Daily Light Regulates Seasonal Responsesin the Migratory Male Redheaded Bunting(Emberiza bruniceps)

SANGEETA RANI1, SUDHI SINGH1, MANJU MISRA1, SHALIE MALIK1,BHANU PRATAP SINGH2, AND VINOD KUMAR1�1Department of Zoology, University of Lucknow, Lucknow 226 007, India2Department of Science and Technology, New Mehrauli Road,New Delhi 110 016, India

ABSTRACT This study analyzed the role of day length in regulation of seasonal body fatteningand testicular growth in a latitudinal Palaearctic-Indian migrant, the redheaded bunting (Emberizabruniceps). When exposed to increasing photoperiods (hours of light: hours of darkness; 11.5L:12.5D,12L:12D, 12.5L:11.5D, 13L:11D, 14L:10D, and 18L:6D) for 9–12 weeks, buntings responded in aphotoperiod-dependent manner and underwent growth and regression cycle under photoperiods ofX12 hr per day. Also, the response to a long photoperiod of birds that were held under naturalphotoperiods at 271N for 2 years was similar to those who arrived the same year from their breedinggrounds (�401N), suggesting that the experience of higher amplitude day–night (light–dark, LD)cycles during migratory and breeding seasons were not critical for the subsequent response(initiation–termination–reinitiation) cycle. Another experiment examined entrainment of thecircadian photoperiodic rhythm in buntings by subjecting them to T 5 2472 hr LD-cycles with8 hr photophase and to T 5 22 and 24 hr with 11 hr photophase. The results showed a reduction incritical day length under T 5 22 hr LD-cycle. In the last experiment, we constructed an actionspectrum for photoperiodic induction by exposing birds for 4.5 weeks to 13L:11D of white (control),blue (450 nm), or red (640 nm) light at irradiances ranging from 0.028 to 1.4 W m�2. The thresholdlight irradiance for photoinduction was about 10-fold higher for blue light, than for red and whitelights. These results conclude that the daily light of the environment regulates the endogenousprogram that times seasonal responses in body fattening and testicular cycles of the redheadedbunting. J. Exp. Zool. 303A:541– 550, 2005. r 2005 Wiley-Liss, Inc.

INTRODUCTION

In long day breeding birds, increasing daylengths of spring and summer induce gonadalrecrudescence and time the post-reproductivegonadal regression and photorefractoriness (Ni-cholls et al., ’88; Dawson et al., 2001; Deviche andSmall, 2001). Day lengths that cause photoperiodicinduction and time photorefractoriness are spe-cies-specific; this reflects adaptiveness to a photo-periodic environment that a species inhabits.Experimentally, subjecting individuals for manyweeks to a range of photoperiods can show this. Ina study on Japanese quail (Coturnix c. japonica),Follett and Maung (’78) found that the rate oftesticular growth was slower by 50% in birdsexposed to 12 hr of light per day (12L:12D)compared with 14 or longer light hours per day;a near full growth occurred in 13L:11D. It willbe of interest to examine induction–regression

response to a range of photoperiods of severalmore species to understand how closely theendogenous time-keeping system is coupled today lengths in the wild.

A photoperiodic response results when daylength interacts with the underlying circadianphotoperiodic response system such that dailylight extends into the photoinducible phase ofthe circadian photosensitivity rhythm; in a natur-al lighting environment, the photoinducible phaseoccurs some 12 hr after dawn each day (Pitten-drigh, ’72; Kumar and Follett, ’93; Dawson et al.,

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jez.a.187.

Received 23 August 2004; Accepted 19 April 2005

Grant sponsor: Department of Science and Technology, New Delhiand Council of Scientific and Industrial Research, New Delhi.�Correspondence to: Vinod Kumar, Department of Zoology,

University of Lucknow, Lucknow 226 007, India.E-mail: drvkumar11@yahoo.com

r 2005 WILEY-LISS, INC.

JOURNAL OF EXPERIMENTAL ZOOLOGY 303A:541–550 (2005)

2001; Kumar, 2001). This is demonstrated byseveral lighting protocols including a light:dark(LD) cycle with variable periods (T LD-cycle,T 5 period length of an LD cycle; Hamner andEnright, ’67; Farner et al., ’77; Tewary andKumar, ’82; Tewary et al., ’82). T LD-cycles arenon-24 hr and usually contain short light periodsof p8 hr. One interpretation of how photoperiodicinduction occurs under a T LD-cycle containingshort light phase is that by varying T, the phaseangle between an LD cycle and the entrainedphotoperiodic rhythm (i.e., temporal relationshipbetween identical phases of the daily LD cycle andrhythm mediating photoperiodic response) be-comes sufficiently large to bring coincidence of atleast part of the photoinducible phase with thelight period. This interpretation is based onactivity rhythm recordings under inductive TLD-cycles (for details, see Hamner and Enright,’67; Follett and Pearce-Kelly, ’91) implying thatthe same circadian pacemaker system regulatescircadian rhythms of locomotion and photosensi-tivity. This may not always be the case, however,as suggested by results from the Japanese quail(Kumar et al., ’92). Furthermore, Kumar andKumar (’95) show that the photoperiodic induc-tion in the blackheaded bunting (Emberiza mela-nocephala) is not caused by increase in the phaseangle difference as noted above since T 5 22 orT 5 26 hr LD cycles with short photophase werenon-inductive. Rather, a T 5 2472 hr LD-cycleentrained the circadian photoperiodic rhythmand consequently altered critical day length (i.e.,minimum day length that will induce a response)for photoperiodic induction (Kumar and Kumar,’95). This has not been tested, however, in otherspecies.

