Interlocality Variation in Speed of Moult in the Citril Finch Serinus Citrinella

Ibis (2004), 146, 14–17

Blackwell PublishingInterlocality Ltd. variation in speed of moult in the Citril Finch Serinus citrinella

A. BORRAS, T. CABRERA, J. CABRERA & J. C. SENAR* Museu de Ciències Naturals (Zoologia), P° Picasso s/n, Parc Ciutadella, 08003 Barcelona, Spain

The speed of avian moult has been related to body condition, and this to habitat quality. These facts suggest a relationship between habitat quality and moult speed. We looked for evidence of such a relationship in the Citril Finch Serinus citrinella in the Pyrenees, where, because of a Mediterranean inﬂuence, there is an important bioclimatic contrast between north- and south-facing slopes of the same mountain. This contrast strongly affects the general body condition of birds there (higher body mass, fat score and survival on north-facing slopes). Citril Finches in the higher quality area moulted more rapidly (measured as residual wing raggedness) than those in the lower quality area, the two sampling localities being only 5 km apart. As the birds initiated moult at the same time in the two locations, this difference in speed cannot have resulted from a differential effect of photoperiod. This stresses the importance of habitat selection, even at very small scales, for the completion of moult.

The moult speed and extent of a bird is assumed MATERIALS AND METHODS to reflect its body condition (Jenni & Winkler 1994). For example, high-quality Barn Swallows Hirundo rustica, A total of 3936 Citril Finches were captured during which display longer tail ornament, have better body the summer (June–September) from 1991 to 1996, condition, fewer parasites and moult more rapidly during 115 trapping days; 220 of these birds were than poorer-quality birds (Møller et al. 1995). The engaged in moult. Birds were captured at two local- body condition of a bird is related to the quality of ities, just 5 km apart, on Port del Comte mountain its habitat (Wiens 1989). Hence we might predict (Pyrenees, north-east Spain): Bofia (41°10′N, that birds occupying habitats of different quality 1°32′E, 2050 m u.p.s.) faces south with a habitat of should differ in the speed at which they moult their intermixed Mountain Pines Pinus uncinata and feathers. subalpine meadows; Vansa (41°12′N, 1°25′E, 2000 m The Citril Finch Serinus citrinella breeds in the u.p.s.) faces north and its habitat is dominated by boreal mountain zones of western temperate Europe open Mountain Pine forests. Because of a Mediterra- (Cramp & Perrins 1994), and maintains its highest nean influence from the south, there is a high biocli- densities in the Pyrenees mountains (Baccetti & matic contrast between north- and south-facing Märki 1997). Here, and especially in the west, there localities (Gutiérrez 1991); this, together with is an important bioclimatic contrast between north- additional topoclimatic factors, results in Bofia being and south-facing slopes (Gutiérrez 1991). We have hotter and drier than Vansa (Senar et al. 2002). Citril previously found that Citril Finches breeding on Finches at Vansa have a greater body mass, higher these two slopes, even when on the same mountain, fat scores and rely more heavily on pine seeds as a differ in body condition parameters (body mass and (highly energetic) food source than birds at Bofia, fat score) and survival rate (Senar et al. 2002). The which use a higher proportion of grass seeds (Senar aim of this study is to test for a locality effect in speed et al. 2002). Vansa is consequently defined here as of of moult (residual wing raggedness) related to these higher quality than Bofia. differences in habitat. Age (yearlings vs. adults) and sex of the birds were determined according to Svensson (1992) and Jenni and Winkler (1994). Remex moult was recorded *Corresponding author. according to Ginn and Melville (1983). Old feathers Email: [email protected] were given a score of 0, new fully grown feathers

