Moult and Ageing of European Passerines

Moult and Ageing of European Passerines

First published 1994 by Academic Press Limited This edition published 2011 by Christopher Helm Publishers, an imprint of Bloomsbury Publishing Pic, 36 Soho Square, London W1D 3QY Copyright © 1994 Academic Press Limited The right of Lukas Jenni and FLaffael Winkler to be identified as the authors of this work has been asserted by them in accordance with the Copyright, Designs and Patents Act 1988. ISBN (print) 978-1-4081-5554-7 ISBN (e-pdf) 978-1-4081-5555-4 A CIP catalogue record for this book is available from the British Library All rights reserved. No part of this publication may be reproduced or used in any form or by any means — photographic, electronic or mechanical, including photocopying, recording, taping or information storage or retrieval systems — without permission of the publishers. This bo ok is produced using paper that is made from wood grown in managed sustainable forests. It is natural, renewable and recyclable.The logging and manufacturing processes conform to the environmental regulations of the country of origin. Printed in Great Britain by Martins the Printers, Berwick upon Tweed 10 9 8 7 6 5 4 3 2 1

Visit www.acblack.com to find out more about our authors and their books.You will find extracts, author interviews and our blog, and you can sign up for newsletters to be the first to hear about our latest releases and special offers

Moult and Ageing of European Passerines Lukas Jenni Schweizerische Vogelwarte Sempach Swiss Ornithological Institute

Raffael Winkler Naturhistorisches Museum, Basel Natural History Museum, Basel

Photographs by

Thomas Degen with additional material by Paul Mosimann, Christian Berger and the authors

CHRISTOPHER HELM LONDON

This page intentionally left blank

Contents Preface

ix

PARTI Chapter 1 The function and consequences of moult

Chapter 2

.,1 1.2

Functions of the plumage Plumage maintenance and the need for plumage renewal 1.2.1 Feather maintenance and wear 1.2.2 Adjustments to the plumage 1.3 The costs of moult 1.3-1 Plumage efficiency during moult 1.3.2 Energetic, nutritional and metabolic demands of moult 1.4 Fitting moult into the annual cycle

1 2 2 2 4 4 4 5

The terminology of feathers, plumages, moults and age classes

7

2.1

7 7 8 8 8 9 9 9 9

2.2

2.3

Chapter 3

1

Arrangement of the feathers 2.1.1 Flight feathers 2.1.2 Wing-coverts Plumages, feather generations and moults 2.2.1 Concepts of moult and plumage terminologies 2.2.2 General terms 2.2.3 Moult terms 2.2.4 Terms for plumages, feathers and feather generations Age classes

The moult of adults

11

3.1 3.2

11 12 12 12 12 12 14 14 15 16 16 17 17 17 17 17 17 18 18 18 20 20 20 20 21 22

3.3

Introduction to the moult strategies Sequence of moult 3.2.1 Basic sequence of the complete moult Flight feathers Body-feathers and wing-coverts 3.2.2 Functional aspects of the basic sequence of moult 3.2.3 Variations and exceptions to the basic moult sequence Primaries Secondaries Tertials Rectrices Wing-coverts Variation in the relationships between flight feather tracts Moult strategies 3.3.1 Complete postbreeding moult in the breeding area: Moult strategies 1 and 2 Complete postbreeding moult in the breeding area Partial prebreeding moult 3.3.2 Complete moult in the non-breeding area: Moult strategy 3 Partial moult before autumn migration Suspension of the complete moult within the non-breeding area Additional prebreeding moult Conclusions 3.33 Seasonally divided moult of remiges: Moult strategy 4 Seasonally divided primary moult Seasonally divided secondary moult 3.3.4 Partial and complete biannual moult of remiges: Moult strategies 5 and 6

vi

Contents

3.4

Chapter 4

3.3.5 Conclusions Timing and duration of the complete moult 3.4.1 Timing and duration of the complete postbreeding moult in the breeding area Reduction of moult duration Overlap between breeding and moult Overlap between moult and autumnal activities Arrested moult Moult duration Timing of moult 3.4.2 Timing of moult in trans-saharan migrants

22 23 23 23 24 24 25 25 25 26

The moult during the first year of life

29

4.1 4.2

29 29 29 29 31 31 31 31 32 32 32 32 32 32 34 37 37 38 38 39 40 40 41 41

4.3

4.4

4.5 4.6

Introduction to the moult cycles The juvenile plumage 4.2.1 Completion of the juvenile plumage after fledging 4.2.2 Structure of the juvenile plumage 4.2.3 Coloration of the juvenile plumage Sequence of postjuvenile moult Complete postjuvenile moult Partial postjuvenile moult 4.3.1 Sequence within wing-feather tracts during partial postjuvenile moult Wing-coverts and alula Tertials Secondaries Rectrices 4.3.2 Relationship between wing-feather tract renewal during partial postjuvenile moult 4.3.3 Sequence of eccentric and other partial primary moults Partial postjuvenile moult in the breeding area 4.4.1 Variation in extent, timing and duration: experimental evidence of control of postjuvenile moult 4.4.2 Interspecific variation in timing and extent Timing and duration Extent 4.4.3 Intraspecific variation in timing and extent Variation with hatching date Differences between populations Differences between the sexes and effects of energetic and nutrient stress Intraspecific variation in the extent of postjuvenile moult during the course of the non- breeding season and between wintering sites 4.4.4 Partial postjuvenile primary moult in the breeding area Complete postjuvenile moult in the breeding area Moults during the first year of life in trans-saharan migrants 4.6.1 Partial postjuvenile and partial prebreeding moult excluding remiges: Moult cycles 5 and 6 4.6.2 Complete moult in the non-breeding area: Moult cycles 9, 10, 11, 14 and 15 4.6.3 Partial moult including remiges in the non-breeding area: Moult cycles 7, 8, 12 and 13 First partial prebreeding moult including secondaries First partial prebreeding moult including primaries Incomplete first prebreeding moult 4.6.4 Conclusions

41 43 44 45 45 45 46 46 46 46 47

PART II Chapter 5

Ageing European passerines

49

5.1 5.2

49 49 49 50 50 50 50 51 51

Ageing criteria in live birds Ageing using plumage characters 5.2.1 Recognition of juvenile feathers Structure and shape Coloration Wear Growth bars and fault bars 5.2.2 Differences in extent of moult 5.2.3 Differences between postjuvenile and subsequent feather generations

Contents 5.3

Chapter 6

General ageing criteria in European passerines based on moult 5.3.1 Species with a complete postjuvenile moult in the first summer/autumn: Moult cycle type 1 5-3.2 Species with a partial postjuvenile/complete postbreeding moult in the breeding area: Moult cycle type 2 5-3.3 Species with a partial postjuvenile/complete postbreeding moult in the breeding area and a partial prebreeding moult in winter/spring: Moult cycle type 3 5.3.4 Species with a complete moult in the non-breeding area

vii 52 52 52 54 57

Species accounts

61

Presentation of the data and directions for use Material Relevance of the data Presentation and analysis of the data Procedure of ageing Species accounts Riparia riparia, Sand Martin Hirundo rustics Swallow Delicbon urbica, House Martin Anthus campestris, Tawny Pipit Anthus trivialis, Tree Pipit Anthus pratensis, Meadow Pipit Anthus spinoletta spinoletta, Water Pipit Motacilla flava, Yellow Wagtail Motadlla cinerea, Grey Wagtail Motacilla alba alba, White Wagtail Troglodytes troglodytes. Wren Prunella modularis, Dunnock Erithacus rubecula, Robin Luscinia megarhynchos, Nightingale Luscinia svecica, Bluethroat Phoenicurus ochruros, Black Redstart Phoenicurusphoenicurus, Redstart Saxicola rubetra, Whinchat Oenanthe oenanthe, Wheatear Turdus torquatus. Ring Ouzel Turdus merula, Blackbird Turduspilaris, Fieldfare Turdusphilomelos, Song Thrush Turdus iliacus, Redwing Turdus viscivorus, Mistle Thrush Acrocephalus scirpaceus, Reed Warbler Hippolais icterina, Icterine Warbler Sylvia curruca, Lesser Whitethroat Sylvia communis communis, Whitethroat Sylvia borin, Garden Warbler Sylvia atricapilla. Blackcap Phylloscopus collybita, Chiffchaff Phylloscopus trochilus, Willow Warbler Muscicapa striata, Spotted Flycatcher Ficedula hypoleuca^ Pied Flycatcher Parus ater, Coal Tit Parus caeruleus, Blue Tit Parus major, Great Tit Sitta europaea, Nuthatch Oriolus orioluSj Golden Oriole Lanius collurio, Red-backed Shrike • Garrulus glandarius, Jay Fringilla coelebs. Chaffinch Fringilla montifringilla, Brambling Serinus serinus, Serin Serinus citrinella, Citril Finch Carduelis chloris, Greenfinch Carduelis carduelis, Goldfinch

61 61 61 61 62 63 64 65 66 68 71 73 76 80 83 87 89 91 94 95 96 99 101 104 107 109 111 113 115 116 118 119 121 123 127 130 133 136 138 141 146 148 150 152 153 154 156 158 161 163 166 168 171

viii

Contents

Carduelis spiniiS) Siskin Carduelis cannabina^ Linnet Carduelisflammea cabaret, Redpoll Loxia curvirostra> Crossbill Pyrrhulapyrrhula, Bullfinch Coccothraustes coccotkraustes, Hawfinch Emberiza citrinella^ Yellowhammer Emberiza ciay Rock Bunting Emberiza hortulana. Ortolan Bunting Emberiza schoenidus, Reed Bunting

APPENDIX

The use of skull pneumatization for ageing The process of skull pneumatization Recognition of skull pneumatization Skull pneumatization scores Age determination by skull pneumatization Explanations of the graphs

173 177 180 182 188 189 191 193 195 198

201 201 201 203 203 203

References

209

Scientific names with their English, German, French, Italian and Spanish translations

219

Species Index

222

Explanations

224

Preface This book has two complementary aims. First, it presents an up to date summary of the moult patterns of European passerines with due emphasis on the extent of partial moults. This information is then used to demonstrate, with the aid of photographs, detailed ageing criteria for 58 species. Although not an exhaustive reference work on all aspects of moult and ageing in European passerines, the book is thought to be a valuable complement to other guides, both through its extensive photographic references and detailed discussion of the extent, sequence and strategies of moult. Part I gives an overview of the various moult patterns and strategies of European passerines based on our own observations and on data from the literature. Until now, most moult studies have investigated the timing, duration and sequence of moult by concentrating on actively moulting birds, and these aspects thus dominate the moult literature (summarized in e.g. Bub 1981, 1984, 1985, Bub & Herroelen 1981, Kasparek 1981, Ginn & Melville 1983, Glutz & Bauer 1985, 1988, 1991, Bub & Dorsch 1988, Cramp 1988, 1992, Rymkevich 1990, Svensson 1992, Cramp & Perrins 1993). For most species, the process of moult and how it is fitted into the annual cycle is well described, although the emphasis tends to be on the complete moult of the adults. In order to further our understanding, we particularly investigated the extent of partial moults, which have been much less thoroughly studied. Few species have been examined in detail in this respect and the information is often not readily available to the non-specialist. Recently, a detailed synopsis on the timing and duration of moult of passerines in NW Russia, including data on the extent of moult, was published by Rymkevich (1990). Since the extent of a partial moult can only be recognized after the moult is completed, we do not focus on the process of moult, but on the result of this process. This requires the recognition of the different feather generations of birds which have completed their moult, a technique widely used for ageing. Partial moults are of particular interest because they vary between and within species not only in timing and duration, but particularly in extent, a parameter clearly visible until the next moult. In contrast to timing and intensity, the extent of moult has direct effects beyond the actual moult period, such as on the quality of the plumage and its appearance. Thus, timing, duration, sequence and extent of moult are all important features in studies of the ecological functions of moult. We hope that the detailed description of partial moults given in this book will stimulate further studies in this area of investigation. Various moult strategies and extents of moult result from the amount of time available for moult and on how moult is fitted into the annual cycle. Until now, moult strategies have usually been described only by reference to the moult of the remiges. We have tried to describe the moult strategies of European passerines based on the seasonal occurrence and extent of moults of the entire plumage, which may provide a more realistic insight into the inter- and intraspecific variations in how moult fits into the annual cycle. Our account also highlights how incomplete is our knowledge of the moult strategies of trans-saharan migrants. The sequence of moult was generally thought to be relatively uniform among passerines and to show special adaptations in a few species only. However, as shown in chapters 3 and 4, a variety of moult sequences actually occur, In this book, we provide an up to date account of these less well known moult sequences, supplemented by our own observations.

Part II provides a general introduction to ageing by plumage characters followed by the species accounts. A knowledge of the moult cycle of a species and the variation in extent of its moult, is a prerequisite for ageing using plumage characters. We have therefore tried to explicitly deduce ageing criteria from the observed moult patterns. This results in the recognition of four main moult cycle types among the passerines breeding in Europe and a description of their corresponding general ageing criteria (chapter 5). We hope that this procedure, together with the information given in the first part of the book, will enable one to more easily age birds in active moult, to appreciate unusual and undescribed moult patterns and their implications for ageing and to develop plumage ageing criteria for species not previously studied in detail. The species accounts provide detailed information on the range of extent of moult for 58 species and explain how this information can be used for ageing. Thus, we avoid giving mere rules based on the most frequent moult patterns. We are well aware that the data given on the extent of moult relate mainly to central Europe and that the restricted number of species treated prohibits the use of this book as a general ageing guide to European passerines (cf. Svensson 1992). However, the main aim is to convey, with the examples given and with the general ageing criteria provided in chapter 5, how feather generations can be recognized and how plumage ageing criteria can be deduced. The colour photographs illustrate the variation in moult patterns and the recognition of feather generations and, hence, ageing criteria. They replace tedious descriptions of feather wear and colour differences and should also speed up the process of acquiring a feel for the clues to ageing a particular species. However, many ageing criteria rely on the recognition of very slight differences in colour and wear, and some experience in recognizing them is certainly required. In the colour photographs, we concentrated on the wing since most plumage ageing criteria are to be found there. Almost the whole range of moult extent is reflected since the marginal coverts are among the first, and the secondaries among the last tracts to be moulted. There are a number of ageing criteria other than plumage characters, e.g. iris colour, colour of the inside of the upper mandible and skull pneumatization. In the section 'best criteria' we mention such other useful ageing criteria, especially if they are more reliable and more easily recognized than plumage characters. However, with the exception of skull pneumatization, we cannot treat them in detail, because we only rarely tested their reliability quantitatively. This is the aim of an ongoing project by Swedish colleagues (Karlsson etal. 1985). Within the section 'best criteria', we systematically provide the period of validity of skull pneumatization for ageing based on our own data presented in the Appendix. In contrast to North American practice, skull pneumatization is not very widely used for ageing birds in the Old World, although it is a very reliable criterion for many species in summer and autumn, when it offers the possibility to check the reliability of newly discovered or ambiguous ageing characters. Many of the ageing criteria and unusual moult patterns described here could only be verified by skulling. Also, certain ageing criteria mentioned in the literature could be shown to be unreliable. The moult data and the colour photographs presented here were collected at several bird ringing stations in Switzerland in autumn and spring, on Ventotene Island, Italy in spring and from the collection of the Natural History Museum, Basel (details see chapter 6). Our interest in moult arose in 1972, but systematic data collection started in 1980.

x

Preface

In total, moult data have been taken from about 140,000 birds by the authors and by Paul Mosimann, Christian Berger, Markus Leuenberger and Thomas Degen. Skull pneumatization was recorded for all birds observed in the summer/autumn. All the photographs were taken from live birds, except about 20 which were from wing preparations in the Natural History Museum, Basel. A special method for taking photographs of wings was developed by Thomas Degen. The feathers of the right wing were carefully arranged with a pair of fine tweezers. Then, the wing was placed on an oblique board covered with a standard grey paper. A small piece of double-sided adhesive tape, 1.5 mm thick, held the wing just distally of the wrist. Photographs were taken on 35 mm slide film, 64 ISO, with a reflex camera mounted on a tripod and equipped with a ring flash. This relatively simple technical equipment had the advantage that it was easy to transport and to install, even in the field. A disadvantage was the inability to ensure exact equivalence of the colours between the slides. For our purpose, however, absolute colours are not so important as relative differences in colour between feathers on the same wing. This work would have been impossible without the tremendous help of many people and we are greatly indebted to them all: Christian Berger, Thomas Degen, Markus Leuenberger and Paul Mosimann showed great enthusiasm and perseverence in the collection of moult

data and photographs during several years. Others too numerous to name assisted at the bird ringing stations in Switzerland although Hildegard Messerknecht and Roland Ammann deserve our special thanks. Ernst Sutter laid the foundations of the carefully documented skin collection in the Natural History Museum, Basel, and provided the data for Garrulus glandarius. Fernando Spina, head of the Italian Ringing Scheme, enabled Raffael Winkler and Paul Mosimann to work on Ventotene Island. Peter Keusch, John Attard Montalto and Frank Neuschulz made it possible to catch and take photographs of Emberiza hortulana, Cisticola juncidis and Sylvia nisoria in their study areas. Markus Leuenberger and Thomas Degen helped with the analysis of the data. Les Underbill carried out the analyses of the dependence of moult extent on skull pneumatization scores and feather-length. Unpublished data were provided by B. Bruderer, C. de la Cruz, A.A. Dhondt, T. Fransson, C, Gauci, H. Leuzinger, A. Lindstrom, G. Nikolaus, D.J. Pearson, W. Priinte, F. Spina, J. Sultana and M. Widmer. T. Wesofowski translated important literature from Russian. E. Gwinner, D.G. Homberger, A. Lindstrom and D.J. Pearson provided valuable comments on earlier drafts of parts of the manuscript. Andrew Richford, editor and ornithologist at Academic Press, was a constant source of advice and encouragement.

PARTI CHAPTER 1

The Function and Consequences of Moult 1.1 Functions of the plumage The process of moult serves to keep the plumage in good condition and adjust its characteristics to the changing needs of a bird during its lifetime. Therefore, understanding moult requires a knowledge of the various functions of the plumage, Feathers form a flexible and light-weight protective barrier against mechanical impact, solar radiation and water. In addition, the plumage is the main layer of thermal insulation and plays an important role in thermoregulation. Short-term adjustments to the degree of insulation can be effected by changes in the position of the feathers. Long-term adjustments can be made by changing the number, mass or length of the feathers, by feather loss or by growing additional or differently structured feathers during moult (see section 1.2.2). The unique function of feathers is that they enable birds to fly by forming the major part of the airfoil-like wings and by clothing the body in an aerodynamically favourable shape. The plumage is also largely responsible for the appearance of the bird, giving it its colour and shape and thus is important for visual communication and camouflage. Specially coloured parts or adornments have often evolved which may be apparent only in one sex or only during a certain time of the year. Many other functions of the plumage or feathers only appear in certain taxa or species (mainly non-passerines), e.g. regulation of buoyancy (in waterfowl), production of sound (e.g. in snipe), deadening of sound produced by the wings (in owls and nightjars), transport of water (in sandgrouse), aid in hearing (in owls), mechanical support (in woodpeckers and treecreepers), tactile sensitivity (e.g. in kiwis), provision of nesting material (in certain waterfowl) and help in digestion (in grebes). Some functions of the plumage are in potential conflict with each other. Thus, the best plumage is a well-balanced compromise between opposing requirements which may differ between species, sexes, age classes and seasons. For instance, flight feathers are under strong selection pressure for optimum flight performance, and this generally impedes the evolution of elaborate adornments. There is a general conflict between the need for crypsis and conspicuous signal coloration. Compromise solutions to this conflict include: distinctive colour patterns which are unusual or easily distinguished close to, but are cryptic at long range, as opposed to conspicuous coloration which is easy to see at a distance (Butcher & Rohwer 1989); reduction of conspicuous coloration to certain feathers which are normally hidden and are presented only on certain occasions; changing the colour of the plumage seasonally through the abrasion of cryptic feather fringes or by moulting twice a year; confining conspicuous coloration to one sex or to certain age groups only. The compromise between conspicuousness and crypsis or between adornments and flight capability has been interpreted as a way in which $ can honestly advertise their own quality to potential mates or rivals by displaying a genuine handicap which they clearly have overcome (e.g. long tail-feathers which reduce flight capacity or foraging efficiency; coloration which is conspicuous to predators; Zahavi 1975, 1977). A further conflict may arise since feathers exposed to intense wear should not be light in colour (Burtt 1986), because a high melanin content (which gives black, brown, redbrown and yellow colours) increases their durability (Fig. 1 and 2). This may conflict with the colour requirements of camouflage or

display. On the other hand, plumage coloration and heat gain from solar radiation are more or less independent, because colour is only one of several factors determining the penetration of radiation into the plumage (Walsberg 1983). Very few studies have tried to use an integrated approach to understand how the various requirements determine the properties of the plumage. In a pioneering study, Burtt (1986) examined how feather durability, energy balance, communication behaviour and camouflage requirements interacted to determine the plumage colour of New World Parulidae. Our incomplete knowledge of the significance of plumage properties makes it difficult to always understand the timing, extent and sequence of moult of a particular species or individual.

Fig. 1. Carduelis carduelis ad $, 22 October. The whole plumage is fresh, the tertials, secondaries and primaries having unworn, white tips.

Fig. 2. Carduelis carduelis ad 6*, 25 April. The plumage is worn. Note how abrasion has especially affected the less resistant white tips of the exposed tertials and outer primaries. However, the white tips remain on the protected inner primaries and secondaries.

2

The Function and Consequences of Moult

1.2 Plumage maintenance and the need for plumage renewal Full-grown feathers are dead structures consisting mainly of avian keratin. Keratin is one of the most durable biological materials, with great strength, flexibility and resistence to hydrolytic, protein-digesting enzymes and bacteria. However, unlike other keratin structures like hair and claws, they cannot be renewed continuously from the base. Consequently, unsuitable feathers must be replaced totally. Furthermore, feathers can only be replaced by pushing out the old feather long before the new one is fully grown and functional. This is a major disadvantage of feathers since it results in a significant reduction in plumage function when many feathers must be replaced simultaneously. The regular replacement of all or part of the plumage is called moult. There are two main reasons for plumage renewal. (1) Damaged or lost feathers need to be replaced in order to maintain the function of the plumage. (2) A plumage unsuitable for ensuing purposes must be replaced in order to adjusts to the new requirements.

also on the behaviour of the bird (see Fig. 4—6). Generally, the more loosely textured feathers of the juvenile plumage are more prone to abrasion than the feathers of subsequent generations (Fig. 7). Exposed feathers abrade and bleach more readily than concealed feathers (Fig. 4-7). In many species, the tertials, rectrices, inner, but not the innermost tenth, greater coverts and the tips of the outer primaries are especially exposed, The secondaries, inner primaries and primary coverts, outer greater coverts and the carpal covert are normally well protected (Fig. 4—7). This difference in exposure may lead to situations in which feathers generated more than six months apart can no longer be distinguished by the degree of wear (see p, 66, 76, 124). Durability also depends on pigmentation. Feathers, or parts of feathers, carrying melanin pigments are more durable than those without melanin (Burtt 1986, Fig. land 2). The fact that the extent of wear differs with habitat, climate, weather and exposure of the feather always has to be kept in mind when assigning feathers to different feather generations and in reconstructing the extent of moult in birds outside the moulting season.

1.2.1 Feather maintenance and wear

1.2.2 Adjustments to the plumage

Feathers are regularly maintained by a variety of comfort behaviours such as preening, scratching, shaking, bathing and drying, oiling with the secretion from the preen gland, dusting, sunning and possibly anting (Simmons in Campbell & Lack 1985). Inevitably, however, feathers are lost or irreparably damaged through mechanical abrasion, photochemical processes, ectoparasites, bacteria and fungi. If not pathological, we subsume all these factors under the more general term 'wear' (other authors use the term 'abrasion* in the same general sense; e.g. Ginn & Melville 1983, Rogers 1990). Whole feathers may be lost through traumatic mechanical interaction with vegetation, the nest, conspecifics or predators. Feathers may also be lost under the stress of'fright moult' (Dathe 1955, Stiefel in Bub 1985, Lindstrom & Nilsson 1988) and in forming the brood patch (Schifferli 1981). Abrasion, caused by rubbing against other feathers, objects in the environment and particles in the air, damages the structure of the feathers by breaking off the barbicles and barbules and by cracking and splitting the barbs and the rachis (Burtt 1986). This impairs or may completely prevent the cohesion of the vane. Abrasion of the exposed and often looser textured feather fringes diminishes the surface area and may substantially change the shape of the feather (Fig. 5). Abrasion also affects the colours of the plumage. The feathers of many birds have dark centres and light fringes. Through wear, the plumage increasingly takes on the colour of the feather centres and may become less cryptic. Feathers are damaged by sunlight: ultraviolet light alters the physical structure of keratin and pigments leading to bleaching. Ectoparasites such as lice (Mallophaga) and feather mites (Acari) actually eat feather material. Regular moult may help to reduce the load of ectoparasites, as suggested by the lower number of Mallophaga found in an American Emberizine bunting (Ammospiza caudacuta) which moults twice a year, compared with a similar species (Ammospiza maritimd) moulting only once a year (Post &Endersl970). Wear depends on habitat, weather, climate and season. Birds sliding through dense, hard vegetation such as reeds, grass, sedges or thorny bushes, or exposed to sand and wind suffer more abrasion than aerial and perching birds. Birds exposed to intense sunlight suffer more bleaching than those living in shady forests (cf. Fig. 3 with Fig. 4-6). Bleaching may be more pronounced in tropical winter quarters than in the temperate breeding area. Changes in wing-length indicate that abrasion may be stronger during the breeding season (due to activities in the nest and intensive foraging) than outside the breeding season (van Balen 1967, Flegg & Cox 1977, Francis & Wood 1989). Wear affects different feathers differently depending on the exposure, shape, structure and pigmentation of the feather in question and

Birds may need to adjust their plumage in response to seasonal changes, age and changing environmental conditions. For some time, studies have centered on the description of changes in plumage coloration according to age and season, but only recently has it been realized that the actual plumage structure may also change. Many passerine species have a distinct juvenile plumage (see section 4.2) which is usually replaced with a more colourful adult plumage within the first year. In some sexually dimorphic passerines, first-year cT acquire a more 9 -like plumage rather than the adult- cT plumage, although they may in fact breed. This delayed plumage maturation has been explained in various ways (see e.g. Lawton & Lawton 1986, Rohwer & Butcher 1988, Butcher & Rohwer 1989). In a few species, the adult plumage may also become slightly more colourful with increasing age (e.g. Ficedula hypoleuca, Winkel et aL 1970; Saxzcola rubetra> Schmidt & Hantge 1954). As mentioned earlier, the requirements of camouflage and display may be in conflict. In many species, their relative importance differs depending on season or on the age of the bird. Therefore, it may be advantageous for birds to change their appearance during the year, or as they age. Apart from temporary exposure of normally concealed colourful feather patches or conspicuous structures, there are two ways in which appearance may be altered. First, abrasion is a limited, but nevertheless striking method of changing plumage colour. Well known examples are the light fringes on the body-feathers of species such as Fringilla coelebs, F. montifringilla, Carduelis cannabina, Sturnus vulgaris etc., which are worn away during winter to disclose the colourful or shiny breeding plumage in spring. Abrasion of these fringes is often facilitated by their light colours, which contain little or no melanin. The second method of changing the appearance is by moulting. It is the only way to profoundly change appearance with age or seasonally and to introduce special feather structures. Species having more than one moult per year usually change between a colourful 'display' plumage worn during the mating season, and a more cryptic, dull 'eclipse' plumage worn during the rest of the year (see section 3.3.1). Many species change from a cryptic juvenile plumage to a more colourful adult plumage (see section 4.2.3). Less well studied are changes with age and with environmental conditions of plumage structures important in behaviour and display. Many plumage characters which play a role in sexual selection, aggressive encounters and establishing dominance in non-breeding flocks are less prominent in first-year birds than in older ones. It has been shown recently that the extent of an apparent handicap (the length of the tailfeathers in S of the African sunbird Nectariniajohnstoni) varies yearly with differing food availability and weather conditions (Evans 1991). It

Plumage maintenance and the need for plumage renewal

Fig, 3. Sylvia atricapilla ad 9 , 2 5 April. The whole plumage is postbr. This species shows hardly any signs of wear in spring, although the plumage was acquired about nine months previously.

3

Fig. 5. Calandrella brachydactyla, 23 April. The whole plumage is postbr or postjuv. The feathers, acquired about nine months previously, are considerably abraded. In this species living in a sunny, dusty environment and running though grassy vegetation, wear has primarily affected the tertials, the innermost greater coverts and the outer webs of the inner secondaries, which are most exposed when the wing is closed. The primaries and outer secondaries are protected by the long tertials, and the central and outer greater coverts and the primary coverts by the inner greater coverts, and are thus hardly worn.