Apart from duration, the wavelength and in-tensity (irradiance) of daily light also influence thephotoperiodic response (Rani et al., 2002). Longlight wavelengths and increasing light intensity,until threshold intensity is reached, acceleratephotoperiodic induction (see, especially, Benoit,’64; Oishi and Lauber, ’73; Kumar and Rani, ’96).Interestingly, in the European starling (Sturnusvulgaris) exposed to 18L:6D, reductions in lightintensity of the photophase attenuate photoper-iodic induction as if birds are exposed to relativelyshorter stimulatory photoperiods (Bentley et al.,’98). Our findings on buntings also show thatcircadian photosensitivity rhythm is affected byboth the light wavelength and intensity (Kumarand Rani, ’96; Rani and Kumar, ’99, 2000; Raniet al., 2001). Recently, we have reported that

photoperiodic response system of the blackheadedbunting is highly responsive to changes in wave-length and intensity of light (Malik et al., 2004;Misra et al., 2004).

A latitudinal migrant species can be a usefulmodel to answer the questions related to photo-periodism, because it experiences varying photo-periodic conditions each year—one at the breedingground, another at the wintering ground, andstill another (consistently varying day lengths)—during migration between the breeding andwintering grounds. For example, Palaearctic-In-dian avian migrants fly south in the autumn (fallor autumnal migration) to their winteringgrounds, and then return north in the spring(spring or vernal migration) to their breedinggrounds. We have been studying photoperiodismin two such migrant species, the blackheadedbunting (E. melanocephala) and the redheadedbunting (Emberiza bruniceps), that breed insummer in west Asia and east Europe (�401N)and overwinter in India (251N). Both species arehighly photosensitive: long photoperiods (16L:8D)induce body fattening and gonadal growth anddevelopment, and subsequently spontaneous re-gression and photorefractoriness follow if thebirds are exposed to stimulatory photoperiods forca. 12 weeks or more (Misra, 2002; Misra et al.,2004).

In this paper, we report results from threeexperiments performed on redheaded bunting(E. bruniceps). The first experiment determinedphotoperiodic induction and regression in photo-sensitive unstimulated birds exposed to a range ofphotoperiods from 11.5 to 18 hr light per day.A subset of this experiment also examinedwhether birds that were maintained on naturalphotoperiods at 271N for 2 years, and hence didnot have any experience of the photoperiodicconditions of breeding grounds and several lati-tudes during migration, respond similar to thosearrived afresh from the breeding grounds. Thesecond experiment investigated whether photo-periodic induction under T 5 2472 hr LD-cyclescan change the critical day length as a result of theentrainment to it. The last experiment studied thedynamics of the photoperiodic system by con-structing an action spectrum response for photo-periodic induction. Birds were subjected to shortand long light wavelengths at several energylevels, similar to that recently reported for theblackheaded bunting (Malik et al., 2004). Anaction spectrum for photoperiodic induction isparticularly useful since it reflects the cumulative

S. RANI ET AL.542

response of the photoperiodic clock (photorecep-tor, oscillator, and output) to daily light.

MATERIALS AND METHODS

The redheaded bunting is a long distancePalaearctic-Indian latitudinal migrant that over-winters in India. Buntings arrive in India inOctober/November, overwinter and return to theirbreeding grounds in late March/April. The experi-ments were done on male birds that were procuredin February from the overwintering flock at 251N.They were held under short days (8L:16D), unlessotherwise stated, until exposed to an experimentallighting condition ensuring their photosensitivityto light. Under such short days, buntings remaingonadally unstimulated; we have referred to theseshort-day-treated birds as photosensitive birds.

Experiment 1: effect of increasingphotoperiods

Experiment 1 contained three sub-experiments(A, B, and C). The first two sub-experimentsinvestigated whether day length regulates season-ality in body fattening and testicular growth anddevelopment in the redheaded bunting. Experi-ment 1A examined response of body fattening andtesticular recrudescence to increasing photoper-iods corresponding to natural day lengths (NDL)that a bunting experiences during its springmigration from 251N to �401N, and its breedinglocation, �401N (Misra et al., 2004). We startedthis experiment in the third week of June 2000 onphotosensitive buntings (i.e., birds maintainedon 8L:16D and hence with small unstimulatedtestes), the time around which they would nor-mally be at their breeding grounds with largegonads. Groups of birds (N 5 5–7 per group) wereexposed for 9 weeks to 11.5, 11.75, 12, 12.25, 12.5,and 13 hr light per day at an intensity of �450 lux.They were examined for changes in body mass andtesticular size at 1.5- and 3-week intervals,respectively.

Experiment 1B compared response of buntingsto two long photoperiods: one photoperiod wasclose to that buntings would experience as NDL at271N (Misra et al., 2004) and the other was longerthan what buntings experience even at theirbreeding grounds. This experiment began in thesecond week of April 2001, the time around whichbuntings begin to recrudesce their testes (Fig. 1D).Groups of photosensitive birds (N 5 6–7 per group)were exposed to 14L:10D and 18L:6D for 12 weeksat an intensity of �450 lux. Data from a group of

birds kept on NDL and observed along withexperimental groups served as control. Changesin body mass and testicular size were recordedat intervals of 2 and 3 or 4 weeks, respectively(Fig. 1C and D).

