there was a significant interaction between season and sex [F(5,18) ¼ 3.727, p ¼ 0.017]. There was no main difference across seasons [F(5,18) ¼ 2.119, p ¼ 0.11], nor was there a main effect of sex [F(1,18) ¼ 0.781, p ¼ 0.388; Fig. 3E]. The interaction between sex and season can be explained by the fact that in females, there were no significant seasonal changes, whereas in males there were. April males had a significantly larger nxiits than males at other times of the year [F(5,12) ¼ 8.189, p ¼ 0.001]. Comparing adults only (in both manners explained for HVC) brought the significance level of the interaction down to p ¼ 0.075, probably due to a decrease in sample size. The other analyses on adults only did not result in any significant differences. Power Analyses. Nonsignificant results can be obtained because there is no effect of season to be detected, or because the analysis lacks the power to actually detect such an effect. In order to distinguish these two possibilities, we performed a power analysis (Cohen, 1988) to verify that we indeed had enough statistical power to detect seasonal changes, if they had existed in our sample. In order to perform a power analysis, one has to define the size of the effect one is looking for. We based our effect size on the two published studies on the song system of Parids: a laboratory study on black-capped chickadees (MacDougall-Shackleton et al., 2003), and a field study on blue tits (Caro et al., 2005). The significant effect sizes in these studies ranged from f ¼ 0.56 for the changes in HVC in black-capped chickadees to f ¼ 0.77 for RA in the blue tits. These are large effect sizes by general standards (Cohen, 1988), but actually rather small compared to the effect sizes obtained in other songbird species in the past (ranging from 1.0 to 2.5; calculations not presented). The power values obtained for our analysis to detect these effect sizes vary between 80 for HVC, based on black-capped chickadees, and 98 for RA based on blue tits. A power value of 80 is generally considered high enough to be able to detect an effect, if the effect is actually present in the population (Cohen, 1988). Because our power calculation included all birds in our sample, we also reran the analysis for the adult birds only. In this case, we calculated the power to find differences between the males in breeding condition (large testes) and those in nonbreeding condition (small testes). As outlined above, we performed these analyses and did not find any differences for any of the song control system nuclei. Power values varied from 66 to 91, still acceptably high. Other Brain Areas. Results for hippocampal formation (HF), total telencephalon (Tel), nucleus rotundus (Rt), and entopallium (E) have been reported previously (Smulders et al., 1995). However, because we have added two new birds, and because in the original analysis we adjusted the volumes to account for differences in weight loss during cryoprotection, we report on these analyses here again. Total telencephalon volume is larger in October than at other times of the year [F(5,42) ¼ 8.122, p < 0.001] and is larger in juveniles than adults [F(1,42) ¼ 12.346, p ¼ 0.001], but shows no sex differences [F(1,42) ¼ 1.099, p ¼ 0.301]. The HF also is larger in October than at any other times of year, even when total telencephalon is included as a covariate [season, F(5,39) ¼ 8.43, p < 0.001; Tel, F(1,39) ¼ 22.315, p < 0.001], and is larger in adults than in juveniles [F(1,39) ¼ 41.450, p < 0.001], but again no sex differences [F(1,39) ¼ 0.054, p ¼ 0.818]. Rt has no sex or age differences, but is slightly larger in August than at any other times of the year [F(5,42) ¼ 3.241, p ¼ 0.015]. E is larger in males than in females [F(1,41) ¼ 7.763, p ¼ 0.008], but no different across seasons or ages when taking Tel into account [Tel, F(1,41) ¼ 8.563, p ¼ 0.006]. Finally, we also measured the lingual part of nxii, which innervates the tongue, rather than the trachea and syrinx. nxiil does not differ between ages, sexes, or seasons, nor are there any significant interactions among these variables. DISCUSSION Like in virtually all other songbird species studied, we found that the song system nuclei in black-capped chickadees are larger in males than in females. Unlike most other studies on temperate zone songbirds, but

