THE INTERACTION OF ENVIRONMENTAL, INTERNAL AND BEHAVIOURAL FACTORS IN THE ROOK, CORVUS F. FRUGILEGUS LINNAEUS

Summary.1 Seasonal changes in the behaviour, sexual cycle, moult and external environment of the Rook ( Corvus f. frugilegus ) were studied in relation to each other over a period of twelve years in the climatically mild Truro area of South‐west England where the species is sedentary and undergoes a prolonged autumnal sexual season. 2 There is an internal rhythm of reproduction, based on cyclical changes in the gonads (and no doubt in the adenohypophysis) which controls the primary (autumnal) sexual season. This is extinguished during the winter. The secondary, productive, “spring” sexual season is under the control of external stimuli. 3 Both sexes begin the “spring” gametogenesis early in February when the weather is still as cold as in mid‐winter, sunshine is not considerably increased, and food is still relatively difficult to find. Reproductive behaviour is inhibited by unusually cold weather, but in the mild Cornwall climate initial ovulations fell within the period 8th‐16th March during five years. The breeding season is “timed” to coincide with the seasonal harvest of earthworms (on which the young are fed) which begins before ovulation. 4 The date of ovulation, and therefore actual reproduction (as distinct from the onset of gametogenesis), may be governed by a combination of factors including temperature, sunshine and food availability. The species breeds up to three weeks later following a line East across the colder European land‐mass. 5 At no period of the year is either ovary or testis without quantities of cholesterol. In the testis this is contained in the lipid Leydig cells and when these become seasonally exhausted at the end of spermatogenesis there is a massive cholesterol‐positive steatogenesis in the seminiferous tubules. This tubule material disappears at the onset of the following gametogenesis, by which time the new generation of Leydig cells has become heavily charged with cholesterol and is again in a process of secretion. During the immediate postnuptial period the spermatogenetic mechanism is temporarily thrown completely out of action (Pl. 1, fig. 2). The ovary, on the other hand, is never without some active follicles. 6 Secretory components of the ovary (i.e. those structures that exhibit an unmistakable gain, followed by depletion, of cholesterol‐positive lipids) are as follows:a.  Ex‐follicular gland‐cells (Text‐fig. 4–F, G, H). These possibly arise from fibroblasts that migrate from the theca to the lumen of the atretic follicle. As the follicle disintegrates they are freed in groups into the stroma in which they are dispersed by pressures exerted by the enlargement of neighbouring follicles. These are already present in the stroma of young only three months old. (These are perhaps ostrogenic in function). b.  Female interstitial cells (Text‐fig. 4‐O). These are morphologically like the above, but arise from connective tissue‐cells in the ovarian stroma and therefore seem to be homologous with male Leydig cells. c.  Glandular cells, similar in appearance to the two types above, occasionally arise in hyaline scars (Text‐fig. 4‐M, N) resulting from a peculiar kind of relatively non‐lipoidal atresia (4‐L). Although they originate from the remains of follicles, these gland‐cells are formed at a period when such remnants are isolated and inactive stromal structures and may be therefore allied to (b) above. d.  Amorphous non‐cellular aggregations of cholesterol‐positive lipid that arise in atretic follicles (Text‐fig. 4‐E, F) and are gradually expended. Such atresias occur in nestlings. (The possibility that these produce progestin is considered.) e.  Thecal gland‐cells (Text‐fig. 4‐B, D. E). At least some of these are incorporated from the stroma as either (a) or (b) above during follicle maturation. These almost certainly produce oestrogen.During both spring and autumn sexual seasons increased numbers of atresias occur and thus the ovary is recharged with the products of (d) and (a). The greater the seasonal sexual excitement the more of all the above structures arise, later to expend their cholesterol in the presumed elaboration of steroid hormones.7 No specific and persistant corpus luteum forms after ovulation although some luteal cells arise and soon disappear. The discharged follicle is immediately invaded by phagocytes, together with much connective tissue which forms one or more bridges (Pl. 3, figs. 2, 3) so that plurality scars may arise. Therefore it is impossible to determine avian age by scar count. 8 The “spring” sexual resurgence (in late winter), and that of autumn, both occur in females at periods of increased follicle proliferation, numerous lipoidal atresias and generation of stromal gland‐cells (see 6 above). Comparable behaviour in males occurs twice a year when the Leydig cells exhibit secretory activity. Both sexes then employ a peculiar slow shoulder‐movement component in all displays that, by observation, seem to be of sexual significance. Such movement does not occur in males during the post‐nuptial refractory period when the newly arisen Leydig cells appear still to be non‐secretory. In addition to sexual display and stick‐carrying, the reorganization of winter flocks and the assumption of winter roosting patterns also coincide with the seasonal development of the male interstitium (but also with the period when adults are freed from parental duties). 9 The annual (post‐nuptial) moult begins when the male seminiferous tubules are clotted with newly formed lipids and when the interstitium is in its early regenerative, non‐secretory phase. This is probably coincidental: moult occurs at the same period in yearling non‐breeders in which no such rehabilitation occurs. Similarly, the onset of moult coincides with the suppression of sexual behaviour, but there is no evidence that the two factors are related: feather replacement is still proceeding strongly early in August when, with renewed interstitial activity, autumnal sexual display begins and gathers momentum. 10 The male post‐nuptial refractory period ends in late July or early August when the Leydig cells show signs of secretory activity and there automatically follows a resurgence of sexual behaviour which by late‐September and October is as pronounced as that of the pre‐building period early in spring. 11 In October, days are still longer, and air temperatures higher, than at the late winter period when gametogenesis is initiated. Food, including earthworms, is still plentiful: conditions for reproduction seem good. Yet gametogenesis (except in about 13.6 per cent of adult males and rarely indeed in females) does not occur although autumn breeding is occasionally recorded. The most obvious environmental differences between autumn and spring are a decrease in hours of sunshine and a progressive reduction in daylength. The second factor may influence the bulk of the population. Conversely, the late‐winter initiation of the “spring” gametogenesis is probably stimulated by progressive light increment rather than daylength per se.12 Inhibition of autumn reproduction conserves the breeding potential until the following spring. Yet autumn sexual behaviour strengthens the pair‐bond and is demonstrably valuable in forming new pairs and in establishing territory long before the beginning of the secondary (productive) sexual season.