Experiment 1C asked whether buntings de-prived of experiencing day lengths that they wouldnormally experience during their migratory jour-ney and at breeding grounds would exhibit normalseasonal cycles in body mass and testes. Wecompared responsiveness of buntings that weremaintained at natural photoperiods of 271N sinceMarch 2000 for 2 years, and so had not experi-enced changing photoperiods of different latitudesthrough which they migrate or breed, with thosethat were procured in February 2002 from freshlyarrived overwintering flock. In mid March 2002,groups (N 5 7–9) of birds from both stocks, the onewhich was held in the aviary since February 2000and the other which was held since February 2002,were subjected to 16L:8D for 3 weeks. Their bodymass and testis size were recorded at the begin-ning and at the end of the experiment.

Experiment 2: effects of T LD-cycles

T LD-cycles refer to non-24 hr photoperiodswhose period length (T) is varied to test the rangeof the entrainment of endogenous circadianrhythm. Several lines of evidence strongly suggestthe involvement of circadian rhythm in photoper-iodic time measurement of the redheaded bunting(Rani and Kumar, 2000), and so we investigatedwhether varying T of an LD-cycle by 2 hr(T 5 2472 hr; T 5 22 or 26 hr) will influence theentrainment and photoinduction of the circadianphotosensitivity rhythm. This experiment specifi-cally asked the question as to whether photoper-iodic induction under T 5 2472 hr LD-cycles isthe result of change in the critical day length. Itbegan in the second week of October 2000, andwas completed in two subsets. In experiment 2A,we investigated whether exposure to T 5 22 or26 hr will produce phase angle between theendogenous circadian photoperiodic rhythm andT of the LD-cycle large enough so that the shortphotophase given in a T LD-cycle extends into thephotoinducible phase and results into photoinduc-tion. Birds (N 5 7–8 per group) were exposed to8L:14D, 8L:16D, and 8L:18D. A T 5 2472 hr LD-cycle paradigm appears to entrain the circadianlocomotor rhythm in this species (Rani, ’99). Inexperiment 2B, we investigated whether photo-induction under a T LD-cycle is the result of the

SEASONAL RESPONSES IN THE MIGRATORY MALE REDHEADED BUNTING 543

entrainment of circadian photoperiodic rhythmand the consequent change of critical day length.Here, birds (N 5 5–6 per group) were exposed toT 5 22 and 24 hr LD-cycles with 11 hr photophase(11L:11D, 11L:13D). The intensity during lightphase was at �450 lux. Observations on body massand testis size were taken at the beginning andend of the 3-week experiment.

Experiment 3: effects of light wavelengths

We studied differential effects of light wave-length on photoperiodic induction by constructingan action spectrum for photoperiodic responseunder a short and a long light wavelength to

address the question whether photoreceptorsmediating photoperiodism in the redheaded bunt-ing have differential spectral sensitivity. Birdswere exposed for 4.5 weeks to long photoperiods(13L:11D) at intensities ranging from 0.028 to1.4 W m�2, with photophase in blue (450 nm), red(640 nm), or white (control) light. This was a two-part study; each part used six groups (N 5 8 pergroup) of photosensitive birds. In the first part,beginning in the second week of April 2001, birdswere subjected to 13L:11D of blue, red, and whitelights at 0.028 and 0.28 Wm�2 (one group perlight condition). The minimum light applied at0.028 W m�2 was designated as 1� level irradi-ance, which served as the reference for higher

Fig. 1. Mean (7sem) body mass and testis volume of the redheaded bunting in response to different photoperiodicconditions. Left: (A and B; Experiment 1A): effect of changing photoperiods: photosensitive birds (N 5 5–7) were exposed todifferent photoperiods (ranging from 11.5, 11.75, 12, 12.25, 12.5 and 13 hr light per day) for 9 weeks. Note: (i) a fasterphotoinduction in 13L:11D and 12.5L:11.5D, significantly different from other groups in testis response, and (ii) a similarity inthe induction–regression response curve. Middle: (C and D; Experiment 1B): effect of long photoperiods: photosensitive birds(N 5 6–7 per group) were exposed to 14L:10D and 18L:6D for 12 weeks at an intensity of �450 lux. A group of birds kept onnatural day lengths (NDL) and observed along with experimental groups served as control. Note the similarity in peak responseand also in the induction–regression response curves in artificial and natural long photoperiods. Right: (E and F; Experiment1C): test of photoresponsivness: two groups of birds (N 5 7–9) that were either held captive in an outdoor aviary (under NDL) at271N for 2 years (since February 2000) or caught from a freshly arrived batch of birds (February 2002) were exposed to longphotoperiod (16L:8D). Note a small group difference in testis size at the beginning of the experiment. Significance of difference,Po0.05; �; group difference, ��, different from initial observation.

S. RANI ET AL.544

light irradiances used in the experiment. In thesecond part beginning in the first week of June2001, birds were exposed to red and white lights at2.5 and 5� levels, and to blue light at 25 and 50�levels. Thus, we had four results from each lightcolor: red and white lights—1, 2.5, 5, and 10�levels (0.028, 0.07, 0.14, 0.28 W m�2, respectively);blue light—1, 10, 25, and 50� levels (0.028, 0.28,0.7, 1.4 W m�2, respectively). Observations onbody mass and testis size were recorded at thebeginning, and after 4 weeks (body mass, day 29)or at the end (testes, day 32) of the experiment.

Food and water were provided ad libitum. Birdswere fed on seeds of Setaria italica. In an LD cycle,birds were held in groups of three or fourindividuals per cage (size: 45� 25� 25 cm) withinlight-tight boxes (size: 138� 60�56 cm) providingwhite or colored light produced by fluorescenttubes (Philips, India). The dark phase of an LDcondition always meant a very dim light at�0.01 W m�2. Varying intensities of white andcolored light were obtained by covering thefluorescent light tubes (14 W Philips) with neutraldensity and colored cinemoid filters, respectively,obtained from Rosco filters (Blanchard Works,London, UK). The transmission peaks for blue andred filters were 450 and 640 nm, respectively.