Chickadee Song System Seasonal Patterns 7 consistent with another study on wild-caught blackcapped chickadees (Phillmore et al., 2006), we did not find evidence for seasonal changes in the volumes of any of the song control areas, with the possible exception of nxiits in males. The sex differences in the song system nuclei of black-capped chickadees are robust and reliable. The ratios of male to female song system nuclei, however, are relatively small, compared to some other species. Gahr and colleagues (1998) compared the ratios for different species and claimed that there was no relationship between the sexual dimorphism in singing behavior, and that in song system anatomy. Nevertheless, their table suggests that dueting species or species in which males and females sing approximately equal amounts have smaller M/F ratios than species in which the females do not sing at all. In a metaanalysis, MacDougall-Shackleton and Ball (1999) confirm this impression, showing that there is a relationship between the degree of song system dimorphism and song output dimorphism. Although it was originally reported that the fee-bee song is almost exclusively sung by males (Ficken et al., 1978), our results are consistent with more recent findings on chickadee song behavior, which show that female black-capped chickadees sometimes sing full-blown fee-bee songs (Smith, 1991) when separated from their mate in the egg-laying season. Males and females also employ a faint fee-bee in interactions around the nest (Ficken et al., 1978; Smith, 1991; K. A. Otter, personal observation). However, comparing our results to the tables produced by MacDougall-Shackleton and Ball (1999), the chickadee song system is closer in sexual dimorphism to dueting species than to species in which the females occasionally sing. This suggests that the females singing behavior may not be a sufficient explanation for the low sexual dimorphism. We do not see any seasonal changes in the song system of wild black-capped chickadees, and neither do Phillmore and coworkers (2006), even though seasonal changes in the song system of Parids have been reported before. Blue tit HVC and RA increase in size from winter to the breeding season in the field (Caro et al., 2005), and when black-capped chickadees are exposed to artificial light cycle changes in the laboratory, photostimulated birds also have a larger song system than photorefractory and photosensitive birds (MacDougall-Shackleton et al., 2003). In both of those cases, the size of the seasonal effect was smaller, however, than the effect size typically reported for other temperate songbirds in the wild [e.g., for HVC, blue tit: f ¼ 0.63; black-capped chickadee: f ¼ 0.56; white-crowned sparrow (Zonotrichia leucophrys nuttalli): f ¼ 2.58 (Brenowitz et al., 1998); rufous-sided towhee (Pipilo erythrophthalmus): f ¼ 1.1 (Brenowitz et al., 1991); dark-eyed junco (Junco hyemalis): f ¼ 1.79 (Gulledge and Deviche, 1997)]. This might indicate that seasonal changes in parids in general are smaller than in other songbirds studied to date. It is possible that the seasonal changes in the wild in black-capped chickadees are so small that we do not have the necessary power to detect them. But even if that is the case, that is an interesting finding that needs explanation. At first glance, our results, as well as those of Phillmore and colleagues (2006), seem contradictory to the results found in the laboratory with photoperiod manipulation (MacDougall-Shackleton et al., 2003). However, other such seemingly contradictory findings exist. The phenomenon of seasonal changes in the song system was first described in domestic canaries (Nottebohm, 1981), and has been replicated since (Kirn et al., 1991). However, a study of wild canaries found no changes in song system anatomy across seasons, despite obvious changes in the composition of the song and in hormone titers (Leitner et al., 2001b). Wild canaries change their syllable repertoire from the nonbreeding to the breeding season, even though repertoire size stays the same (Leitner et al., 2001a). Songs also become longer during the breeding season (Leitner et al., 2001a), just like we have observed with the fee-bee song of the chickadees. We also observed a lower variability in song length during the breeding season than during the nonbreeding season in the chickadees. Such a change in stereotypy has previously been suggested to be related to seasonal changes in the song system (Smith et al., 1997), but we do not see this. It is possible that the changes in the length of the fee-bee song and its variability are not really seasonal changes, but are due to other differences between our two song samples. Breeding season song was collected during dawn chorus, whereas winter song was collected in response to playback. This different context, as well as the fact that during dawn chorus the birds sing for much longer intervals than in a territorial intrusion simulation, may explain the subtle differences in song between the two seasons. However, even if the differences in song structure in our samples are not seasonal in nature, there are still large differences in the amount of singing between the seasons (Smith, 1991; Phillmore et al., 2006). The total amount of singing performed by the bird is known to directly affect the size of the song system (Sartor et al., 2002; Sartor and Ball, 2005). This effect probably works through an increase in the neurotrophic factor BDNF, which is upregulated in