Fig. 4. Sylvia canttllamiA T and 56 are postjuv, the rest of wing juv. The feathers acquired 8-12 months previously are considerably worn. In this sit-and-wait predator, hunting prey mainly by flying into low vegetation, wear has mainly affected the tips of the remiges, including the secondaries, these feathers being those exposed when the wing is open. The relatively short tertials and the greater coverts are less worn, both because they are less exposed on the open wing and because they have been renewed in summer/autumn.

seems that conditions prevailing during moult or some clue to the likely future conditions perceived at the time of moult, determines the extent of the handicap a see part II), In Sturnus vulgaris, however, the remiges to be moulted last (the outermost primaries) or first (the innermost primaries since moult is eccentric) maybe retained (Schleussner etal. 1985, Evans 1986, Schleussner 1990, Meijer 1991); in the case of an eccentric primary moult, the suggested sequence is descendant according to Schleussner (1990), but ascendant according to Evans (1986). As far as is known, birds with a seasonally divided primary moult usually follow the normal descendant sequence and resume primary moult at the point of interruption (suspended primary moult; see section 3.3.3). However, Sylvia c. communis may retain some central primaries before the autumn migration and may show various irregular moult sequences (T. Fransson in lift., G. Nikolaus in litt.\ see p. 123 and section 3.3.3 for a possible explanation).

Sequence of moult

15

Fig. 12. Ficedula albicollis ad 6 after partial prebr moult, 12 May. MaC and MeC prebr. GC 1 postbr, 2—10 prebr. T prebr. S 1—4 postbr, 5—6 prebr. Rest of wing postbr. In certain species performing an extensive prebreeding moult which includes all tertials, the innermost secondaries may also be moulted.

Fig. 13. Percentage of individuals which have moulted a given tertial after completion of a partial moult, of those individuals (N) which have moulted at least one tertial (own data). The individuals are subdivided into those with all three tertials renewed (white), those with only one tertial renewed (hatched) and those with two tertials renewed (dotted). In all Motacillidae and in Phylloscopus collybita (prebreeding moult), adults and second-year birds are included. In Sylvia borin and Muscicapa striata (postbreeding moult) only adults are shown.

Fig. 11. Percentage of adults which have retained a given secondary after an otherwise complete moult (hatched) or have moulted a given secondary after completion of a partial moult (dotted), of those individuals (N) with at least one retained (hatched) or moulted (dotted) secondary. In spring birds Q£ Muscicapa striata, it is difficult to judge whether or not S 6 has been retained (cf. p. 139). Sources: Carduelis spinus, Ficedula hypoleuca, Emberiza hortulana, Sylvia c. communis spring: own data; Phylloscopus trochilus; Mead & Watmough 1976, Swann & Baillie 1979, own data; Sylvia c. communis autumn: Mead & Watmough 1976, Swann & Baillie 1979, T. Fransson in litt., own data; Sylvia nisoria: Hasselquist etal. 1988, Lindstrom et al. 1993; Muscicapa striata: Mead & Watmough 1976, Hansen 1985, Rymkevich 1990, own data.

Secondaries Exceptions to the normal ascendant sequence of a complete secondary moult are few, but frequent in partial secondary moult. In the course of a complete moult, a few individuals of many species may shed S 6 before S 5- The secondary moult of Locustella luscinioides may show a

variety of sequences (Steiner 1970, Thomas 1977, Miiller 1981). In Cinclus cinclus, the secondaries are moulted in a variable sequence, but S 6 usually before S 4 (Richter 1954, Stresemann & Stresemann 1966, Galbraith et aL 1981). Convergent secondary moult was found in all individuals of Muscicapa striata (Diesselhorst 1961, Stresemann 1963b, Williamson 1972) and Cisticola juncidis (sequence 1—2—6—3—5^; Gauci & Sultana 1981) and in 2.5% ofSturnus unicolor (Peris 1988). Species which normally renew all the secondaries during the complete postbreeding moult may exceptionally retain one or a few of the last secondaries (usually S 6 or S 5—6, e.g. Motacilla flava, Anthus trivialis, Fringilla montifringilla, Serinus citrinella> Carduelis flammea, see part II). Species in which a small percentage of the individuals retain one or a few secondaries during the postbreeding moult usually start with S 1, but may also renew S 6 in addition to the outer secondaries (9.3% of the individuals retaining at least one secondary in Carduelis spinus^ 7.0% in Ficedula hypoleuca^ Fig. 11). In Phylloscopus trochilus^ this percentage is about 40% (Fig. 11, cf. Norman 1991b) and in Emberiza hortulana, which usually retains part or all of the secondaries during the postbreeding moult, 23% moult S 6. Sylvia

16

The Moult of Adults

nisoria, which usually retains all or most of the secondaries, may moult S 1 or, if more secondaries are moulted, sheds S 1+2+6 almost simultaneously, followed by S 5, although other (irregular) sequences occur as well and rarely all the secondaries are moulted following the basic sequence (Hasselquist et al 1988, Rymkevich 1990, A. Lindstrom pers. comm.), In Sylvia communis, the moult wave starting at S 6 may regularly proceed further, resulting in a convergent sequence (Fig. 11). Long-distance migrants with a partial postbreeding moult including the tertials may, along with the tertials, also replace S 6 (e.g. Anthus campestris, Locustella naevia, Oriolus oriolus, Lanius collurio, see part II). Species which renew the secondaries during an extensive partial moult do not usually follow the basic ascendant sequence. During the partial prebreeding moult, the renewal of secondaries seems to proceed strictly descendantly in Ficedula hypoleuca (all birds which renewed at least one secondary moulted S 6, Fig. 11) and probably also in F. semitorquataand F. albicollis (Fig. 12, Roselaar in Cramp & Perrins 1993). A descendant sequence during the prebreeding moult was also observed in actively moulting Sylvia curruca (see p. 121). Other species which regularly renew tertials during the prebreeding moult may, along with the tertials, also replace S 6 (Anthus trivialis, A. spinoletta, Motacilla flava, Phylloscopus colly bita, see part II; Sylvia cantillans, own data). In Sylvia c. communis, the sequence during the prebreeding moult may be descendant, convergent or eccentric (31% renewed all the secondaries, 13% only S 6, 16% two to five innermost secondaries, 13% one or two innermost as well as one to three outermost and 3% one central secondary, N=39, own data). A study of actively moulting Sylvia nisoria in Kenya revealed that secondary moult usually proceeds convergently starting either with S 6 or S 5 and S 1 or S 2, but irregular sequences may occur as well (Hasselquist et aL 1988, Lindstrom et al 1993, A. Lindstrom pers. comm.). Hence, the central secondaries are moulted by most individuals, while the outer- and innermost may be retained (Fig. 11). The secondary moult of Anthus richardi may also show irregularities (Stresemann & Stresemann 1968a). Muscicapa striata may retain the central secondaries during the prebreeding moult and may moult some central secondaries during the postbreeding moult. To summarize, species retaining some secondaries during an otherwise complete moult have the tendency to moult their secondaries convergently, retaining the well protected central secondaries. During the complementary partial moult, the sequence may be either strictly descendant, convergent, eccentric or variable.

Tertials Although the most frequent sequence of the tertial moult is T 8—9—7, it is often varied (usually to 8-7—9), especially in species which shed them at short intervals. In those species which moult the tertials during the course of a partial prebreeding or postbreeding moult, T 8 is the most frequently moulted, in particular if only one tertial is renewed (Fig. 13). This suggests a divergent sequence within the tertials, although some individuals may renew only T 9 or T 7- In Motacilla, alba, individuals renewing only T 7 are relatively frequent. Motacilla flava usually moults all the tertials during the partial prebreeding moult, and the most frequent sequence is T 8-7—9 (Wood 1976), but the few individuals known to have renewed only one tertial all replaced T 7 (see p. 77). Rymkevich (1990) reports that Sylvia nisoria may renew only T 7 or T 7-8 during the postbreeding moult.

Rectrices During a complete moult, the rectrices may be shed almost simultaneously or in an irregular sequence (e.g. many individuals of Motacilla flava, M, alba, Anthus spinoletta, A, trivialis, A. richardi, Cinclus cinclus,

Fig. 14. Percentage of individuals which have moulted a given rectrix after completion of an extensive partial prebreeding or postbreeding moult, of those individuals (N) which have moulted at least one rectrix (own data). The individuals are subdivided into those with all six rectrices renewed, those with only one rectrix renewed, those with R 1 and R 2 renewed, those with R 1 and R 6 renewed and those with other combinations. In all Motacillidae and in Phylloscopus collybita (prebreeding moult), adults and second-year birds are included. In Sylvia borin (postbreeding moult) only adults arc shown.

Prunella modularis, Luscinia megarhynchos> L. luscinia, Oenanthe oenanthe, Turdus torquatus, Acrocephalus dumetorum, Regulus regulus, R. ignicapillus, Sturnus vulgaris, S. unicolor, Plectrophenax nivalis, Calcarius lapponicus), in a convergent sequence within each half of the tail (1-6-2-5-3-4, often modified: Motacilla alba, M. flava; Glutz & Bauer 1985), R 6 before R 1 (Motacilla cinerea) or R 6 before R 5 or R 4 (Hirundo rustica, Ptyonoprogne rupestris, Anthuspratensis). In the two Certhia species, R 1 is shed only after the simultaneously shed R 2-6 are well grown. As in woodpeckers, this can be explained by the importance of the supporting function of the tail (Stresemann & Stresemann 1966). In Muscicapa striata, moult of the rectrices proceeds centripetally from R 6 (Diesselhorst 1961, Stresemann 1963b,

Moult strategies Williamson 1972). A centripetal sequence is also found in some individuals of Sturnus unicolor (Peris 1988), while S. vulgaris may show nearly centripetal, alternating or irregular sequences (Bahrmann 1964, 1970). In the course of a partial prebreeding moult, the rectrices are usually renewed in the normal centrifugal sequence (e.g. Sylvia, borin and Phylloscopus colly bita, Fig. 14). Exceptions are again most frequent among the Motacillidae. Observations on birds whose prebreeding moult is completed (Fig. 14) suggest that rectrix moult usually starts with R 1. In Anthus trivialis, it is usually followed by R 2, rarely by R 6. In A. pratensisy A. spinoletta, Motacilla cinerea, M. alba and possibly A campestris (Stresemann & Stresemann 1968a), R 1 is usually or always followed by R 6, suggesting a convergent sequence within each half of the tail as Wood (1976) found in M. flava. In M. alba, R 6 may even be renewed on its own (Fig, 14). In Sylvia nisoria during the postbreeding moult, some individuals moult only R 1—2, others R 1-2+5-6 (Rymkevich 1990).

Wing-coverts Few studies have investigated the moult sequence of the wing-coverts during a complete moult and, as described above, there appears to be some variation between species, especially in the greater covert moult sequence. During a partial moult, the greater coverts are usually renewed descendantly, except that GC 10 is generally out of sequence as in the partial postjuvenile moult (see wing schemes in part II). However, in some species, the greater coverts are often moulted in an irregular sequence (e.g. Motacilla flava, Sylvia borin, S. c. communis, S. cantillans).

Variation in the relationships between flight feather tracts During a complete moult, the relationship between the renewal of the flight feather tracts exhibits some variation within and among species. The most divergent patterns are shown by species which accelerate the primary moult by having five to seven primaries growing simultaneously and in which the innermost primaries are shed at very short intervals. In these species, the onset of secondary moult is delayed relative to primary moult. In Luscinia luscinia^ the innermost four primaries are shed within four days. S 1 is shed only after P 7, and S 2 only after P 9. The retained secondaries, therefore, form a functional part of the wing while up to seven primaries are growing simultaneously, and are mainly renewed after the primaries are grown (Berger 1967). In other species with rapid moults, up to six primaries are growing simultaneously and secondary moult starts when P 6 or P 7 is shed, but is finished together with or shortly after primary moult (Oenanthe oenanthe, Williamson 1957b; Plectrophenax nivalis, Stresemann & Stresemann 1970a, Green & Summers 1975; Calcarius lapponicus, Francis et al 1991). As in Luscinia luscinia, the moult of the remiges of Cinclus cinclus does not proceed successively, but in three phases during which the inner primaries, the outer primaries and the secondaries are shed, the secondaries being moulted in an irregular sequence (Richter 1954, Galbraith et aL 1981). However, the total moult duration is not reduced in this species (mean of 70 days for primary moult and 15 additional days to complete secondary moult; Galbraith et al. 1981). This peculiar pattern is probably an adaptation for underwater locomotion, for which it appears to be essential to maintain an uninterrupted surface to the wing (Galbraith et al. 1981). Motacilla alba spreads the moult of the rectrices over almost the entire duration of the primary moult, thus the function of the tail is less impaired than in other species (Jukema & Rijpma 1984). Birds which suspend primary moult maintain the basic relationship

17

between primary and secondary renewal, whereas birds interrupting secondary moult, but renewing all the other feathers, interrupt the shedding of the secondaries at an earlier stage than would be expected in relation to the stage of primary moult. It appears as though the number of secondaries to be retained is predetermined.

3.3 Moult strategies 3.3.1 Complete postbreeding moult in the breeding area: Moult strategies 1 and 2 Complete postbreeding moult in the breeding area The majority of European passerines, i.e. all the sedentary and migrant species which winter in cold and temperate climates, as well as some of the long-distance migrants, perform a complete moult between the breeding season and the autumn migration or the onset of winter (groups 1 and 2 in Table 2). This strategy has several advantages. The plumage is renewed after the season of most intense wear and during a warm period which usually provides predictable and high food resources which both allow successful postfledging development of the young and a period during which the adults have no other major commitments and can take time to fit in their moult. Birds wintering in cold and temperate climates can renew their plumage before the cold period when they will need good insulation and during which food availability is reduced or unpredictable. The maintenance of social hierarchies or winter territories and good powers of flight for hard weather movements also require sound feathers. Apart from Loxia curvirostra with its particular moult strategy, we know of no species wintering in Europe that delays its postbreeding moult into the winter. Resident birds of temperate regions generally have more time available for moulting than do migrants. Passerines breeding in the arctic and wintering in cold or temperate climates as well as those longdistance migrants moulting completely in late summer have to squeeze in moult between a late breeding season and an early autumn migration. Adaptations in timing and duration of the complete postbreeding moult in relation to the breeding season, migration and onset of winter are discussed in section 3.4.1.

Partial prebreeding moult About one third of the species performing a complete postbreeding moult in the breeding area renew part of the plumage again during late winter and spring (moult strategy 2, species see p. 59). The timing, extent and duration of this partial prebreeding moult are poorly documented. Migrants normally undergo this moult in the wintering area, but some may finish moult only during the spring migration (e.g. Anthus spinoletta, Herremans 1987; Motacilla flava, Serra 1992; Phylloscopus collybita, own obs.). In Anthus spinoletta (Herremans 1987) and Motacilla flava (Wood 1976), Locustella • luscinioides, Acrocephalus schoenobaenus, A. arundinaceus, A. paludicola and A. scirpaceus (Pearson 1971, 1973, 1975a, Hanmer 1979, Aidley & Wilkinson 1987, Roselaar in Cramp 1992, DJ. Pearson in lift., part II). For A schoenobaenus and A. scirpaceus, this moult seems to occur in those individuals which moult completely during the early winter, in the northern part of the Afrotropics, but not in those which moult completely later on during the winter, further south (Pearson 1973, Roselaar in Cramp 1992; cf. section 3.4.2 and p. 118).

Seasonally divided primary moult

Fig. 15. Acrocephalus schoenobaenus 2y/ad after complete prebr moult which has been suspended within the primaries in the non-breeding area, 5 May. This bird suspended primary moult after the renewal of P l^t and resumed it later. P 1—4 are more bleached than P 5—10. S 1—6 were probably replaced during the course of the moult of P 5-10. The tertials and wing-coverts are difficult to assign. These feathers may have been moulted together with P 5—10, or later during a separate moult of body-feathers, except for the conspicuously older GC 1-2.

Conclusions It appears that trans-saharan migrants moulting completely in Africa may moult before autumn migration in Europe (a partial moult), in the northern Afrotropics (a partial or complete moult, or a complete moult suspended) and further south (a complete or partial moult). There is much inter-, but also intraspecific variation in the moult strategy. The question is whether the partial moult before the autumn migration represents a separate moult or whether it is the beginning of the complete moult carried out mainly in Africa. In the first case, the interpretation would be that of a partial postbreeding moult in the breeding area followed by a complete prebreeding moult in the non-breeding

Among the various seasonally divided moult patterns, suspension of the primary moult has been known for a long time, mainly from nonpasserines (e.g. Stresemann & Stresemann 1966). The complete moult starts in the breeding area, where some of the primaries are moulted, is suspended during the autumn migration and resumed at the point of interruption in the non-breeding area, thus conserving the basic sequence throughout (complete postbreeding moult suspended within primaries, group 5 in Table 2). Primary moult suspension is the normal strategy in Anthus campestris and A. richardi (Stresemann & Stresemann 1968a, own obs.), while a small percentage of both species perform a complete moult in the breeding area. Furthermore, primary moult suspension occurs regularly in a considerable proportion of Sylvia communis icterops (which normally moult completely in the nonbreeding area), in a small percentage of S. c. communis and Hirundo rustica (the first normally moulting completely in the breeding, the second in the non-breeding area), rarely in Riparia riparia, Delichon urbica, Muscicapa striata and Lanius senator? which normally moult completely during the winter, and in Sylvia cantillans (Sylvia communis and hirundines see part II; S. cantillans, Dowsett 1971; Muscicapa striata, Mead & Watmough 1976; Lanius senator, Svensson 1992). Birds in spring with clear signs of primary moult resumption (inner primaries older than outer; Delichon urbica, Fig. 63; Sylvia communis, see p. 124; S. cantillans, own obs; Anthus campestris, Fig. 70) confirm that the postbreeding moult has been suspended. All of these species,

Moult strategies except perhaps Hirundo rustica and Riparia riparia, renew at least some body-feathers, and sometimes some wing-coverts (Anthus campestris) and tertials (Sylvia communis, S. cantillans) twice a year, thus performing a partial prebreeding moult. Adult Sylvia c. communis which interrupt the primary moult before autumn migration, may retain the central primaries and thus show irregular moult sequences (T. Fransson in //#., G. Nikolaus in litt., see p. 123). There are no data on the subsequent moult of such birds in Africa. Since only second-year birds, and no adults, occur in spring with a complementary primary moult pattern (see p. 124), the moult of the inner and outer primaries during the postbreeding moult could be interpreted as being the completion of an eccentric primary moult which took place during the first prebreeding moult. In this case, the incomplete moult of the primaries during the postbreeding moult would not constitute the suspension of a normal postbreeding moult, but rather be a complementary facet of the foregoing first prebreeding moult, similar to a seasonally divided secondary moult (see below and section 3.3.5).

Seasonally divided secondary moult Birds which interrupt moult within the secondaries (groups 3, 4, 6, 12, 14 in Table 2) show rather variable and often complicated moult patterns. Two main types may be distinguished: birds which moult almost completely in the breeding area (groups 4 and 12) and those moulting almost completely in the winter quarters (groups 6 and 14). Group 4 comprises birds which undergo an extensive postbreeding moult in the breeding area, but which retain all or part of the secondaries, and occasionally some wing-coverts, tertials and rectrices (Sylvia nisoria, Hasselquist et al. 1988, Rymkevich 1990; S. communis, S. curruca, Ficedula hypoleuca> Emberiza hortulana see part II). In the nonbreeding area, these birds perform an extensive partial prebreeding moult which includes some or all of the wing-coverts, tertials and rectrices, and usually also renewal of some or all of the secondaries (postbreeding moult interrupted within secondaries, partial moult in the non-breeding area including secondaries, group 4 in Table 2, Fig. 16). As shown in section 3.2.3 for Sylvia nisoria, S. communis and Ficedula hypoleuca, the moult sequence of those secondaries lost in the

Fig. 16. Sylvia nisoria ad after partial prebr moult, 30 May. MaC mostly prebr. MeC prebr or postbr. GC 1+4 postbr, rest prebr. CC probably postbr. Al 1 prebr, 2—3 postbr. P and PC postbr. T and S 1-6 prebr. Adult Barred Warblers moult their primaries during the postbr moult in the breeding area together with all the tertials, part or all of the body-feathers, part of the wingcoverts, usually the inner rectrices and occasionally one or a few secondaries. During the prebr moult in Africa, they moulr most or all of the secondaries, all the recrrices and tertials and apparently part of the wing-coverts and bodyfeathers, but only rarely single primaries.

21

breeding area is ascendant or convergent, while secondaries moulted in the non-breeding area are renewed in a variety of sequences (usually descendant or convergent). Hence, the renewal of the secondaries in winter proves not to be a clear resumption of the postbreeding moult. The two partial secondary moults seem either to be indiscriminately interlaced or the secondary moult is shifted entirely to the prebreeding moult. Indeed, some of the secondaries retained during the postbreeding moult are occasionally not renewed in the non-breeding area so that they have to last for one and a half or two years (S. nisoria^ Hasselquist et aL 1988; S. communis Fig. 343, F. hypoleuca Fig. 417). On the other hand, some secondaries may be moulted twice a year (complete postbreeding moult, partial prebreeding moult including secondaries, group 3 in Table 2). The fact that first winter individuals of these species start renewing their secondaries during the first prebreeding moult in winter (see section 4.6.3) suggests that winter replacement of the secondaries prepares them for possible retention in the following postbreeding moult, and does not represent a resumption of a secondary moult suspended during autumn migration as in the case of a suspended primary moult (Hasselquist et al. 1988, Lindstrom et al. 1993). Emberiza hortulana might differ from the other species of group 4; the first winter birds examined did not renew any secondaries in Africa (see p. 195), and the secondary moult during winter may often be a resumption of the suspended secondary moult started in autumn. The moult strategy of the birds in group 4 (Table 2) shows additional features which are difficult to interpret at present and which are not presented explicitly in Table 2. In spring, Sylvia nisoria, S. cantillans and S. c. communis may show differently worn wing-coverts within the same wing (own obs., p. 124). It is not yet clear whether they have been moulted at different times during a protracted prebreeding moult or whether there is an additional, possibly overlapping, moult (cf. Roselaar in Cramp 1992 and p. 124). In Sylvia nisoria, some adults retain P 10 during the postbreeding moult (Rymkevich 1990) and others were found to have renewed the outer primaries during the prebreeding moult in Africa (Hasselquist et al. 1988, Lindstrom et al. 1993). Rarely, some adults renew all the secondaries in the breeding area (Rymkevich 1990), thus possibly performing a complete postbreeding moult. A postbreeding moult interrupted within the secondaries is found in several other species (group 12 in Table 2). In most, it is likely that the retained secondaries are moulted in the non-breeding area in the course of a partial prebreeding moult (although caged Sylvia hortensis did not; Berthold & Querner 1982a), thus they might actually belong in group 4. A moult pattern related to group 4, but seasonally reversed, is shown by some Muscicapa striata and Oriolus oriolus (see part II). These birds perform an almost complete moult in Africa, during which they may retain some secondaries, and a partial postbreeding moult of limited extent in Europe, where they may replace some secondaries (prebreeding moult interrupted within secondaries, partial postbreeding moult including secondaries, group 6). It is not known whether or not the two moults are actually complementary. Similarly, a few Sylvia borin moult one or a few secondaries in the breeding area. Since spring birds with old secondaries have been observed, it is possible that these feathers are retained during the prebreeding moult (seep. 128). Some Phylloscopus bonelli have also been found with partially renewed secondaries in the autumn, but with the other remiges unmoulted (Mead & Watmough 1976) and very likely also belong to group 6, although it is not known whether or not these secondaries are moulted again during the prebreeding moult. Occasionally, Lanius collurio and Locustella naevia may be found with renewed innermost secondaries in autumn (Swann 6c Baillie 1979, p. 154, Fig. 17), but it is not known whether these feathers are renewed again during the prebreeding moult (group 14 in Table 2). A few Acrocephalus schoenobaenus and Phylloscopus sibilatrix were observed with old secon-

22

The Moult of Adults

Fig. 17. Locustella naevia ad after partial postbr moult, 1 September. MaC mostly prebr, MeC and GC mostly postbr. T postbr. S 1—5 prebr, 6 postbr. Rest of wing prebr. This bird had also renewed the whole body-plumage and R 1-6, Some long-distance migrants with a complete prebr moult in the nonbreeding area may previously perform a rather extensive postbreeding moult in the breeding area. It remains unknown whether or not the feathers renewed in Europe are moulted again in Africa.

Fig. 18. Phylloscopus sibilatrix 2y/ad after an almost complete prebr moult, 28 April. Whole wing prebr except S 5. The feather generation of S 5 cannot be assigned. It may be a juvenile feather (in this case the bird is a 2y), a feather from the last complete prebr moult a year earlier or even a feather renewed in the course of the partial postbr moult during the foregoing summer. daries (mostly S 5, once S 1—2 and S 2) in spring (own data, Fig. 18, not included in Table 2). Whether or not this is related to a postbreeding moult including the secondaries is also unknown.

3.3.4 Partial and complete biannual moult of remiges: Moult strategies 5 and 6 A few individuals of Sylvia borin and Lanius collurio have been found to renew some primaries in the breeding area. They most likely renew them again in the non-breeding area, in the course of a complete moult later on (postbreeding moult arrested within primaries, complete prebreeding moult, group 9a), since no spring birds have been found showing signs of primary moult resumption. The large number of S. borin examined in the spring tends to confirm this interpretation (see p. 128). Only a few spring individuals of L. collurio were, examined, but a moult of this type would make them comparable to other shrikes which also have an extensive postbreeding and a complete prebreeding moult (see Stresemann & Stresemann 1972a).

Locustella fluviatilis moults the outer one to five (usually three) long primaries, part or all of the alula feathers, tertials> greater, median and marginal coverts in NE Africa, then migrates further south where it performs a complete moult (Pearson & Backhurst 1976, 1983, Tucker 1978). This species carries out both the postbreeding moult arrested within primaries and the complete prebreeding moult in the nonbreeding area (group 9b in Table 2) and may start the postbreeding moult in the breeding area, if the renewal of some body-feathers by some individuals in the breeding area is interpreted as the beginning of the partial moult (Glutz & Bauer 1991, Roselaar in Cramp 1992). There are records of a number of species renewing some primaries in the breeding area (postbreeding moult interrupted within primaries, group 13 in Table 2; Williamson 1968, Mead & Watmough 1976, Thomas 1977, Montalto 1988, Spina 1990, Nikolaus & Pearson 1991, Roselaar in Cramp 1992), but it has never been shown whether they belong to group 5 (suspended primary moult) or group 9a (arrested primary moult). Adult Locustella luscinioides have been shown to renew only the central or outer primaries (Thomas 1977). Similarly, some Locustella certhiola and L. lanceolata, wintering in tropical Asia, have been found with renewed outer primaries in the autumn, although it is not known whether or not such birds renew the outer primaries again during the prebreeding moult, like L. fluviatilis (Glutz & Bauer 1991, Roselaar in Cramp 1992). Only one trans-saharan migrant, Phylloscopus trochilus, regularly performs two complete moults each year (biannual complete moult, group 7). One other Palearctic passerine, Lanius cristatus, shows a similar moult pattern, but the complete postbreeding moult may be suspended during autumn migration or may occur after leaving the breeding grounds. Furthermore, this species may show divergent complete primary moult, arrested primary moult, incomplete secondary moult and irregularities in the moult sequence (Stresemann & Stresemann 1971, Neufeldt 1981, Roselaar in Cramp & Perrins 1993). Three more Palearctic passerines, Locustella certhiola, Lanius tigrinus and Pericrocotus divaricatus, are reported as having two complete moults annually (Stresemann & Stresemann 1971, 1972b, 1976). However in L. tigrinus and P. divaricatus, complete moult of the secondaries during the postbreeding moult has not been documented. Thus, they cannot be classified with certainty and may in fact belong to group 8 (postbreeding moult arrested within secondaries, complete prebreeding moult), together with those Phylloscopus trochilus retaining some secondaries during the postbreeding moult. In Z,. certhiola, it is not known whether the postbreeding primary moult is completed or whether the inner primaries are usually retained (similar to L. fluviatilis}, and whether birds performing an (almost) complete postbreeding moult have a complete or partial prebreeding moult (Roselaar in Cramp 1992). Some individuals of species which normally carry out their complete moult in Africa have been found to moult completely in Europe (group 11 in Table 2; Sylvia borin see part II, caged Lanius collurio, Kramer 1950). However, it is not known whether or not these individuals also perform a second complete moult in the non-breeding area, and thus belong to group 7. In Locustella luscinioides, this seems improbable according to some authors (e.g. Glutz &: Bauer 1991, Roselaar in Cramp 1992) and, consequently, individuals moulting completely in the breeding area might belong to group 1, even though a second complete moult cannot be excluded.