Light intensity measurements reflect light illu-mination at perch level within the cage. In nature,sunlight is available at much higher intensity thanthe artificial light used in photoperiodic studiesunder laboratory conditions. However, naturallight to which a bird species could adapt mightnot be the sunlight of an open sky (Thorington,’80), since birds usually inhabit shady areas wherelight intensity is substantially reduced. Further-more, several results including ours on buntingshave shown that the photoperiodic inductionunder white light above 100 lux faithfully reflectsthe response that is found in aviary birds undernatural lighting conditions (Misra, 2002; Misraet al., 2004). Automatic time switches controlledtimes of light-on and light-off. Temperature wasnot strictly regulated. Our photoperiodic boxes arehowever well aerated through inlets and outletsconnected to air circulators, and so the tempera-ture inside them does not vary more than 11C–21Cfrom the room temperature, which varies from281C in March to 321C in April to 351C in June to281C in October.

As indicated above, the induction and subse-quent regression of a photoperiodic response weremeasured by recording changes in the body massand testis size at the beginning and the end of the

experiment, and also at appropriate intervalsduring the experiments (see also figures). Bodymass was measured using a top pan balance to anaccuracy of 0.1 g. The testicular response wasassessed by laparotomy under local anesthesia(Kumar et al., 2001). The dimensions of the lefttestis were recorded, and testis volume wascalculated from 4/3 pab2, where a and b denotehalf of the long and short axes, respectively. Wealso had a subjective grading of the testis size toexplain the gonadal response: TV 5 0.33 too2.35 mm3, no response; 2.35 to o9.82 mm3,initiation of response; TV 5 9.82 to o18.86 mm3.small response; 18.86 to o41.9 mm3, moderateresponse; 41.9 mm3 and above, full response(Kumar et al., 2002).

The data are presented as means and SEs. Theywere analyzed statistically using one-way analysisof variance (ANOVA) with or without repeatedmeasures (RM), as appropriate, followed by a post-hoc Newman–Keuls test, if ANOVA indicated asignificance of difference. Various groups werealso compared by two-way ANOVA. A paired t-testwas used to compare before and after treatmentmeans of the same group. Comparison of twodifferent groups at a single time-point was doneusing the Student’s t-test. Significance was takenat Po0.05.

RESULTS

Experiment 1: effect of increasingphotoperiods

Experiment 1A

Birds fattened and gained in body mass in allphotoperiods (Po0.05, one-way RM ANOVA), butthe increase was most rapid in the 13L photoper-iod, and slowest in the 11.5L photoperiod. In fact,the amplitude of the response curve for body massunder 11.5L was small, although it was statisti-cally significant (F6,30 5 2.501, P 5 0.0441; one-way RM ANOVA). Birds of 11.5L were signifi-cantly leaner (Po0.05, one-way ANOVA, New-man–Keuls post-hoc test) than those of 12.25L,12.5L and 13L photoperiods after 4.5 weeks ofexposure (cf. Fig. 1A). Overall, the maximummean gain in body mass after 6 weeks of exposurein all photoperiods was as follows: 11.5L:12.5D,2.2270.61 g; 11.75L:12.25D, 4.3571.19 g; 12L:12D,4.6070.92 g; 12.25L:11.75D, 5.8371.9 7g; 12.5L:11.5D, 6.2371.40 g; and 13L:11D, 8.0971.69 g.After 7.5 weeks, body fat showed a decline and birdshad lost weight in all photoperiods (cf. Fig. 1A).

SEASONAL RESPONSES IN THE MIGRATORY MALE REDHEADED BUNTING 545

There was not much further decrease in bodymass by the end of the experiment. Two-wayANOVA indicated a significant effect of thephotoperiod (F5,210 5 8.130, Po0.0001) and theduration of exposure (F6,210 5 11.81, Po0.0001)but not of the interaction between the photoperiodand the duration (F30,210 5 0.492, P40.05).

Testes underwent a growth–involution cycle inall except 11.5L and 11.75L photoperiods (Fig. 1B).Although at the end of the first 3-week exposure,testes were stimulated in all groups, the rate oftestis growth progressively increased. In 11.5L, forexample, there was only an initiation of testicularresponse (mean TV 5 3.0370.46 mm3). In 11.75L,on the other hand, four of six birds had initiationand two had attained small testicular response(mean TV 5 8.2972.36 mm3). In four other photo-periods, mean TV at the end of first 3 weeks wasas follows: 12L:12D, 14.5472.70 mm3; 12.25L:11.75D, 17.1872.05 mm3; 12.5L:11.5D, 25.9574.73 mm3; 13L:11D, 26.8972.89 mm3. Thus, birdsof 12.5L and 13L had a significantly largerresponse (Po0.05, one-way ANOVA) than that ofbirds in other photoperiods. Birds of 11.5L had thesmallest response and the other three groups hadintermediate response. The peak testis response in12L and 12.25L was attained at the same time(after 6 weeks) as in 12.5L and 13L photoperiods.In 11.5L and 11.75L photoperiods, testes were stillincreasing in size at the end of the experimentwhen testes were found regressing in birds underthe rest of the photoperiods. Two-way ANOVAindicated a significant effect of the photoperiod(F5,120 5 12.29, Po0.0001), the duration of expo-sure (F3,120 5 40.92, Po0.0001), and the interac-tion between the photoperiod and the duration(F15,120 5 2.795, P 5 0.0010).