8 Smulders et al. the song system by singing (Li et al., 2000; Li and Jarvis, 2001). This factor is known to contribute to the survival of new neurons in HVC (Rasika et al., 1999), which in turn may affect the rest of the song system (Brenowitz and Lent, 2001). The fact that we found a seasonal increase in nxiits volume in males suggests that its response to seasonal cues may not come from the telencephalic song system (Brenowitz and Lent, 2000), but may instead come from its target structure, the syrinx, which is well known to be sensitive to testosterone and to change seasonally, at least in other species (Luine et al., 1980; Luine et al., 1983; Bleisch et al., 1984). It remains to be determined whether syrinx mass changes seasonally in chickadees as well. The question therefore remains: why are there no seasonal changes in the chickadee song system? We suggest three possible and nonexclusive explanations. First, it could be that seasonal changes in the song system do occur within individual males, but that we fail to detect it at the population level. It is possible that male chickadees in the field are less synchronized in their breeding conditions than they are when the photoperiod is artificially manipulated. This would lead to grouping birds that are and are not in breeding condition together in the same groups, and preclude the detection of seasonal changes that do occur within each individual. However, we also split the birds by the best measure of breeding condition we have (testis size), and if anything, breeding birds had slightly smaller song systems than nonbreeding birds. A second explanation is a mechanistic one. We found clear seasonal changes in the hippocampal formation of our sample of birds, with more neurons in the hippocampus in the October sample than at other times of the year (Smulders et al., 2000). This is presumably due to an increased incorporation of newly generated neurons in the autumn (Barnea and Nottebohm, 1994; Hoshooley and Sherry, 2004). It is possible that the evolutionary pressures that led to the regulation of neuronal incorporation into the HF in the autumn, had to tap into the same molecular mechanisms as are used to regulate neuronal incorporation in HVC in other songbirds. This may have led to a mechanistic shift from regulating new neurons in HVC to the HF, leading to a lack of seasonal changes in HVC, and in turn in the rest of the song system (Brenowitz and Lent, 2001). Because of its reliance on food hoarding in the autumn, this songbird lineage may therefore have lost seasonal changes to the song system. A final possible explanation is a functional one. If the amount of singing influences the sizes of the song system nuclei, and these nuclei do not change seasonally, then maybe the fee-bee song is not the only vocalization controlled by the song system. This would also explain the relatively small sexual dimorphism alluded to above. Because of their winter social organization, chickadees have a richer vocal repertoire than many other oscine song birds (Hailman et al., 1987). This repertoire includes many vocalizations that cannot be classified as song, but are nevertheless experience-dependent and plastic throughout the lifetime of the individual, as well as across the year. These vocalizations are also less sexually dimorphic than most songs. Two examples are the gargle vocalization (a short-distance aggressive vocalization) and the chick-a-dee vocalization (a flock-cohesion contact call; Ficken et al., 1978). Both vocalizations have learned components to them that remain plastic in adulthood, and both are produced by males and females (although the gargle more by males; Mammen and Nowicki, 1981; Ficken and Weise, 1984; Ficken et al., 1987; Shackleton et al., 1992; Kroodsma et al., 1995; Hughes et al., 1998). Because the song system is believed to be involved in the acquisition, production, and modification of learned vocalizations, it is possible that it is involved in the learning and production of these other vocalizations as well, as it is in the zebra finch long call (Simpson and Vicario, 1990). Only by directly interfering with these structures and investigating the effects of this interference on the learning and production of these nonsong vocalizations can we test this hypothesis. In conclusion, we find that the chickadee song system is unexpectedly stable in size across the seasons, even though singing (sensu stricto) is a seasonal behavior, as it is in other songbirds. The song system is sexually dimorphic in black-capped chickadees, but the magnitude of this dimorphism is similar to that in duetting species. Of the three possible explanations, the one that best explains both the lack of seasonality and the small sexual dimorphism is the hypothesis that the chickadee song system controls more than just the traditionally recognized fee-bee song, but may also control other complex social vocalizations. This hypothesis remains to be tested. André Dhondt, Barbara Finlay, Scott MacDougal- Shackleton, and Elizabeth Adkins-Regan, as well as two anomymous reviewers, provided useful comments on earlier versions of this manuscript. REFERENCES Arai O, Taniguchi I, Saito N. 1989. Correlation between the size of the song control nuclei and plumage color change in orange bishop birds. Neurosci Lett 98:144 148.