3.3.5 Conclusions As shown in the above, a thorough description of the sequence of plumages acquired by successive moults - the basis of any comparative study on moult - has still not been completed, especially for European trans-saharan migrants. Sadly, this prevents our being able to give reliable plumage and moult cycles for trans-saharan migrants and leads to

Timing and duration of the complete moult the approximate moult strategies presented in Tables 1 and 2. We are well aware that the description of the moult patterns in trans-saharan migrants presented above and in Table 2 is sometimes based on scant or speculative information. More detailed studies, especially of the species mentioned in groups 11-14, will certainly add more species to the groups mentioned here, modify and clarify the classification and interpretation of the moult pattern of certain species, and perhaps reveal additional moult strategies. In this respect, Locustella luscinioides with its wide variety of recorded moult patterns might be a promising species for further studies (Steiner 1970, Stresemann & Stresemann 1970b, Mead & Watmough 1976, Thomas 1977, Miiller 1981, Trias etal. 1982, Aidley & Wilkinson 1987, Bensch etal. 1991, Nikolaus & Pearson 1991). Many unresolved questions have been raised in the preceding sections and the following questions, in particular, merit further study. In certain trans-saharan migrants, it is still unclear whether or not there is a partial moult of the adults before autumn migration and, if so, whether or not this constitutes the beginning of the partial (e.g. Acrocephalus palustris, Locustella fluviatilis) or complete (e.g. Acrocephalus schoenobaenus] moult in Africa. The moult in Africa requires further study in order to see whether it consists of one or two moults (e.g. the postbreeding moult) or the completion of the postbreeding moult and an (overlapping?) prebreeding moult (e.g. Acrocephalus schoenobaenus, A. a. arundinaceus). This question is of special interest in birds with a seasonally divided moult of the remiges, since many Sylvia communis, S. cantillans and S. nisoria present two generations of wing-coverts, apparently acquired during the winter (cf. also Svensson 1992). In this context, the fact that some trans-saharan migrants moult certain feathers three times a year (Motacilla flava and Anthus cervinus, Pearson &C Backhurst 1973, DJ. Pearson in lift.} is significant and may represent the discovery of an additional moult, apart from the postbreeding and prebreeding ones. In summary (Table 1 and 2), European passerines exhibit the whole range from one to two complete moults each year to almost all possible partitions of the moults between the two main moult periods. However, most species and most individuals of a species cluster around particular patterns, justifying a classification into moult strategies. Most European passerines renew the whole plumage during a continuous complete moult, while seasonally divided moult occurs only in a relatively few long-distance migrants. The main moult period for most European passerines is in late summer/autumn, though many longdistance migrants moult in winter in their tropical non-breeding areas. In northern latitudes, late winter/spring seems to be a rather unfavourable season for an extensive moult, probably because of the demands of territory defence and courtship, rather low food resources and the spring migration. No Palearctic passerine is known to perform a complete moult during spring, except when still in its tropical wintering area and with the exception of Carduelis spinoides which has an unusually late breeding season in August in the Himalayas and moults completely in May and June after spring migration (Whistler 1940). Based on intraspecific variation, transitions between the main moult strategies are recognizable along two axes (Table 1 and 2): on the diagonal a shift of the main moult from summer (strategy 1 and 2) to winter (strategy 3) or vice versa and on the lower horizontal axis an extending of the postbreeding moult to a second complete moult. A seasonal division of moult appears to occur in two. ways. The first is to suspend the complete postbreeding moult for autumn migration or within the non-breeding area, usually conserving the basic moult sequence (classical suspended moult, group 5 and lOa in Table 2). Intraspecifically, this occurs in some species which normally moult completely in winter, but already start before autumn migration (e.g. hirundines) as well as in Anthus campestris which may moult completely before the autumn migration. It might have evolved by extending the normal onset of body-feather moult before autumn migration to some

23

remiges (hirundines) or by postponing the completion of the postbreeding moult to winter (A. campestris). The primary moult of Sylvia communis and S. cantillans are exceptional in showing the whole range from a complete moult in the breeding area to a complete moult in winter (see p. 123). The second way of dividing moult seasonally seems not to be by a clear suspension, but rather through an incomplete moult before the autumn migration followed by an incomplete moult in winter, which are only more or less complementary to each other and show changes to the basic moult sequence (interlaced postbreeding and prebreeding moult). This type of seasonal division might have evolved from the replacement of secondaries in preparation for a possible retention during the next moult (e.g. Sylvia c. communis, S. curruca and Ficedula hypoleuca occurring in group 3 and 4 of Table 2) and appears to be the normal strategy in S. nisoria. Whether the seasonally reversed pattern (group 6) is a classical suspended or an interlaced moult is unknown. Thus, given our present state of knowledge, moult strategy 4 may actually comprise two quite different strategies, i.e. the classical suspended moult and the interlaced postbreeding and prebreeding moult. In Sylvia communis, both strategies have certainly been found (seep. 123). A trend towards two moults each year appears in species with a complete prebreeding moult which renew extensive parts of their plumage during a postbreeding moult. Intraspecifically, this probably occurs in Sylvia borin (occurring in groups 10, 9 and 11) and possibly in some of the species of groups 12 and 13. As shown by Phylloscopus trochilus and some shrikes (Stresemann & Stresemann 1971), it is the postbreeding moult which tends to be incomplete. The converse trend, a complete postbreeding moult and a tendency towards a complete prebreeding moult has not been developed very far (group 3).

3.4 Timing and duration of the complete moult 3.4.1 Timing and duration of the complete postbreeding moult in the breeding area Broadly, the postbreeding moult is fitted in between the end of the reproductive season and the beginning of autumn migration, autumnal territorial behaviour or the seasonal reduction in food availability. Consequently, the time available for the postbreeding moult varies considerably among species and individuals and may be very short. Four different means of fitting moult into a short period exist.

Reduction of moult duration This is achieved mainly by reducing the intervals between successive feather loss, resulting in an increased number of simultaneously growing feathers and, consequently, in a reduction of flight capability (Fig, 19 and 20). Rapidly moulting birds become reluctant to fly or even more or less flightless (e.g. Oenanthe oenanthe, Williamson 1957b; Luscinia luscinia, Berger 1967; Plectrophenax nivalis, Stresemann & Stresemann 1970a, Green & Summers 1975; L. svecica, Sylvia c. communis, Phylloscopus trochilus, Haukioja 1971; Calcarius lapponicus, Francis et al. 1991). Birds moulting very quickly also show an increased growth rate of the individual primaries. While primary growth rates of small passerines with moult durations of 50—90 days range from 2.1 to less than 3.4 mm per day (Zeidler 1966, Newton 1967, 1969, Dhondt 1973, Ojanen & Orell 1982, Winkler & Winkler 1985), the mean daily growth rates in the two fastest moulting birds (27—37 days) are 4.2 and 5 mm per day (Luscinia luscinia, Berger 1967; Plectrophenax nivalis, Stresemann & Stresemann 1970a).

24

The Moult of Adults

Fig. 19. Serinus citrinella ad 6 in complete postbr moult, 23 August. P 1-4 full-grown, 5-6 growing, 7-10 old. S 1 growing, 2-6 old. T 7+9 growing, 8 full-grown. CC and GC 1-6+8-9 full-grown, 7+10 growing. Al and MeC old. MaC growing or full-grown, undermost row old. Example of a slow moulting short-distance migrant with two simultaneously growing primaries and almost fully renewed greater coverts, the median coverts being still old (c£ Fig. 20).

In these cases, moult regularly overlaps with feeding the young or even with incubation. In large species, such as the Corvidae, which generally have long moult durations, overlap between breeding and moult is also frequent (Gwinner 1966, Kalchreuter 1969, Dorka 1971, Holyoak 1974, Seel 1976, Winkler etai 1988). There are, however, small species with long moult periods, in which moult also overlaps with feeding, incubation or laying (e.g. Ptyonoprogne rupestris, Stresemann & Stresemann 1966; Sylvia melanocephala, Cisticola juncidis, Gauci & Sultana 1979, 1981; Parus major, P. caeruleus> Flegg & Cox 1969, Dhondt 1973; P. montanus, Orell & Ojanen 1980; Sitta europaea, Matthysen 1986; Passer domesticus* Alonso 1984). In these species, the maintenance of flight capacity (especially in Ptyonoprogne rupestris) or the reduction of energy and nutrient demands during moult is apparently paramount. Generally, overlap between breeding and moult is more common in S than in ? and usually connected with a reduction or abandonment of parental care. During breeding, usually only one or a few innermost primaries are moulted, and rarely some tertials, rectrices and secondaries (Orell & Ojanen 1980). Moult progresses considerably more slowly than normally (van Hecke 1980, Boddy 1983) and, in Pyrrhula pyrrhula, does not include the body-feathers (Newton 1966). As shown by P. pyrrhula, Phylloscopus trochilus and Passer montanus^ it may be so slow that the next primary is only shed after the preceding one is full-grown, thus giving the impression of a suspended moult (Newton 1966, Kasparek 1979b, van Hecke 1980, Norman 1990b, Fig. 21). Primary moult may also start during or soon after breeding, but be suspended if another breeding attempt is made (van Laeken & Caekebeke 1982, Boddy 1983, 1992, Harper 1984).

Overlap between moult and autumnal activities Intense moult during migration does not occur in European passerines and was never observed among the many migrants we caught on Col de Bretolet. However, migration during the final stages of moult (P 9 or S 5 and 6 are still growing) occurs regularly (see part II, Evans 1966, Haukioja 1971, Hyytia & Vikberg 1973) and moult during the beginning of the autumn migration has been suggested for Motacilla flava

Fig. 20. Sylvia curruca ad in complete postbr moult, 25 August. P 1 fullgrown, 2—6 growing, 7—10 old. S 1 growing, 2—6 old. T 7+9 growing, 8 fullgrown. Al old. CC full-grown. GC all growing. MeC all growing. MaC old, growing or full-grown. Example of a fast moulting long-distance migrant with five growing primaries and simultaneously growing greater and median coverts (cf.Fig. 19)

Overlap between breeding and moult In species whose moult extends over a medium or long period, moult usually starts as soon as the young of the last brood become independent. However, late breeders may commence moult earlier; while caring for fledged young or even feeding nestlings (e.g. Carduelis flammea^ Evans 1966; Pyrrhula pyrrhula > Newton 1966). Overlap between breeding and moult is a regular phenomenon in species with short moult periods, such as long-distance migrants and birds breeding in the far north (e.g. Anthus trivialis, Luscinia megarhynchos, Phoenicurus phoenicurus^ Turdus iliacus, Sylvia curruca, Ficedula hypoleuca> Phylloscopus trochilus, Carduelis flammea, Pinicola enucleator, Calcarius lapponicus, Plectrophenax nivalis; Creutz 1955, Haukioja 1971, Haukioja & Kalinainen 1972, Hussell 1972, Green & Summers 1975, van Hecke 1980, Tiainen 1981, Ojanen & Orell 1982, Boddy 1983, Ginn & Melville 1983, Rvmkcvich 1990, Underbill etaL 1992),

Fig. 21. Anthus trivialis ad 9, with broodpatch, in suspension of recently started postbr moult, 10 August. P 1-2 are full-grown, P 3-10 still old. There are no feathers growing either on the body or wings. This bird is probably still breeding or caring for fledged young and has either temporarily stopped its postbr moult or is moulting very slowly.

Timing and duration of the complete moult

(Hereward 1979, Dittberner & Dittberner 1987). Ptyonoprogne rupestris apparently moults slowly during the autumn migration (Elkins & Etheridge 1977). In Phylloscopus trochilus and perhaps in many other long-distance migrants, moult seems to be timed so as to finish before the endogenously programmed start of migration, thus often overlapping with breeding (Tiainen 1981, Bensch et al. 1985, Underfill! rf*/. 1992). It also seems that moult is reduced in intensity, or avoided entirely, while defending territories in the autumn (Parus major, Dhondt 1973, 1981; Sitta europaea, Matthysen 1986; Passer montanus, Myrcha & Pinowski 1970) and while storing food for the winter (Parus montanus, Orell & Ojanen 1980). Dhondt (1973) suggests that <S Parus major in Belgium slow down their moult as an adaptation to the overlapping of moult and autumnal territorial behaviour by starting earlier (often during second broods) and finishing slightly later than the 2. Similarly, Gauci & Sultana (1981) argue that the moult of c? Cisticola juncidis is considerably longer than that of $ (92 versus 67 days) to allow for territorial defence against early-hatched first-year cJ, which are ready to breed even at this early age.

Arrested moult Some individuals do not complete their postbreeding moult, but retain some unmoulted feathers. This may happen 'accidentally' in almost every species, but seems to be more frequent in birds with short moult periods (long-distance migrants, northern populations, late breeders) and in Carduelis spinus whose breeding season is very prolonged. In most species which arrest the flight feather moult, the feathers to be moulted last (S 6, S 5—6, alula) are those retained (Sturnus vulgaris see section 3.2.3). In those species with one moult annually, these retained feathers are probably not moulted until the next complete postbreeding moult. In species which perform an extensive prebreeding moult, the retained feathers are usually moulted in the winter (see section 3.3.3 and 3.3.4).

Moult duration The timing and duration of the complete postbreeding moult in the breeding area has been studied in a large number of species and populations, usually with reference to the primary moult • (summarized in Ginn & Melville 1983). The start, duration and end of the moult, as well as the time needed for growth of the individual feathers, can be estimated from a sample of free-living birds caught only once, from recaptures, from captive birds or by counting the daily growth bars of shed feathers (Green 1974, Winkler et al 1988). Considerable statistical problems are involved when estimating the timing and duration of moult from birds caught only once (discussed by e.g. Evans 1966, Newton 1966, 1967, Pimm 1976, Kasparek 1980, Summers et aL 1983, Winkler et aL 1988) and the best analysis currently available (Underhill & Zucchini 1988, Underfill! et al. 1990) has not yet been used widely. Hence, many published estimations of moult duration are likely to be biased and not comparable. Nevertheless, the following generalizations can be made regarding the mean durations of primary moult (data from Ginn & Melville 1983, supplemented by more recent publications). In European passerines, the duration of primary moult depends only roughly on the size of the feathers. Large birds generally have longer primary moult durations than small birds (e.g Corvus corax 140-145 days, C. corone and C, frugilegus 105-172 days, smaller Corvidae 92-182 days). Thrush-sized birds have only slightly longer moult durations than smaller passerines (large thrushes and Sturnus vulgaris 70—100 days; Turdus torquatus alpestris, T. philomelos and T. iliacus

25

50—60 days) and relatively short durations in N Scandinavia (T. iliacus at 70° N 40 days, T. pilaris 51 days). Resident populations of small passerines generally take 60-85 days for primary moult, exceptions being Prunella modularis in Great Britain with only 54—60 days (Ginn 1975), Panurus biarmicus with only 45-55 days (Ginn & Melville 1983) and Cisticola juncidis in Malta with 67-92 days (Gauci & Sultana 1981). In partial migrants, primary moult lasts between 50 and 85 days and in short-distance migrants 45—60 days, exceptions being Motadlla altavndi 68-73 days (Ginn & Melville 1983) and Carduelis flavirostris with 75 days (Ginn & Melville 1983). Long-distance migrants (30-50 days) and the two arctic breeders Calcarius lapponicus and Plectrophenax nivalis (28—40 days) have the shortest primary moult durations yet recorded for European passerines.

Timing of moult In birds which moult over a medium or long period, there is considerable variation in the onset and duration of moult between individuals within a population, between years and between populations. In these species, birds breeding late (late breeders, years with a late breeding season, northern populations) usually start moult later than early breeders, but normally moult faster and may compensate almost totally for any seasonal delay (e.g. Parus major, P. caeruleus, Flegg & Cox 1969; Passer domesticus, Zeidler 1966, Haukioja & Reponen 1968; P. montanus, Deckert 1962; Fringilla coetebs, Dolnik & Blyumental 1967, Dolnik & Gavrilov 1980; Carduelis flammea, Boddy 1983; Pyrrhula pyrrhula^ Newton 1966; Emberiza schoeniclus, Kasparek 1980; Sturnus vulgarity Meijer 1991). For example, in Finland moult is shortened in Parus major by 8 days, in Phylloscopus collybita by 10 days and in Motadlla alba by 20 days compared to Great Britain or central Europe (Orell & Ojanen 1980, Ginn & Melville 1983), However, the onset of moult may be earlier and moult duration longer in northern than in southern populations if northern populations have only one instead of two broods annually (Emberiza schoeniclus, Kasparek 1980). In birds with short moult periods (long-distance migrants), moult is more synchronized within populations and moult duration hardly differs between southern and northern populations, moult having apparently already been shortened as much as possible (Phylloscopus trochiluS) Motadlla flava^ Phoenicurus phoenicurus, Saxicola rubetra^ Sylvia c. communis\ Ginn & Melville 1983, Underhill etaL 1992). In the majority of species studied, <J commence moult earlier than $, and reduce their share of parental care if they still have dependent young. 9 then have shorter moult durations in order to catch up (e.g. Orell & Ojanen 1980, Francis etaL 1991). In some species, no difference in timing between the sexes was observed (Sylvia melanvcephala^ Gauci & Sultana 1979; Passer hispaniolensis, Alonso 1984; Carduelis flammea in Great Britain, Evans 1966; Sitta europaeay Matthysen 1986; Motadlla flava, Hereward 1979; Fringilla montifringilla^ Ottosson & Haas 1991). Non-breeders (often second-year birds) and failed breeders usually start moult earlier than successful breeders (e.g. Corvidae, Bahrmann 1958, Kalchreuter 1969, Dorka 1971, Holyoak 1974, Seel 1976, Winkler etaL 1988; Pyrrhulapyrrhula, Newton 1966; Sitta europaea> Matthysen 1986; Cinclus cinclus, Richter 1954). The timing and duration of the complete postbreeding moult are therefore intimately related to both the preceding breeding season and the oncoming autumnal activities (e.g. migration, territorial behaviour, food storage etc.). Thus, trade-offs between flight capacity, energy and nutrient demands (moult duration), the length of the breeding season and period of parental care (overlap between moult and breeding) can be expected to operate. Until now, very few studies have investigated the effects of moult on other events in the life of birds in detail. For instance, Dhondt (1973, 1981) suggests that S Parus major may either start the postbreeding moult during the first brood and finish it before the autumnal territorial contests begin, or raise a second brood before

26

The Moult of Adults

moulting more slowly during autumnal territorial behaviour, in which case possibly shorter wings are grown. Bensch etal (1985) found that 5 Phylloscopus trochilus with small broods begin their moult just after hatching while ? with large broods start moult only once the young reach independence. The possible disadvantages faced by 9 with large broods who moult late might be to retain some secondaries or to moult at a faster rate. Thus, there may be a trade-off between clutch size and moult.

3.4.2 Timing of moult in tram-saharan migrants Various explanations have been proposed to account for the timing of moult in trans-saharan migrants. Pearson (1973) noticed that longdistance migrants which reach equatorial or southern latitudes frequently moult completely in Africa (see also Ginn & Melville 1983). He suggested that more time is available for the moult in Africa than between the breeding season and autumn migration and that the rapid and demanding spring migration can be undertaken with the benefit of new feathers. Swann & Baillie (1979) also pointed out that since the breeding season in E Europe is delayed, the time period available for a premigratory moult in eastern breeding populations is even shorter (e.g. in Sylvia communis icterops). Bensch etal. (1991) suggest that a tropical climate may also be helpful in reducing heat loss during feather replacement, though Aidley & Wilkinson (1987) argue that low temperatures are the reason for moult suspension in Acrocephalus schoenobaenus in Nigeria. Alerstam & Hogsted (1982) proposed that species whose nonbreeding area is larger than the breeding area may be expected to moult under the more relaxed conditions of the non-breeding area. The large variation in moult patterns suggests that the timing and duration of moult in any particular trans-saharan migrant is actually determined by the interaction of several factors, such as the timing of the end of the breeding season (which varies geographically and between individuals, e.g. failed breeders, those with second broods), the timing of autumn migration, the degree of flight capability to be maintained during moult, the seasonality and predictability of the environment as well as the inter- and intraspecific competition in both the breeding and the non-breeding area. The outcome of these highly complex trade-offs varies considerably between and within species, populations and individuals, as well as between years (e.g. Hyytia & Vikberg 1973, Berthold & Querner 1982a). Although our knowledge of the ecology of trans-saharan migrants is still limited, some broad trends can be proposed. In the northern tropics, migrants arrive at the end of the summer rains to find a rich abundance of insect life which, however, soon starts to decline (Aidley & Wilkinson 1987, Bensch et al. 1991). Thus, migrants wintering in subsaharan Africa, north of about 10° N, are faced with a short period favourable for moulting. Migrants wintering entirely in this area should either moult in the breeding area or try to arrive in the wintering area as soon as possible and moult there rapidly, or else split their moult into a first part in the breeding area and a second in the wintering area. Indeed, Fig. 22 shows that in the N African steppe and savanna belt (north of about 10° N in the west, extending further south in E Africa) over 70% of wintering European passerine species moult in the breeding area. The minority which perform a complete moult in this summer rainfall area moult before the end of November and as rapidly as do some long-distance migrants in N Europe (Aidley & Wilkinson 1987, Bensch etal 1991, Hedenstrom et al. 1993). Moreover, all those migrants which regularly divide their moult seasonally winter in the steppes and savannas of N and NE Africa (Anthus campestris, Locustella lu$cinioides> Sylvia communis communis and some Sylvia c. icterops, S. nisoria, S. hortensis, S. cantillans, Lanius nubicus, L. senator niloticus, Emberiza hortulana, E. caesia). Thus, a seasonally divided moult of the remiges might serve to fit moult into the two short moult periods before and after autumn migration. However, as shown by Locustella luscinioides and first winter

Fig. 22. Percentage of the total number of European passerine species present which perform a complete moult in their wintering grounds in the various parts of trans-saharan Africa. Species and their wintering distributions are taken from Moreau (1972).

Lanius senator in N Ghana, species moulting only part of the remiges may maintain a higher degree of flight capability by moulting more slowly (Bensch etal. 1991). Birds which winter around the equator and further south do not experience a prolonged drought during their stay (Moreau 1972, Pearson & Lack 1992), and hence generally encounter a prolonged favourable period for moulting. Fig. 22 shows that the proportion of European passerine species undergoing a complete moult in the wintering area increases markedly towards the southwest. Moult duration there is generally longer than in Europe or N Ghana (summarized in Ginn & Melville 1983, Dowsett-Lemaire & Dowsett 1987, Bensch et al 1991). A direct intraspecific comparison of moult duration is provided by Phylloscopus trochilus which takes 37 days to perform a complete moult in the breeding area, but 50—68 days in southern and central Africa (Underhill etaL 1992). The pattern of the rainy seasons in Africa profoundly influences the seasonal distribution, migratory patterns and moult periods of European migrants (Pearson 1990, Pearson & Lack 1992). The moulting season in E Africa generally coincides with the wet season and gets progressively later towards the south (Pearson 1973, Hanmer 1979). Many migrants follow the rains as they move south during the autumn and winter and regularly have two moult periods, one in autumn somewhere in NE Africa and a second after about December in the southern part of Africa. Thus, they may either divide their complete moult into two temporally separate phases (e.g. group lOa in Table 2) or may perform an additional partial moult either in NE Africa (e.g. Locustella fluviatilis, Acrocephalus palustris which moult completely in southern Africa) or in southern Africa (e.g. those Acrocephalus schoenobaenus moulting completely in NE Africa, see section 3.3.2). E and W Africa, therefore, provide different conditions for moult. In E Africa, migrants can successively exploit areas both

Timing and duration of the complete moult

north and south of the equator over a large N—S range, all of which provide a wide variety of habitats, especially savannas which are favoured by many migrant species. In W Africa, they are squeezed into the narrow strip north of the equator between the Sahara and the rain forest, from which they have little access to areas with late winter rains. A situation similar to that in E Africa is found on the Indian subcontinent where several Palearctic migrants (e.g. Acrocephalus dumetorum, A. agricola, Sylvia hortensis, Hippolais caligata) apparently exploit the temporarily abundant food resources in N India after the summer rains for a complete moult and then migrate further south in early winter when this area dries out (Gaston 1976). Similarly, Emberiza a. aureola. moults completely in E China during autumn, before migrating to SE Asia with a fresh plumage (Stresemann & Stresemann 1969a). The timing of the breeding season also affects the timing of moult in trans-saharan migrants. Sylvia c. icterops breeds later than S, c. communis (Swann &: Baillie 1979) and, in addition, encounters no drought period in its southern African wintering grounds. Both the time constraint in its breeding area and the conditions in its wintering area favour its moulting predominantly in Africa. The converse is true of Lanius senator. L. s. senator, wintering in W Africa, breeds about one month later than L s. niloticus of the Near East which winters in NE Africa. While the nominate subspecies moults the remiges in the wintering area, niloticus moults the primaries in the breeding area, but retains some or all of the secondaries (Nikolaus & Pearson 1991). There are, of course, exceptions and special cases to the very general pattern outlined above: some species (Sylvia cantillans, Hippolais pallida, Phylloscopus bonelli> Lanius senator senator) moult in the Sahel area and appear to be well adapted to dry habitats. 5. cantillans continues moulting even when Acrocephalus schoenobaenus suspends moult (Aidley & Wilkinson 1987). Other species performing a complete moult in the northern tropics include warblers of the genera Acrocephalus and Locustella which breed very late in the season and might be forced to moult in Africa by time constraints in the breeding area. Furthermore, their usual wetland and seasonally flooded habitats tend to persist some time into the dry-season, even in Africa. Ficedula

27

hypoleuca, F. albicollis, Luscinia megarhynchos and L. luscinia winter mainly south of 10° N, but perform a complete moult in the breeding area. It would be interesting to compare the ecological conditions for these species during the non-breeding season in the breeding and nonbreeding area. The hirundines need adequate powers of flight during their entire life, and must moult slowly over the entire non-breeding season starting with moult of the body-feathers, and rarely the innermost primaries, in the breeding area. Their moult takes longer than any of the other small European passerines (121—185 days; see part II and Ginn& Melville 1983). In summary, the timing of moult of trans-saharan migrants is still poorly understood, despite the general trends summarized above. In particular, it is still unclear as to why there are so many different seasonally divided moult strategies. Although most species cluster into distinct moult categories, moult cycles can be highly variable within a species and apparently adapted to the ecological conditions specific to a population or individual. Thus, the moult and plumage cycles of trans-saharan migrants cannot be considered as inflexible regimes, but rather as highly adaptive processes, sensitive to immediate environmental influences. Precisely how environmental conditions influence prebreeding and seasonally divided moults has hardly been studied. Berthold & Querner (1982a) concluded that the seasonal division of moult of Sylvia hortensis is not endogenously controlled, but is influenced in some way by the preceding breeding season (but probably not simply controlled by the end of breeding). T. Fransson (in lift.) found that one individual Sylvia c. communis retained one, four and five secondaries after three consecutive postbreeding moults. Studies on the control of moult in trans-saharan migrants while in Africa are few. They seem to utilize favourable environmental conditions during late summer in Europe, during autumn in the northern tropics and during winter in equatorial and southern latitudes (D. J. Pearson in lift.) where they may perform one to three partial and/or one or two complete moults which may be suspended during autumn migration or within Africa. How these, apparently individually variable, moult strategies are controlled and how, if ever, they can be related to moult cycles invites further exploration.

This page intentionally left blank

CHAPTER 4

The Moult During the First Year of Life 4.1 Introduction to the moult cycles At fledging, all young passerines wear the juvenile plumage, which differs from that of adults in two respects: first, it consists of fewer feathers and these are more loosely textured; second, in many species, its coloration differs. When the juvenile plumage is fully grown, it is fresh, while that of the adults is usually worn and in need of replacement. Consequently, the plumage cycle of juveniles is out of phase with that of the adults and must be brought into line with the adult cycle without replacing feathers too often or keeping them too long. In addition, environmental factors bind the juveniles to the same two general moult periods, in late summer/autumn and winter, as the adults. The moults during the first year of life must, therefore, address these issues and bring the juvenile plumage into the adult form. Depending on the moult strategy of the adults and the relative nature of the juvenile and adult plumages, the various European passerines display a range of moult cycles, which are summarized in Table 3. Note that the cycles numbered 1-15 are a product of the juvenile moult pattern and the various adult moult strategies described in section 3.3Most species whose adults perform a complete postbreeding moult in the breeding area change the juvenile body-feathers from some weeks to a few months after fledging. The juvenile remiges are retained slightly more than a year, until the next postbreeding moult when young birds enter the adult moult cycle (moult cycles 5 and 6). The postjuvenile moult is of variable extent and may include some primaries and secondaries (moult cycles 3 and 4). Only a few species renew the entire juvenile plumage soon after fledging with a complete postjuvenile moult (moult cycles 1 and 2). In species with a partial prebreeding moult, first-year birds also perform this moult, thus replacing the postjuvenile body-feathers within a few months of their first growing them (moult cycles 2, 4 and 6). In species whose adults perform a complete moult in a tropical wintering area, the partial postjuvenile moult is either of limited extent or virtually absent and the first-year birds usually rapidly align their moult with the adult pattern by also performing a complete moult in the non-breeding area (moult cycles 9 and 15). In species whose adults may show a seasonally divided moult of the remiges, the first-year birds perform a postjuvenile moult of limited extent and may also renew part of the remiges during an extensive first prebreeding moult (moult cycles 7, 8 and 12). Exceptions to the above and more complex moult cycles (moult cycles 10, 11, 13 and 14) are treated in section 4.6. In summary, most juvenile body-feathers are replaced soon after fledging, or at most after about six months in some long-distance migrants. The juvenile remiges, however, are usually-kept either for about a year, or for about six months in certain long-distance migrants. Less frequently, they may be shed after only a few weeks or months in species which perform a complete postjuvenile moult, or be retained for more than a year in species which migrate to tropical Asia (see section 4.6.2) and possibly also in some species which exhibit a seasonally divided moult of the remiges (see section 4.6.4). The juvenile wing-coverts, tertials and rectrices may be moulted together with the body-feathers or else with the remiges.