Experiment 1B

As expected, buntings exposed to both longphotoperiods (14L and 18L) underwent a completeinduction–regression cycle, as they did in NDL atthis time of the year (Fig. 1C and D). That is, theyfirst fattened and gained in body mass andrecrudesced testes, and then underwent sponta-neous depletion in fat stores and loss in body mass,and testicular regression (Fig. 1C and D). This wassimilar to the induction–regression cycle in bodymass and testes in birds under NDL, observedfrom mid-April to mid-July. The induction–regres-sion response curves of all the three groups weresignificant (Po0.05, one-way RM ANOVA), andthe peak response was similar in artificial and

natural long photoperiods. However, the shapeof the response curve of NDL was less steep (cf.Fig. 1C and D). There was a significantly higherbody mass and slightly stimulated testes (Po0.05,one-way ANOVA; Newman–Keuls post-hoc test) ofNDL birds at the beginning of the experimentbecause of the initiation of photostimulation inthese birds in response to gradually increasing daylengths; those exposed to artificial day lengthscame from short days. By the end of the experi-ment, however, all birds were regressed.

Experiment 1C

There was a significant gain in body mass(Po0.0001, one-way RM ANOVA) and full enlar-gement of testes after 3 weeks of exposure to16L:8D in both the groups, the one which wascaptive for 2 years at 271N and the other whichwas procured from the freshly arrived flock(Fig. 1E and F). At the beginning of the experi-ment, both groups had similar body mass buttestes in captives were slightly stimulated andhence larger than the freshly arrived birds(Po0.05, Student’s t-test). Captives fattened ra-pidly as made evident by the observation after 10days of long days when they weighed significantlyhigher (Po0.05, Student’s t-test) than freshlyarrived birds (gain in body mass on day 11:captives—27.6072.83%, N 5 7; fresh birds—16.1072.81%, N 5 8). The gain in the second halfof the experiment (day 11–day 21) for captive andfresh birds was �13.6% and �23.3%, respectively.Thus, at the end, both groups had similar fat storesand gain in body mass (captives: 41.275.93%,N 5 7; fresh birds: 39.473.45%, N 5 8) and theirbody mass was not different (Fig. 1E and F).

Experiment 2: effect of T LD-cycles

Figure 2 shows the results. Birds did not fattenand testes were unstimulated when birds wereexposed to T 5 22, 24, and 26 hr LD-cycles with8 hr photophase (Fig. 2A and B). All birds hadnormal body mass and testes were regressed at thebeginning, which they maintained throughout theexperiment. There was no significant difference inbody mass on day 0 of the experiment (F2,19 5

2.388, P 5 0.1188, one-way ANOVA). However,there occurred a small but significant decrease(Po0.05, paired t-test) in birds of 8L:14D. Hence,at the end of the experiment, there was a groupdifference (F2,19 5 8.299, P 5 0.0026, one-wayANOVA) birds of T 5 22 weighed significantlylesser (Po0.05; Newman–Keuls post-hoc test)

S. RANI ET AL.546

than those of T 5 24 and 26 hr LD-cycles (Fig. 2A).Two-way ANOVA also indicated a significant effectof the photoperiod (F2,38 5 10.41, P 5 0.0002), butneither of the duration of exposure (F1,38 5 0.523,P 5 0.436) nor of the interaction between the photo-period and the duration (F2,38 5 1.810, P 5 0.1774).

When photophase was extended to 11 hr, threeof six birds under T 5 22 (11L:11D) fattened(change in body mass 5 13.7274.90%), but birdsunder T 5 24 (11L:13D) did not fatten but ratherlost a little (change in body mass 5�4.7871.79%).This resulted in a significant difference (Po0.05,Student t-test) in body mass between 11L:11D and11L:13D birds at the end of the experiment (Fig.2C). Similarly, testes were significantly enlarged(Po0.05, Student t-test) under 11L:11D than thatunder 11L:13D in which slight initiation occurredin two of five individuals (Fig. 2D).

Experiment 3: effect of light wavelengths

Figure 3 summarizes the results. White and red(640 nm) lights produced a similar response. At

0.028 W m–2 (1� level) light irradiance, controlsunder white light had variable fattening and gainin body mass (mean Dbody mass 5 0.871.0 g;N 5 8). Whereas six individuals had small gainsfrom 3.4% to 22%, two individuals in the samegroup lost weight by 11%–12%. Testes werehowever slightly initiated (TV 5 5.24 mm3) in onlyone of eight birds. In comparison, birds under red(640 nm) light at this irradiance level exhibitedbetter response. The mean gain (N 5 8) in bodymass and testis volume was 3.2171.12 g and8.9573.2 mm3, respectively. Increasing irradianceto 2.5� level (0.07 W m–2) led to photostimulation

Fig. 2. Mean (7sem) body mass and testis volume of theredheaded bunting in response to different T LD-cycles. Birdswere exposed to 8L:14D, 8L:16D and 8L:18D (Experiment 2A:left panel—A and B; N 5 7–8) or to 11L:11D, 11L:13D(Experiment 2B: right panel—C and D; N 5 5–6). Significanceof difference, Po0.05; �, group difference.

Fig. 3. Response of the redheaded bunting to differentlight wavelengths and intensities. Birds (N 5 8 each condition)were exposed to 13L:11D (L: in white, blue [450 nm], or red[640 nm] color). The gain in body mass recorded after 4 weeksis plotted as the per cent change in body mass (A), and thetestis growth measured after 4.5 weeks is represented as thetestis volume (B). The minimum light intensity applied was at0.028 W m�2, and this was designated as 1� ; in the figure it isshown as 1 and its multiples. The white and red lights wereapplied at 1, 2.5, 5 and 10� level irradiances, and the bluelight was applied at 1, 10, 25 and 50� level irradiances. X-axisis in log scale. Significance of difference 5 Po0.05; �,compared with the initial value of the same group, a,compared to value(s) of the other group(s) at the sameobservation.