Chickadee Song System Seasonal Patterns 9 Barnea A, Nottebohm F. 1994. Seasonal recruitment of hippocampal neurons in adult free-ranging black-capped chickadees. Proc Natl Acad Sci U S A 91:11217 11221. Benton S, Nelson DA, Marler P, Devoogd TJ. 1998. Anterior forebrain pathway is needed for stable song expression in adult male white-crowned sparrows (Zonotrichia leucophrys). Behav Brain Res 96:135 150. Bernard DJ, Ball GF. 1995. Seasonal changes in the volume of song control nuclei in free-living male and female European starlings. Soc Neurosci Abstr 21:961. Bernard DJ, Wilson FE, Ball GF. 1997. Testis-dependent and -independent effects of photoperiod on volumes of song control nuclei in American tree sparrows (Spizella arborea). Brain Res 760:163 169. Bleisch W, Luine VN, Nottebohm F. 1984. Modification of synapses in androgen-sensitive muscle. I. Hormonal regulation of acetylcholine receptor number in the songbird syrinx. J Neurosci 4:786 792. Brenowitz EA, Arnold AP. 1986. Interspecific comparisons of the size of neural song control regions and song complexity in duetting birds: evolutionary implications. J Neurosci 6:2875 2879. Brenowitz EA, Baptista LF, Lent K, Wingfield JC. 1998. Seasonal plasticity of the song control system in wild Nuttall s white-crowned sparrows. J Neurobiol 34:69 82. Brenowitz EA, Lent K. 2000. Intracerebral implants of testosterone induce seasonal-like growth of adult avian song control circuits. Soc Neurosci Abstr 470:14. Brenowitz EA, Lent K. 2001. Afferent input is necessary for seasonal growth and maintenance of adult avian song control circuits. J Neurosci 21:2320 2329. Brenowitz EA, Margoliash D, Nordeen KW. 1997. An introduction to birdsong and the avian song system. J Neurobiol 33:495 500. Brenowitz EA, Nalls B, Wingfield JC, Kroodsma DE. 1991. Seasonal changes in avian song nuclei without seasonal changes in song repertoire. J Neurosci 11:1367 1374. Caro SP, Lambrechts MA, Balthazart JB. 2005. Early seasonal development of brain song control nuclei in male blue tits. Neurosci Lett 386:139 144. Cohen J. 1988. Statistical power analysis for the behavioral sciences. Hillsdale, NJ: L. Erlbaum Associates. 567 p. DeVoogd TJ, Houtman AM, Falls JB. 1995. White-throated sparrow morphs that differ in song production rate also differ in the anatomy of some song-related brain areas. J Neurobiol 28:202 213. DeVoogd TJ, Nixdorf B, Nottebohm F. 1985. Formation of new synapses related to acquisition of a new behavior. Brain Res 329:304 308. DeVoogd TJ, Szekely T. 1998. Causes of avian song: using neurobiology to integrate proximal and ultimate levels of analysis. In: Pepperberg I, Kamil A, Balda R, editors. Animal cognition in nature. New York: Academic Press, p 337 380. Dloniak SM, Deviche P. 2001. Effects of testosterone and photoperiodic condition on song production and vocal