Mass-change during moult in the Great Tit Parus major

Transcription

1 Bird Study ISSN: (Print) (Online) Journal homepage: Mass-change during moult in the Great Tit Parus major A. G. Gosler To cite this article: A. G. Gosler (1994) Mass-change during moult in the Great Tit Parus major, Bird Study, 41:2, , DOI: / To link to this article: Published online: 25 Jun Submit your article to this journal Article views: 183 View related articles Citing articles: 24 View citing articles Full Terms & Conditions of access and use can be found at Download by: [ ] Date: 17 December 2017, At: 10:29

2 Bird Study (1994) 41, Mass-change during moult in the Great Tit Parus major A.G. GOSLER Edward Grey Institute of Field Ornithology, Department of Zoology, South Parks Road, Oxford OX1 3PS, UK Downloaded by [ ] at 10:29 17 December 2017 Changes in mass during moult were investigated by analysis of data recorded on standard moult cards by members of the British Trust for Ornithology Ringing Scheme. In the analysis of post juvenile moult, all available cards that presented wing length, mass and time of capture were used. The analysis of adult post-nuptial moult was restricted to cards from a single year. Juvenile mass correlated strongly with body size, time of day and the number of greater coverts being replaced. After controlling for the effects of size and time of day, mass was found to increase with date except in those individuals in which all juvenile greater coverts were retained. In these birds residual mass declined with date. In adult males, mass increased with both date and moult scores of primary, secondary and tertial remiges and of rectrices but because of the strong correlations between moult scores and date, the true relationship was not clear. In adult females, mass declined with date before increasing again. Unlike the males, female moult scores did not predict body mass. However, the timing of moult was a strong predictor of body mass. The differences between age and sex groups are discussed in relation to the nutritional demands of moult. T he moult period can be a stressful time for many birds. This is not only because it is nutritionally demanding for both energy and proteins providing specific amino acids" but also because it may compete in time with other activities such as breeding and migration so that its specific food requirements may not be at their most abundant at the time that moult occurs 4-7 To overcome this, many species have evolved variable moult strategies that relieve the individual from the commitment of a complete moult each year. For example, many migratory passerines arrest remige moult before migrating in Autumn rather than delaying their departure until the moult is completed; 7'8 many others delay moult entirely until they are within the winter quarters, 9 or, in far-northern populations, may attempt to moult so rapidly that they become temporarily flightless. 10 Furthermore, variation in moult within a species may reflect differences in the costs of moult between various sections of the population. Hence, while adult passerines generally have one complete moult each year the post-juvenile (PJ) moult is frequently reduced in extent (usually excluding the remiges) and may vary between individuals in its completeness. Common examples are the incomplete and variable moult in juveniles of greater coverts, tertials, alula and in some species also the tail. 5,9 Age differences in the extent of moult may reflect the poorer foraging skills of juveniles, so adding to the relative cost to them of a complete moult. 11 Such a situation is found in Great Tit populations in northern Europe. Here, adults undergo a complete moult after breeding (postnuptial or PN moult). This usually commences with the loss of the first (innermost) primary remex. At about the time that the third to fifth remex is dropped, secondary, tertial and tail moult all commence. Several studies give detailed descriptions of the moult sequence and the nutritional conflict between breeding, moult and territoriality in this species. 5'6,12-15 In contrast, the PJ moult is partial and never includes remiges. It usually includes all body feathers and rectrices but the moult of the greater coverts is often incomplete, as is the replacement of the alula. The proportion of juveniles which fail to complete greater covert moult varies greatly between populations, between years within populations, between sexes and in relation to the date of fledging.12