4,2 The juvenile plumage 4.2.1 Completion of the juvenile plumage after fledging At fledging, the first set of juvenile body-feathers is usually full-grown or almost so, but the remiges, rectrices and underwing-coverts are still growing and large parts of the body remain bare. An additional set of juvenile body-feathers grows after fledging in most passerine species (Rymkevich 1990, Dorsch 1993), and this has been described in detail for four species of Phylloscopus warblers (Gwinner 1969), Sylvia borin, S. atricapilla (Berthold et aL 1970) and Acrocephalus melanopogon (Leisler 1972). In some long-distance migrants, this second set is already growing during the last days in the nest. These feathers grow predominantly at the edges of the existing feather tracts and cover any extensive bare parts, especially on the breast and belly, but they also grow within the feather tracts of the nape, head and undertail-coverts. This completion of the juvenile plumage should not be confused with the postjuvenile moult which usually starts in the centre of the body-feather tracts. The growth of the second set of juvenile bodyfeathers and the postjuvenile moult overlap in long-distance migrants, but are separated by about three weeks in Phylloscopus coliybita (Fig. 23; Gwinner 1969, Gwinner et ai. 1971) and probably by more than a month in Parus caendeus (Bensch & Lindstrom 1992). Species with an accelerated juvenile development (Sylvia borin, Phylloscopus trochilus, P. bonelli, P. sibilatrix) do not renew the second set of juvenile feathers during the postjuvenile moult. Species with a slower postjuvenile development (short-distance migrants and residents) do so while growing a third set of additional body-feathers (S. atricapilla, P. coliybita and Acrocephalus melanopogorr, Gwinner 1969, Berthold et aL 1970, Gwinner etal. 1971, Leisler 1972, Dorsch 1993). However, latehatched short-distance migrants and residents may retain the second set (Rymkevich 1990).

4.2.2 Structure of the juvenile plumage In most species, the juvenile body-feathers are more loose in texture than those of adults, especially on the nape and undertail-coverts (Fig. 24). Compared with the adult body-feathers, there are fewer, more widely spaced, barbs and fewer barbules with booklets interlocking adjacent barbs (Gohringer 1951). Only a small area at the tip of the juvenile body-feathers is firmly interlocked. Thus, the individual juvenile feathers are lighter in weight and the total mass of the bodyplumage less than in adults. As shown for Sylvia atricapilla, the total number of body-feathers may also be smaller (Berthold & Berthold 1971). In a few species, the juvenile body-feathers are of relatively firmer texture than those of most other species (e.g. Locustella naevia, Acrocephalusschoenobaenus, Roselaar in Cramp 1992; hirundines). Whether the juvenile remiges of European passerines are also more loosely textured than those of the adults has not been studied, but inspection by eye, at least, does not show conspicuous differences (in contrast to tropical species, Fogden 1972). However, the juvenile flight feathers, especially the rectrices, are often narrower and more pointed

30

The Moult During the First Year of Life

Table 3. Moult cycles during the first one and a half years of life in European passerines. Moults indicated in bold are different in extent from those shown by adults at the same season. Moults in parenthesis occur in some individuals only. Partial moult denotes a moult without renewal of the secondaries and primaries. The extent of the various moults are simplified (see text for details). Moult cycle 9 may consist of various cycles (see text and section 3.3.2). For reference, the moult strategy of the adults is indicated (see Tables 1 and 2). Some individuals may switch from one moult cycle to the other (not indicated). Other moult cycles of transsaharan migrants (see section 4.6) and moult cycles ofLoxia curvirostra (see p. 182) and Cisticolajuncidis (see section 4.4.4) are not included. Moult cvcle

First summer/ autumn

First winter/ spring

Second summer/ autumn

Second winter/ spring

Postjuvenile rnoult

First preb reeding moult

First postbreeding moult

Second prebreeding moult

1

complete

—

—^- complete

2

complete

——^- partial

—^~ complete

3

partial + P

—

—^ complete

4

partial + P

—

5

partial

—

6

partial

—

7

partial

——^- partial + S

—^- complete, except S — —^- partial + S

8

partial

——^- partial + P eccentric

—>• P suspended

9

(partial)

——^- (partial) complete (partial) —^- (partial)

^- partial

—^- complete

—^- partial

—>- partial

—^- complete ^~ partial

—^- complete

Examples

*- e.g. Passer domesticus species see p. 53

1

^- Acrocephalus melanopogon

2

*- e.g. Carduelis chloris species see p. 55

1

*- e.g. Sylvia meianocephala species see p. 59

2

^ —>- partial

Moult strategy of adults

e.g. Erithacus rubecula species see p. 55

>• e.g. Motacillaflava species see p. 59 ^ e.g. Sylvia nisoria species see p. 59

1 2 4 group 3+4 4 group 5

—^- resumption + partial

^- e.g. Sylvia communis species see p. 59

—^- (partial) complete (partial)

^~ e.g. Acrocephalus scirpaceus species see p. 59

3

10

—>> partial (+ P eccentric)

—>- complete

^- Carpodacus erythrinus

3

11

—^- partial

—^ complete —^- partial

^- Emberiza a. aureola

3

—^- complete except S

^> e.g. Oriolus oriolus

4 group 6 3

^- partial

12

partial

——^ complete except S

—^- partial + S

13

partial

——>- partial + P eccentric

—^- partial + P suspended

14

partial

——*-

—^- partial

—^- partial + P —*~ complete

^ Locustella fluviatilis

5

15

partial

——^- complete

—^ complete

—^- complete

^* Phylloscopus trochilus

6

^ complete

•*- resumption

^> Lanius senator

Fig. 23. The progress of juvenile plumage development and the postjuvenile moult in four species of Phylloscopus warbler, showing the differences between the species. t^sSS^I the main period during which the second set of juvenile feathers develops. These feathers develop as shown: '(//////A on the back, upper belly, crown and nape, I>0\\\\NI on the chin and cheek, I I on the belly and flanks. ^BH shows the period during which the postjuvenile moult occurs. (After Gwinner 1969).

Fig. 24. Phylloscopus collybita in juvenile plumage (left), 5 August, and after postjuvenile moult (right), 19 September. The body-feathers of the juvenile plumage are more loosely textured than those of the first non-breeding plumage and all subsequent plumages.

than those of the adults. They are also slightly shorter on average, with the exception of the juvenile outermost primary which in many species is longer than in adults (this is very conspicuous in Panurus biarmicus). Thus the wings of juveniles are often slightly more rounded than those of the adults (Alatalo etal. 1984, Winkler & Winkler 1985). There appear to be two possible explanations for the loosely textured body-feathers of juveniles which are difficult to distinguish from each other. First, if the different colour of the juvenile plumage is required for only a short time and needs to be changed soon in order to fulfil

other behavioural functions (see section 4.2.3), there might be no need to develop a durable juvenile plumage. Second, nestlings which grow a loose juvenile plumage may spend less energy in feather growth and may thus fledge earlier or have more energy to direct into other body structures than nestlings which form a durable plumage (see also Fogden 1972). This would also prompt replacement of the imperfect juvenile feathers soon after fledging. This second hypothesis is supported by the fact that species with a juvenile plumage similar in colour to that of the adults also have a loosely textured juvenile

The juvenile plumage

31

plumage and that most passerines complete the juvenile plumage after fledging by adding a second set of juvenile feathers.

4.2.3 Coloration of the juvenile plumage In most European passerines, the juvenile plumage is cryptic and inconspicuously coloured. In many species, it differs markedly from that of the adults, but usually only on those feathers which are moulted soon after fledging such as the body-feat hers, wing-coverts and in a few species also the tertials and rectrices. The remiges are generally the same colour as in the adults. The juvenile body-plumage generally lacks conspicuous coloration (e.g. Erithacus rubecula, Regulus spp., Carduelis carduelis, Pyrrhula pyrrhula; Fig. 25 and 26). It is often spotted, streaked or barred (e.g. Turdidae, Muscicapidae, Laniidae, Prunella modularisy Cinclus cinclus), without gloss (e.g. Sturnus vulgaris, S. unicolor)y less intensely coloured (e.g. Motacilla spp., Parus major, P. caeruleus), and in sexually dimorphic species is more like the adult $ than the adult 6 plumage (e.g. Monticola saxatilis, Turdus merula, Sylvia atricapilla, Panurus biarmicuSy Passer domesticus, Fringilla coelebs^ Carduelis cannabina, C. flammea, Emberiza schoenidus). An exception is Tichodroma muraria whose juvenile wing-coverts are conspicuously coloured like the postbreeding plumage of the adults. The juvenile body-plumage is similar in coloration to that of the adults only in species in which the adult plumage is also inconspicuous and cryptic (e.g. Sylvia borin and most warblers of the genera Acrocephalus, Phylloscopus and Hippolais). The feathers acquired at the postjuvenile moult are generally adultlike in colour, although they may be less bright or have broader fringes (e.g. the head-feathers in Sylvia atricapilla and Fringilla montifringilla). Exceptions are some cJ Phoenicums ochruros, Carpodacus erythrinus and Oriolus oriolus which take on a 9 -like plumage in their first year, Lanius collurio and L. nubicus in which the postjuvenile feathers differ considerably from those of the adults and form a second juvenile plumage (see p. 154) and Sylvia nisoria in which the postjuvenile body-feathers and wing-coverts differ slightly but consistently from those of adults. The juvenile secondaries and primaries, which are kept for longer, are always the same colour as the adults with the exception of very few

Fig. 25. Regulus regulus in juvenile plumage (left), 7 August, and after postjuvenile moult (c?, right), 5 September. The juvenile body-feathers of many passerines lack conspicuous coloration, as for example the black and yellow crown in this Goldcrest. After the postjuvenile moult, the body-feathers are generally coloured as in the adult and express any sexually dimorphic characters, as for example the orange centre of the crown in this S Goldcrest.

Fig. 26. Carduelis carduelis in juvenile plumage (left), 24 August, and after postjuvenile moult (c£, right), 31 October. The juvenile plumage lacks the conspicuous colour pattern of the head. During the postjuvenile moult, the adult colour pattern is acquired.

Fig. 27. Tichodroma muraria after postjuvenile moult, 14 September. In most passerines, the juvenile remiges already show the colour pattern of the adults. species (e.g. Bombycillagarrulus). Conspicuous patterns on the remiges, for instance the yellow marks on the remiges of Carduelis carduelis (see p. 172) or the red on the remiges of Tichodroma muraria (Fig. 27), are already fully developed in juveniles. The sex of juveniles can therefore be determined in species with sexually dimorphic remiges (e.g. Turdus merula, Coccothraustes coccothraustes, Carduelis cannabina, C. Moris}. Two adaptive advantages have been attributed to the particular coloration of the juvenile plumage. First, the general cryptic effect of the juvenile plumage can be related to their lack of experience and consequently a greater vulnerability to predators. Second, the lack of conspicuous and distinctive patterns on the juvenile plumage may serve in intraspecific signalling. Many juveniles fledge while the adults are still territorial and producing second or substitute broods. A distinct juvenile plumage signals their status and avoids either aggressive or amorous responses from adults during the difficult first phase of independence. Lack (1943) showed that stuffed Erithacus rubecula in juvenile plumage are hardly attacked by adults during the breeding season while those in adult plumage are. However, we know of no behavioural studies investigating the effect of the juvenile plumage during the postbreeding season of European passerines.

4.3 Sequence of postjuvenile moult Complete postjuvenile moult The complete postjuvenile moult generally follows the basic sequence described in the discussion of the complete adult moult (section 3.2). No differences have been reported between the sequence of the complete postjuvenile moult and the postbreeding moult of conspecific adults (e.g. Zeidler 1966, Spitzer 1972, Pearson 1975b, Alonso 1984, Peris 1988). However, in Sylvia melanocephala, the renewal of the secondaries and tertials in juveniles starts later relative to the progress of the primary moult than in adults (Gauci & Sultana 1979).

Partial postjuvenile moult The partial postjuvenile moult starts on the upper- and underparts and soon includes the marginal coverts and feathers of the head. The under-

32

The Moult During the First Year of Life

and uppertail-coverts, as well as the median covens, are shed next, followed by the greater coverts and later by the alula, tertials and rectrices. The first areas to complete the moult are the marginal and median coverts, the tail-coverts and parts of the body. The head feathers are usually the last to be moulted. The tertials and rectrices are often not full-grown until much later than the greater coverts, so that in many species the final extent of the postjuvenile moult can only be determined when all the feathers on the wing are full-grown. The sequence of the postjuvenile moult of the body-feathers varies somewhat between species, and detailed descriptions are available for only a limited number (e.g. Sylvia borin and S. atricapilla^ Berthold et aL 1970; Sylvia spp. and Phylloscopus spp. Norman 1981, 1990a, 1991a, Bensch & Lindstrom 1992; Anthus spinoletta, Prunella modularis, Turdus torquatus alpestrisy Winkler & Jenni in Glutz & Bauer 1985, 1988; Pyrrhula pyrrhula, Newton 1966). In species and individuals with a protracted moult duration, the different feather tracts enter the postjuvenile moult successively and fewer feathers grow simultaneously than in species and individuals whose moult duration is short. In the latter, many feather tracts commence moult simultaneously (e.g. Dolnik &: Blyumental 1967, Berthold etal. 1970) and individuals may have well over half of the body-feathers growing simultaneously (e.g. a regular feature in Sylvia curruca). The progress of the formation of the second set of juvenile feathers and of the postjuvenile moult can be used to estimate the age (in days) of juveniles up to the end of the moult (e.g. Bensch & Lindstrom 1992), provided that differences in the onset, speed and sequence of moult between early- and late-hatched birds are taken into account (see sections 4.4.1 and 4.4.3).

4.3.1 Sequence within wing-feather tracts during partial postjuvenile moult

by T 9 and later by T 7 (divergent sequence; Parus caeruleus, Flegg & Cox 1969; Phylloscopus collybita, Norman 199la; Motacilla cinerea*, Herremans 1988a). This divergent sequence is also suggested by the frequency distributions of renewed tertials in about half of the species examined (Fig. 28). However, considerable individual variation exists and T 7 or T 9 may occasionally be renewed alone. In the remaining half of the species, the frequency distributions suggest a descendant sequence of tertial moult (T 9—8—7), though again, there is considerable individual variation. Why the tertials should moult either descendantly or divergently is not, at present, clear.

Secondaries In some species which regularly renew the tertials during the partial postjuvenile moult, occasionally S 6, rarely S 5—6 or S 4—6, may be included as well (Motacilla alba, Phoenicurus ochruros, Sylvia atricapilla, Phylloscopus collybita, Parus caeruleus, P. major, Fringilla coelebs, Carduelis carduelis, C, chloris, C. spinus, Serinus serinus, Emberiza schoeniclus, see part II, Mester & Priinte 1982; Sylvia undata, S. melanothoraxy S. rueppelli, Roselaar in Cramp 1992, Svensson 1992). In Saxicola torquata and Cyanopica cyana up to five secondaries may be renewed (Gwinner et aL 1983, Fracasso 1985, de la Cruz et aL 1991, 1992). It seems that the renewal of some secondaries occurs as an extension of the tertial moult and some individuals with renewed secondaries also have eccentrically renewed primaries (see section 4.3.3). During the course of an extensive renewal of the primaries, the outermost secondaries are renewed as well as the innermost ones in Carduelis carduelis and Saxicola torquata (Mester & Priinte 1982, Fracasso 1985; see section 4.3.3).

Rectrices Within the feather tracts of the wing, the moult sequence of the partial postjuvenile moult may deviate from the basic sequence of the complete moult of adults.

Wing-coverts and alula The moult of the marginal coverts starts in their proximal part. The undermost row is the last to moult and is the most likely to be retained. The median coverts are moulted simultaneously, or else in a variety of sequences, In many species, the greater coverts which are to be renewed are shed almost simultaneously or in groups. Basically, a partial greater covert moult follows the descendant sequence starting with GC 9, i.e. the outer greater coverts remain unmoulted. GC 10 is usually moulted out of sequence after the adjacent greater coverts are already wellgrown. Thus, in species which renew only a few greater coverts, GC 10 is frequently retained out of sequence (e.g. Motacillidae, Fringillidae, see part II). However, in some other species, only GC 10 or GC 9+10 arc moulted and individuals with only a renewed GC 9 are virtually unknown (e.g. Phoenicurus phoenicurus^ Sylvia borin, Ficedula hypoleuca). Species which only occasionally moult their greater coverts sometimes show irregular patterns (especially frequent in Emberiza hortulana), although this is generally less frequent than in those adults which moult some greater coverts during a partial postbreeding moult. If only one alula feather is moulted, it is usually the smallest, or if two are moulted, the two smallest, thus conserving the descendant sequence.

Judging by the frequency distributions of renewed feathers after the completion of partial postjuvenile moult, rectrix renewal usually starts with the innermost one (Fig. 29). In about half of the species examined, the basic centrifugal sequence is followed quite strictly. As in the complete postbreeding and partial prebreeding moult, the three Motacilla species moult the rectrices convergently within each half of the tail (see also Herremans 1988a for Motacilla cinerea) and R 6 may occasionally be renewed alone. A convergent sequence of rectrix moult is also suggested for the majority of Carduelis carduelis (67% of the individuals with two new rectrices show renewed R 1 +6,33% R 1 +2; see also Mester & Prtinte 1982). In other Fringillidae and in Emberiza schoeniclus^ R 6 and R 2 are moulted in similar frequencies, but more often than R 3—5, suggesting a tendency for a convergent sequence (see also Westphal 1976). In all the species examined, considerable individual variation exists and it is often difficult to decide whether or not an unusual pattern is due to accidental rectrix loss or a genuine peculiarity. The rectrices are among those juvenile feathers which may already show signs of wear at fledging and which may be very worn after only a few months, if they have not been renewed. Thus, in those species which regularly moult only some of the rectrices, it might be an advantage to moult the most exposed central and outermost ones. However, those species which only rarely moult the rectrices and those which regularly renew them all (Parus major) have apparently not evolved this moult sequence.

Tertials

4.3.2 Relationship between wing-feather tract renewal during partial postjuvenile moult

The sequence of tertial moult during partial postjuvenile moult is very variable within and among species. In many, T 8 is shed first, followed

During partial postjuvenile moult, the relationship between the sequence and extent of feather renewal in the wing-feather tracts varies

Sequence ofpostjuvenile moult

33

Fig. 28. Percentage of individuals which have moulted a given tertial, after completion of the partial postjuvenile moult, divided into those with one, two or all three tertials renewed. Only those birds which moulted at least one tertial are included (N= sample size: own data).

between species. The postjuvenile moult of the wing usually starts with the marginal coverts, followed by the median coverts, the greater coverts and carpal covert, the alula, the tertials and rectrices. In Parus major, moult of the tertials and rectrices starts about three weeks after the beginning of postjuvenile moult (Dhondt 1973) and in Phylloscopus collybita about 11-12 days after (Norman 1991a). We chose to take the number of renewed greater coverts as a basis for comparing the extent ofpostjuvenile moult among the different wingfeather tracts. Generally, the extents of the postjuvenile moult of the various wing-feather tracts are correlated among each other, both within and among species. Within a species, individuals with lots of renewed greater coverts are more likely to have renewed feathers of the other tracts than individuals with only a few moulted greater coverts (data for the different species see part II). However, this interdependence is not very strict and a large number of different moult patterns are seen within a species. Interspecifically, some species renew some or all tertials, rectrices, carpal covert or alula feathers while moulting comparatively few greater coverts, while in other species birds with some of these feathers renewed have usually moulted all the greater coverts. Anthus trivialis, A.

pratensis and A. spinoletta may occasionally moult a tertial even when no greater covert is renewed and very often do so when only one greater covert is moulted, while e.g. Carduelis Moris only moults the tertials when all ten, rarely nine, greater coverts are renewed (see part II). Fig. 30 shows the mean number of postjuvenile tertials, rectrices, carpal coverts and alula feathers moulted in relation to the mean number of renewed greater coverts for 56 species. On average at least one tertial is renewed by species with more than 8.5 greater coverts moulted. Species with less than 8.5 greater coverts renewed usually moult the tertials only sporadically with the exception of three Anthus species, two Motacilla species, Phylloscopus collybita, Troglodytes troglodytes^ Turdtis merula, Garrulusglandariusand three Fringillidae (Fig. 30a). The relationship between the extent of the rectrix, carpal covert and alula feather moult and the greater covert moult is basically similar (Fig. 30b—d). Roughly the same species renew these feathers when few greater coverts are moulted, with the exception of the Anthus species. The juvenile marginal and median coverts are partly retained in those species which show only a limited postjuvenile moult (see part II). Apart from the marginal and median coverts, the innermost greater coverts (except GC 10) are most exposed when the wings are closed. It

34

The Moult During the First Year of Life

Fig. 29. Percentage of individuals which have moulted a given rectrix, after completion of the partial postjuvenile moult, divided into those with all six rectrices renewed, those with only one rectrix renewed, those with R 1 and R 2 renewed, those with R 1 and R 6 renewed and those with other combinations of renewed R. Only those birds which moulted at least one rectrix are included (N=sample size: own data).

therefore seems adaptive that, during a partial moult, all species renew the greater coverts in a descendant sequence, so that the innermost ones are moulted most frequently. The importance of the renewal of the other feathers of the wing relative to the extent of greater covert moult is shown in Fig. 30. The regular renewal of the tertials when, on average, less than about six greater coverts are moulted, may be related to the marked protective function of the tertials in these species (e.g. Motacillidae) and to their divergent moult sequence (see above). Why Fringilla coelebsand Sylvia curruca^ which, on average, moult more than eight greater coverts, only very rarely renew tertials remains unclear. The rectrices are regularly renewed together with a few greater coverts in Motacilla cinerea and M. alba which have the longest tails among small European passerines. Why some species renew the carpal covert and the alula feathers when only a few greater coverts are moulted is also a mystery. The frequent renewal of the conspicuously coloured alula feathers by Garrulus glandarius might be related to their signal function. The degree of regularity in their barred pattern may serve to

highlight the feather growth bars and, hence, display regular growth, good condition and high 'quality' of an individual bird (Hasson 1991).

4.3.3 Sequence of eccentric and other partial primary moults Some first-year birds renew some of their primaries during an extensive partial postjuvenile or partial first prebreeding moult (see sections 4.4.4 and 4.6.3). The focus and sequence of partial primary moults are still poorly known. Theoretically, partial primary moults may be generated by (a) a descendant sequence starting with P 1 (basic sequence interrupted at some stage), (b) a descendant sequence starting with a central primary, (c) an ascendant sequence starting with a central primary, (d) an ascendant sequence starting with P 9 or P 10, (e) a divergent sequence starting with a central primary, (f) a convergent sequence or (g) two or three foci and various sequences. In most cases, partial primary moult does not start with the innermost primary, but includes some central or outer primaries and has

Sequence ofpostjuveniie moult

35

Fig. 30. Mean number of moulted tertials (a), rectrices (b), carpal coverts (c) and alula feathers (d) in relation to the mean number of renewed greater coverts, after completion of the partial postjuvenile moult for 56 species (for data see part II). Key species are indicated by their abbreviated scientific names (see Fig. 34, p. 39).

been referred to as eccentric primary moult (Fig. 31).-In practice, this term has been used to describe a moult pattern of old inner and new central or outer primaries (cases b-e above). However, we suggest that the term eccentric primary moult should only be used for those partial primary moults which start somewhere in the centre of the primaries (cases b, c, e, possibly g) and not for those which follow a descendant sequence starting with P 1 (case a) or an ascendant sequence starting with P 9 or P 10 (case d). However, an ascendant partial primary moult, starting with P 9 or P 10 (case d), can only be distinguished from an eccentric primary moult which includes the outermost primaries when the focus or sequence can definitely be determined in actively moulting birds. The sequence of eccentric primary moult Is usually assumed to be descendant and in most cases, this assumption would be correct as confirmed by most studies on actively moulting birds, which start from a focus distally from P 1 (Sylvia melanocephala, Gauci & Sultana 1979; S. melanothorax, Roselaar in Cramp 1992; S. hortensis, Williamson 1968; Lanius senator, Bensch et al. 1991; L. collurio, Gwinner & Biebach 1977; L. isabeliinus, Stresemann & Stresemann 1972a; Carduelis carduelis^ Mester & Priinte 1982, W. Priinte pers. comm.; C. Moris, Westphal 1976, own obs.; Loxia curvirostra, Herremans 1988b). This finding is backed up by studies on Amercian passerines also (Miller 1928, Michener & Michener 1940, Taylor 1970, George

1973, Lloyd-Evans 1983, Stangel 1985, Rohwer 1986) as well as for Australian Meliphagidae (Dow 1973, Paton 1982). However, exceptions to the descendant eccentric primary moult can be found. Evans (1986) lists three first-year Sturnus vulgaris apparently moulting ascendantly from P 6 or 7 to P 1. There are also a few cases which suggest that an ascendant sequence may produce the pattern characteristic of an 'eccentric' moult. A small percentage of first-year Cyanopica cyana replace some outer primaries, very probably ascendantly starting with P 10 (de la Cruz et al. 1992, C. de la Cruz in lift.; Fig. 32). A first-year Certhia brachydactyla also moulted its primaries ascendantly (P 10—8 half grown, P 7—6 in pin, P 5—1 old, possibly complete moult?; Copete & Senar 1990). Adult Locustella fluviatilis renew the outermost primaries ascendantly (see section 3.3.1). The peculiar primary moult pattern occurring in some first-year Cisticola juncidis, described as similar to an eccentric moult, actually consists of a divergent complete primary moult following an interrupted descendant primary moult started with P 1 (Gauci & Sultana 1981, C. Gauci in lift., cf. section 4.4.4). The examination of the distributions of renewed primaries in birds which have completed a partial primary moult reveals additional peculiarities of the moult sequence (Fig. 32). In all the Fringillidae, P 6 is renewed most often, suggesting at first sight a moult focus at P 6. However, Loxia curvirostra and some Carduelis chloris and C. spinus

36

The Moult During the First Year of Life

Fig. 31. Sylvia cantillans 2y 9 after partial prebreeding moult, 28 April. During the course of an extensive prebreeding moult, this bird renewed primaries 6-9 eccentrically. Note that the corresponding primary coverts have been retained.

may replace only one or a few inner primaries. Moreover, in C carduelis and, to a lesser extent, in C, Moris and C spinus, the more primaries that are renewed, the more proximal is the innermost renewed primary. If it is true that the sequence is invariably descendant, the focus of primary moult would, therefore, generally depend on the number of primaries to be renewed. This would agree with the finding that the focus of a basically descendant primary moult of adult cT Sturnus vulgaris shifts distally with increasing levels of experimentally administered testosterone which suppresses moult (Schleussner 1990). Thus, skipping the innermost primaries produces an eccentric moult where the focus is the more distal, the fewer the primaries that are to be renewed. This is precisely what seems to happen in the partial primary moult of Lanius senator (Fig. 32) for which a descendant sequence has indeed been observed (Bensch etal. 1991). In most individuals of the Fringillidae, the renewed primaries form an uninterrupted block of central primaries which may extend to P 9 or P 10. However, in some individuals some primaries not adjacent to a renewed primary are moulted, thus indicating two foci of primary moult (8.8% of Carduelis Moris showing partial primary moult, 8.7% of C cannabina, 5.4% of C. spinus, 0% of C. carduelis, 8.6% of Loxia curvirostra\ included in Fig. 32). These primaries are often either the innermost or the outermost (e.g. P 1-2+4-8 renewed in a Carduelis spinus, P 1+5-7 in a C Moris, P 3-7+9 in a C. Moris}. Furthermore, a few individuals of these species were found to have interrupted a primary moult which probably started with P 1 (not included in Fig. 32; see part II and Fracasso 1985 for Saxicola torquata). In Sylvia species which moult only some of the primaries, P 7—9 are renewed most often, but the moult patterns are much more variable than in the Fringillidae. Individuals with only one renewed primary suggest that the focus may be at any primary (Fig. 32). Furthermore, individuals frequently have old primaries in between new ones (26.4% of those showing partial primary moult in Sylvia melanocephala, 22.0% in S. communisznd 12.0% in S. cantillans). Again, in addition to the central primaries, the outermost or innermost ones may often be renewed (e.g. P 1-2+6-8 renewed in a S. communis, P 1+7—9 in a S. melanocephala). In some S. communis the primaries of three different blocks may be renewed (e.g. P 1 +6-7+9-10). Rarely, individuals of these Sylvia species may.moult only one to three innermost primaries (not included in Fig. 32). Thus it seems as if there might be a proximal, a central and a distal area within the primaries, each with a variable moult focus which can be activated alone or in combination with others. At present, nothing is known about the sequence of a partial primary moult which starts at two or three foci. There are a number of other characteristics of eccentric and other partial primary moults. As a rule, the primary coverts of the renewed primaries are either not moulted or are replaced independently of the

Fig. 32. Percentage of individuals which have moulted a given primary after completion of partial postjuvenile or first prebreeding primary moult out of those individuals (N)r which have moulted at least one primary. Individuals with only the innermost primaries renewed (descendant partial primary moult starting with P I ) are excluded, individuals showing two or three blocks of renewed primaries, including the innermost, are included. Except in Cyanopica cyana, the individuals are divided into those which renewed only one primary (hatched) and those which moulted more primaries (dotted). In Loxia curvirostra, Sylvia melanocephala and S. cantillans, P 10 was not assessed. Sources: Carduelis Moris: Westphal 1976, Mester & Priinte 1982, own data; C. cannabina: Mester & Priinte 1982, own data; C. spinus: own data; C. cardueli$\ Mester & Priinte 1982, own data; Loxia curvirostra: Herremans 1982, 1988b, own data; S. melanocephala: Gauci & Sultana 1979, Table 3 and own data; S. cantillans and S. c. communis: own data; Cyanopica cyana: de la Cruz et ai 1992; Lanius senator: Ullrich 1974, Roselaar in Cramp & Perrins 1993, own data, including six birds which have moulted all primaries.