control region volumes in adult male dark-eyed juncos (Junco hyemalis). Horm Behav 39:95 105. Ficken MS, Ficken RW, Witkin SR. 1978. Vocal repertoire of the black-capped chickadee. Auk 95:34 48. Ficken MS, Weise CM. 1984. A complex call of the blackcapped chickadee (Parus atricapillus). I. Microgeographic variation. Auk 101:349 360. Ficken MS, Weise CM, Reinartz JA. 1987. A complex vocalization of the black-capped chickadee. II. Repertoires, dominance and dialects. Condor 89:500 509. Gahr M, Sonnenschein E, Wickler W. 1998. Sex difference in the size of the neural song control regions in a dueting songbird with similar song repertoire size of males and females. J Neurosci 18:1124 1131. Gammon DE, Baker MC. 2004. Song repertoire evolution and acoustic divergence in a population of black-capped chickadees, Poecile atricapillus. Anim Behav 68:903 913. Gammon DE, Baker MC, Tipton JR. 2005. Cultural divergence within novel song in the black-capped chickadee (Poecile atricapillus). Auk 122:853 871. Gulledge CC, Deviche P. 1997. Androgen control of vocal control region volumes in a wild migratory songbird (Junco hyemalis) is region and possibly age dependent. J Neurobiol 32:391 402. Hailman JP, Ficken MS, Ficken RW. 1987. Constraints on the structure of combinatorial chick-a-dee calls. Ethol 75: 62 80. Hill BG, Lein MR. 1987. Function of frequency-shifted songs of black-capped chickadees. Condor 89:914 915. Horn AG, Leonard ML, Ratcliffe L, Shackleton SA, Weisman RG. 1992. Frequency variation in songs of blackcapped chickadees (Parus atricapillus). Auk 109:847 852. Hoshooley JS, Sherry DF. 2004. Neuron production, neuron number, and structure size are seasonally stable in the hippocampus of the food-storing black-capped chickadee (Poecile atricapillus). Behav Neurosci 118:345 355. Hughes M, Nowicki S, Lohr B. 1998. Call learning in black-capped chickadees (Parus atricapillus): the role of experience in the development of chick-a-dee calls. Ethol 104:232 249. Johnson F, Bottjer SW. 1995. Differential estrogen accumulation among populations of projection neurons in the higher vocal center of male canaries. J Neurobiol 26:87 108. Kirn JR, Alvarez-Buylla A, Nottebohm F. 1991. Production and survival of projection neurons in the high vocal center of adult male canaries. J Neurosci 11:3167 3187. Kirn JR, Clower RP, Kroodsma DE, DeVoogd TJ. 1989. Song-related brain regions in the red-winged blackbird are affected by sex and season but not repertoire size. J Neurobiol 20:139 169. Kroodsma DE, Albano DJ, Houlihan PW, Wells JA. 1995. Song development by black-capped chickadees (Parus atricapillus) and carolina chickadees (P. carolinensis). Auk 112:29 43. Leitner S, Voigt C, Gahr M. 2001a. Seasonal changes in the song pattern of the non-domesticated island canary (Serinus canaria), a field study. Behaviour 138:885 904.