3 The evidence is now considerable that the extent of PJ moult in the Great Tit is related to the availability of resources, and subsequently the physiological condition of the individual, at the time of moult. For example, first-year Great Tits trapped in winter in Wytham Woods, Oxford that had incomplete greater covert moult were in significantly poorer condition in terms of a pectoral muscle score (a measure of protein reserves) than those that had completed moult of this tract. It was suggested that both measures were indicators of an individual's ability to obtain food. 16 Hence, a bird that had poor access to resources during winter may have been similarly limited during moult. Furthermore, the muscle score correlated negatively with the number (rather than simply the presence) of unmoulted greater coverts, suggesting a graded response to stress induced by resource-limitation. While a relationship has been described between the extent of moult and muscle condition some months after moult, 16 it is not clear whether this is causal or whether both are merely correlates of resource availability. This paper considers this relationship further by examining variation in mass during the moult period in both juveniles and adults. This is the first time that mass change during moult has been described in this species. In addition, the sequence of PJ moult is described, since little has been published on this and existing studies have differed in their conclusions. 14 METHODS This study is based chiefly on the analysis of data collected by members of the British Trust for Ornithology Ringing Scheme and contributed to the BTO Moult Enquiry between 1973 and ,17,18 Data were recorded in a standardized way on printed recording cards supplied to ringers by the BTO. 18'5 Cards were only completed for birds that were undergoing moult so that no biometric data (e.g. mass) are available from the moult period for birds that were not moulting. Prior to 1971, biometrics other than moult details were not requested by the Moult Enquiry, so that no cards returned before that year could be used here (see below for biometrics). Only data referring to the primary, secondary and tertial remiges, rectrices, greater (secondary) coverts and alula have been Condition and moult of Great Tits 147 used, 18 although the actual set of tracts used must depend upon whether the analysis refers to PJ or PN moult. The state of replacement of each feather was scored by the observer on a 6- point scale where 0 represents an old feather and 5 a fully grown new feather. The state of moult in any tract was summarized by summing these scores across all feathers in the tract. Hence, a maximum score of 50 (5 x 10 feathers) can be attained for the primaries of one wing, 45 for the greater coverts, 60 for the whole tail and so on.5'14'19 During PJ moult, the greater coverts are generally dropped and replaced simultaneously as a block such that any retained juvenile coverts are the outermost. Therefore, in these analyses I have used the number of greater coverts replaced in addition to the combined moult score of feathers in the greater covert tract. Observation dates have been taken as the number of days after 1 June. Wing-length was measured by the maximum chord method. 9 Recent studies (to be published elsewhere) suggest that this measurement is highly repeatable between British ringers. In the analysis of PJ moult this has been taken as a measure of body size. However, as the wing point is formed by the seventh primary in the Great Tit, wing-lengths of birds that had already dropped this feather (in practice, those with primary scores greater than 30) could not be used in the equivalent analysis of PN moult. This might introduce a bias in the analysis of size-corrected mass in adults since later-moulting birds must be excluded from any analysis. Mass was recorded to 0.1 g with a x 50 g 'Pesola' spring balance, and the time of weighing was recorded as the last full hour passed. The data on PJ moult (181 cards) span the full period The present analysis of PN moult, however, is based on 179 cards (106 male, 73 female) from As the data are not complete on all cards, sample sizes varied between analyses, but this is indicated where appropriate. No significant differences between years were found in wing-length, mass or the relationship between moult scores, hence between-year effects have been ignored throughout in the analysis of PJ moult. Statistical analyses were carried out by the use of the correlation, stepwise multiple regression and general linear modelling (GLM) procedures available on Minitab Release Because of the problems inherent in the regression of

4 148 A.G. Gosler moult scores on date,13,2122 and hence the unreliability of residual moult scores derived from such regression, to control for date, moult scores were expressed as deviations from the mean moult score for the date on which they were recorded (by day or, in the analysis of female PN moult where sample size was usually too small on any one day, by the week). This avoids the need to fit a pre-determined ad hoc model. A negative value indicates a bird with a lower than average moult score for the day (or week) on which it was captured. Finally, new data on the relationship between fat reserves and mass are presented to allow some interpretation of mass changes during moult. These data were collected during routine trapping (in winter, October-March inclusive and when feeding well-grown young, May-June) as part of the Edward Grey Institute's long term population study of the Great Tit in Wytham Woods, near Oxford. 