Partialpostjuvenile moult in the breeding area corresponding primaries. This often helps to distinguish a partial postjuvenile primary moult from the suspended moult of adults, which generally includes the corresponding primary coverts in sequence with the primaries. The extent of an eccentric or other partial primary moult is not symmetrical on both wings in some 60% of cases. A partial postjuvenile moult which includes some primaries differs from the primary moult of adults in that it starts with the body-feathers and wing-coverts, not the primaries. The primaries are usually replaced together with the tertials, or shortly after when the wing-coverts and alula feathers are already growing or full-grown (Gauci &C Sultana 1979, own obs.). An eccentric moult is always combined with an extensive postjuvenile or prebreeding moult which includes all the bodyfeathers, marginal, median, greater and carpal coverts as well as at least part of the tertials, rectrices and alula. In many instances, some secondaries are also moulted, usually the innermost ones in a descendant sequence along with the tertials (Carduelis spinus> C. carduelis, C. Moris, Cyanopica cyana\ see part II and Westphal 1976, Mester & Priinte 1982, de la Cruz et al. 1992). In Carduelis carduelis and Saxicola torquata, the outermost secondaries may rarely be moulted during the course of an extensive partial primary moult (Mester & Priinte 1982, Fracasso 1985). During the first prebreeding moult, Lanius senator moults five to nine primaries and one to six secondaries (mean 3.5, all six secondaries by eight birds, N=53, own data). S 6 is always renewed and usually also some adjacent inner secondaries. In nine birds one or a few outermost secondaries were shed as well, suggesting a basically descendant sequence starting with S 6, and in a few birds, a convergent sequence. Most Sylvia melanocephala, S, communis and S. cantillans which moult some of the primaries also renew some or all of the secondaries (S. melanocephala: mean number of secondaries moulted 2.0, range 0-5, N=21; S. cantillans: mean 2.2, range 0-6, N=62; S, communis: mean 2.6, range 0—6, N=38; own data; see also Gauci & Sultana 1979, p. 124 as well as S. melanothorax, Roselaar in Cramp 1992). In all three species, one to four innermost secondaries are usually moulted, as well as sometimes one to three outermost ones, suggesting a convergent sequence (see Gauci & Sultana 1979 and p. 124). Descendant, convergent and sometimes irregular sequences in partial secondary moult are also observed in adults (see section 3.2.3).

4.4 Partial postjuvenile moult in the breeding area With the exception of a few species, European passerines perform a partial postjuvenile moult in their first summer/autumn (moult cycles 3-9, 12-15 in Table 3). Whether the moult of some long-distance migrants after autumn migration or during a stopover should be regarded as part of, or as the entire, postjuvenile moult is discussed in section 4.6.2. In this section, only partial postjuvenile moults in late summer/autumn in the breeding area are treated. The main function of a partial postjuvenile moult appears to be the replacement of the loosely textured and often differently coloured juvenile body-feathers (see section 4.2) and of those feathers which would be too worn if retained until the next moult. Factors discriminating against a postjuvenile moult of the remiges soon after fledging include reduced flight capability and increased energetic and nutrient demands during the sensitive first phase of independence, as well as the lengthening of the postjuvenile moult duration. On the other hand, retaining the remiges for a year or more means that durable remiges must be grown during the nestling phase. Indeed, only birds with a suitably long period of time potentially favourable for moult renew some or all of the primaries, mostly during the terminal phase of the moult (see sections 4.4.4 and 4.5). The timing, duration and extent of the partial postjuvenile moult are extremely variable among and within species, as we shall see in the

37

remainder of this section. Its extent varies between species from virtual absence to include all the body-feathers, greater coverts, tertials, rectrices, alula, carpal covert (moult cycles 5-15 in Table 3) as well as part of the secondaries, primaries and primary coverts (moult cycles 3 and 4 in Table 3, treated in section 4.4.4). Even within a species, the postjuvenile moult may show all the transitional stages between a partial moult including only part of the wing-coverts and a complete moult (see sections 4.4.4 and 4.5). The categories of postjuvenile moult shown in Table 3 are thus not clearcut, and transitions between them occur. Because observations on the control of postjuvenile moult can largely explain the observed inter- and intraspecific variation, this information is given beforehand in the next section.

4.4.1 Variation in extent, timing and duration: experimental evidence of control of postjuvenile moult In many European passerines, moult is controlled by an endogenous, circannual rhythm which is more pronounced in tightly scheduled, long-distance migrants than in short-distance migrants (reviewed in Gwinner 1986). This rhythm is synchronized with the natural year by seasonal changes in the photoperiod and experimental changes in photoperiod can modify the annual moult programme (e.g. Dolnik & Gavrilov 1980, Noskov & Rymkevich 1985, Gwinner 1986). In birds kept under natural daylength, there are pronounced differences between species in the timing, duration and extent of the postjuvenile moult, as observed in free-living conspecifics. These species-specific differences are less pronounced, but still apparent when the birds are kept under a constant photoperiod (reviewed in Gwinner 1986). As shown in four species of Phylloscopus and two species of Sylvia warblers, postjuvenile moult starts earlier, overlaps more with the growth of the second set of juvenile body-feathers, is of shorter duration, ends earlier, includes more simultaneously growing feathers, is of lesser extent and is less variable in its timing in the early departing, long-distance migrants P. bonelli, P. sibilatrix, P. trochilus and S. borin than in the late departing, short-distance migrants P. collybita and S. atricapilla (Gwinner 1969, Berthold etal. 1970, Gwinner etaL 1971). The timing, extent and duration of the postjuvenile moult also differ between populations of the same species. In Sylvia borin, S. atricapilla and Phylloscopus trochilus kept under natural photoperiods, postjuvenile moult starts earlier and is generally of shorter duration in northern than in southern populations. These differences persist when birds from different populations are held under the same constant photoperiod, suggesting an underlying genetic control of postjuvenile moult (Gwinner etaL 1972, Berthold etaL 1974, Berthold 1977, Gwinner 1979, Berthold & Querner 1982b). Indeed, hybrids of two populations of S. atricapilla and of Saxicola torquata with different natural timing of the postjuvenile moult showed an intermediate timing and moult duration. The postjuvenile moult is thus partly controlled by an innate, genetically fixed time prograjnme (Berthold & Querner 1982b, Gwinner & Neusser 1985). This ultimate genetic time programme can be modified by the prevailing photoperiod. In several species, it can be shown that latehatched individuals held under natural daylength, moulted at an earlier age and more rapidly than early-hatched conspecifics of the same population; likewise, birds held under a constant short daylength moulted at an earlier age and more rapidly than birds of similar hatching date held under long daylength (e.g. Berthold et al 1970, 1972, Gwinner et al 1971, Dolnik & Gavrilov 1980). Noskov & Rymkevich (1985) found this dependence of postjuvenile moult on photoperiod in almost all of the 42 European passerine species examined. According to their analysis, postjuvenile moult proceeds normally only within a certain interval of daylength, specific to the species and population. If the daylength is too short for a given species and population, moult stops.

38

The Moult During the First Year of Life

The start of moult is largely determined by the hatching date. In species with a long breeding season, photoperiod is capable of modifying the onset of moult to a greater extent than in species with a short breeding season. For instance in Emberiza citrinella, the maximum, probably genetically controlled, age at which postjuvenile moult starts is 60 days, which cannot be increased by increasing daylength. Under short days of 14 hours, postjuvenile moult starts at the age of 20 days. In E. hortulana? however, the maximum age at which postjuvenile moult starts under a long day regime is only 25 days; the minimum, prompted by short days is 18 days (Noskov &: Rymkevich 1985; see also Gwinner etal. 1971). These experimental results show that the controlling programme is broadly genetically fixed and adapted to the population. It can be modified to a certain extent by photoperiod (though other factors which may modify postjuvenile moult are treated below). Late-hatched birds moult young and fast and can at least partly catch up and finish postjuvenile moult only shortly after their early-hatched conspeciflcs. The following two sections will show that the partial postjuvenile moult of free-living birds is also governed by these rules,

4.4.2 Interspecific variation in timing and extent Timing and duration The postjuvenile moult may start immediately after fledging (e.g. Sylvia nisoria, Glutz & Bauer 1991; Ficedula hypoleuca, Muscicapa striata, Rymkevich 1990) or more than two months after leaving the nest (e.g. Cardueiis spinus, lovchenko & Smirnov 1987; Loxia curvirostra, see p. 182; Carpodacus erythrinus and Emberiza a. aureola see

section 4.6.2). Intense postjuvenile moult does not generally overlap significantly with autumn migration or the cold season, but overlaps with low intensity moult are more frequent than in the postbreeding moult of adults. In many species, low intensity postjuvenile moult may continue during the first stages of migration, as for instance in Luscinia svecica (Lindstrom et al. 1985), Acrocephalus scirpaceus (Herremans 1990b), Phylloscopus trochilus (Norman 1981) and Motacilla flava (Dittberner & Dittberner 1987). In the long-distance migratory hirundines, moult of the body-feathers occurs at least during the first part of the autumn migration (see part II) and Ptyonoprogne rupestris moults slowly during the entire autumn and winter (Elkins & Etheridge 1977). Emberiza schoenicluswzs found in active body-feather moult as late as December in S France (Bell 1970). Despite these overlaps between moult and autumn migration, the postjuvenile moult is generally fitted within the period between fledging and migration or the onset of the cold period (species delaying postjuvenile moult until after autumn migration are discussed in section 4.6.2). The data available at present reveal that the age at which postjuvenile moult starts is linearly related to the time available, expressed as the difference between the mean fledging date and either the peak autumn migration date in Central Europe, or the onset of winter (Fig. 33a). Hence, species with a long period available for moult start postjuvenile moult at a later age. The duration of the postjuvenile moult is also related to the time available, but levels off at about 56 days (Fig. 33b). Thus, none of the species included in Fig. 33 extends the postjuvenile moult over more than 45-70 days, even though more time may be available. Hence, the age at which postjuvenile moult is completed is never more than about 90-130 days (Fig. 33a). The slopes of the relationships presented in Fig. 33 show that as more time is available for moulting, the onset of moult is only slightly postponed

Fig. 33. Onset and completion (a) and duration (b) of partial postjuvenile moult of European passerine species during summer/autumn in relation to the time potentially available for moult, i.e. the time between fledging and peak autumn migration or the onset of winter. Data on the onset and end (relative to the bird's age in days since hatching) and duration of postjuvenile moult are taken from the summaries in Ginn & Melville (1983), Glutz & Bauer (1985, 1988, 1991) and Rymkevich (1990, data from early-hatched birds) as well as from Norman (1981, 1990a, 1991a), Boddy (1983), Rymkevich (1983), Dittberner & Dittberner (1987, 1989), lovchenko & Smirnov (1987) and Rymkevich &c Pravosudova (1987). The time available for moulting between fledging and peak autumn migration is the peak date of autumn migration (Jenni 1984; accurate to within five days) minus the peak date of laying (accurate to within five days), minus the mean incubation and nestling period (Glutz & Bauer 1985, 1988, 1991 and other references; if possible using data from the area where the moult data were collected). For mainly sedentary species, the peak autumn migration was repkced by the onset of winter, arbitrarily set at the beginning of November. Since data on laying as well as on postjuvenile moult are often not very accurate, the graphs presented can only give a general picture. Linear regressions indicated are: onset of postjuv moult: y=0.330x+278, r=0.81, N=44, P P.cae = P. caeruleus, P.maj = P.major, P.mod = Prunella modularis, P.col = Phylloscopus collybita, P.och = Phoenicurus ochruros, P.pho = P. phoenicurus, P.tro = Phylloscopus trochilus, R.rip = Riparia riparia> S.atr = Sylvia atricapilla, S.bor = S. borin, S.cit = Serinus citrinella, S.com = Sylvia communis, S.cur = S. curruca, S.eur = Sitta europaea, S.rub = Saxicola rubertra, S.ser = Serinus serinus, T.ili = Turdus iliacus, T.mer = T. merula, T.phi = T. philomelos, T.pil = T. pilaris, T.tor = T, torquatus alpestris, T.tro = Troglodytes troglodytes, T.vis = Turdus viscivorus.

40

The Moult During the First Year of Life

marginal and median coverts (e.g. Sylvia nisoria, Rymkevich 1990; Lanius collurio, Oriolus oriolus, many Emberiza hortuiana and Anthus campestris, see part II; possibly some other species, e.g. certain Acrocephalus and Hippolais warblers, in which juvenile and postjuvenile body-feathers are difficult to distinguish). In other long-distance migrants, all the body-feathers, marginal and median coverts are regularly moulted, but none (e.g. most Oenanthe oenanthe) or only the innermost one or two greater coverts (e.g. most Phoenicurus phoenicurus, Sylvia borin, Muscicapa striata, Ficedula hypoleuca). Only a few long-distance migrant species regularly moult all the body-feathers, several of the greater coverts and occasionally some tertials or rectrices (e.g. Motacilla flava, Luscinia megarhynchos, Sylvia curruca, cf. Fig. 30). Short-distance migrants and sedentary species with a longer potential moult period (more than 100 days in Fig. 34) usually moult all the body-feathers, the marginal and median coverts as well as part of the greater coverts. In some species, some tertials are also regularly moulted (e.g. Sylvia atricapilla) as well as some rectrices (e.g. Phylloscopus collybita, Emberiza schoenidus, Carduelis chloris, Coccothraustes coccothraustes). In Parus major, the partial postjuvenile moult regularly includes all the wing-coverts, tertials and rectrices. Some species occasionally also renew the innermost, rarely the two or three innermost, secondaries (e.g. Motacilla alba, Phoenicurus ochruros, Sylvia atricapilla, Phylloscopus collybita, Parus major, P, caeruleus, Fringilla coelebs, Carduelis carduelis, C. chloris, C. spinus, Serinus serinus, Emberiza schoenidus, see part II; Sylvia undata, S. melanothorax, S. rueppelli, Roselaar in Cramp 1992, Svensson 1992). Saxicola torquata and Cyanopica cyana occasionally renew up to five secondaries (Gwinner et al, 1983, Fracasso 1985, de la Cruz ^rf/. 1991, 1992). However, there are a number of species with a rather long potential postjuvenile moult period in summer/autumn which only moult a few feathers (Fig. 34). At present, it is difficult to explain why some species do not moult as many feathers as do others with a comparable potential moult period. Possible explanations may include the following factors. In some species, it may be less important for a first-year bird to completely attain the adult coloration. The retained juvenile feathers might be durable enough to last until the next moult, either because they face little exposure to wear or because durable feathers could be grown in the nest. However, in species with long moult periods (> 100 days in Fig. 34), there is no clear relationship between the extent of the postjuvenile moult and the occurrence of a prebreeding moult, unlike long-distance migrants (see section 4.6.1). In some species, a tight energy and nutrient regime might prevent an extensive postjuvenile moult. In this respect, it is noteworthy that the alpine and subalpine species Serinus citrinella, Carduelis flammea, Turdus torquatus and Anthus spinoletta moult fewer feathers than comparable species occurring mainly in lowlands. Their harsher montane environment may impose more stringent energy and nutrient stresses. The occasional renewal of the innermost one or few secondaries during a partial postjuvenile moult in summer/autumn is difficult to explain (in contrast to the first prebreeding moult, see section 4.6.3). It constitutes an extension of the renewal of the tertials, perhaps because S6 wears while it takes over the protective function of the longest tertial during its renewal.

4.4.3 Intraspecific variation in timing and extent Variation with hatching date Many studies on free-living birds confirm the experimental evidence (see section 4.4.1) that birds hatched late in the season start their postjuvenile moult at an earlier age, moult more rapidly and renew fewer feathers than early-hatched young of the same population. Late-hatched birds of species which generally start postjuvenile moult at an early age, can advance the onset of moult by only a few

days relative to early-hatched young (Berthold et al. 1970, Noskov & Rymkevich 1985, Rymkevich 1990). In species which generally start their postjuvenile moult at a later age, late-hatched birds can advance the moult by several weeks and may start it before the juvenile plumage is completed, as do long-distance migrants (e.g. Pyrrhula pyrrhula, Newton 1966; Fringilla coelebs, Dolnik & Gavrilov 1980; Carduelis flammea, Boddy 1983; Turdus iliacus, Chochlowa et al. 1983 after Glutz & Bauer 1988; Sylvia atricapilla, Berthold et al. 1970; Carduelis flammea, Boddy 1983; Parus major, Dolnik & Blyumental 1967, Dhondt 1973; see Rymkevich 1990 for further species). The duration of the postjuvenile moult of late-hatched birds is generally considerably shorter than in early-hatched birds (e.g. Emberiza schoenidus, Haukioja 1969, Bell 1970; Emberiza spp., Rymkevich 1983; Turdus iliacus, Chochlowa et aL 1983 after Glutz & Bauer 1988; Fringilla coelebs, Dolnik & Blyumental 1967, Dolnik & Gavrilov 1980; Sylvia borin and S. atricapilla, Berthold et al. 1970; see also Rymkevich 1990 for many species), but not in Sylvia nisoria with its already very short moult duration (Rymkevich 1990). Early-hatched Motacilla alba, for instance, moult within 60 days, late-hatched within 42 days (Baggott 1970), early-hatched Sylvia communis within 45 days, late-hatched within 36 days (Stolbowa & Musaew 1987 after Glutz & Bauer 1991). By starting the postjuvenile moult earlier and by moulting more rapidly, late-hatched birds catch up with early-hatched birds either partly (e.g. Emberiza schoenidus, Bell 1970, Parus major, Dhondt 1973; Sylvia borin and S. atricapilla, Berthold et al. 1970) or completely (e.g. Turdus iliacus, Chochlowa et al. 1983 after Glutz & Bauer 1988; Fringilla coelebs, Dolnik & Gavrilov 1980). The shorter moult duration of late-hatched birds is achieved first by moulting more feather tracts and feathers simultaneously (e.g. Dolnik & Blyumental 1967, Berthold et al. 1970, Chochlowa et al. 1983 after Glutz & Bauer 1988). Second, the extent of the moult is reduced, something most easily observed in species which normally include some feathers of the wing in the moult (see part II). A reduction in the extent of greater covert moult of late-hatched birds was observed in Pyrrhula pyrrhula (Newton 1966), Sylvia atricapilla (Norman 1990a, Rymkevich 1990), S. communis (Norman 1990a), Erithacus rubecula> Sylvia curruca, Ficedula hypoleuca, Fringilla montifringilla (Rymkevich 1990), in the extent of alula and greater covert moult in Parus major (Dhondt 1973, Reith & Schmidt 1987, Rymkevich 1990, Gosler 1991) and in the extent of greater covert, tertial and rectrix moult in Motacilla alba (Baggott 1970, Rymkevich 1990), M. flava (Dittberner & Dittberner 1987), Troglodytes troglodytes (Hawthorn 1974), Parus caeruleus, Carduelis chloris, C. spinus and C. carduelis (Rymkevich 1990). Early-hatched Sylvia melanocephala perform a complete postjuvenile moult, or include many remiges, while late-hatched young do not generally moult any remiges (Gauci & Sultana 1979). On the other hand, Rymkevich (1990) reports a similar extent of moult in early- and late-hatched Sylvia borin and S. communis which generally perform a postjuvenile moult of limited extent within a short period of time. There are very few studies investigating the effect of variation in timing and extent of postjuvenile moult within a population on future events in the life of the bird and on its survival. For instance, Dhondt (1973) observed that late-hatched Parus major are still moulting when autumnal territorial behaviour begins and are more likely to emigrate. He suggests that the disadvantages of moulting might cause subordination during autumnal contests in this species. Furthermore, one might suppose that an extensive postjuvenile moult may be advantageous in two ways. First, in many species it reduces the juvenile plumage characters, which might be of behavioural advantage in contests and, in species without an extensive prebreeding moult, in the first reproductive season. Second, an extensive renewal of the juvenile feathers might significantly increase the structural quality of the plumage.

Partialpostjuvenile moult in the breeding area

Differences between populations Apart from the experimental evidence on Sylvia borin, S, atricapilla and Phylloscopus trochilus (see section 4.4.1), there is very little data on the timing and duration of postjuvenile moult in free-living populations in different geographical areas. It has been shown for Fringilla coelebs that the postjuvenile moult is shorter in northern than in southern populations (Dolnik & Blyumental 1967, Gavrilov 1979 after Ginn & Melville 1983). Russian Phylloscopus collybita commence postjuvenile moult at 25—45 days old and do not renew the second set of juvenile feathers (Rymkevich 1990), In contrast, British birds are 53 and captive German birds 68 days old at the start of moult and both renew the second set of juvenile feathers (Gwinner 1969, Norman 1991a). On the other hand, many species are known to differ in the extent of postjuvenile moult between different populations. Northern populations usually perform a less extensive moult than more southern ones, while populations occurring at similar latitudes do not differ in the extent of the moult (see Motacilla flava, M. alba, Erithacus rubecula, Turdus merula, Phylloscopus trochilus, Pants major, P. ater, Carduelis Moris, C. carduelis, C. spinus, Serinus citrinella, Emberiza schoeniclus in part II). While this seems to be an adaptation to the shorter time available between fledging and migration in northern populations, it is unclear to what extent the photoperiodic and innate genetic differences (cf. section 4.4.1) contribute respectively to the lesser extent of postjuvenile moult of northern populations. Despite their living at different latitudes, the extent of the postjuvenile moult of Turdus torquatus torquatus and T. t. alpestris does not differ (p. 107), possibly due to their similar timing of breeding and, consequently, similar period available for moult in both the Alpine and the Scandinavian subspecies. Some observed differences in the extent of postjuvenile moult between populations do not fit the general latitudinal pattern. For example, why should Parus major and P. caeruleus moult fewer feathers on the wing in Great Britain than in central Europe? Furthermore, slight differences in the extent of moult of Parus major have been found between nearby sites (Reith & Schmidt 1987): urban birds renewed more alula feathers and greater coverts than rural birds, and birds hatched at higher altitudes moulted alula feathers less frequently than lowland birds. This can be explained, at least partly, by differences in hatching dates between the sites.

Differences between the sexes and effects of energetic and nutrient stress $ apparently perform a more extensive postjuvenile moult in summer/autumn than ? in at least 13 species of European passerines. In d of species with a generally restricted moult, more juvenile greater coverts are lost (Phoenicurusphoenicurus, Ficedula hypoleuca\ see part II, Karlsson et ai 1986b). 3 of species with a moderate extent of moult replace more greater coverts, tertials, rectrices, carpal covert or alula feathers (Motacilla alba, Sylvia atricapilla, Parus major, Carduelis carduelis, C. spinus, C. cannabina, Serinus citrinella, Pyrrhula pyrrhula, Loxia curvirostra, Emberiza schoeniclus', see part II, Newton 1966, Baggott 1970, Dhondt 1973, Pettersson 1981, Herremans 1982, 1991, but not 1988b, Reith & Schmidt 1987, Gosler 1991). In cJ Cyanopica cyana, whose moult is extensive, more juvenile primaries, secondaries, primary coverts and rectrices are lost (de la Cruz et al. 1991). Of the species examined (see part II), the mean difference between cT and $ is 0.3—1.0 greater coverts, 0.2—0.6 tertials and 0.16—0.5 rectrices. Haukioja(1969) suggests that the postjuvenile moult of 2 Emberiza schoeniclus takes about 10 days less than in d. In three other species, 6 moult slightly more feathers on the wing than $, but the difference is not significant (Parus caeruleus, Carduelis Moris, Serinus serinus; part II, Frelin 1977). In six species, there is no difference in the extent of moult between the

41

sexes (Erithacus rubecula, Turdus torquatus alpestris, T. merula, T. pilaris, Fringilla coelebs, F. montifringilla; part II, Baillie & Swann 1980, Karlsson et al. 1986a). No European passerine is known in which $ perform a more extensive postjuvenile moult than d. Why there should be a difference in the extent of postjuvenile moult between the sexes is largely unexplained. It appears that a sexual difference in moult extent occurs in species in which the greater coverts and tertials are generally more brightly or conspicuously coloured. Thus, d with important signal coloration on the greater coverts and tertials might gain more by investing in postjuvenile moult than $?, in order to achieve a more adult-like plumage for contests during winter and the following breeding season. However, this explanation can hardly apply to Ficedula hypoleuca and Phoenicurusphoenicurus, in which the average difference between cT and ? is very small and concerns inconspicuously coloured greater coverts (see part II). Furthermore, d performing a prebreeding moult of greater extent than the postjuvenile moult do not gain any obvious advantage for the next breeding season from moulting more feathers during the postjuvenile moult. Gosler (1991) showed that the extent of greater covert moult in ? Parus major is correlated with a measure of their protein reserves in the following winter. He concluded that the extent of greater covert moult is determined directly by protein stress, or by a correlate of protein stress, during moult which also has an effect on protein reserves during winter. One can argue that such a correlate might be hatching date resulting in a shortened postjuvenile moult due to short daylength; protein stress due to the subdominant status of late-hatched birds may also play a role. If protein stress or energetic constraints during moult directly affect the extent of postjuvenile moult, this may also explain the differences in the extent of moult between years (as observed in Parus major and P. caeruleus•; Spencer & Mead 1978a, Reith & Schmidt 1987), between nearby populations (as observed in Parus major, Reith & Schmidt 1987), between habitats and between dominant and subdominant individuals and sexes. There is, however, no conclusive study so far known to us which clearly demonstrates energetic or nutrient effects on the extent of postjuvenile moult in passerines.