SEASONAL RESPONSES IN THE MIGRATORY MALE REDHEADED BUNTING 547

in all but one bird of both white and red lights.Birds were slightly fattened with a gain in bodymass of about 23%–24%, and testes were enlargedalbeit to a submaximal size in more than half thebirds of the group. A higher light irradiance at 5 or10� level (0.14 and 0.28 W m–2) did not increasethe mean response although variability in res-ponse within the group was relatively reduced(cf. Fig. 3A and B).

In contrast, there was no photoinduction inbirds exposed to the 450 nm light wavelengths(blue light) at 1� level at which birds had no fatdepots (mean Dbody mass 5�0.370.7 g; N 5 8)and the testes were small and unstimulated. At10� level (0.28 W m–2) a few birds had small fatdepots (mean Dbody mass 5 2.670.8 g; N 5 8;Po0.05, paired t-test) and testes were small andunstimulated in all but two of eight birds, whichshowed small initiation. A 25� level irradiance(0.70 W m–2) induced about 18% gain in body mass(mean Dbody mass 5 4.070.8 g; N 5 8; Po0.05,paired t-test). Testes were also recrudesced andthe group exhibited small to half maximalresponse (Fig. 3B). Birds exposed to 50� level oflight irradiance exhibited relatively larger testesbut did not show any marked change in bodyfattening as compared with 25� group (cf. Fig. 3Aand B).

DISCUSSION

The results show that the redheaded buntingcan distinguish small increments in day lengths. A12 hr photoperiod can be regarded as critical daylength for photoperiodic induction for buntings,since they had submaximal response with largeintra-group variation under o12 hr photoperiodsand had full response with small intra-groupvariation under 412 hr photoperiods (cf. Fig. 1Aand B). Thus, these results under artificial condi-tions correspond to those buntings exhibit in thewild. However, there exists a small difference inthe response between laboratory and naturalphotoperiodic conditions, which may be attributedto the nature of the photoperiodic stimulusavailable. In nature, transitions from dark to lightand back to dark every day occurs in several steps:first, there is a gradual increase in light intensityfrom the onset of the morning twilight to themiddle of the day, and then a gradual decrease inlight intensity from the middle of the day to theend of the evening twilight period. In such an LDcycle, which corresponds to a saw-tooth form, thewhole daylight period may not be photoinductive

since light intensity early in the morning and latein the day may lie below threshold intensity forphotoperiodic induction. Compared with this, alaboratory light regimen with a one-step transi-tion from dark to full light intensity and back todark constitutes a square-wave form providinglight above threshold intensity for photoperiodicinduction throughout the day. Hence, it is notsurprising if an 11.5L artificial photoperiod(square-wave LD) produces a response equivalentto that seen in birds held captives under naturalphotoperiods (saw-tooth LD cycle) at 271N by mid-April (�12.6 hr). Similarly, a larger responseunder 13L and longer photoperiods (Fig. 1A andB) could be comparable to the response under daylengths that buntings normally experience athigher latitudes during migration or at theirbreeding grounds. A complete response cycle(initiation–termination) occurring in buntingsexposed to long photoperiods (14L:10D and18L:6D; Fig. 1C and D) is similar to responsecycles reported under long days in many otherphotoperiodic species (Nicholls et al., ’88; Dawsonet al., 2001; Deviche and Small, 2001).

Interestingly, the photoperiod-induced responsecycle in the redheaded bunting appears flexible tophotoperiodic variations of the environment, asshown by the results of experiment 1. A completeresponse (initiation–termination–reinitiation) cycleoccurs in birds maintained on a relatively loweramplitude day–night cycle at 271N (Fig. 1). Hence,experience of the large amplitude day–night cyclesat breeding grounds and during migration,and possibly light-dependent magnetoreceptionfound critical for other seasonal behaviors in fewbirds (for references, see Wiltschko and Wiltschko,2001), is not necessary for regulation of photo-period-induced seasonal responses in the red-headed bunting. A similar finding has beenrecently reported for the blackheaded bunting(Misra et al., 2004). Thus, plasticity in responsive-ness to a wider range of photoperiodic variationscould be the general property of the avianphotoperiodic response system.

Experiment 2 shows that there was no photo-induction in T 5 22 or 26 LD-cycles when photo-phase measured 8 hr, but a T 5 22 applied with an11 hr photophase became inductive. If interpretedas suggested by the classical experiments ofHamner and Enright (’67) and Farner et al.(’77), these data mean that 22 cycles of an8L:14D or 8L:18D LD-cycle did not induce phaseangle sufficiently large (did not widen the tempor-al relationship between identical phases of the

S. RANI ET AL.548

daily LD cycle and entrained photosensitivityrhythm) to lead coincidence of light with thephotoinducible phase of the circadian photoper-iodic rhythm, but an 11L:11D did. However,we consider this unlikely. Instead, we proposethat an LD-cycle entrained circadian photosensi-tivity rhythm and as a consequence the critical daylength was altered. The data presented in Figure2C and D support this. The critical day length thatlies close to 12 hr in T 5 24 appears reduced closeto 11 hr when buntings are entrained to T 5 22;hence, 11L:11D and 12L:12D were almost equallyinductive (Fig. 2). In another elegant study onblackheaded buntings, using skeleton photoperiodparadigm (an LD cycle with two light periods),Kumar and Kumar (’95) have shown that criticalday length is altered to 11 and 13 hr under T 5 22and 26 hr, respectively. They found that a 1-hrlight pulse extending up to mid point of the T 5

22 hr (6L:4D:1L:11D) or T 5 26 hr (6L:6D:1L:13D)induced photoperiodic induction equivalent to thatinduced by a similar paradigm under T 5 24 hr(6L:5D:1L:12D).