12'23'24 Since May 1982, I have recorded a score of the visible tracheal pit fat of each bird as follows: 0: no visible fat; 1: a trace of fat; 2: base of tracheal pit obscured by fat, but less than half full; 3: pit at least half full but still below the level of the clavicle; 4: fat filling the tracheal pit but not bulging; 5: fat filling tracheal pit, bulging and overlying pectoral muscle. Additional half score values (e.g. 2.5, 3.5 etc.) were also interpolated. This system was first described by McCabe 25 and is now widely used to give an indication of total body fat Mass was recorded as above. RESULTS Post-juvenile moult Sequence of moult Flegg & Cox13 stated that PI moult started with the dropping of the central pair of rectrices. Dhondt 14 disputed this, saying that tail moult did not commence until some 10 days into moult which usually started with the loss of some feathers on the breast and then humerus (secondary coverts) and back. Although the analysis of body moult is not included here the present data support Dhondt's argument, since greater covert moult is under way by the start of tail moult, and of those birds in which greater covert replacement occurred only 10% had failed to complete this by the time that a tail score of 24 (out of 60) was reached (Fig. 1). Moult of the alula is undertaken less often than that of the greater coverts. When it does occur, however, its replacement is co-ordinated with the greater covert moult. Of 136 birds in which a score for the alula was recorded, 56 (41%) were replacing part or all of the alula. This is greater than the 8% described in Hertfordshire, 13 but is the same as that described for the Belgian population. 14 Of the 56 birds showing alula moult, only 2 were moulting fewer than all the greater coverts (significantly fewer than random x 2 = 25.7, df = 1, P <0.001). Changes in mass through the moult period Mass correlated significantly with wing length (r = 0.585, df = 179, P <0.001), indicating that the latter may be used as an indicator of size, and time of capture (r = 0.324, df = 179, P <0.001) indicating that birds were heavier later in the day. This increase in mass with time of day has been noted many times in Great Tit NM 40* Tail moult score Figure 1. The relationship between greater covert and tail moult scores for 145 juvenile Great Tits. The broken vertical line marks a tail moult score of 24 (completed moult scores 60) by which time 90% have already completed replacement of the greater coverts.

5 samples measured in winter and while breeding. Gosler16 showed that this component of variation in mass could be explained largely in terms of daily changes in the amount of stored fat. This is the first time that the same daily mass cycle has been described for Great Tits Day (June 1=1) Condition and moult of Great Tits 149 undergoing moult. The high masses observed in some individuals (Fig. 2A) can best be explained in terms of considerable fat storage (see below and Fig.5). In a stepwise multiple regression in which wing-length and time were entered as predictors they explained 40.9% of the variance in mass. Overall, no significant change in mass was recorded in relation to date through the moult period. If the variance in mass due to winglength and time of day was first removed by regression, the date explained 2.8% of the residual variance in mass and this was statistically significant (r = 0.168, df = 178, P = 0.024) suggesting a weak improvement in body condition through the moult. In contrast to this general trend (and even to birds replacing just 2, 3 or 4 coverts), the 26 birds in which no greater covert moult was recorded showed a strongly significant decline in mass with date (r = 0.539, df = 24, P = 0.005, Fig. 2A) although when corrected for wing-length and time of day, their mass showed no significant relationship with date. These results, however, may have resulted from sampling bias. A sample of birds showing no greater covert loss, trapped early in the season, may include some individuals that will moult but which have not yet started. Since females are significantly more likely than males to retain juvenile greater coverts, it is possible that a juvenile sample from early in the season may include some males, while later in the season most (if not all) birds lacking greater covert moult would be female. Since males are significantly larger and heavier than females (usually some 4% larger 12) such a bias could lead to the apparent change in mass (or lack of Day (June 1=1) Figure 2. Mass change during moult in juvenile Great Tits. (A) The relationship between mass and date. A, birds undergoing greater covert moult. (- - -), regression of mass on date: Y = X, F = 3.59, df = 1,153, n.s., birds which did not replace greater coverts. ( ), significant decline in mass with date in these birds: Y = X, F = 9.69, df = 1,24, P= (B) The relationship between mass standardised for wing-length and time of day with date. Only birds trapped after day 60 were used to remove sex bias (see text). Symbols are as (A). Regression statistics are for moulting birds: Y = X, F = 5.98, df = 1,141, P = 0.016; for non-moulting birds: Y = X, F = 5.21, df = 1,23, P = 0.032).