10 Smulders et al. Leitner S, Voigt C, Garcia-Segura LM, Van t Hof T, Gahr M. 2001b. Seasonal activation and inactivation of song motor memories in wild canaries is not reflected in neuroanatomical changes of forebrain song areas. Horm Behav 40:160 168. Li D-F, Fu L-B, Zhang J-L. 1996. Seasonal differences of the volume of the forebrain song control nuclei in songbird (Emberiza rutila). In: Avian brain and behaviour meeting. Tihany, Hungary: Hungarian Academy of Sciences. Li X, Jarvis E. 2001. Sensory- and motor-driven BDNF expression in a vocal communication system. Soc Neurosci Abstr 538:8. Li XC, Jarvis ED, Alvarez-Borda B, Lim DA, Nottebohm F. 2000. A relationship between behavior, neurotrophin expression, and new neuron survival. Proc Natl Acad Sci U S A 97:8584 8589. Luine VN, Harding CF, Bleisch W. 1983. Specificity of gonadal hormone modulation of cholinergic enzymes in the avian syrinx. Brain Res 279:339 342. Luine VN, Nottebohm F, Harding CF, McEwen BS. 1980. Androgen affects cholinergic enzymes in songbird syringeal motor neurons and muscle. Brain Res 192:89 107. MacDougall-Shackleton SA, Ball GF. 1999. Comparative studies of sex differences in the song-control system of songbirds. TINS 22:432 436. MacDougall-Shackleton SA, Hernandez AM, Valyear KF, Clark AP. 2003. Photostimulation induces rapid growth of song-control brain regions in male and female chickadees (Poecile atricapilla). Neurosci Lett 340:165 168. Mammen DL, Nowicki S. 1981. Individual differences and within-flock convergence in chickadee calls. Behav Ecol Sociobiol 9:179 186. Margoliash D. 1997. Functional organization of forebrain pathways for song production and perception. J Neurobiol 33:671 693. Mennill DJ, Ratcliffe LM. 2004a. Do male black-capped chickadees eavesdrop on song contests? A multi-speaker playback experiment. Behaviour 141:125 139. Mennill DJ, Ratcliffe LM. 2004b. Overlapping and matching in the song contests of black-capped chickadees. Anim Behav 67:441 450. Mennill DJ, Ratcliffe LM, Boag PT. 2002. Female eavesdropping on male song contests in songbirds. Science 296:873. Nordeen KW, Nordeen EJ. 1997. Anatomical and synaptic substrates for avian song learning. J Neurobiol 33:532 548. Nottebohm F. 1981. A brain for all seasons: cyclical anatomical changes in song control nuclei of the canary brain. Science 214:1368 1370. Nottebohm F, Nottebohm ME, Crane LA. 1986. Developmental and seasonal changes in canary song and their relation to changes in the anatomy of song-control nuclei. Behav Neural Biol 46:445 471. Otter K, Chruszcz B, Ratcliffe L. 1997. Honest advertisement and song output during the dawn chorus of blackcapped chickadees. Behav Ecol 8:167 173. Otter K, Ratcliffe L. 1993. Changes in singing behavior of male black-capped chickadees (Parus atricapillus) following mate removal. Behav Ecol Sociobiol 33: 409 414. Otter KA, Ratcliffe L, Njegovan M, Fotheringham J. 2002. Importance of frequency and temporal song matching in black-capped chickadees: Evidence from interactive playback. Ethol 108:181 191. Phillmore LS, Hoshooley JS, Sherry DF, MacDougall- Shackleton SA. 2006. Annual cycle of the black-capped chickadee: seasonality of singing rates and vocal-control brain regions. J Neurobiol 66: Rasika S, Alvarez-Buylla A, Nottebohm F. 1999. BDNF mediates the effects of testosterone on the survival of new neurons in an adult brain. Neuron 22:53 62. Ratcliffe L, Otter K. 1996. Sex differences in song recognition. In: Kroodsma DE, Miller EH, editors. Ecology and evolution of acoustic communication in birds. Ithaca, NY: Cornell University Press, p 339 