Variation in speed of moult in the Citril Finch 15

a score of 5, and growing feathers a score from 1 to 4. The primary score is the sum of the scores for primaries, and the secondary score is the sum of the scores for secondaries. Raggedness is defined as the size of the gap in the wings caused by missing and growing feathers (Haukioja 1977); a raggedness score was defined as the complement of primary and secondary scores, i.e. a bird with a moult score of 4 was given a raggedness score of (5 − 4) = 1, and both old and new, fully developed feathers were scored as 0. The wing raggedness score is the sum of the scores for primaries and secondaries. The speed of moult was estimated as the residual from a model I linear Figure 1. Variation in moult speed of Citril Finches according to regression of raggedness on primary moult score, locality (n = 220 birds). according to Bensch and Grahn (1993) (taking moult scores > 5 and < 45; r = 0.45, F1,331 = 84.4, P < 0.001). Because moult speed did not vary across because Citril Finches desert the Bofia area before years (Vansa data, F3,153 = 1.29, P = 0.28, control- the end of moult. Taking only data from Vansa ling for age and sex), we pooled data for the different birds, we found that yearlings ended primary moult years. The timing of moult onset at the two localities 13.6 days later than adults (Table 3). No differences was estimated according to Rothery and Newton between the sexes were detected (Table 3). (2002) from the probit regression (Hooke–Jeeves method & quasi Newton, implemented in the DISCUSSION program STATISTICA) of moult state (0 = not started moult, 1 = started moult) on number of days from Because of an associated reduction in wing area, 1 June. Values reported are means ± se. moult typically impairs flight ability and increases predation risk (Swaddle et al. 1996, 1999, Swaddle & Witter 1997). We should therefore expect birds RESULTS to increase moult speed in order to reduce the moult- Adult Citril Finches moulted more quickly than ing period. Given that moult is energetically highly yearlings (Table 1). No sex effect was detected. demanding (Walsberg 1983a, 1983b), we should also Locality affected moult significantly, with birds expect birds in higher quality areas to moult more at Vansa moulting more rapidly than birds at Bofia rapidly than birds in lower quality areas. Our data (Table 1, Fig. 1). The date of primary-moult onset on the Citril Finch clearly support this relationship differed neither between the two localities nor between moult speed and habitat quality: birds between sexes (Table 2). However, on average, at Vansa, a high-quality area (Senar et al. 2002), yearling birds started moult 1.4 days before adults moulted more quickly than birds at Bofia, just 5 km (Table 2). We were unable to compare the timing of away. This difference cannot have resulted from a the end of primary moult between the two localities differential effect of photoperiod (Morton & Morton 1990, Dawson 1994, 1998, Dawson et al. 2000), because birds at the two localities started to moult Table 1. Three-way ANOVA on the variation in moult speed of the Citril Finch according to locality (Vansa vs. Bofia), age (adults vs. more or less simultaneously. yearlings) and sex. A prediction from this relationship between habitat quality and speed of moult is that birds F1,212 P should select ‘good’ areas in which to moult. This is well known to be so in geese, ducks and waders Locality 13.21 < 0.001 Age 4.35 0.04 (Salomonsen 1968, Jehl 1990, Kjellén 1994), but it Sex 0.13 0.72 has increasingly been found in recent years that Locality × Age 0.51 0.48 some long-distance migrating passerines may also Locality × Sex 0.01 0.92 gather at special areas to moult (Ellegren & Staav × Age Sex 0.05 0.82 1990, Thompson 1991, Young 1991). Given that Locality × Age × Sex 0.03 0.87 differences in habitat quality, which affect moult,

16 A. Borras et al.

Table 2. Timing of the start of primary moult according to locality (Boﬁa vs. Vansa), sex and age (yearlings vs. adults). Comparison tests, locality: P = 0.24, sex: P = 0.73, age: P = 0.01. N: sample size, a: intercept and b: slope for the probit regression model relating presence/ absence of moult for each bird to number of days from 1 June, according to Rothery and Newton (2002).

Na bDays from 1 June (se) Start date

Locality Bofia 714 −4.53 0.085 53.4 (11.8) 24 July Vansa 1273 −3.99 0.076 52.7 (13.2) 23 July

Sex Males 1208 −4.07 0.076 53.3 (13.1) 24 July Females 773 −4.59 0.086 53.1 (11.6) 24 July

Age Yearlings 1189 −4.56 0.088 52.1 (11.4) 23 July Adults 783 −4.09 0.076 53.5 (13.1) 24 July

Table 3. Timing of the end of primary moult according to locality (only data from Vansa are provided because birds desert the Boﬁa area before the end of moult), sex and age (yearlings vs. adults). Comparison tests, sex: P = 0.35, age: P < 0.0001. N: sample size, a: intercept and b: slope for the probit regression model relating presence/absence of moult for each bird to number of days from 1 June, according to Rothery and Newton (2002).

Na bDays from 1 June (se) End date

Locality Vansa 454 8.43 −0.065 129.9 (15.4) 8 October

Sex (Vansa) Males 219 9.29 −0.072 129.8 (14.0) 8 October Females 231 8.45 −0.066 128.5 (15.2) 7 October

Age (Vansa) Yearlings 144 8.30 −0.059 140.3 (16.9) 19 October Adults 299 10.35 −0.082 126.7 (12.2) 5 October

