There appears a close correspondence between the gonad development cycle of house sparrow and annual variations in day length (cf. Figs. 1 and 2). Increasing photoperiods (≥12 h per day) of spring months (March, April) triggered the gonadal growth, but then gonads regressed during summer months (June or July) when day lengths were still much longer than of spring months. Similar to that described in the annual reproductive cycles of many species including temperate house sparrows [8, 27, 37], this post-reproductive refractory period suggests that the physiological mechanisms controlling reproductive cycle in the subtropical house sparrows undergo dramatic changes in their response to day length. Interestingly, a study of Misra et al.  on blackheaded bunting (Emberiza melanocephala) showed the evidence of seasonal changes in photoresponsiveness even when birds were maintained on a non-stimulatory short photoperiod, which is widely used to ensure photosensitivity in a photoperiodic species. In buntings maintained since February on 8L:16D and subjected to 16L:8D from March to August, the magnitude of testicular response declined as short days progressed until July when the response was restored.
Figure 2 shows that the gonad development and molt cycles of house sparrow at 27°N, 81°E closely compare with those of its population living at high latitudes. A study on sparrows at 52°N reported that testes grew in size steadily until May, remained large in June and were fully regressed by late August . In this study, one-fifth of sparrows began molt of their primary flight feathers by late July, all were molting a month later, and the molt was complete by early November. Compared to this, our sparrows had largest gonads in May and were fully regressed in July. Molt began in June, progressed steadily and was almost complete by late September. Thus, sparrows at 27°N appeared exhibiting a photoperiodic strategy in regulation of their reproductive cycle, similar to their high latitude conspecifics (Figs. 2a,b). This is interesting since a previous study on Indian house sparrows at 22°N, 88°E reported a longer breeding season lasting for 10 months . The difference in reproductive cycle of these two Indian studies could be attributed to the difference in the food availability as a consequence of the differences in changes in temperature over seasons at these two latitudes, similar to that reported for high latitude birds [34, 35, 40, 41]. In Threadgold's study also, sparrows of 34°N had longer breeding season than those of 52°N . Whatever explanation is offered, these findings tend to suggest that actual breeding strategy in birds is modified by local conditions of the given latitude.
A comparison of gonad development cycle between the wild and captive sparrows revealed that in captives (i) the attainment of peak testicular growth was delayed, (ii) the amplitude of follicular growth was attenuated, and (iii) overall duration of the gonadal growth phase was longer (Figs. 2a,b). These differences could have occurred due to one or all of the following reasons. (1) The two conditions were different in terms food availability. Whereas wild sparrows had free access to all kinds of food in its environment, captive sparrows had fixed diet of seeds of Setaria italica and Oriza sativa. Several reports suggest the effects of food availability on the gonadal development [40, 40–42]. (2) Absence of supplementary factors including opportunity of pair-formation to captives modified their gonad development cycle [1, 43]. (3) Confinement within the aviary (size = 3.0 × 2.5 × 2.5 m3) may have been stressful; this in turn affected the activity of hypothalamo-hypophyseal-gonadal axis, and hence the timing and magnitude of gonadal growth and development in captives [44, 45].
Results of the series B experiments further support the idea of seasonality in photoresponsiveness of house sparrows (Fig. 3). A 16L photoperiod did not prevent the collapse of large gonads in May birds or re-initiate recrudescence of regressed gonads in June and July birds, but caused recrudescence in September sparrows. This suggested that there was a seasonality in loss and gain of photoresponsiveness in sparrows. A comparable situation exists in several photoperiodic species including temperate zone grey partridge [Perdix perdix, ] and subtropical Indian weaver bird (Ploceus philippinus ) and brahminy myna (Sturnus pagodarum ). A study of Dawson  on house sparrows at 52°N, nevertheless, provided slightly different results. In this study, sparrows shifted to 18L:6D in June did not regress their testes at least for the next 25 days; in fact, by mid-July testes under 18L were significantly larger than those in natural day length at this time. The difference between our results and those of Dawson  could be due to difference in the amplitude of photoperiodic cycle which sparrows experienced at these two latitudes as well as due to local conditions such as temperature and food availability. It is known that ambient temperature significantly affects the photoperiodic induction of gonadal development and regression in birds [49, 50].