355. Ratcliffe L, Weisman RG. 1985. Frequency shift in the FeeBee song of the black-capped chickadee. Condor 87: 555 556. Reiner A, Perkel DJ, Bruce LL, Butler AB, Csillag A, Kuenzel W, Medina L, Paxinos G, Shimizu T, Striedter G, Wild M, Ball GF, Durand S, Güntürkün O, Lee DW, Mello CV, Powers A, White SA, Hough G, Kubikova L, Smulders TV, Wada K, Dugas-Ford J, Husband S, Yamamoto K, Yu J, Siang C, Jarvis ED. 2004. Revised nomenclature for avian telencephalon and some related brainstem nuclei. J Comp Neurol 473:377 414. Rucker M, Cassone VM. 1991. Song control nuclei of the house sparrow (Passer domesticus) are photoperiodic and photorefractory. Soc Neurosci Abstr 17:1051. Sartor JJ, Ball GF. 2005. Social suppression of song is associated with a reduction in volume of a song-control nucleus in European starlings (Sturnus vulgaris). Behav Neurosci 119:233 244. Sartor JJ, Balthazart J, Ball GF. 2005. Coordinated and dissociated effects of testosterone on singing behavior and song control nuclei in canaries (Serinus canaria). Horm Behav 47:467 476. Sartor JJ, Charlier T, Pytte CL, Balthazart J, Ball GF. 2002. Converging evidence that song performance modulates seasonal changes in the avian song control system. Soc Neurosci Abstr 781:10. Schlinger BA. 1997. Sex steroids and their actions on the birdsong system. J Neurobiol 33:619 631. Shackleton SA, Ratcliffe L. 1993. Development of song in hand-reared black-capped chickadees. Wilson Bull 105: 637 644. Shackleton SA, Ratcliffe L. 1994. Matched counter-singing signals escalation of aggression in black-capped chickadees (Parus atricapillus). Ethol 97:310 316. Shackleton SA, Ratcliffe L, Weary DM. 1992. Relative frequency parameters and song recognition in black-capped chickadees. Condor 94:782 785. Shiflett MW, Gould KL, Smulders TV, DeVoogd TJ. 2002. Septum volume and food-storing behavior are related in parids. J Neurobiol 51:215 222.

Chickadee Song System Seasonal Patterns 11 Simpson HB, Vicario DS. 1990. Brain pathways for learned and unlearned vocalizations differ in zebra finches. J Neurosci 10:1541 1556. Smith GT. 1996. Seasonal plasticity in the song nuclei of wild rufous-sided towhees. Brain Res 734:79 85. Smith GT, Brenowitz EA, Beecher MD, Wingfield JC. 1997. Seasonal changes in testosterone, neural attributes of song control nuclei, and song structure in wild songbirds. J Neurosci 17:6001 6010. Smith GT, Brenowitz EA, Wingfield JC, Baptista LF. 1995. Seasonal changes in song nuclei and song behavior in Gambel s white-crowned sparrows. J Neurobiol 28:114 125. Smith S. 1991. The black-capped chickadee: behavioral ecology and natural history. Ithaca, NY: Cornell University Press. 362 p. Smulders TV. 2002. Natural breeding conditions and artificial increases in testosterone have opposite effects on the brains of adult male songbirds: a meta-analysis. Horm Behav 41:156 169. Smulders TV, Sasson AD, Devoogd TJ. 1995. Seasonal variation in hippocampal volume in a food-storing bird, the black-capped chickadee. J Neurobiol 27:15 25. Smulders TV, Shiflett MW, Sperling AJ, DeVoogd TJ. 2000. Seasonal changes in neuron numbers in the hippocampal formation of a food-hoarding bird: the blackcapped chickadee. J Neurobiol 44:414 422. Weisman R, Ratcliffe L, Johnsrude I, Hurley TA. 1990. Absolute and relative pitch production in the song of the black-capped chickadee. Condor 92:118 124. Wild JM. 1997a. Functional anatomy of neural pathways contributing to the control of song production in birds. Eur J Morphol 35:303 325. Wild JM. 1997b. Neural pathways for the control of birdsong production. J Neurobiol 33:653 670.