Results from experiment 3 clearly suggest thatlight wavelengths influence photoperiodic regula-tion of physiological response in the redheadedbunting. Long light wavelengths (red light) weremore effective as they produced photoperiodicresponse at a very low light intensity (Fig. 3).This is consistent with several other avian studiesin which red light is reported to induce a largerphotoperiodic response (Benoit, ’64; Oishi andLauber, ’73; Rani et al., 2002). Interestingly, whilered light is most effective for photoperiodicinduction in birds, it is non-inductive for stimula-tion of photoperiodic response in mammals (Raniet al., 2002). Is it because of the difference inspectral sensitivity of photoreceptors mediatingphotoperiodism between the two vertebrategroups, birds and mammals? This has not beeninvestigated. We know nonetheless that unlike inmammals, which perceive light only through eyes(retina), birds have multiple sites of light percep-tion including the retina, the pineal, and thehypothalamus (Oliver and Bayle, ’82; Rani et al.,2002).

The action spectrum, which addresses thespectral sensitivity of pigment(s) mediating photo-periodic response (Rani et al., 2002), for a long-dayresponse of the redheaded bunting (Fig. 3) issimilar to that of the blackheaded bunting (Maliket al., 2004). Both the redheaded and blackheadedbuntings partly overlap their breeding and winter-ing grounds, and exhibit similar photoperiodic

responses except that the magnitude of bodyfattening is lower in the redheaded bunting (Aliand Ripley, ’74; Misra, 2002). One conclusion fromthe data shown in Figure 3 is that the redheadedbunting has a dynamic photoperiodic responsesystem, which responds differently to light wave-lengths at least until light is applied at higherillumination. Short light wavelengths are to beapplied at a higher illumination than the long lightwavelengths despite the fact that the avian brainphotoreceptor photopigments mediating photoper-iodism appear to be maximally sensitive to lightwavelengths around 500 nm (short wavelengths,e.g., green light; Foster and Follett, ’85; Fosteret al., ’85). This is consistent with the hypothesisthat the short light wavelengths penetrate lesseffectively across brain tissues, as compared withlong light wavelengths, and so net light energyreaching the hypothalamic photoperiodic responsesystem is relatively smaller at short light wave-lengths (Vriend and Lauber, ’73; Veen et al., ’76;Hartwig and Veen, ’79; Oliver and Bayle, ’82;Foster and Follett, ’85; Foster et al., ’85). In anycase, it appears that the changes in irradiancelevels of daylight, which may be achieved bychanges in spectral distribution especially duringtwilight periods, are detected by the birds’ timemeasurement system (Roenneberg and Foster,’97) and used to synchronize circadian functionswith the 24 hr day–night cycle of the environment(Krull, ’76). It is probable that the circadiansystems use different classes of photoreceptorswith varying spectral sensitivities in the morningand evening to sample the relative amounts ofshort and long light wavelength radiations, as hasbeen argued recently by Malik et al. (2004).

In conclusion, the redheaded bunting usesphotoperiods to regulate its seasonal responses.Apart from the duration of daily light, the lightintensity and light wavelength help determine thelight energy required by the circadian processesmediating seasonal responses in this species. Thisappears to have adaptive implications. For exam-ple, while photoperiodic entrainment may beachieved early in the day at relatively low lightenergy levels, the photoperiodic induction occursduring long day lengths of spring and summerwhen later in the day plenty of light at a higherenergy level is available.

LITERATURE CITED

Ali S, Ripley SD. 1974. Handbook of birds of India andPakistan. Bombay: Oxford University Press.

SEASONAL RESPONSES IN THE MIGRATORY MALE REDHEADED BUNTING 549

Benoit J. 1964. The role of the eye and the hypothalamus inthe photostimulation of gonads in the duck. Ann New YorkAcad Sci 117:204–216.

Bentley GE, Goldsmith AR, Dawson A, Briggs C, PembertonM. 1998. Decreased light intensity alters the perception ofday length by male European starlings (Sturnus vulgaris).J Biol Rhythms 13:148–158.

Dawson A, King VM, Bentley GE, Ball GF. 2001. Photoper-iodic control of seasonality in birds. J Biol Rhythms 16:365–380.

Deviche P, Small T. 2001. Photoperiodic control of seasonalreproduction: neuroendocrine mechanisms and adaptation.In: Dawson A, Chaturvedi CM, editors. Avian endocrinology.New Delhi: Narosa Publishing House, p 113–128.

Farner DS, Donham RS, Lewis RA, Mattocks PW, Darden TR,Smith JP. 1977. The circadian component in the photoper-iodic mechanism of the house sparrow, Passer domesticus.Physiol Zool 50:247–268.

Follett BK, Maung SL. 1978. Rate of testicular maturation, inrelation to gonadotrophin and testosterone levels, in quailexposed to various artificial photoperiods and to natural daylengths. J Endocr 78:267–280.

Follett BK, Pearce-Kelly AS. 1991. Photoperiodic induction inquail as a function of the period of the light–dark cycle:Implications for models of time measurement. J BiolRhythms 6:331–341.

Foster RG, Follett BK. 1985. The involvement of a rhodopsin-like photopigment in the photoperiodic response of Japanesequail. J Comp Physiol A 157:519–528.