Intraspecific variation in the extent of postjuvenile moult during the course of the non-breeding season and between wintering sites The extent of completed postjuvenile moult often shows striking seasonal trends, especially during autumn migration (e.g. Herremans 1988a, 1991, graphs in part II). Such birds with different moult extents migrate through a study site at different times. There is generally a decrease in the extent of greater covert moult during autumn migration in all species examined (see part II, Hereward 1979, Karlsson et al. 1986a, Herremans 1988a, 1991, Pettersson et al. 1990), except in Phoenicurus phoenicurus, Phylloscopus trochilus and Ficedula hypoleuca, which always show a limited extent of greater covert moult. Judging by the discussion so far, this decrease may be due to the later migration of late-hatched birds or birds from northern origin, or of $, all of which usually perform a less extensive moult, or a combination of factors (cf, Herremans 1988a). Furthermore, in some species, differences in the extent of postjuvenile moult have been found between populations in different wintering sites (e.g. Swann & Baillie 1979, Pettersson et al. 1990). The question is whether these differences can be related to the various possible causes of reduced moult extent, thereby giving indications of the sexual, age-specific or geographical composition of migrant or wintering bird populations. Of the factors known to influence the extent of moult (hatching date, geographical area and sex), geographical area and sex can be excluded when investigating a sexable species of limited distribution. This is the case in the Alpine population of Serinus citrinella. This species shows a

42

The Moult During the First Year of Life

distinct decrease in the extent of greater covert moult (Fig. 512) at the Alpine pass of Col de Bretolet during the course of autumn migration, which takes place from mid-September to early November (Jenni 1984). The number of moulted greater coverts correlates in both sexes with a measure of postjuvenile developement, derived from the degree of skull pneumatization and assumed to be related to hatching date (Table 4), but not with body mass or wing-length. This suggests that hatching date is an important factor causing the variation in the extent of greater covert moult and that early-hatched birds migrate through the study site earlier than late-hatched birds. Birds examined on migration or in the wintering areas may also vary in the extent of their moult due to differences between geographical populations. This can be assessed when an independent characteristic indicating the geographical origin of migrants is available. In Sylvia atricapilla, wing-length (or alternatively feather-length of P 8, Jenni & Winkler 1989) increases from southwest to northeast (Klein etai 1973, own unpubl. data). The number of moulted greater coverts is again correlated with the measure of hatching date in both sexes, but also negatively with feather-length, suggesting that both hatching date and geographical origin partly explain the variation in moult extent (Table 4). According to the analysis of wing-length, northern populations of S. atricapilla move through central Europe during the early part of the autumn migration, before most of the central European populations have started (Klein etal. 1973, Turrian & Jenni 1989). Thus, in Sylvia atricapilla, the number of moulted greater coverts decreases during autumn migration (Fig. 356) despite the early passage of northern birds, which have slightly fewer greater coverts moulted. This indicates that hatching date is probably the predominant factor determining the decrease of greater covert moult during autumn migration. In the study

Table 4. Mean number of moulted greater coverts of S and 9 Serinus citrinella and Sylvia atricapilla for selected dates at which skull pneumatization score 2 (see appendix) was reached (date of skull 2) estimated from a generalized linear model analysis. From retraps, the intervals between successive skull pneumatization scores was estimated to be 23 and 16 days for the two species, respectively. Using this information, the date at which skull pneumatization score 2 was attained was estimated for each individual and was assumed to be a measure of hatching date. For Serinus citrinella, analysis by generalized linear models revealed that the number of moulted greater coverts is significantly dependent on the date of skull 2 (t = -11.79) and sex (t = -7.08), but not on body mass or length of P 8 (as a measure of wing-length, Jenni & Winkler 1989) (N= 695). For Sylvia atricapilla, the number of moulted greater coverts was significantly dependent on the date of skull 2 (t = -10.85), sex (t = -3.54) and feather-length of P 8 (t = -2,44) (N=658). Serinus citrinella Date of skull 2

21 May 20 June 20 July 19 August 8 September

Mean number of moulted greater coverts

8

9

9.2 8.6 7.8 6.8 5.5

8.8 8.1 7.1 5.8 4.4

Sylvia atricapilla Date of skull 2

Mean number of moulted greater coverts Feather-length 50 mm

16 June 9 July 3 August 27 August 10 September

9.9 9.8 9.5 9.1 8.2

9.8 9.7 9.3 8.7 7.7

Feather-length 60 mm

9.8 9.5 9.0 8.2 6.8

9.7 9.3 8.7 7.6 6.0

by Herremans (1991), a similar relationship between the extent of greater covert moult and wing-length was found in $ S. atricapilla, although the relationship with skull pneumatization was only very weak (skull data not transformed to give an indication of hatching date). In Troglodytes troglodytes, the number of greater coverts moulted also decreases during the autumn migration (Fig. 155), while wing-length increases. Furthermore, there is a significant relationship between the number of greater coverts moulted and wing-length, northern populations having on average slightly longer wings than southern populations. This suggests that northern populations migrate through Switzerland later and have fewer greater coverts moulted than do southern populations (Jenni & Winkler 1983). In the absence of any information on hatching date, it remains unclear whether geographical origin genuinely influences the extent of greater covert moult or whether it is an artefact of the effect of hatching date due to the later breeding of northern populations. A reduced extent of postjuvenile moult in late-hatched compared to early-hatched British Wrens was suggested by Hawthorn (1971, 1974). Even when skull pneumatization can be used to assess the hatching date and wing-length to suggest the geographical origin, it remains difficult to judge the relative importance of hatching date and geographical origin as factors causing a change in moult extent. This is because skull pneumatization and wing-length are only approximate indicators of hatching date and geographical origin and are probably accurate to different degrees. From the available evidence, however, it is probable that late-hatched birds migrate later in autumn than earlyhatched birds, both within and between populations. Of the species wintering in Switzerland, the number of postjuvenile greater coverts is lower during the beginning of winter at least compared to the end of autumn migration in Troglodytes troglodytes (see Jenni & Winkler 1983), Erithacus rubecula, Turdus pilaris, Fringilla coelehs, F. montifringilla, Carduelis Moris and C. spinus (see part II). It is lower than the average during autumn migration in Turdus merula and Pyrrhulapyrrhula (see part II). This suggests that birds wintering in Switzerland are mainly of northern origin and/or late-hatched. In these species, the number of postjuvenile greater coverts increases during the spring, probably due to the return of early-hatched birds, or of those from central European populations. In birds caught during spring migration, no seasonal trend of the extent of postjuvenile moult has been observed (Erithacus rubecula, Karlsson et ai 1986a, p. 91; Sylvia atricapilla, p. 130). Thus, there seems to be no sequential passage of second-year birds correlated to hatching date as in the autumn. Furthermore, the extent of postjuvenile moult during spring is approximately equal to the mean over the entire autumn migration in all species examined (Troglodytes troglodytes, Erithacus rubecula, Phoenicurus ochruros, Turdus merula, T, pilari$> T. philomelos, Sylvia atricapilla, Fringilla coelebs, Carduelis chloris, C, spinus, Pyrrhula pyrrhula; see part II). In wintering populations, the extent of postjuvenile moult may differ between sites. In Turdus merula wintering in Scotland, birds in rural roosts showed a less extensive postjuvenile moult than those in urban roosts. This was attributed to there being more continental immigrants in rural roosts (Swann & Baillie 1979). Wintering Parus ater from England also showed differences in the extent of moult, probably due to a lesser extent of moult in northern and eastern populations (Christmas etal. 1989). Pettersson etal. (1990) showed that Erithacus rubecula wintering in the E Mediterranean had fewer greater coverts moulted than birds wintering in the W Mediterranean. Differences between years in the extent of postjuvenile moult have been observed in some migrants (Motacilla flava, Hereward 1979; Parus major, Pettersson 1981; Parus ater, p. 146). This may be due to different hatching dates between years, other differences between years acting on the same population or a variation in geographical origin of the migrants. Annual variations and differences between nearby sites (sometimes due to differences in the timing of breeding) further

Partialpostjuvenile moult in the breeding area

complicate the use of moult extent in determining the origin of birds (see e.g. Pettersson 1981, Reith & Schmidt 1987).

4.4.4 Partialpostjuvenile primary moult in the breeding area Partial renewal of the primaries during the postjuvenile moult in the breeding area (moult cycles 3 and 4 in Table 3) has only recently been studied in European passerines. Its occurrence and significance, therefore, are still poorly known. In all species in which a postjuvenile moult includes part of the primaries, most, or at least a significant proportion of individuals perform a partial postjuvenile moult which does not include any primaries (moult cycles 5 and 6 in Table 3), or a complete postjuvenile moult (moult cycle 1 in Table 3). In Sylvia melanocephala, S. melanothorax, S. conspicillata, Carduelis chloris, Loxia curvirostra and possibly others, the entire range of possibilities from a partial postjuvenile moult of relatively limited extent to a complete postjuvenile moult is displayed. Thus, the partial renewal of the primaries during the postjuvenile moult can be regarded as a transition or compromise between a less extensive partial and a complete postjuvenile moult. However, the usually eccentric renewal of primaries (see section 4.3.3) is a distinct characteristic of this moult strategy. In certain species which usually moult most or all of the greater coverts, tertials and rectrices, some individuals have been found to also renew some primaries. Partial primary moult has been recorded in 68% of Carduelis carduelis and in three out of seven C cannabina on the Balearic Islands (Mester & Priinte 1982), in 57% of those Sylvia melanocephala not performing a complete postjuvenile moult in Malta (Gauci & Sultana 1979), in 13% of Cyanopica cyana in S Spain (de la Cruz et al. 1992), in 10-12% of Carduelis chloris in central Europe (Westphal 1976, p. 168), in 8% of Loxia curvirostra, 4% of Carduelis cannabina and 1.1% of C. spinus in Switzerland (mostly migrants, see part II), in 3% of Sylvia atricapilla on spring migration in Italy (p. 130) and exceptionally in Motacilta alba, Carduelis carduelis, Serinus serinus and Emberiza schoenidus in Switzerland (mostly migrants, see part II). Furthermore, partial primary moult was observed in Saxicola torquata (Italy, Fracasso 1985), Sylvia conspicillata^ S. melanothorax (Roselaar in Cramp 1992, Svensson 1992), Remiz pendulinus (Greece, Kasparek 1981), Lanius excubitor (southern populations, Roselaar in Cramp & Perrins 1993), Emberiza cirlus (Italy, Winkler & Jenni 1987), probably in Certhia brachydactyla (Spain, Copete & Senar 1990) and possibly in Sylvia undata and other sedentary Mediterranean Sylvia warblers (Gauci & Sultana 1979). Most of these birds moult the primaries eccentrically, but Saxicola torquata, Carduelis spinus and C. carduelis occasionally show a descendant partial primary moult starting with PI, and Cyanopica cyana and Certhia brachydactyla an ascendant partial primary moult (see section 4.3.3). As might be expected, eccentric primary moult occurs in birds with a long potential period available for moult. They usually start breeding early in the season in central Europe (especially Carduelis spinus and Loxia curvirostra) or belong to Mediterranean populations with an earlier breeding season than in central or N Europe (see e.g. p. 171 for geographical variation in the occurrence of eccentric moult in Carduelis carduelis). In Sylvia melanocephala^ early-hatched birds perform a complete, or almost complete, postjuvenile moult, while late-hatched birds show a partial postjuvenile moult which includes a few primaries at most (Gauci & Sultana 1979). In Carduelis chloris; only the earlyhatched birds moult part of the primaries, and they prolong the usual moult duration of 50-70 days up to 95-100 days (Rymkevich 1990). Thus, partial primary moult depends on hatching date and, consequently, geographical area. It entails a longer moult duration than a partial moult without the primaries, as also shown for Carpodacus mexicanus (Michener & Michener 1940). The available data show that partial primary moult occurs in a wide variety of species. Further study, especially in the Mediterranean area,

43

will no doubt reveal it in more species and perhaps shed light on its advantages. A partial primary moult starting with P 1 and ending somewhere in the centre of the wing can be interpreted as an attempt to perform a complete primary moult which is, however, interrupted before completion. In this case, only the protected innermost primaries are renewed, and this type of partial primary moult occurs rarely. Eccentric partial primary moult differs in that a moult focus in the centre of the wing is activated only once in the life of the bird and presents a compromise between a complete primary moult and no renewal of the primaries. It renews the most exposed primaries, or those that partly cover the most exposed ones, so maximizing the effectiveness of the moult for minimum effort. An early hatching date and consequently long potential moult period are prerequisites for a partial primary moult. Furthermore, early-hatched birds would have worn the juvenile primaries for longer than late-hatched ones and, therefore, may need this reinforcement of the wing-tip before the first complete postbreeding moult. However, in some sedentary species with early hatching dates and long potential moult periods, a partial primary moult has not yet been observed (e.g. Parus spp., Sitta europaea, Troglodytes troglodytes, Cinclus cinclus, Erithacus rubecula, Turdus spp.). Reinforcement of the wing-tip may be important in birds which skulk in dense vegetation, as do some Mediterranean warblers, or which are exposed to bright sunlight (see Mester & Priinte 1982 for Carduelis spp.). Partial primary moult not only reinforces the juvenile wing, but may also change its shape and surface area (and consequently alter the wing-loading) since the newly grown primaries are usually longer than their juvenile predecessors (as shown for Vireo griseus, George 1973, and Icteria virens, Phillips 1974). Whether this results in significant benefits is not yet known. Partial primary moult has also been observed by Evans (1986) in Sturnus vulgaris, which usually perform a complete postjuvenile moult. According to this author, 45-5% of first-year birds moulted some of the primaries, some eccentrically, some apparently descendantly, but it is unclear whether this proportion also includes birds with an interrupted secondary moult (see also Williamson 1961 and Scott 1965). Most examples of partial primary moult given by Evans (1986) are from July and August, so may refer to birds which suspended moult temporarily. Partial primary moult was more frequent in island than in continental populations, but the island populations studied were also more northern in location. Furthermore, a partial primary moult was more frequent in juveniles and 2, which also moulted later, than in adults and c?. Therefore, the partial primary moult of Sturnus vulgaris could be due either to a suspension of moult for summer migration (Zwischenzug) or to the time constraints on northern and late moulting birds. Whether protein constraints, as suggested by Evans (1986), might be an additional factor remains to be shown. Interrupted primary moult (between one and seven primaries renewed) was also observed in some Passer domesticus (Harper 1984), though whether this represents a partial, a very slow or a temporarily suspended moult remains to be shown. A particular case is that of Cisticola juncidis (Fig. 35). While a complete postjuvenile moult is the normal moult strategy (Thomas 1979, Gauci & Sultana 1981), many cT from early broods may interrupt the moult for breeding activities (which can occur within a few weeks of fledging in this species; Ueda 1985). These birds renew up to six primaries descendantly, starting with P 1, as well as part of the body-feathers and wing-coverts. After interruption, the primary moult is resumed descendantly, usually from the point of interruption, but probably all of these birds moult the recently renewed inner primaries again soon after this resumption in an ascendant sequence, as well as all the other feathers (Gauci & Sultana 1981, C. Gauci in litt). Another peculiarity of the complete moult of this species is that at least 74% of juveniles and adults renew the central tertial (and occasionally others) and at least 55% renew up to five greater coverts (occasionally all) a second time at the end of the primary moult (Gauci & Sultana 1981).

44

The Moult During the First Year of Life

Fig. 35. Cisticolajuncidis ly, 23 November. P 1+3-4 postjuv, 2+5-10 juv (on the other wing P 5 growing). S 1 almost full-grown postjuv, 2-5 juv, 6 in pin and hidden. T, GC, CC, MaC and MeC postjuv. Al 1 postjuv, 2-3 growing postjuv. PC juv. P 1 is slightly older than primaries 3-4. Thus, this bird probably interrupted primary moult after renewing P 1, resumed it later with P 3 and is now moulting slowly and including the secondaries in a convergent sequence, typical of this species (cf. section 3.2.3).

4.5 Complete postjuvenile moult in the breeding area Only a few European passerines regularly perform a complete postjuvenile moult in the late summer/autumn (species see Table 5, p. 53). In Sylvia melanocephala part of the population do so (Gauci & Sultana 1979, Vowles & Vowles 1987) and in Loxia curvirostra and Carduelis Moris a few individuals do (see part II). Complete postjuvenile moult may perhaps be observed in some individuals of other species (see e.g. Carduelis spinus and C. carduelis in part II; Sylvia conspicillata, Gauci & Sultana 1981; S. melanothorax, Roselaar in Cramp 1992). These complete postjuvenile moults follow the basic sequence described in chapter 3.2. Whether or not the complete moult of some first-year long-distance migrants, during winter in Africa, actually represents a true complete postjuvenile moult is discussed in section 4.6.2. Here, only species performing their complete postjuvenile moult in late summer/ autumn, in the breeding area, are discussed (moult cycles 1 and 2 in Table 3). Most species start the complete postjuvenile moult when one to two months old (Alaudidae, Glutz & Bauer 1985; Panurus biarmicus, Pearson 1975B; Passer spp., Heinroth & Heinroth 1926, Deckert 1962; Montfringilla nivalis, Winkler &: Winkler 1985; Sturnus vulgaris, Lundberg & Eriksson 1984, Rymkevich 1990). Cisticola juncidis starts this moult later at two to three months of age (Gauci & Sultana 1981) and Acrocephalus melanopogon later still, from two to four and a half months of age (Leisler 1972). Thus, the juvenile primaries are usually only worn for several weeks after fledging. For instance in Montifringilla nivalis, P 1 is shed as soon as 12 days after the juvenile P 9 is full-grown (Winkler & Winkler 1985). Compared with the postbreeding moult of adults, the postjuvenile moult is less synchronized within a population, especially in multibrooded species (e.g. Bahrmann 1964, Alonso 1984, Peris 1988), and may start slightly later than in adults (e.g. Miliaria calandra, Gauci & Sultana 1983; Passer montanus, Aegithalos caudatus, Ginn & Melville 1983; Sturnus vulgaris, Bahrmann 1964; S. unicolor, Peris 1988). The complete postjuvenile moult is usually roughly similar in duration to that of the postbreeding moult of adults, or slighdy longer, especially in early-hatched young (Passer montanus Deckert 1962; P. domesticus, Zeidler 1966; 2 Cisticola juncidis, Gauci & Sultana 1981, $ see p, 25), although it is shorter in Sylvia melanocephala (Gauci &

Sultana 1979). It usually lasts from about 60—80 days (Alauda arvensis, Ginn & Melville 1983, Glutz & Bauer 1985; Eremophila alpestris, Glutz & Bauer 1985; Cisticola juncidis, Sylvia melanocephala, Gauci &: Sultana 1979, 1981; Miliaria calandra, Gauci & Sultana 1983, Ginn & Melville 1983; Passer domesticus, Zeidler 1966, Ginn & Melville 1983, Alonso 1984; P. montanus, Ginn & Melville 1983; P. hispaniolensis, Alonso 1984). A shorter duration is reported only in Panurus biarmicus (42-56 days depending on hatching date, Spitzer 1972, Buker et ai 1975, Pearson 1975b), Acrocephalus melanopogon (47.5 days, Leisler 1972) and late-hatched (50 days, Zeidler 1966) and Finnish (52-64 days, Dyer et al. 1977) Passer domesticus. Sturnus vulgaris takes about 80 days for its primary moult (Ginn & Melville 1983), but the total moult duration of captive birds is reported to range from 87-130 days (Lundberg & Eriksson 1984, Rymkevich 1990). Aegitalos caudatus takes 75-100 days to moult (Ginn & Melville 1983), while Montifringilla nivalis has an exceptionally long moult duration of 108 days (93 days for the primaries), at least in captivity, which sets the end of postjuvenile moult as late as midNovember to early December (Winkler & Winkler 1985). This might be an adaptation to its harsh mountain environment. As with the partial postjuvenile moult, late-hatched birds perform a complete postjuvenile moult at an earlier age and more rapidly than early-hatched birds (Acrocephalus melanopogon, Leisler 1972; Panurus biarmicus, Pearson 1975b; Cisticola juncidis, Gauci & Sultana 1981; Passer domesticus, Zeidler 1966; P. montanus, Deckert 1962, Myrcha & Pinowski 1970, Sutter 1985; P. hispaniolensis, Alonso 1984; Sturnus vulgaris, Rymkevich 1990). Similarly, Finnish Passer domesticus moult more quickly than more southerly populations (Zeidler 1966, Dyer et ai 1977, Ginn & Melville 1983, Alonso 1984). In Sturnus vulgaris held under constant daylength, birds from a sedentary N Norwegian population started to moult at a similar age, but more rapidly than birds from a migratory Swedish population, suggesting an endogenous difference in moult duration (Lundberg & Eriksson 1984). Why some species should perform a complete postjuvenile moult has hardly been studied and remains obscure. The presence of a complete posrjuvenile moult appears to stem from phylogenetic relationships. It is remarkable that all European members of the Alaudidae, Passeridae, Sturnidae, Timaliidae, Aegithalidae and the genus Cisticola perform a complete postjuvenile moult: they are all predominantly tropical taxa whose rropical representatives also normally perform a complete postjuvenile moult. A long period available for moult is certainly a prerequisite for this particular habit. However, the age at which this moult commences (30-60 days) is basically similar to that of short-distance migrants and sedentary species which perform a partial postjuvenile moult, and its duration (60-80 days) is only about two weeks longer (cf. Fig. 33). Another prerequisite supposes a rich and easily accessible food supply which provides sufficient energy and nutrients and can be utilized despite a reduction of flight capability during the first months of independence. However, it remains unclear whether or not species performing a complete postjuvenile moult do in fact generally have access to richer and more accessible food resources than species performing a partial moult. In P. biarmicus and A. melanopogon, limited flight requirements in their dense marshy habitat may facilitate their particularly rapid complete postjuvenile moult (Pearson 1975b). CJ. Mead (in Ginn & Melville 1983) pointed out that the nestlings of many species moulting the juvenile plumage completely are reared on vegetable food and proposed that such a diet might be deficient in certain amino acids, causing the juvenile feathers to be less durable. However, the majority of species with a complete postjuvenile moult actually feed their nestlings largely on arthropods. Furthermore, it has still not been shown whether the juvenile remiges are really less durable than those of adults and whether the remiges of species with a complete postjuvenile moult are less durable than those of species moulting partially (cf. Fogden 1972 for tropical species).

Moults during the first year of Life in trans-saharan migrants

4.6 Moults during the first year of life in trans-saharan migrants The moult cycles during the first year of life of all sedentary species and short-distance migrants (all of which moult as adults according to moult strategy 1 or 2, see Tables 1 and 2) are rather uniform and well known (moult cycles 1-6 in Table 3, sections 4.4 and 4.5). Long-distance migrants, whose adults perform a complete postbreeding moult before autumn migration, also follow moult cycle 5 or 6 (Table 3). As with the moult strategies of adults, the moult cycles during the first year of life of long-distance migrants, whose adults perform a complete moult in the tropics or a seasonally divided moult, are more complex and also far less well known than those of sedentary species and short-distance migrants. The information on their moult cycles is scarce since birds in their first year of life have only rarely been distinguished from adults and many published reports cannot be assigned to age groups and have had to be omitted from both the account of the moult strategies of adults and this section. As with the adults, we therefore concentrate on those European passerines wintering in Africa in the remainder of this section.

4.6.1 Partialpostjuvenile and partial prebreeding moult excluding remiges: Moult cycles 5 and 6 All species which perform a partial prebreeding moult as adults after the complete postbreeding moult in the breeding area (moult strategy 2 in Table 1) also show a partial prebreeding moult which follows the partial postjuvenile moult (moult cycle 6 in Table 3). Until now, there has been little information as to whether the timing, duration and extent of the prebreeding moult during the first winter is similar to that of adults. In all species examined, we found that second-year birds perform a more extensive prebreeding moult than adults (Anthus trivialis, A. spinoletta, Motacilla flava, M. alba, Saxicola rubetra, Sylvia c. communis, Emberiza hortulana, details see part II), similar to the case in two American species (Cannell et al. 1983, Greenwood et al. 1983). Ficedula hypoleuca moults similar numbers of greater coverts and tertials during the first winter, but significantly more secondaries than during later winters (p. 142). Long-distance migrants without a prebreeding moult, or with one of very limited extent, perform a postjuvenile moult which includes at least all the body-feathers, marginal and median coverts, usually part of the greater coverts and occasionally some tertials (e.g. Luscinia spp., Phoenicurus phoenicurus, Oenanthe pleschanka, O. hispanica). In contrast, long-distance migrants with a considerable prebreeding moult perform a postjuvenile moult in the breeding area of limited extent, during which part of the juvenile body-feathers may even be retained (see p. 39). In these species, the rest of the juvenile bodyfeathers and at least part of the juvenile wing-coverts are replaced during the partial prebreeding moult during the first winter, which is considerably more extensive than the postjuvenile moult (e.g. Ficedula hypoleuca, Anthus trivialis). Thus, the partial prebreeding moult during the first winter might have the additional purpose of making up a postjuvenile moult of very limited extent before autumn migration and, therefore, might be more extensive than in the adults.

4.6.2 Complete moult in the non-breeding area: Moult cycles 9, 10, 11, 14 and 15 Species which show a complete moult in Africa as adults (moult strategy 3, 5 and 6 in Tables 1 and 2) usually also moult completely during their first winter (moult cycles 9, 14 and 15 in Table 3). So far, the only known exceptions are Lanius s. senator, some L L isabellinus and L. minor and possibly, rarely L, collurio, as well as those species in

45

which some of the adults also do not moult completely (see section 4.6.3). Furthermore, two species (Carpodacus erythrinus, Emberiza a. aureola) wintering in tropical Asia apparently perform only a partial moult during their first winter, but show a complete moult as adults (Stresemann & Stresemann 1969a, Bozhko 1980; moult cycle 10 and 11 in Table 3). In C. erythrinus, this partial moult includes the bodyfeathers, wing-coverts, tertials, rectrices and occasionally the outer primaries, and is only rarely complete (Bozhko 1980), In both C. erythrinus and E. a. aureola, the adults and first-year birds do not moult before the onset of autumn migration, while E. a. ornata does so (Stresemann & Stresemann 1969a, Rymkevich 1983, 1990, Strom 1991). Thus, in C erythrinus and E. a. aureola the postbreeding and postjuvenile moult seem to be delayed until the arrival in the first staging area in late autumn. Additionally, E. aureola shows a partial prebreeding moult restricted to the feathers of the head, before spring migration (moult cycle 11 in Table 3). The moult preceding the autumn migration of first-year birds which perform a complete moult in Africa, is of limited extent or may be completely suppressed (e.g. Acrocephalus palustris according to Dowsett-Lemaire 1981; A. schoenobaenus, Griill & Zwicker 1982; A. paludicola, Roselaar in Cramp 1992; Locustella luscinioides, Miiller 1981). It usually includes only apart of the body-feathers, marginal and median coverts and, rarely, some greater coverts, at most all of the bodyfeathers, marginal and median coverts as well as the innermost greater coverts (e.g. Sylvia borin). As in adults, it is not clear whether this very restricted premigratory moult is actually a postjuvenile moult in its own right, or the beginning of the complete postjuvenile moult which takes place mainly in Africa. First-year Hirundo rustica, Delichon urbica and Riparia riparia perform the complete moult only in Africa, but may start to renew the juvenile body-feathers (and rarely some primaries, not shown in Table 3) before autumn migration. The current interpretation is that the complete moult of first-year birds in Africa is actually the postjuvenile moult which may already start before autumn migration (Glutz & Bauer 1985, Roselaar in Cramp 1988, part II). This agrees with the view that the postbreeding moult of the adults is delayed until winter. An additional, prebreeding moult of both age classes was found in Delichon urbica (Broekhuysen 1953) and possibly Riparia riparia (Pearson 1971). The same interpretation may apply to some of the Acrocephalus, Locustella and Hippolais warblers and to Anthus campestris (cf. Roselaar in Cramp 1988, 1992) as well as to certain long-distance migrants wintering in tropical Asia (e.g. Carpodacus erythrinus, Emberiza a. aureola, E. bruniceps, E. melanocephala-, Stresemann & Stresemann 1969a, Rymkevich 1983, 1990) which either perform a very restricted moult or no moult at all before autumn migration. However, in other species moulting more extensively before the autumn migration, the feathers renewed in autumn are very probably renewed again during the complete moult in Africa (e.g. Sylvia borin) so that the premigratory moult may represent the actual postjuvenile moult. As discussed earlier (sections 3.3.3 and 3.4.2), adults which moult completely in winter may suspend the moult in Africa and show an additional partial moult before spring migration. Since first winter birds have rarely been recorded separately from adults, it remains to be seen whether the two age classes differ in their moult cycles while in Africa (see e.g. Pearson 1973 for a possible age related difference in Acrocephalus scirpaceus), In species whose adults show an extensive partial or complete moult in the breeding area, or in NE Africa, before a complete moult further south (moult strategy 5 or 6 in Tables 1 and 2), first-year birds perform either a postjuvenile moult of only limited extent, or hardly any moult at all, before autumn migration. Adult Acrocephalus palustris moult the body-feathers in NE Africa before the complete moult in southern Africa, while first-year birds do not show this body-feather moult, but renew their plumage for the first time in southern Africa (Pearson & Backhurst 1976, Pearson 1982, 1989). Similarly, adult Locustellafluviatilis moult the body-feathers and outer primaries in NE Africa during

46

The Moult During the First Year of Life

the autumn (moult strategy 5 in Table 2), while first-year birds do not (Pearson & Backhurst 1983; moult cycle 14 in Table 3). The postjuvenile moult of Sylvia borin, whose adults may occasionally moult almost completely in the breeding area (moult strategy 5 in Table 2), is always of limited extent. Phylloscopus trochilus, Lanius tigrinus and L. cristatus, which perform (almost) two complete moults annually as adults (moult strategy 6 in Table 1), moult only the body-feathers in their first summer/autumn (p. 136, Stresemann & Stresemann 1971; moult cycle 15 in Table 3). First-year birds of some species exhibit the complete moult in Africa later than the adults (e.g. Hirundo rustica in Zambia and Zaire, but not in southern Africa, Herroelen 1960, de Bont 1962, Broekhuysen & Brown 1963, Francis 1980; Acrocephalus schoenobaenus, Bensch et ai 1991; Sylvia borin and possibly A scirpaceus, Pearson 1973; Lanius collurw, B. Bruderer pers. comm.). This earlier moult of adults might be related to their earlier autumn migration, possibly helped by the absence of a complete premigratory postbreeding moult. However, both first-year and adult Phylloscopus trochilus moult at the same time (Pearson 1973), coinciding with a complete postbreeding moult in adults, and also show similar timing of the autumn migration.