may occur very close together, as found in our REFERENCES study, we should also expect habitat selection in Baccetti, N. & Märki, H. 1997. Citril Finch. In Hagemeijer, W.J.M. relation to moult to appear in non-migratory but non- & Blair, M.J. (eds) The EBCC Atlas of European Breeding territorial species, such as cardueline finches (e.g. Birds: 711. London: T. & A.D. Poyser. Nakamura 1979). We believe that this merits further Bensch, S. & Grahn, M. 1993. A new method for estimating research. individual speed of molt. Condor 95: 305–315. Cramp, S. & Perrins, C.M. 1994. The Birds of the Western We thank two anonymous referees and the Editor for Palearctic, 8. Oxford: Oxford University Press. comments on an earlier draft of this paper. We are also Dawson, A. 1994. The effects of daylength and testosterone on the initiation and progress of moult in Starlings Sturnus grateful to Xavi Colomer, Daniel Mañas and Carles vulgaris. Ibis 136: 335–340. Serrasolses for field assistance, and Enric Hernández Dawson, A. 1998. Photoperiodic control of the termination of and Joan Parramón for allowing us to capture birds on their breeding and the induction of moult in House Sparrows property. This paper is a contribution to DGICYT research Passer domesticus. Ibis 140: 35–40. project DGICYT BOS 2000-0141 from the Spanish Dawson, A., Hinsley, S.A., Ferns, P.N., Bonser, R.H.C. & Research Council, Ministerio de Ciéncia y Tecnología. Eccleston, L. 2000. Rate of moult affects feather quality: a We would like to acknowledge the support and facilities mechanism linking current reproductive effort to future survival. provided by Jordi García-Petit, director of the Parc Proc. R. Soc. Lond. B 267: 2093–2098. Natural Cadí-Moixeró and Reserva Nacional de Caça de la Ellegren, H. & Staav, R. 1990. Ruggningsflyttning hos blåhaken Cerdanya-Alt Urgell (Departament de Medi Ambient, Luscinia s. svecica. Vår Fågelvärld 49: 80–86. Generalitat de Catalunya). Ginn, H.B. & Melville, D.S. 1983. Moult in Birds. Tr ing: BTO.

Variation in speed of moult in the Citril Finch 17

Gutiérrez, E. 1991. Climate tree-growth relationships for Pinus Swaddle, J.P., Williams, E.V. & Rayner, J.M.V. 1999. The effect of uncinata Ram. in the Spanish pre-Pyrenees. Acta Oecologica simulated flight feather moult on escape take-off performance 12: 213–225. in Starlings. J. Avian Biol. 30: 351–358. Haukioja, E. 1977. Flightlessness in some moulting passerines Swaddle, J.P. & Witter, M.S. 1997. The effects of molt on the in Northern Europe. Ornis Fennica 48: 101–116. flight performance, body mass, and behavior of European Jehl, J.R. 1990. Aspects of the molt migration. In Gwinner, E. Starlings (Sturnus vulgaris): an experimental approach. Can. (ed.) Bird Migration: 102–115. Berlin: Springer Verlag. J. Zool. 75: 1135–1146. Jenni, L. & Winkler, R. 1994. Moult and Ageing of European Swaddle, J.P., Witter, M.S., Cuthill, I.C., Budden, A. & Passerines. London: Academic Press. McCowen, P. 1996. Plumage condition affects flight perform- Kjellén, N. 1994. Moult in relation to migration in birds – a ance in Common Starlings: implications for developmental review. Ornis Svecica 4: 1–24. homeostasis, abrasion and moult. J. Avian Biol. 27: 103– Møller, A.P., Magnhagen, C., Ulfstrand, A. & Ulfstrand, S. 111. 1995. Phenotypic quality and molt in the Barn Swallow, Thompson, C.W. 1991. The sequence of molts and plumages Hirundo rustica. Behav. Ecol. 6: 242–249. in Painted Buntings and implications for theories of delayed Morton, G.A. & Morton, M.L. 1990. Dynamics of postnuptial plumage maturation. Condor 93: 209–235. molt in free-living mountain White-crowned Sparrows. Condor Walsberg, G.E. 1983a. Ecological energetics: what are the 92: 813–828. questions? In Brush, A.H. & Clark, G.A. (eds) Perspectives in Nakamura, H. 1979. Summer concentrations and moult in the Ornithology: 135–164. Cambridge: Cambridge University Press. Oriental Greenfinch Carduelis sinica. Tor i 28: 1–28. Walsberg, G.E. 1983b. Avian ecological energetics. In Farner, D.S., Rothery, P. & Newton, I. 2002. A simple method for estimating King, J.R. & Parkes, K.C. (eds) Avian Biology, Vol. 7: 161– timing and duration of avian primary moult using field data. 220. New York: Academic Press. Ibis 144: 526–528. Wiens, J.A. 1989. The Ecology of Bird Communities, Vol. 1 Salomonsen, F. 1968. The moult migration. Wildfowl 19: 5 –24. Foundations and Patterns. Cambridge: Cambridge University Senar, J.C., Conroy, M.J. & Borras, A. 2002. Asymmetric Press. exchange between populations differing in habitat quality: Young, B.E. 1991. Annual molts and interruption of the fall a metapopulation study on the Citril Finch. J. Appl. Stat. 29: migration for molting in Lazuli Buntings. Condor 93: 236– 425–441. 250. Svensson, L. 1992. Identification Guide to European Passerines. Stockholm: L. Svensson. Received 22 October 2001; revision accepted 4 November 2002.