6 150 A.G. Gosler change in the residual) observed here, although overall no significant correlation was found between wing-length and the number of juvenile greater coverts moulted. To test for such a bias, I looked at the correlation between residual mass (after correction for wing and time as above) and date in only those birds trapped after day 60 (30 July) by which time moult should have been apparent in all those juveniles that would moult. This residual mass still declined significantly with date in the non-moulting group and differed significantly from the birds that underwent moult (interaction: F = 5.75, df = 1,143, P = 0.018) which showed a more strongly significant increase in mass with date (see Fig. 2B). This suggests that those juveniles that failed to moult any greater coverts may have done so because they were, in some way, in poorer condition and even declined in condition during the moult period. 3.0^ i Changes in mass in relation to moult score Once initiated, moult of any tract proceeds in a more or less determinate fashion so that moult scores are strongly correlated with date, therefore the analysis of mass in relation to moult was carried out on the date-standardized moult scores. Mass also was expressed as a residual, as described above. This residual mass was entered as the dependent variable in a stepwise multiple regression with the standardized moult scores for greater coverts, tail and alula as predictors. Only the standardized tail score correlated significantly with the residual variance in mass (r = 0.217, df = 119, P = 0.017). There is, therefore, some evidence that tits that were more advanced in tail moult than average for a given date were relatively heavy for their size. Figure 3 shows the relationship between residual mass and the number of greater coverts being replaced. Birds undergoing replacement of a few greater coverts were significantly lighter than those undergoing either no moult or a full moult of the greater covert tract. A quadratic regression fitted to these data explained 8.5% of the variance in residual mass and was strongly significant (F = 7.23, df = 2,156, P = 0.001). This was still true if observations made before day 60 (F = 8.32, df = 2,144, P <0.001, r2 = 10.4%), or even 75 (F = 3.99, o^ -2.0 v No greater coverts replaced Figure 3. The relationship between residual mass (mass corrected for body size and time of day by regression) and greater covert moult in 160 juvenile Great Tits. Note that birds undergoing a partial moult of the greater coverts were significantly lighter than those that completed moult of the tract (Y = X+0.047X2, F = 7.23, df = 2,156, P = 0.001). They were also lighter on average than birds which moulted none of the coverts, suggesting that the latter group benefited somewhat from the strategy, but see text and Fig. 2. df = 2,106, P = 0.021, r2 = 7%) were omitted to eliminate possible biases described earlier. Post-nuptial moult Unlike the data on PJ moult all of the cards used for PN moult included the bird's sex, therefore analyses were carried out separately for each sex. However, the fact that wing length is unreliable as a measure of size in more than half the birds sampled (102 birds had primary moult scores greater than 30) means that mass cannot be controlled for variation in body size within sexes. Although this may be a serious drawback, there is no a priori reason to expect sampling to have been biased in respect of body size. As the sequence of feather replacement observed in the present data set agreed well with that already described elsewhere, 5'12-15 it will not be described further here.