4.6.3 Partial moult including remiges in the non-breeding area: Moult cycles 7> 8y 12 and 13

24% of Sylvia c. communis replace pan, rarely all, of the primaries during their first winter, usually eccentrically (moult cycle 8 in Table 3; probably similar in Lanius nubicus, Roselaar in Cramp & Perrins 1993). Thus, the replacement of the outer primaries especially, during the first winter could also be interpreted as a preparation for a possible primary moult interruption during the following postbreeding moult. In Sylvia c. communis, the peculiar primary moult pattern of some birds during the postbreeding moult (see p. 123) may actually be complementary to an eccentric first prebreeding moult (see section 3.3.3). However, an eccentric renewal of primaries in the first winter was also found in Sylvia nisoria (Hasselquist etaL 1988, Lindstrom et al. 1993), S. cantillans (Roselaar in Cramp 1992, own data; Fig. 31) and S. hortensis (Williamson 1968, Roselaar in Cramp 1992), all species in which primary moult suspension in adults is exceptional, or absent (except in captive S. hortensis). Therefore, many individuals of these Sylvia species probably renew some of the primaries twice during the first year of life, i.e. during an extensive first prebreeding and during the first postbreeding moult. As discussed in the preceding section on secondary moult, it remains unclear whether or not a partial primary moult during the winter is part of a continuation of the postjuvenile moult (cf. Roselaar in Cramp 1988, 1992, Cramp & Perrins 1993) and whether it is temporally separated from the actual prebreeding moult.

First partial prebreeding moult including secondaries

Incomplete first prebreeding moult

All species which may renew the secondaries during a partial prebreeding moult as adults (group 3 and 4 in Table 2) may also do so during their first winter (Sylvia nisoria^ Hasselquist et al. 1988, Lindstrom et al. 1993; S. cantillans, own data; S. communis, S. curruca, Ficedula hypoleuca see part II; S. hortensis, Williamson 1968, Berthold &: Querner 1982a; probably Lanius nubicus, Roselaar in Cramp & Perrins 1993; moult cycle 7 in Table 3). First winter Ficedula hypoleuca renew the secondaries even more frequently than do the adults (p. 142). As discussed in section 3.3.3, the first winter renewal of the secondaries in these species may be interpreted as a preparation for a possible retention of secondaries during the subsequent postbreeding moult. Additionally, it may also provide a more adult-like plumage, especially in cT F. hypoleuca. The renewal of the secondaries occurs during an extensive prebreeding moult which includes the body-feathers, most wing-coverts, the tertials and often the rectrices. As observed in actively moulting S. nisoria in Kenya, the secondaries are usually moulted convergently, though irregular sequences occur as well (A. Lindstrom pers. comm.). First winter S. c. communis and F. hypoleuca appear to generally moult the secondaries descendantly (see part II). Replacement of some secondaries during the first winter also occurs in individual Anthus campestris which suspend the postbreeding moult as adults (p. 66). Emberiza hortulana may be an exception to this moult cycle (and is not shown in Table 3), since the second-year spring birds examined had not renewed any secondaries (p. 195). Thus, this species might retain the juvenile secondaries, which are often not moulted during the postbreeding moult, for one and a half or two years. Second-year spring birds of Sylvia nisoria, S, cantillans, S. communis and S. curruca often show differently worn non-juvenile wing-coverts within the same wing (own obs., see part II). It is not yet clear whether they have been moulted at different times during a protracted prebreeding moult or whether there is an additional, possibly overlapping, moult (cf. section 3.3.3). In the latter case, the first 'prebreeding' moult in Africa might be interpreted as a continuation of the postjuvenile moult (cf. Roselaar in Cramp 1992).

The strategy of some adults whereby some secondaries are retained during an otherwise complete prebreeding moult before renewing some secondaries during the partial postbreeding moult (group 6 in Table 2) is also followed during the first year of life by an even higher percentage of Oriolus oriolus (p. 153; moult cycle 12). Because the ages of individual Muscicapa striata observed in the non-breeding area were not determined, it remains to be shown whether or not both age classes of this species follow the same moult cycle. Whether the secondary moult is resumed at the point of interruption during the postbreeding moult is not known (cf. p. 139) and the reasons for this particular moult strategy remain unclear. In certain long-distance migrants of the Laniidae, the reduction of the first prebreeding moult may involve some primaries as well as some secondaries, while adults perform a complete prebreeding moult. This is the common strategy of Lanius senator senator and L s. badius. Firstyear birds show a postjuvenile moult of limited extent in the breeding area. During the winter, they do not moult completely (cf. Lanius collurw), but retain up to five of the innermost primaries (Fig. 32), none to six of the secondaries (see also section 4.3.3), none to all primary coverts and rarely some alula feathers (Ullrich 1974, Svensson 1992, Roselaar in Cramp & Perrins 1993, own data; Fig. 36). The rest of the plumage is renewed. However, among 53 second-year spring birds, three renewed all the primaries (one of them also replaced all the secondaries, and PC 7 and 8 were the only juvenile feathers retained, own data, Fig. 37). Thus, a complete moult during the first winter possibly occurs as well. During the first postbreeding moult, secondyear birds often start to renew the juvenile innermost primaries as well as the juvenile primary coverts while still in the breeding area. They may then show three generations of primaries simultaneously, e.g. P 2 juvenile, P 3-10 prebreeding, P 1 postbreeding (Ullrich 1974, Svensson 1992). Primary moult is probably then suspended until arrival in the African winter quarters (moult cycle 13 in Table 3). In contrast to all other European passerines, this species takes two years to enter the adult moult cycle with its partial postbreeding moult before the autumn migration and subsequent complete moult in Africa (moult strategy 3 in Tables 1 and 2). Similarly, some first winter Lanius i. isabellinus may also probably retain the innermost primaries during the first prebreeding moult (Stresemann & Stresemann 1972a). However, the peculiar start of

First partial prebreeding moult including primaries Among the two species which regularly suspend primary moult as adults, Anthus campestris may renew the central primaries (p. 66) and

Moults during the first year of life in trans-saharan migrants

47

4.6.4 Conclusions

Fig. 36. Lanius senator badius 2y 9 after partial prebreeding moult, 22 April. Unlike the adults, most first winter Lanius s. senator and L s. badius do not perform a complete moult in Africa, but retain primaries, secondaries, primary coverts and rarely alula feathers. This bird retained primaries 1—3, secondaries 1—5, all primary coverts and alula 2,

Fig. 37. Lanius senator senator 2y <S after almost complete prebreeding moult, 5 May. Rarely, first winter L. s. senator and L. s. badius moult all remiges in Africa. This bird can be recognized as 2y only by the retained primary coverts 7-8. primary moult in the breeding area during the first postbreeding moult, shown by Lanius senator> has not been observed in these species (which is not included in Table 3). In Lanius cristatus, which usually shows two complete moults each year, the subspecies lucionemis generally retains the innermost primaries during the prebreeding moult, although it is not clear whether this occurs only in first winter birds (Stresemann & Stresemann 1971). Among 15 captive Lanius collurio, two birds only renewed the four or five outermost primaries during their first prebreeding moult (Gwinner & Biebach 1977). Thus, in the Laniidae incomplete primary moult during the first year of life appears to be a regular phenomenon, especially in the non-breeding area (but see Miller 1928 for Z, ludovicianus in the breeding area). However, only Lanius senator appears to start and then suspend the next primary moult during the first postbreeding moult. The partial moult of the remiges in the Laniidae may serve to shorten the total moult duration in first winter birds which (at least in Lanius isabellinus and L. collurio, Stresemann & Stresemann 1972a, B. Bruderer pers. comm.) starts later than in adults. However, Bensch et aL (1991) mention that first winter Lanius senator moult more slowly, thus might better maintain their capacity for flight during their first stay in Africa than the adults. First-year Locustella luscinioides often renew only the outer or central primaries during the winter (Roseiaar in Cramp 1992). How this relates to the complex and variable moult strategy of the adults of this species remains to be shown.

The moult cycles of trans-saharan migrants during their first year are, like the adult cycles, poorly understood. For instance, it is often not clear whether the restricted postjuvenile moult that precedes autumn migration represents the beginning of the complete moult performed mainly in Africa or whether it is a postjuvenile moult in its own right (consequently, moult cycle 9 in Table 3 may actually consist of a variety of moult cycles). This problem confuses the classification and nomenclature of plumage and moult cycles between species, one of the earliest aims of the study of moult. In particular, we can still not be certain whether the first moult after hatching, designated the postjuvenile moult, is homologous among all species, although it may have this connotation for some readers. It must be reiterated that the terms given to plumages and moults in this book do not indicate homologies, but that they merely refer to the occurrence of moult relative to the breeding season and autumn migration in a strongly seasonal environment (see section 2.2.1). The moult of migrants in Africa requires further study and the collection of separate data for adults and first-year birds. The moult of first-year birds in Africa might, in some species, actually consist of two (overlapping?) moults, similar to that of adults (see section 3.3.5; Motacilla flava and Anthus cervinus, Pearson & Backhurst 1973, D.J. Pearson in //'#.; possibly Sylvia communis> S. cantillans, S. nisoria, see section 4.6.3). There may thus be three moult phases during the first year in certain species (see Rohwer 1986, Thompson 1991 and Young 1991 for American Passerina species). In summary, most trans-saharan migrants enter the moult cycle of the adults during their first winter. Exceptions are some Laniidae with a reduced first prebreeding moult, Locustella fluviatilis and Acrocephaluspalustris which omit the partial moult in NE Africa in the first autumn, and some species which perform a more extensive first prebreeding moult than the adults (e.g. some Sylvia warblers). Furthermore, Carpodacus erythrinus and Emberiza a. aureola show only a partial moult in their first winter, not a complete moult as do the adults. There appears to be a relationship between the extent of the moult before autumn migration and the extent of the first moult in Africa. Species with a restricted premigratory moult are more likely to show an extensive or complete first prebreeding moult. The retention of remiges during the first postbreeding moult, due to a seasonally divided moult, in first-year birds seems to result from their moulting these feathers half a year in advance, during an extensive first prebreeding moult. However, it is still not known whether this is always the case or whether some birds may retain some juvenile remiges for one and a half or two years (probably some secondaries in Emberiza hortulana and in some Sylvia communis and S. nisoria\ possibly secondaries and primaries in Anthus campestris\ some primaries in Lanius senator]. There seem to be four types of partial moult of remiges during the first winter, which are related to the moult strategies of the adults. (1) Birds which renew some secondaries during an extensive partial prebreeding moult (moult cycle 7), probably as a preparation for a possible retention of some secondaries during the subsequent postbreeding moult. (2) Birds which retain some secondaries during an otherwise complete first prebreeding moult and usually replace them during the subsequent postbreeding moult (moult cycle 12), as do some of the adults. The reasons for this moult cycle remain obscure. (3) Birds which renew part of the primaries during an extensive first partial prebreeding moult (moult cycle 8) perhaps to prepare for a subsequent primary moult suspension, but which may renew primaries for other reasons also. (4) Certain long-distance migrant Laniidae which reduce the complete prebreeding moult during their first winter to a partial moult, possibly to shorten the moult duration, or to reduce wing raggedness during moult, while the adults generally perform a complete moult.

This page intentionally left blank

PART II CHAPTER 5

Ageing European Passerines 5.1 Ageing criteria in live birds The most common aim of ageing is to assign individual birds to age classes based on years. In passerines, birds in their first year of life are to be distinguished from older (adult) birds (normally the terminology referring to the calender year is applied, see section 2.3). Older age classes can only be distinguished in a few European passerine species. During the first months of life, the age of a bird may be estimated more precisely (in days) using the progress of plumage developement and postjuvenile moult (see section 4.3) or of skull pneumatization (see section 4.4.3), techniques which will not be discussed further here. In part II of this book, ageing refers to the distinction of birds during their first year of life from older birds. Although any character which changes with age can be used as an ageing criteria, for live birds examined in the field, such criterion are limited to external features which can be examined by eye, such as plumage characters (see sections 5.2 and 5.3), skull pneumatization (see appendix) and the appearance of bare parts. Each potential ageing criterion must be tested for reliability by some independent means of ageing. The use of ringed birds of known age as a reference is infallible, but limited to retraps, which are not usually numerous. In summer/autumn, skull pneumatization is among the most important independent ageing criteria applicable to the majority of birds (see appendix). The reliability of any ageing criterion depends partly on the way the character changes with age. Many bare parts change gradually with increasing age, such as the colour of the iris (e.g. Kuschert 1980, Karlsson et al. 1985, 1988, Leverton 1987, Shirihai 1988, Gargallo 1992), orbital ring (Gargallo 1992), eye-lid (Reichholf-Riehm 1977), inside of the upper mandible (e.g. Hogstad 1971, Karlsson et al. 1986a,b, 1988), tongue (e.g. in Sturnus vulgaris, Svensson 1992), bill (e.g. in Ptyonoprogne rupestris, Svensson 1992), tarsus and feet (e.g. Karlsson et al 1988) as well as the shape of the bill (e.g. Bub 1985, Svensson 1992). Such characters are most useful for separating firstand second-year birds from adults when the colour change occurs not too soon after fledging, but before the first breeding season. Inevitably, there is a certain transitional range within a character during which a bird cannot be reliably attributed to either of the two age classes. This is accentuated by individual variation in hatching date and the speed of the colour change. Thus, characters which change gradually with age must be examined for ease of recognition, reliability and consistency among individuals. In certain species, the age classes may be easily separated at a certain season (e.g. Leverton 1987, Karlsson et aL 1988, Gargallo 1992, p. 101), while at other seasons, some of the individuals are impossible to age or are assigned to the wrong age class (e.g. Kuschert 1980, Pettersson 1983, Karlsson etal. 1986a,b, 1988, see also p. 89). In some species, the age difference in the colour of the bare parts is confounded by differences between the sexes (e.g. iris colour, Buxton 1947, Neuschulz 1988) or seasons (e.g. bill colour in Turdus merula, Miles 1971, Glutz & Bauer 1988). In the case of skull pneumatization (see appendix) and tongue spots (in the genera Cettia, Locustella, Acrocephalus, Hippolais\ Kuschert 1980, Karlsson et aL 1988, Svensson 1992), a character typical of young birds disappears with time. Again, these characters are most useful when the change does not occur too soon after fledging, but before the first

breeding season. Usually, such characters are easy to recognize. Provided that all individuals lose the juvenile character (i.e. non-pneumatized skull windows, tongue spots), young birds can be recognized with certainty. Adults, however, can only be determined before the time when the first juveniles attain the adult state. Thus, it is vital to know whether some adults may retain the juvenile state as well as when and how synchronized is the disappearance of the juvenile character from young individuals. Tongue spots may be present in some adults (e.g. Kuschert 1980, Karlsson et aL 1988) and a few species regularly retain non-pneumatized skull windows as adults (see appendix). Most plumage characters indicative of age concern differences between juvenile feathers and those of subsequent feather generations. Such characters change suddenly when these feathers are moulted and there can thus be no doubt about a bird's age, provided that all individuals moult in the same way and at a similar, known, age. Such plumage characters must be checked for ease of recognition and for time, constancy and synchronization of the change among individuals. Other plumage features vary continuously (e.g. wing-length, relative or absolute length of certain feathers, shape of rectrices, shape and structure of primary coverts) and individual variation produces a smaller or larger overlap in the distributions of the character between the age classes. In most species, linear measurements, especially the length of feathers or the wing, show too much overlap between the two age classes to be a useful criterion (e.g. Alatalo et aL 1984, Winkel & Winkel 1992, but see Nakamura 1990). On the other hand, the length of primary 10 relative to the primary coverts is a very useful ageing character in Panurus biarmicus. The shape of the rectrices and primary coverts may be a helpful character in certain species (e.g. Laaksonen & Lehikoinen 1976, Amann 1980, Pettersson 1983, Karlsson et aL 1986a), while in others it assigns many individuals to the wrong age class (e.g. Sylvia atricapilla own obs.; Regulus spp., Jackson 1992, own obs.) (see also Svensson 1992). In many species, there is no easily recognizable ageing criterion applicable to all individuals. In such cases, we do not recommend the use of single characters or dichotomous ageing keys, but strongly advise the examination of several characters on the same bird. If any one of the presence-absence characters can be assessed with certainty, it should be given preference over the other characters. Consequently, we stress the recognition of juvenile feathers and skull pneumatization as ageing criteria in the species accounts and give indications regarding the ease of recognition and constancy of these characters.

5.2 Ageing using plumage characters Ageing by plumage characters is based on the recognition of juvenile feathers and/or on recognizing differences in the extent of moult. In both cases, a thorough knowledge of the moult cycles and of the variation in the extents of moults is a prerequisite. In a few species, differences between postjuvenile and adult feathers exist.

5.2.1 Recognition of juvenile feathers In most European passerines, the juvenile feathers differ in structure and often also in coloration from subsequent feather generations. The

50

Ageing European Passerines

feathers which replace the juvenile plumage, however, are generally adult-like in colour and structure (see section 4.2). Thus, all juvenile plumage features are generally lost when the last juvenile feather has been moulted. Intermediate 'immature' plumages acquired during moults in the first year of life are exceptional or differ only slightly from the adult plumage (see section 5.2.3), in contrast to many large nonpasserines.

Structure and shape In most species, the juvenile body-feathers are more loosely textured, especially those of the neck, mantle and tail-coverts (see section 4.1, Fig. 24—26). In a very few species only, the juvenile body-feathers are indistinguishable from the adult in texture (e.g. hirundines). With the exception of a few species, first-year birds replace all their body-feathers during the postjuvenile moult. The difference in texture between retained juvenile and adult wing-coverts, rectrices and remiges, however, is, if any, slight and difficult to recognize. The juvenile primary coverts, rectrices and remiges are often narrower and more pointed than the corresponding feathers of subsequent generations. This may be a useful ageing criterion in some species (e.g. Paruspalustris, /*. montanus, P. cristatus\ Laaksonen & Lehikoinen 1976, Amann 1980), but in many, there is considerable overlap in shape between juvenile and adult feathers. If two feather generations are present within the same feather tract, they may differ in shape and length, which facilitates the recognition of moult limits. In certain species, there are conspicuous differences in shape and length between juvenile and postjuvenile/postbreeding primary 10 (e.g. Panurus biarmicus, Pica pica)) central primaries (Coccothraustes coccothraustes) and rectrices (Hirundo rustica, Aegithalos caudatus).

Coloration In species which exhibit conspicuous colour differences between the juvenile and subsequent plumages (e.g. Erithacus rubecula}^ birds in the juvenile plumage and first-year birds which retain some juvenile bodyfeathers after the postjuvenile moult are easily recognized (e.g. juvenile body-feathers in Anthus campestris, retained juvenile uppertail-coverts in Muscicapa striata, Ficedula hypoleuca). In many species, colour differences between the wing-coverts, rectrices and remiges of the juvenile and subsequent feather generations are important ageing criteria. In many cases, however, these differences are slight or virtually nonexistent, but may be affected by wear which often intensifies them. In good light, differences in the colour intensity, brilliance or gloss of the feather centres, as well as in the colour and colour pattern of the fringes can be detected and are important characters for recognizing juvenile feathers and moult limits. Many examples are given in the species accounts. In a few species, the colour pattern of the juvenile flight feathers is different from that of the adults (e.g. tertials of Ficedula hypoleuca^ outer primaries of Bombycilla garrulus, primary 9 of Pica pica, central primaries in Coccothraustes coccothraustes', outer rectrices in some Sylvia warblers).

Wear The juvenile feathers grow at a time when the adults have usually worn their feathers for between half and almost a full year. Thus at fledging, the juvenile feathers are fresh (with the exception of some possible wear whilst in the nest), while those of the adults are usually worn. This difference in wear is often a useful ageing criterion in summer in species in which the juvenile plumage is similar in coloration to that of the adults (e.g. Acrocephalus scirpaceus) and in species with a complete postjuvenile moult, provided that some outer primaries are not yet moulted. In most species which moult completely in winter, this differ-

ence in wear is an important ageing criterion until the next complete moult (see section 5.3.4). As the feather fringes wear off, the colour pattern of certain feathers may change with time. In most species, the adults renew their plumage soon after the breeding season. Consequently, there is only a difference of up to several weeks in wear between juvenile and postbreeding feathers after the summer/autumn moult period. However, since the juvenile feathers are less durable and since they are slightly older than those of the adults, a difference in wear may still be obvious and becomes progressively more pronounced. The precise degree of feather wear depends on the species, individual and exposure of the particular feather (see section 1.2.1). Thus, recognizing feather generations by their degree of wear requires a lot of experience, unless two feather generations can be compared within the same feather tract. The utility of recording the extent of wear in moult studies is discussed in Rogers (1990) who provides a classification of feather wear.

Growth bars and fault bars Feathers grow day and night at a similar rate (Murphy & King 1986). During the night, when most birds are fasting, the feather material deposited is of slightly different quality than during the day. This results in a pattern of alternating day-night growth bars on the vane which is not normally visible on casual inspection, but may be seen in reflecting light in some species and feathers, especially the rectrices and remiges (Michener & Michener 1938, Wood 1950, Stiefel in Bub 1985, Winkler et al, 1988). Under certain circumstances, feather growth is disturbed resulting in more severe alterations in structure and pigmentation and, consequently, more or less conspicuous growth bars of different width and/or of structurally inferior quality (e.g. Stiefel in Bub 1985, Murphy et ai 1989; Fig. 38-41). If such disturbance of feather growth is severe, the resulting low-quality bars are called fault bars. If visible, irregularities in growth bars are equal on all feathers grown simultaneously (Michener & Michener 1938; Fig. 38—40). However, the conspicuousness of irregularities in growth bars often varies among simultaneously grown feathers, and can even be absent on some. The reasons for variations and irregularities in growth bars are still in dispute. Some authors believe that fault bars are due to nutritional stress (summarized in Machmer et al. 1992), but others have shown that they may be caused by stress situations such as handling, perhaps amplified by malnutrition (King & Murphy 1984, Murphy etaL 1988, 1989, Machmer et al. 1992). The widths of the regular growth bars certainly reflect the growth rates of the feathers and may be influenced by nutrition (e.g. Michener & Michener 1938, Grubb 1989). If the same pattern of irregularities in growth bars is present on several feathers of the same or adjacent feather tracts, the relative timing of growth of these feathers can be assessed. In some cases, conclusions about the bird's age may be drawn, but such conclusions need to be drawn with care. Since the feather-tip grows first and more material added at the base, feathers grown at the same time show fault bars at approximately the same distance from the tip. This is the case in juvenile feathers which are grown simultaneously in the nest (Fig. 38—39). If the feathers are grown sequentially within a tract, the fault bar pattern shows progressively further away from the feather-tip and, depending on the shedding intervals, may be present on only a few feathers (Fig. 40). This is the pattern produced by a sequential moult, typical of adults in many species. An instructive example is the blue and black barring on the wing of Garrulus glandarius which seems to represent growth bars and whose pattern of irregularities is a good ageing criterion (see p. 156). However, great care is needed when applying this technique (cf. Svensson 1992). Growth bar patterns may differ in conspicuousness

Ageing using plumage characters

51

Fig. 38. Sylvia curruca ly after partial postjuv moult, 25 August. Fault bars are present across all primaries and secondaries at similar distances from the tip. This indicates that these feathers have grown simultaneously, as is typical of juvenile feathers.

Fig. 40. Turdus merula ad 9 after complete postbr moult, 17 October. A distinct pattern of fault bars is visible on primaries 1 and 2 which is further away from the tip on P 2 than on P 1, indicating that these feathers have grown sequentially. On the other primaries and on the secondaries the pattern of growth bars is faint or rather regular, but slightly displaced against each other.

Fig. 39. Fringilla coelebs ly 9 after partial postjuv moult, 20 September. The distinct fault bar pattern crosses all primaries, secondaries and tertials at similar distances from the tip, indicating that these feathers have grown simultaneously. Fault bars are also present on the juvenile primary coverts, alula feathers 2—3 and greater covert 1, but not on the postjuvenile greater coverts and alula feather 1.

Fig. 41. Sylvia atricapilla ad 6 after postbr moult, 9 May. This adult shows prominent fault bars on secondaries 2-6 and tertials 7-8 at similar distances from the tip on the right wing only. These feathers have apparently been lost accidentally and re-grown simultaneously. In cases of accidental feather loss, fault bars on the replacement set give no clues to age.

among the feathers of a tract or may be absent on some feathers within a tract. Many birds renew feathers simultaneously during a normal moult (especially rectrices, see section 3.2.3). Feathers may have been lost accidentally, in which case the replacement set has grown simultaneously (Fig. 41). Indistinct irregularities in growth bars may appear by chance at similar distances from the tip. In most individuals, any irregularities in growth bars are confused and one should refrain from constructing ageing arguments from them. Of course, all the feathers of first-year birds which have performed a complete postjuvenile moult show a sequential fault bar pattern.

5.2.2 Differences in extent of moult In many species, the extent of the moults during the first year of life differs from that of the corresponding moults of adults (see chapter 4). For instance, in many species adults perform a complete moult while first-year birds show a partial moult during summer/autumn. In this case, it is sufficient for ageing to recognize whether one or two feather generations are present, without the need to identify the precise feather

generations (e.g. juvenile feathers). In a few species, both first-year and adult birds perform a partial moult, but of consistently different extent thus indicating their age (e.g. Muscicapa striata, see p. 138; Locustella fluviatilis and Acrocephaluspalustris in NE Africa, see section 4.6.2). Differences in the extent of moult between age classes can also be of use when determining the age of birds in active moult, provided that the sequence is known (see sections 3.2 and 4.3).

5.2.3 Differences between postjuvenile and subsequent feather generations In a few species, the feathers which replace the juvenile feathers are conspicuously different from those of subsequent feather generations (e.g. Lanius collurio, see p. 154; Oriolus oriolus, Svensson 1992) and provide a convenient ageing criterion. In Sturnus vuigaris and S. unicolor, the shape and colour of the postjuvenile rectrices, throat and mantle feathers differ on average from those of adults which allows one to age most individuals (Hiraldo & Herrera 1974, Williams 1991, Svensson 1992).

52

Ageing European Passerines

In some species, there are, on average, slight differences between the postjuvenile and subsequent feather generations in the intensity and brightness of colour or gloss and in the width of the light feather fringes (e.g. Eremophila alpestris, Motacilla flava, Luscinia svecica, Phoenicurus phoenicurus, Saxicola rubetra, Sylvia atricapilla, Ficedula parva, Cinclus cinclus, Fringilla montifringilla, Carduelis flammea\ Fig. 42). However, the overlap between first/second-year birds and adults is usually too large to provide reliable ageing criteria for individual birds, Furthermore, in some of these species, the age differences in coloration are confounded by sex differences.

former have been allocated to moult cycle type (3), the latter to moult cycle type (2), (3) and (4), respectively. In the Tables 5—8, all species of passerines breeding in Europe are listed, except a few breeding in the far east of Europe. For other species, occurring in Europe, the indications on moult in Svensson (1992) and in the Handbooks by Glutz & Bauer (1985, 1988, 1991) and Cramp (1988, 1992, Cramp & Perrins 1993) will allow their allocation to one of the moult cycle types. Each species has been entered in only one of the Tables 5-8, although it may moult according to two moult cycle types. Such cases and other exceptions are indicated in the Tables.

5.3 General ageing criteria in European passerines based on moult

5.3.1 Species with a complete postjuvenile moult in the first summer/autumn: Moult cycle type 1

As shown in the preceding section, ageing on plumage characters relies primarily on the recognition of juvenile feathers and on differences between age classes in the extents of corresponding moults. How long juvenile feathers remain on the bird and how different are the extents of corresponding moults between age classes, depend primarily on the moult cycle (see chapter 4 and Table 3). Therefore, the general plumage ageing criteria are basically similar for birds moulting according to the same moult cycle. In the remainder of this section, we group the European passerines according to their moult cycles and indicate the general ageing criteria they have in common. Being aware of the moult cycle of a species when ageing it has several advantages: (1) it may allow ageing of birds in active moult, something generally neglected in existing ageing guides; (2) it allows a better understanding of unusual or previously unknown moult patterns for the species, and of their implications for ageing; (3) it may allow ageing of species whose ageing criteria have not been studied in detail; (4) it allows easier detection of previously undescribed plumage ageing criteria. However, since in certain species, not all individuals moult according to the-same moult cycle, a knowledge of the possible intraspecific variation in moult cycles, as presented in chapters 3 and 4 and Table 2, is indispensable. For the purpose of ageing, four main types of moult cycles can conveniently be distingushed. They are based on the number and extents of moults up until the last juvenile feather has been replaced: (1) In a few European passerines, all juvenile feathers are replaced in the first summer during a complete postjuvenile moult (section 5-3.1). (2) In many species, part of the juvenile feathers are moulted during the postjuvenile moult in the first summer/autumn and the remainder during the first complete postbreeding moult in the second summer/autumn (section 5.3.2). (3) A number of species present the same moult cycle as in type (2), but with an intervening partial prebreeding moult which may or may not include some of the retained juvenile feathers (section 5.3.3). (4) In some long-distance migrants, only a few or no juvenile feathers are moulted during the postjuvenile moult in the breeding area, the remainder being replaced later during a complete moult in the non-breeding area (section 53.4). There are few birds which do not readily fall within one of these four moult cycle types: those with a seasonally divided moult as well as Carpodacus erythrinus, Emberiza aureola and Phylloscopus trochilus. The

Moult cycle: Members of this moult cycle type (Table 5) replace all the juvenile feathers during the postjuvenile moult in their first summer/autumn. Adults also perform a complete moult at about the same time (Fig. 43). Only Acrocephalus melanopogon and Cisticola juncidis have a partial prebreeding moult which occurs in second-year birds and adults and gives no clue to a bird's age. Ageing before the postjuvenile and postbreeding moults: The juvenile plumages are generally easily recognized by the structure and coloration of the body-feathers. Adults in summer have worn their plumage for almost a year and usually show distinct signs of abrasion and bleaching (Fig. 44-47). Ageing during the postjuvenile and postbreeding moults: If still present, the unmoulted feathers can usually be scored as being juvenile or not by their coloration, structure and wear. The presence of growing remiges is not an ageing criterion. In Panurus biarmicus, the juvenile P 10 (the last primary to be moulted) is considerably longer than the primary coverts, while in adults P 10 only reaches the primary coverts. Ageing after the postjuvenile and postbreeding moults: This is generally not possible on plumage characters, except in Sturnus vulgaris and S. unicolor (see Hiraldo & Herrera 1974, Williams 1991, Svensson 1992). However, some birds exceptionally retain individual juvenile feathers. In most species of this moult cycle type, skull pneumatization is valid until well into the autumn, but not in Alauda arvensis, Lullula arborea and possibly other Alaudidae in which the first fully pneumatized first-year birds appear in September/October (Winkler 1979). In Panurus biarmicus^ skull pneumatization is hardly visible in live birds because of the pigmented skin.