Data presented in figure 4a are consistent with the idea that longer the photoperiod, faster is the rate of gonadal growth and subsequent regression; hence, gonadal growth phase becomes narrower [38, 51–53]. In the present study, the rate of photoperiodic induction and subsequent regression was faster in 15L than in 12L photoperiod (cf. Fig. 4a). Although the timings of the peak testicular response were not different between 12L and 15L groups (Fig. 4a), probably because a 4-week interval of first observation was long enough to allow full testicular growth under both the photoperiods, the timing of the onset of testicular regression and molt clearly showed a faster photoinduction under 15L:9D (Fig. 4). Testes regressed (cf. data on week 13, Fig. 4a) and molt began earlier in 15L than in 12L photoperiods (Fig. 4b,c). Thus, as one would expect, molt followed the testicular regression: faster was the testicular regression, earlier was the onset of the molt (cf. Fig. 4a–c). A similar molt pattern has been reported in house sparrows of 52°N subjected to stimulatory photoperiods . Photoperiodic control of molt is also reported in the European starling . A close relationship between the gonadal regression and post-nuptial molt is known in several species . The present results are also comparable to those reported on house sparrows at 52°N exposed to 18L, 16L and 13L photoperiods . In this study, as compared to that in 16L and 18L, peak testis growth in 13L was delayed by 3 weeks. Similarly, testes regressed by 9, 12, 18 weeks in 18L, 16L and 13 L groups, respectively . Figure 5, however, shows that in sparrows exposed to 22L, 18L and 14L, testes recrudesced faster in the 22L photoperiod, as one would expect, but the subsequent regression was not different between the three photoperiods. It appears that the duration of testicular growth phase under stimulatory photoperiods is not altered when daily light period far exceeds the photoperiodic threshold (cf. Fig. 4a, 5). In other words, the extension of the light period into the photoinducible phase of the circadian rhythmicity underlying photoperiodism longer than ~2 h each day does not change the gonadal growth phase in house sparrows. This may be logical given the fact that sparrows at 27°N experience a maximum of about 14 h light per day (sunrise to sunset, Fig. 1), and hence the photoinducible phase of circadian rhythmicity in this species  is daily exposed to only slight greater than 2 h light period. A study of Rani and Kumar  on redheaded bunting (Emberiza bruniceps) provided the evidence that illumination of photoinducible phase beyond a critical duration did not enhance the testicular response.
At this latitude (27°N, 81°E), house sparrows closely share habitat with weaver birds. Although there is a difference of about 4 weeks in peak testicular development between the two species (our unpublished data), both of them initiate their gonad development cycle with increasing day lengths of spring. Hence, they are expected to share similar photoperiodic mechanisms, although sparrow may have slightly lower photoperiodic threshold. Indeed, Indian weaver bird at 25°N is reported exhibiting testicular response to 9L, 12L and 15L photoperiods [47, 58, 59] similar to that is shown by sparrows in the current study (Fig. 4).
On exposure to short photoperiods (≤10 h per day), a long day breeder usually will not show gonadal recrudescence, which is indicative of the importance of photoperiodic cues over endogenous seasonal rhythm in the control of reproductive cycle of a species. However, if a long day breeder exhibits gonadal recrudescence under such short photoperiods, its response may be considered as a consequence of the seasonal rhythm rather than to the photoperiod. Results of the experiment 2 of series III (Fig. 5a) should be viewed against this background. Sparrows exhibited a testicular response, albeit slow and small, to 2L:22D and 10L:14D; however, no response to 6L:18D (cf. Fig. 5a). An earlier study has also reported testicular recrudescence under 1L:23D temperate house sparrows . In the current study, the small induction under 2L and 10L photoperiods could be occurring due to different reasons. A 10 h square-wave of light at bright intensity in 10L:14D regime could be compared to a lighting situation in the natural environment closer to when sparrows would initiate testicular recrudescence. Therefore, sparrows continuously exposed to 10 h light per day exhibited small initiation of response. This is supported by the result under 6L:18D in which no significant enlargement of testes occurred. Several studies have shown spontaneous full growth and regression cycle of gonads under 12 h photoperiods . On the other hand, we interpret small response under 2L:22D as the consequence of the annual rhythm of the growth-regression-regrowth of testes rather than of the photoperiodic condition. It is possible that a short photoperiod like 2L:22D is unable to sustain the entrainment of the endogenous clocks underlying seasonality in house sparrows. The photoperiodic entrainment of the circadian rhythmicity, which can be different under different zeitgeber conditions, sets the timing of the photoinductible phase , and the interaction of the latter with the light regulates gonad development cycle in photoperiodic birds [12, 31–33].
In general, the timing and amplitude of the photoperiodic response differed between male and female sparrows (cf. Figs. 2a,b). Long days induced full testicular growth but only partial ovarian growth. It appears that increasing photoperiods induce initial slow growth phase and not the final fast or exponential growth phase of ovary, which is influenced by supplementary factors [1, 43]; full reproductive competence in females is often determined by the stimuli from mate, nest site and food availability [8, 61–63]. Such a difference in photoperiodic regulation of seasonal cycles between sexes probably reflects an adaptive strategy since it restricts ovulation, and hence fertilization, until the time when the chances are favorable for the survival of the offspring.