7 Condition and moult of Great Tits 151 Downloaded by [ ] at 10:29 17 December 2017 Mass change through the moult period Male mass increased significantly through the moult (r = 0.331, df = 96, P <0.001, Fig. 4). Unlike juveniles, adult male mass was not significantly correlated with time of day. Female mass, in contrast, shows a quadratic relationship to date, with mass dipping before recovering again (F = 3.5, df = 2,65, P = 0.035, Fig. 4). Like the males, mass was not significantly correlated with time of day. The high masses observed in both sexes (in females especially near the start of moult) suggest that birds maintained high fat levels through the moult. However, it is not possible to distinguish mass changes due to changes in fat reserves from those due to changes in muscle condition in these data. 28'16 Some information relating to this question is provided by the relationship between body mass and fat score recorded at Wytham between May 1982 and December Figure 5A shows this relationship for 2246 Great Tits trapped during the winter; Fig. 5B shows the same relationship for 1907 trapped during the breeding season. In each sex and season the relationship was essentially linear. In winter, males were consistently about 1 g heavier than females for a given fat score. This sex difference was reduced to less than 0.5 g when breeding. Although, within a season (breeding or winter), the regression slopes did not differ significantly between sexes (Fig. 5), for each sex the slope was significantly greater in winter than when breeding (males: F = 24.07, df = 1,2166, P = ; females: F = 4.79, df = 1,1981, P = 0.029). While one cannot standardize the masses from the BTO moult dataset for variance due to body size or fat scores, the Wytham data indicate that, irrespective of seasonal differences in the details of the relationship, any Great Tit under 17.5 g is likely to have very reduced fat reserves while any birds over 20 g are likely to be carrying a substantial fat store. Such variation is too great to be explained by changes in muscle mass alone. 16 Hence, at least a part of the mass variation observed in these moulting birds must be due to considerable variation in fat reserves. Mass change in relation to moult score Male mass correlated significantly with all of Day (June 1st = I ) Figure 4. Mass change in relation to date in adult Great Tits in Males (0) showed a significant increase in mass through the moult period (Y = X, F = 11.81, df = 1,96 P = 0.001) while females (A) showed a significant decline in mass after the start of moult followed by a recovery after the end of July (Y = X+0.004X 2, F = 3.5, df = 2,65, P = 0.035). the moult score data (primaries: r = 0.326, P <0.001, secondaries: r = 0.226, P = 0.028, tertials: r = 0.211, P = 0.041, tail: r = P = 0.006, df = 96 throughout). However, because of the strong correlations which inevitably exist between moult scores of the different tracts and date, these results are confounded by date. No significant correlation was found between mass and any of these moult scores after they were transformed to date-standardized scores as described above. Nor was there any significant relationship between these scores and residual mass (controlling wing-length as above) in those males with primary moult scores of less than 30 (n = 18). In females, mass was not correlated significantly with any of the moult scores. However, a strongly significant quadratic relationship existed between mass and the standardised primary moult score (Fig. 6). Females that were advanced or retarded in their primary moult relative to the week in which they were measured were significantly lighter in mass than birds that showed the average state of moult. No such relationship

8 152 A.G. Gosler was found between mass and standardized moult scores of any other tract or in males. DISCUSSION Of the many published studies of passerine moult, remarkably few comment on other o Male Female 2 5A Fat score class changes in body condition suffered by the birds at that time, and even fewer consider which tissues might be involved in such changes. Across different species, both increases 19'3o,31 and decreases followed by recovery 2'29'32'33 in mass have been noted in adults undergoing primary moult. Where reported, loss of mass during moult has been attributed to negative protein balance rather than a loss of fat 4'29'33 The increase in mass in adult Bullfinches Pyrhula pyrhula during moult is attributable largely to an increase in protein mass, but also to increased fat and water. 34 The importance of protein reserves to the moult cycle is also implied by observations of arrested moult. For example, juvenile Starlings Sturnus vulgaris (which normally have a complete PJ moult) may arrest primary moult if the food supply is poor, and this is associated especially with low muscle condition. This occurs less often among adults. 35 The present study also found evidence for the involvement of protein stress in determining the character of moult in the Great Tit (especially in juveniles). As in the Starling, the degree to which individuals were affected depended on their social status (age and sex) since adult males are usually dominant over females and, at the time of moult, both are dominant over juveniles. 12 While juveniles showed both variation in the extent of moult and relationships between this and mass, and adult females showed some effect of moult on mass, adult males appeared comparatively immune to the demands of moult. Juveniles which moulted a few greater coverts were significantly lighter than those Fat score class Figure 5. The relationship between body mass and tracheal pit fat score class in Great Tits trapped at Wytham, Oxford. (A) Winter (October March inclusive). (B) When feeding large young (May June). All data were collected by the author. Separate points are shown for each sex but each point indicates the mean+1 se mass for the fat score class concerned; sample sizes are also indicated. Regression lines are indicated for each sex (unbroken: male, broken: female) in each season. All were significant at P < Winter: male mass = fat, F = 365.4, df = 1,1251, female mass = fat, F = 236.3, df = 1,992; breeding: male mass = fat, F = 41.27, df = 1,915, female mass = fat, F = 72.1, df = 1,989. See text for further explanation.