Foster RG, Follett BK, Lythgoe JN. 1985. Rhodopsin-likesensitivity of extra-retinal photoreceptors mediating thephotoperiodic response in quail. Nature 313:50–52.

Hamner WM, Enright JT. 1967. Relationships betweenphotoperiodism and circadian rhythms of activity in thehouse finch. J Exp Biol 46:43–61.

Hartwig HG, Veen T van. 1979. Spectral characteristicsof visible radiation penetrating into the brain and stimulat-ing extraretinal photoreceptors. J Comp Physiol A 130:277–282.

Krull F. 1976. Zeitgebers for animals in the continuousdaylight of high arctic summer. Oecologia (Berl.) 24:149–157.

Kumar V. 2001. Melatonin and circadian rhythmicity in birds.In: Dawson A, Chaturvedi CM, editors. Avian endocrinology.New Delhi: Narosa Publishing House, p 93–112.

Kumar V, Follett BK. 1993. The circadian nature of melatoninsecretion in Japanese quail (Coturnix coturnix japonica).J Pineal Res 14:192–200.

Kumar V, Kumar BS. 1995. Entrainment of circadian systemunder variable photocycles (T-photocycles) alters the criticaldaylength for photoperiodic induction in blackheaded bunt-ings. J Exp Zool 273:297–302.

Kumar V, Rani S. 1996. Effects of wavelength and intensity oflight in initiation of body fattening and gonadal growth in amigratory bunting under complete and skeleton photoper-iods. Physiol Behav 60:625–631.

Kumar V, Kumar BS, Singh BP. 1992. Photostimulation ofblackheaded bunting: subjective interpretation of day andnight depends both upon photophase contrast and lightintensity. Physiol Behav 51:1213–1217.

Kumar V, Singh S, Misra M, Malik S. 2001. Effects of durationand time of food availability on photoperiodic responses inthe migratory male blackheaded bunting (Emberiza mela-nocephala). J Exp Biol 204:2843–2848.

Kumar V, Singh S, Misra M, Malik S, Rani S. 2002. Role ofmelatonin in photoperiodic time measurement in themigratory redheaded bunting (Emberiza bruniceps) andthe nonmigratory Indian wever bird (Ploceus philippinus).J Exp Zool 292:277–286.

Malik S, Rani S, Kumar V. 2004. Wavelength dependency oflight-induced effects on photoperiodic clock in the migratoryblackheaded bunting (Emberiza melanocephala). Chrono-biol Int 21:367–384.

Misra M. 2002. Photoperiodic control of seasonal responses inmigratory buntings (Emberiza sp.). Ph.D. Thesis. Lucknow,India: University of Lucknow.

Misra M, Rani S, Singh S, Kumar V. 2004. Regulation ofseasonality in the migratory male blackheaded bunting(Emberiza melanocephala). Reprd Nutr Dev 44:1–12.

Nicholls TJ, Goldsmith AR, Dawson A. 1988. Photorefractori-ness in birds and comparison with mammals. Physiol Rev68:133–176.

Oishi T, Lauber JK. 1973. Photoreception in the photosexualresponse of quail. II. Effects of intensity and wavelengths.Am J Physiol 4:880–886.

Oliver J, Bayle JQ. 1982. Brain photoreceptors for the photo-induced testicular responses in birds. Experientia 38:1021–1029.

Pittendrigh CS. 1972. Circadian surfaces and the diversity ofpossible roles of circadian organization in photoperiodicinduction. Proc Nat Acad Sci USA 69:2734–2737.

Rani S. 1999. Light relations of the photoperiodic clock inbuntings (Emberiza sp.). Ph.D. Thesis, Lucknow, India:University of Lucknow.

Rani S, Kumar V. 1999. Time course of sensitivity of thephotoinducible phase to light in the redheaded bunting.Biol Rhythm Res 30:555–562.

Rani S, Kumar V. 2000. Phasic response of photoperiodic clockto wavelength and intensity of light in the redheadedbunting (Emberiza bruniceps). Physiol Behav 69:277–283.

Rani S, Singh S, Misra M, Kumar V. 2001. The influence oflight wavelength on reproductive photorefractoriness inmigratory blackheaded bunting (Emberiza melanocephala).Reprod Nutr Dev 41:277–284.

Rani S, Singh S, Kumar V. 2002. Light sensitivity of thebiological clock. In: Kumar V, editor. Biological rhythms.New Delhi, India: Narosa Publishing House, p 232–243.

Roenneberg T, Foster RG. 1997. Twilight times: light and thecircadian system. Photochem Photobiol 66:549–561.

Tewary PD, Kumar V. 1982. Photoperiodic responses of asubtropical migratory finch, the blackheaded bunting(Emberiza melanocephala). Condor 84:168–171.

Tewary PD, Prasad BN, Kumar V. 1982. Circadian basis ofphotoperiodically induced testicular growth in redheadedbunting (Emberiza bruniceps). Anim Reprod Sci 4:245–249.

Thorington L. 1980. Actinic effects of light and biologicalimplications. Photochem Photobiol 32:117–129.

Veen T van, Hartwig H-G, Muller K. 1976. Light dependentmotor activity and photosensitive behavior in the eel(Anguilla anguilla L.). Evidence for extraretinal andextrapineal photoreception. J Comp Physiol 111:209–219.

Vriend J, Lauber JK. 1973. Light intensity, wavelength andquantum effects on gonads and spleen of the deer mouse.Nature 244:37–38.

Wiltschko W, Wiltschko R. 2001. Light-dependent magneto-reception in birds: the behaviour of European robins,Erithacus rubicula, under monochromatic light of variouswavelengths and intensities. J Exp Biol 204:3295–3302.

S. RANI ET AL.550