9 p 190 ISO Standardized primary score Figure 6. Change in mass in relation to primary moult score corrected for date (week) in adult female Great Tits. Birds moulting close to the mean date for the population were significantly heavier than those that were either advanced or retarded in timing of their moult (Y = X 0.066X 2, F = 4.65, df = 2,65, P = 0.013). These results suggest that the timing of moult is critical for these birds. No such relationship was observed in males. that completed this moult. As this observation still held after removing both a size-dependent (wing-length) and the time-dependent component of mass which reflects daily changes (but not seasonal changes) in fat reserves, 16 this suggests that these birds were probably lighter partly because they had poorer protein reserves. It is interesting that birds which moulted no greater coverts had a slightly greater residual mass than those that moulted just a few coverts, suggesting that they may have benefited a little from their lack of moult. However, that their residual mass declined steadily through the moult period while the moulting birds showed a significant increase may indicate that they were, nevertheless, suffering greater resource stress since a reduction in muscle mass at this time is unlikely to reflect an adaptive mass reduction. 36 Juveniles that were advanced in their tail moult were also relatively heavy and again, the fact that the analysis was conducted after 20 Condition and moult of Great Tits 153 the prior removal of the time-dependent component of mass suggests that protein mass is probably involved. The presence of a relationship between mass and the number of greater coverts replaced in juveniles strongly supports my earlier contention that the retention of unmoulted juvenile greater coverts is a badge of poor condition during moult I have shown elsewhere that adult male mass may increase during the nestling stage of their brood due to an increase in pectoral muscle condition although, compared with earlier breeders, later breeding males tended to have poorer muscle scores. These observations contrasted with the mass and muscle scores of their mates which declined through the breeding attempt. 16 The present data suggest that these patterns of mass change with time are continued into the moult, with males increasing, and females continuing to decline in mass before finally recovering their losses after the end of July (Fig. 4). It is not clear why these differences exist between sexes during the moult, as it was not clear why they occurred during the breeding season, although they suggest that, at least during the early stages, moult may be more stressful for females than for males. Some further support for this argument is provided by the observation that the timing of primary moult appears to have a critical effect on the condition of females (Fig. 6) while it is apparently irrelevent to males. A.A. Dhondt (pers. comm.) suggests that the difference between sexes may also relate to the overlap between moult and the onset of autumn territorial behaviour in males. 15'31 This study indicates the great potential for biometric studies of data collected routinely by members of the British Ringing Scheme. While detailed observations of fat and muscle condition of birds undergoing moult are still required to identify more precisely which tissues are involved in mass change, the Moult Enquiry data clearly represent a valuable resource for initial analysis in this area. For example, in the analysis of PN moult I have considered a single year. Equivalent data are available for the whole period so that the analysis of between-year variation in moult and condition could be considered in future in relation to differences in environmental conditions, as might comparative studies, for example, of migratory and resident species.

10 154 A.G. Gosler Downloaded by [ ] at 10:29 17 December 2017 ACKNOWLEDGEMENTS I am grateful to the British Trust for Ornithology for permission to analyse these data and to Drs S.R. Baillie and P.C. Lack for their time and assistance in making the data available to me. I thank Professor A.A. Dhondt, Professor C.M. Perrins and Dr S.R. Baillie for their constructive comments on an earlier draft of the manuscript. Finally, I should like to thank the many ringers that contributed data to the Moult Enquiry since it too often goes unrecognized that such analyses are possible only after thousands of hours of fieldwork by others. REFERENCES 1. Tollefson, C.I. (1982) In Diseases of cage and aviary birds (Ed. M.L. Petrak). Lea & Febiger, Philadelphia. 2. Murphy, M.E. & King, J.R. (1984) Sulfur amino acid nutrition in the White-crowned Sparrow, 1. Does dietary sulfur amino acid concentration affect the energetics of molt as assayed by metabolized energy? Condor, 86, Murphy, M.E. & King, J.R. (1987) Dietary discrimination by molting White-crowned Sparrows given diets differing only in sulfur amino acid concentration. Physiol. Zool. 60, Murphy, M.E. & King, J.R. (1991) Nutritional aspects of moult. Acta XX Cong. Int. Om Ginn, H.B. & Melville, D.S. (1983) Moult in Birds. BTO, Tring. 6. Orell, M. & Ojanen, M. (1980) Overlap between breeding and moulting in the Great Tit Parus major and Willow Tit P. montanus in Northern Finland. Om. Scand. 11, Mead, C.J. & Watmough, B.R. (1976) Suspended moult of trans-saharan migrants in Iberia. Bird Study, 23, Swami, R.L. & Baillie, S.R. (1979) The suspension of moult by trans-saharan migrants in Crete. Bird Study, 26, Svensson, L. (1984) Identification Guide to European Passerines (4th edn). Naturhistoriska Riksmuseet, Stockholm. 10. Haukioja, E. (1971) Flightlessness in some moulting passerines in Northern Europe. Ornis Fenn. 48, Wunderle, J. (1991) Age-specific foraging proficiency in birds. Curr. Orn. 8, Gosler, A.G. (1993) The Great Tit. Paul Hamlyn, London. 13. Flegg, J.J.M. & Cox, C.J. (1969) The moult of British Blue Tit and Great Tit populations. Bird Study, 16, Dhondt, A.A. (1973) Post-juvenile and postnuptial moult in a Belgian population of Great Tits, Parus major, with some notes on captive birds. Gerfaut, 63, Dhondt, A.A. (1981) Postnuptial moult of the Great Tit in southern Sweden. Orn. Scand. 12, Gosler, A.G. (1991) On the use of greater covert moult and pectoral muscle as measures of condition in passerines with data for the Great Tit Parus major. Bird Study, 38, Snow, D.W. (1965) The moult enquiry. Fourth Report, June Bird Study, 12, Snow, D.W. (1967) A Guide to Moult in British Birds. BTO Field Guide No. 11. BTO, Tring. 19. Newton, I. (1966) The moult of the Bullfinch Pyrrhula pyrrhula. Ibis, 108, Ryan, T.A., Joiner, B.L. & Ryan, B.R. (1985) Mi nitab Handbook. PWS-Kent, Boston. 21. Summers, R. W., Swann, R.L. & Nicoll, M. (1983) The effects of methods on estimates of primary moult duration in the Redshank Tringa totanus. Bird Study, 30, Underhill, L. & Zucchini, W. (1988) A model of avian primary moult. Ibis, 130, Perrins, C.M. (1979) British Tits. Collins, London. 24. Gosler, A.G. (1987) Pattern and process in the bill morphology of the Great Tit Parus major. Ibis, 129, McCabe, T.T. (1943) An aspect of the collector's technique. Auk, 60, Helms, C.W. (1968) Food, fat and feathers. Am. Zool. 8, King, J.R. (1972) Adaptive periodic fat storage in birds. Proc. XV Int. Orn. Congr. Leiden, Blem, C.R. (1991) Avian energy storage. In Current Ornithology, Vol 7 (Ed D.M. Power), pp Plenum Press, New York. 29. Gaston, A.J. (1981) Seasonal breeding, moulting and weight changes among birds of dry deciduous forest of North India. J. Zool., Lond. 194, Bell, B.D. (1970) Moult in the Reed Bunting. Bird Study, 17, Dhondt, A.A. & Smith, J.N.M. (1980) Postnuptial molt of the song sparrow on Mandarte Island in relation to breeding. Can. J. Zool., 58, Evans, P.R. (1966) Autumn movements, moult and measurements of the Lesser Redpoll Carduelis flammea cabaret. Ibis, 108, Seel, D.C. (1976) Moult in five species of corvidae in Britain. Ibis, 118, Newton, I. (1968) The temperatures, weights and body condition of molting Bullfinches. Condor, 70, Evans, P.G.H. (1986) Ecological aspects of wing moult in the European Starling Sturnus vulgaris. Ibis, 128, Freed, L.A. (1981) Loss of mass in breeding Wrens: stress or adaptation? Ecology, 62, (MS received 1 September 1993; revised MS accepted 17 December 1993)