5.3.2 Species with a partialpostjuvenile/Complete postbreeding moult in the breeding area: Moult cycle type 2 Moult cycle: Members of moult cycle type 2 (Table 6) have one complete moult annually just after breeding as adults, while the juveniles perform a partial postjuvenile moult after hatching and before winter or migration (Fig. 48). The only exception is Carpodacus erythrinus which delays the postjuvenile and postbreeding moult until its arrival in the tropics. Extent of moults: In all these species, the postjuvenile moult of bodyfeathers is nearly always complete. On the wing, however, the extent of Fig. 42. 6 Phoenicurusphoenicurus in autumn. The grey, white, black and orange colours on the head and breast are on average more concealed by light fringes in first-year birds (centre and right) than in adults (left). However, only extremes can be aged by the conspicuousness of the coloration. Some adults look like the bird in the centre.

General ageing criteria in European passerines based on moult

summer

autumn/winter

spring/summer

53

autumn/winter

First year oflife

Adults

Fig. 43. Schematic presentarion of the plumage cycle of moult cycle type 1: Species with a complete postjuvenile moult in the first summer/autumn. After the postjuvenile/postbreeding moult, the plumage of both adults and first/second-year birds is composed of one feather generation acquired at about the same time. Only two species perform a partial prebreeding moult (not shown, see Table 5). Key to Fig. 43, 48, 53 and 54: Colours indicate the feather generations acquired by a complete (full arrow) or partial (broken arrow) moult: green = juvenile feathers; red = feathers acquired during the postjuvenile/postbreeding moult in the breeding area; blue = feathers acquired during the prebreeding moult in winter/spring. Dotted areas indicate worn feathers. Table 5. Moult cycle type 1: Species which regularly perform a complete postjuvenile moult in their first summer/autumn. Species in which only some of the individuals perform a complete postjuvenile moult are shown in Table 6 and 7. Chersophilus duponti Melanocorypha calandra Calandrella brachydactyla Calandrella rufescens j Galerida cristata Galerida theklae Lullula arborea Alauda arvensis Eremophila alpestris Cisticola juncidis ' Acrocephalus melanopogon 1

Panurus biarmicus Aegithalos caudatus Sturnus vulgaris Sturnus unicolor Passer domesticus Passer hispaniolensis Passer montanus Petronia petronia Montifringilla nivalis Miliaria calandra

' Some ly perform a partial postjuvenile moult before a complete moult in late autumn (see section 4.4.4). Prebreeding moult occurs in at least some individuals, but its extent is poorly known. ' Prebreeding moult in 2y and ad the postjuvenile moult varies considerably among species, from a few wing-coverts and no greater coverts moulted (e.g. Sitta), through a regular moult of all the greater coverts and usually some tertials as well (Parus major, P. caeruleus, Coccothraustes coccothraustes, Emberiza citrine/la, E. da) to partial, usually eccentric, primary moult (in earlyhatched birds or in Mediterranean populations of some Carduelis species and Loxia curvirostra, see section 4.4). At the extreme, a complete postjuvenile moult may occur (some Loxia curvirostra and

Carduelis Moris]. Within species, the extent of the postjuvenile moult may vary considerably as well. Ageing before the postjuvenile and postbreeding moults: As for members of moult cycle type 1. Ageing during the postjuvenile and postbreeding moults: Birds in active moult of the median coverts, greater coverts and tertials and/or regular, symmetric rectrix moult but without signs of moult of the secondaries and primaries are first-year birds (Fig. 49—50). Birds with growing primaries and/or secondaries following the basic sequence are generally adults (Fig. 52). However, in some species first-year birds may renew the innermost secondaries (see section 4.3.1) and/or some primaries during the partial postjuvenile moult (Fig. 51; see Table 6 and section 4.4.4). Usually partial primary moult is eccentric, i.e. does not start with P 1 as does the complete moult of the adults, but with one of the central primaries (frequently with P 5-7), but may exceptionally also include the innermost primaries (see section 4.3.3). First-year birds moulting some primaries during the postjuvenile moult typically retain the corresponding primary coverts or replace them irregularly, while adults performing the complete postbreeding moult renew the primary coverts together with the corresponding primaries. Partial primary moult may be observed in more species than shown in Table 6, especially in Mediterranean populations. Thus, when examining actively moulting birds of species which may show partial primary moult as first-year birds, the sequence of primary, secondary and primary covert moult must be observed in detail; skull pneumatization (see appendix)

54

Ageing European Passerines

Fig. 44. Passer domesticus ly at the beginning of the complete postjuv moult, 26 August. Whole wing juv, except P 1-2 and PC 1-2 growing. Juvenile remiges are fresh until renewal at the complete moult. Also on the wing, the colour of the juvenile plumage is different from that of subsequent feather generations.

Fig. 45. Passer domesticus 9 after complete moult, 2 May. Compared with the juvenile wing (see Fig. 44), the wing of adults is differently coloured and worn, especially the tertials, inner greater coverts and outer primaries. and, if still present, retained juvenile body-feathers (which are usually easily identified) are other important ageing criteria. In a few species, some first-year birds (see Table 6) perform a complete postjuvenile moult (following the basic sequence) and the general ageing criteria described for moult cycle type 1 apply. Ageing after the postjuvenile and postbreeding moults: In most species, ageing by plumage characters is based on the recognition of whether one (in adults) or two feather generations (the juvenile and postjuvenile in first/second-year birds) are present (Fig. 48). Thus, one has to decide whether or not there are moult limits between juvenile and postjuvenile feathers. This is possible in most species of this moult cycle type during autumn and winter and in some species with only slight feather wear, up to the time of the complete moult (e.g. Pyrrhula pyrrhula). The few first/second-year birds which have performed a complete postjuvenile moult are inseparable from adults on plumage characters, but may be recognized by skull pneumatization. Moult limits are usually best detected within the greater coverts. Birds with no greater coverts moulted may show moult limits within the median and marginal coverts or colour differences between the renewed median coverts and juvenile greater coverts (e.g. Nucifraga caryocatactes, Jenni 1983). Birds with all the greater coverts moulted must be checked for moult limits within the tertials, rectrices and alula feathers or between the tertials and secondaries as well as between the greater coverts, primary coverts, alula feathers and carpal covert. In

Fig. 46. Sturnus vulgaris ly in juvenile plumage, 8 June. Juvenile wing fresh, uniformly grey, with dull fringes and without gloss.

Fig. 47. Sturnus vulgarised. cT after complete postbr moult, 5 May, Compared with juveniles (see Fig. 46), the wing of adults is worn, glossy and with prominent light fringes.

some species, some individuals show moult limits within the primaries and secondaries and may be determined as first/second-year birds. If valid, skull pneumatization will be the useful criterion for birds in which moult limits are difficult to assess. Although individuals presenting moult limits are generally first/second-year birds, one has to bear in mind that adults of almost every species may, exceptionally, retain individual feathers or interrupt moult before it is completed. In these cases the retained feathers are conspicuously bleached (being at least one year old). The feathers retained are always the last feathers to be moulted during the postbreeding moult, i.e. S 6, P 9, 10 and Al. Some adults at the end of the postbreeding moult may be recognized as such by the sheaths still showing on the last feathers moulted (but beware of partial primary moult of first-year birds). Exceptionally, individual juvenile body-feathers are retained.

53.3 Species with a partial postj uvenilelcomplete postbreeding moult in the breeding area and a partial prebreeding moult in winter I spring: Moult cycle type 3 Moult cycle: Adults following this moult cycle type (Table 7) perform a complete postbreeding moult in the breeding area and a partial

General ageing criteria in European passerines based on moult

Table 6. Moult cycle type 2: Species which perform a partial postjuvenile moult in their first summer/autumn and a complete postbreeding moult as adults. Species in which only some of the individuals do so are shown in Table 7. Ptyonoprogne mpe$tris()

Sitta neumayer6

Bombycilla garrulus* _. . . / 6 Linclus cinclus

Certhia familiaris Certhia hrachydactyla2

Troglodytes troglodytes Prunella modularis Prunella collaris Erithacus rubecula Luscinia luscinia Lusciniamegarhynchos Tarsiger cyanums Phoenicurus ochruros' Phoenicurus phoenicurus Saxicola torquata^ Oenamhe leucura Monticola solitarius6 Turdus torquatus Turdus merulalA Turdus pilaris Turdus philomelos Turdus iliacus Turdus viscivorus Regulus regulus Regulus ignicapillus Parus palustris Parus lugubris Parus montanus Parus cinctus Parus cristatus Parus ater Parus caeruleus1 Parus cyanus 3or Sitta whiteheadi Sitta europaea

Remiz pendulinus2 Garrulus glandarius Perisoreus infaustus Cyanopica cyana}'z Pica pica Nudfraga caryocatactes Pyrrhocorax graculus Pyrrhocorax pyrrhocorax Corvus ^onedula Corvus frugtlegus Corvus corone Corvus corax

Fringilla coelebs Fringilla montifringilla Serinusserinus1'2 Serinus citrinella Carduelis chloris] l23 Carduelis carduelis^2 Carduelis spinus^1 C

*rd?dis ™™ahi™2 Carduelis favirostris Carduelis flammea Carduelis hornemanni Loxia leucoptera^ Loxia curvirostra7 Loxia pytyopsittacus* Carpodacus erythrinus23'10 Pinicola enucleator Pyrrhula pyrrhula Coccothraustes coccothraustes

Emberiza citrinella^ Emberiza cia

1

Postjuvenile moult may include secondaries Postjuvenile moult may include primaries, usually eccentrically 3 Postjuvenile moult may be complete 4 Postjuvenile moult usually in late autumn/winter, extent poorly known 3 Postbreeding moult: some feathers may be retained 6 Some individuals moult some body-feathers, occasionally marginal and median coverts in winter/spring. Thus, there might be a restricted prebreeding moult. Moult pattern highly variable 8 Moult pattern may be as complex as in Loxia curvirostra '} Postbreeding moult regularly overlaps with breeding. Very protracted postjuvenile and postbreeding moult lasts well into winter. 111 No moult before autumn migration in ly and ad. Postjuvenile and postbreeding moult take place in autumn/winter in the tropics. :

prebreeding moult in winter/spring (Fig. 53). Birds in their first year of life have a partial postjuvcnilc moult of variable extent in the breeding area, a partial prebreeding moult in winter/spring and attain the full adult plumage by performing a complete postbreeding moult in their second summer/autumn in the breeding area. This moult cycle type is the most complex one, since part of the juvenile feathers can be replaced during the postjuvenile moult in the first summer/autumn or during the prebreeding moult in the first winter/spring and the remainder during the first postbreeding moult in the second summer/autumn. Due to the many possible combinations of feather generations in both second-year and adult birds (Fig. 53), ageing is difficult in many species. Furthermore, in some species a seasonally divided moult of the remiges occurs. Extent of moults: During the postjuvenile moult, all or most of the body-feathers are generally moulted (except m Anthus campestris). The

55

extent of the postjuvenile moult on the wing varies considerably from none (e.g. many Anthus campestris)^ through all the wing-coverts and often the tertials and rectrices (e.g. Emberiza schoenidus), up to a partial primary moult (e.g. Sylvia melanocephala). In some species, some individuals perform a complete postjuvenile moult (some Mediterranean Sylvia warblers). The postbreeding moult of the adults is usually complete. However, adults may interrupt the moult of the remiges before autumn migration more often than in members of moult cycle type 2 (see section 3.3.3, Table 2 and 7). This occurs regularly in Anthus campestris, A. richardi, Sylvia hortensis, S. nisoria, Lanius nubicus, Emberiza hortulana, E. caesia, less frequently in Ficedula hypoleuca and Sylvia c. communis and only exceptionally in other species. The extent of the prebreeding moult varies from an insignificant spotwise moult of the body or head-feathers, through an extensive renewal of the body-feathers, wing-coverts, tertials and rectrices (e.g. Motacilla flava), up to a partial secondary and/or primary moult (usually eccentric). First-year birds with a postjuvenile moult of limited extent tend to replace more wing-coverts during the prebreeding than during the postjuvenile moult and thus replace all the postjuvenile wing-coverts a second time. There is generally such a large overlap in the extent of the prebreeding moult between adults and first/second-year birds, that differences in extents between the age classes (see section 4.6.1) cannot be used for ageing. Ageing before the postjuvenile and postbreeding moults: As for members of moult cycle type 1. Ageing during the postjuvenile and postbreeding moults: As for members of moult cycle type 2. Species which may perform a partial primary moult or a complete postjuvenile moult (common in Sylvia melanocephala) are listed in Table 7. Ageing after the postjuvenile and postbreeding moults: As for members of moult cycle type 2, but restricted to the time between the completion of the postjuvenile/postbreeding moult and the onset of the prebreeding moult. First/second-year birds which have performed a complete postjuvenile moult are inseparable from adults on plumage characters, but may be recognized by skull pneumatization. Those adults which interrupt their complete postbreeding moult before autumn migration are easily recognized by their very worn and bleached retained remiges (usually secondaries) or alula feathers. The retained secondaries are not always the last to be moulted during a complete moult (see Fig. 11). Ageing after the prebreeding moult: (a) Species with a prebreeding moult restricted to the body-feathers (see Table 7): As in the autumn, the wing of adults still consists of one and that of second-year birds still of two feather generations (juvenile and postjuvenile). Hence, the same ageing criteria as for members of moult cycle type 2 apply, provided that the characters have not faded away due to wear. (b) Species with a prebreeding moult including some wing-coverts: Both adults and second-year birds show a moult limit within the wing-coverts due to the prebreeding moult and, in many cases, it will be difficult or impossible to separate the age classes. The only way is to assign those feathers not moulted during the prebreeding moult (remiges, primary coverts, often outer greater coverts and tertials, occasionally median and marginal coverts) to either the juvenile (thus second-year bird), the postjuvenile or the postbreeding feather generation. Juvenile feathers can often (but not always) be distinguished from the corresponding postbreeding feathers of the adults by their more abraded and bleached appearance. Retained postjuvenile feathers, however, are usually impossible to distinguish from the corresponding postbreeding feathers of the adults. If only prebreeding and postjuvenile feathers are present within a feather tract, without any juvenile feathers, ageing may become very difficult. Within the wing-coverts of some second-year birds (e.g. Motacilla alba), the prebreeding moult is less extensive than the postju-

56

Ageing European Passerines summer

autumn/winter

spring/summer

autumn/winter

First year of life

Adults

Fig. 48, Schematic presentation of the plumage cycle of moult cycle type 2: Species with a partial postjuvenile/complete postbreeding moult in the breeding area (see Fig. 43 for details). First/second-year birds retain part of the juvenile feathers until the first postbreeding moult. After the postjuvenile moult, their plumage is composed of two feather generations, in contrast to the single generation of the adult plumage. The extent of the postjuvenile moult is very variable. In a few species, partial postjuvenile primary moult occurs (second row of ly/2y birds).

Fig. 49. Carduelis cannabina ly 8 in partial postjuv moult, 30 August. MaC postjuv, MeC missing or postjuv. GC 1—2 juv, rest growing. Rest of wing juv. During summer/autumn, individuals belonging to moult cycle types 2 and 3 with growing greater coverts, but without signs of moult of secondaries and primaries can be determined as first-year birds.

Fig. 50. Parus caeruleus ly in partial postjuv moult, 13 August. MaC postjuv. MeC juv or missing. GC growing, CC postjuv. Al 1 growing. T 8—9 growing. Rest of wing juv. During summer/autumn, growing greater coverts and tertials, but no growing remiges indicate a first-year bird in species belonging to moult cycle types 2 and 3-

General ageing criteria in European passerines based on moult

Fig. 51. Carduelis chloris ly 6 in partial postjuv moult, 30 August. MaC postjuv. MeC growing. GC 1—9 postjuv, 10 juv. CC postjuv. Al 1—2 shed, 3 juv. T 7+9 juv, 8 growing. P 6 growing, 1—5+7—10 juv. PC and S juv. This bird renews primary 6 eccentrically, but not the corresponding primary covert. This, as well as the facts that primaries 7-9 are not worn and no secondaries are growing, indicates that it is a first-year bird.

venile moult and there are still juvenile coverts present. In these cases, there will be three feather generations among the wing-coverts, e.g. within the greater coverts, from inside to outside, prebreeding, postjuvenile and juvenile coverts. Three feather generations within the coverts is diagnostic of second-year birds. In many species (e.g. some Sylvia warblers) the extent of the prebreeding moult is not well known. Furthermore, the prebreeding moult of some species seems to be subdivided, protracted or to consist of two moults (e.g. Motacillaflava, Sylvia communis, S. cantillans). This may result in the wing-coverts being moulted in an irregular sequence and in a mixture of different degrees of wear spread over the whole feather tract. Such moult patterns are difficult to interpret and no conclusions on age should be drawn. (c) Species with a prebreeding moult including some remiges: Most of these species may interrupt the postbreeding moult before autumn migration and thus show a seasonally divided moult of the remiges. During the winter, adults usually complete the moult of the remiges, but second-year birds may also renew some secondaries or primaries (usually eccentrically). Due to the high variability of these moult cycles (see section 3.3.3 and 4.6.3), it is at present impossible to give any reliable ageing criterion based on the renewal of the remiges. Ageing of some of these birds may be possible by the criteria given above. In the case of partial primary moult, the non-renewal of the corresponding primary coverts usually indicates a second-year bird.

5.3.4 Species with a complete moult in the non-breeding area Moult cycle: All species following this moult cycle type (Fig. 54, Table 8) are long-distance migrants. Most adults and first-year birds perform a partial moult between the end of the breeding season and the onset of autumn migration, and a complete moult in the tropical winter quarters. Within the winter quarters there may be an additional partial moult before or after the complete moult, and the complete moult may be temporarily suspended. Extent of moults: In most species following this moult cycle type, the postjuvenile moult before the autumn migration is of very limited extent (comprising part of the body-feathers and occasionally some wing-coverts) or completely absent (see section 4.4.2). The same holds for the postbreeding moult of adults, but some may include some tertials and rectrices, occasionally some secondaries (e.g. Muscicapa striata, Oriolus oriolus, Phylloscopus bonellt) or the innermost primaries (e.g. hirundines, Acrocephalus arundinaceus, Lanius senator). Among

57

Fig. 52. Sylvia atricapilla ad 9 in complete postbr moult, 1 September. Whilst renewing greater coverts and tertials, adults belonging to moult cycle type 2 and 3 also moult their primaries, following the basic descendant sequence.

first-year birds, only the hirundines may rarely renew some primaries before the autumn migration. Adult Phylloscopus trochilus regularly perform a complete postbreeding moult (sometimes retaining some secondaries) and the same has been observed in a very small percentage of adult Sylvia borin. In the tropics, the adults only of some species show a partial moult in autumn/early winter (e.g. Locustella fluviatilis, Acrocephalus palustris, not shown in Fig. 54). The complete moult may be suspended within the winter quarters, but it is not yet clear whether this occurs in both first/second-year birds and adults. In some species, both age classes may occasionally retain some secondaries (e.g. Muscicapa striata, Oriolus oriolus). Second-year Lanius senator regularly retain the innermost primaries and some secondaries. The occurrence and extent of an additional partial prebreeding moult (e.g. in many Acrocephalus warblers) is poorly known (not shown in Fig. 54). Ageing before the postjuvenile and postbreeding moults: In many species, the juvenile plumage is not easily recognizable by its colour or structure. However, juveniles have a fresh plumage while that of the adults is usually worn (see below). Ageing during the postjuvenile and postbreeding moults in the breeding area: Since the postjuvenile and the postbreeding moults are usually of limited extent, the same ageing criteria as before or after the postjuvenile/postbreeding moult apply (exceptions are Phylloscopus trochilus and some Sylvia borin). Ageing after the postjuvenile and postbreeding moults and before the complete moult in the winter quarters: In most species, ageing relies on the difference in wear between first-year birds and adults. First-year birds have worn their juvenile and postjuvenile feathers for only several weeks or a few months and have a fresher plumage than adults who have worn those feathers not moulted during the postbreeding moult for half a year or more, including the breeding season, and generally show quite abraded and bleached feathers. Abrasion is especially strong in species living in dense vegetation and most evident on the tertials, tips of the outer primaries and wing-coverts. Since both adults and first-year birds may perform a partial moult before the autumn migration, both age classes may show moult limits. Due to the greater difference in age between the two feather generations, however, any moult limits are more prominent in adults than in first-year birds. Furthermore, birds which have moulted some remiges are adults (except in hirundines). In many species, juvenile bodyfeathers are still present and may be easily recognized (e.g. hirundines, Muscicapa striata, Lanius spp.).

58

Ageing European Passerines

A summer

autumn/winter

spring/summer

autumn/winter

First year of life

Adults

B summer First year of life

Adults

autumn

winter

spring/summer

autumn

General ageing criteria in European passerines based on moult

59

Fig. 53. (Opposite) (A) Schematic presentation of the plumage cycle of moult cycle type 3: Species with a partial postjuvenile/complete postbreeding moult in the breeding area and a partial prebreeding moult in winter/spring (see Fig. 43 for details). After the postjuvenile/postbreeding moult, the plumage of first-year birds is composed of two feather generations, while that of the adults generally of only one, as in members of moult cycle type 2. The extent of the postjuvenile moult is very variable, ranging from only a few body-feathers up to a partial postjuvenile primary moult (not shown, see Fig. 48). After the prebreeding moult, the composition of the plumage depends on the extent of the prebreeding and postjuvenile moult. Hence, many combinations of feather generations in both second-year and adult birds are possible, of which only the most typical ones are shown. Adults may show one (upper bird) or two (lower bird) feather generations within the feathers of the body and one (not shown) or two feather generations in the wing. Second-year birds show one or two feather generations within the bodyfeathers, i.e. prebreeding (upper bird) or prebreeding and postjuvenile bodyfeathers (lower bird); two or three feather generations within the wing-coverts, i.e. juvenile and postjuvenile (not shown) or juvenile and prebreeding (upper bird) or postjuvenile and prebreeding (not shown) or juvenile, postjuvenile and prebreeding wing-coverts (lower bird); one or rarely two feather generations within the remiges, i.e. juvenile (upper and lower bird) or juvenile and postjuvenile or juvenile and prebreeding remiges (not shown). (B) An example of a bird showing a seasonally divided moult of remiges, as it occurs in some Sylvia warblers (for other types of seasonally divided moult of remiges see text, Table 2 and 7).

These general rules do not apply to Phylloscopus trochilus in which both adults and first-year birds have fresh plumages (see p. 136) and to the very few Sylvia borin which have performed a complete postbreeding moult (see p. 127). In Locustella fluviatilis and Acrocephalus palustris, adults perform a partial moult in autumn/early winter in NE Africa which includes the outer primaries, part or all of the wing-coverts and the tertials (in L. fluviatilis) and the body-feathers (in A. palustris and L fluviatilis). Firstyear birds do not show this moult, so it represents an ageing criterion after this partial moult and before the complete moult further south (Pearson 1989). Ageing after the complete moult in the winter quarters: Adults and second-year birds are usually inseparable on plumage characters at this time. Both age classes may retain some secondaries (e.g. Oriolus oriolus). Only in Lanius senator and possibly some L. minor, are second-year birds distinguished from adults by the retained innermost primaries, or at least some primary coverts (see Fig. 36 and 37).

Table 7. Moult cycle type 3: Species which perform a partial postjuvenile moult in their first summer/autumn or a complete postbreeding moult as adults in the breeding area and a partial prebreeding moult in winter/spring. In some species, seasonally divided moult of remiges occurs.

Table 8. Moult cycle type 4: Species in which both age classes perform a complete moult in the non-breeding area. In some species, the prebreeding moult may be incomplete.

A nthus campestris' " 5 "'' Anthus trivialis*3'" Anthus pratensis Anthus cervinus Anthus spinoletta'' Motacilla flava'' Motacilla citreola Motacilla cinerea Motacilla alba2'* Luscinia svecica* Saxicola rubetra Oenanthe isabellina** Oenanthe oenantke9 Oenanthe pleschanka* Oenanthe hispanica* Monticola saxatilis9'12 Cettiacetti™ Sylvia sarda™'{2 Sylvia undata23^2 Sylvia conspicillata2™ Sylvia cantillans**7" Sylvia melanocephala2'-* Sylvia melanothorax2'*A Sylvia rueppelli2'* 1

Sylvia honensis6 !' Sylvia nisoria6'" Sylvia curruca^1 Sylvia c. communis67'" Sylvia atricapilla™M Phylloscopus collybita2'*' Ficedula parva Ficedula semitorquata'' Ficedula albicollis'' Ficedula hypoleucab~u Tichodroma muraria^ Lanius excubitor3'™'12 Lanius nubicus6"11 Calcarius lapponicus^ Plectrophenax nivalis* Emberiza cirlus*'& Emberiza hortulana^^ Emberiza caesia*3 Emberiza rustica}1 Emberiza pusillal2 Emberiza aureola**'1* Emberiza schoeniclus2'^

Postjuvenile moult restricted to part of the body-feathers or absent Postjuvenile moult may include secondaries •' Postjuvenile moult may include primaries, usually eccentrically 4 Postjuvenile moult may be complete * Postbreeding moult often suspended within primaries, completion in tropics. A few individuals postpone all primary moult until arrival in the tropics. 6 Postbreeding moult may be incomplete (usually some or all secondaries retained) 7 Postbreeding moult: some primaries may be retained " Prebreeding moult restricted to part of the body-feathers or completely absent " Prebreeding moult includes body-feathers, but only rarely wing-coverts and tertials 10 Prebreeding moult occurs only in some individuals 1 ' Prebreeding moult may include secondaries (in some species primaries) in 2y and ad (in Emberiza hortulana possibly only in ad) '" Prebreeding moult poorly known 13 No moult before autumn migration in ad and ly. Postjuvenile and postbreeding moult takes place in autumn at staging site. :

Riparia riparia1'2 Hirundo rustica12 Hirundo daurica2 Delichon urbica1'2 Cercotrichas galactotes'' Locustella naevia2'* Locustella fluviatilis** Locustella luscinioides" Acrocephalus paludicola Acrocephalus schoenobaenus Acrocephalus agricola Acrocephalus dumetorum* Acrocephalus palustris ^ Acrocephalus scirpaceus Acrocephalus arundinaceus2 Hippolais pallida Hippolais caligata

Hippolais olivetorum Hippolais icterina Hippolais polyglotta Sylvia communis icterops2 Sylvia borin2'17 Phylloscopus trochiloides Phylloscopus borealis Phylloscopus bonelli2^7 Phylloscopus sibilatrix2'1 Phylloscopus trochilus™ Muscicapa striata2'*'7 Oriolus oriolus2 )7 Lanius collurio2^ Lanius minor Lanius senator2* Emberiza melanocephala

1

Postjuvenile moult before autumn migration may rarely include innermost primaries. Postbreeding moult before autumn migration may include innermost primaries (very rarely eccentric primary moult). 3 Postbreeding moult before autumn migration may include secondaries. * Postbreeding moult before autumn migration may be complete. f Partial moult of body-feathers in NE Africa before complete moult further south in ad, 2

6 7

but not in ly.

Partial primary moult in NE Africa before complete moult further south in ad, but not in ly.

Prebreeding moult may be incomplete in 2y and ad: some secondaries retained. Prebreeding moult may be incomplete (probably in 2y): innermost primaries retained. ' Prebreeding moult usually incomplete in 2y, but not in ad: secondaries, primary coverts and innermost primaries retained. !0 Ad have a biannual complete moult (some secondaries may be retained during the post-breeding moult), ly/2y a partial postjuvenile moult and a complete prebreeding moult 1 ' Moult strategy variable and not fully understood. Ageing difficult. 8 J

60

Ageing European Passerines

summer

autumn

winter

spring/summer

autumn

First year of life

Adults

Fig. 54. Schematic presentation of the plumage cycle of moult cycle type 4: Species with a complete moult in the non-breeding area (see Fig. 43 for details). First-year birds retain large parts, occasionally all, of the juvenile plumage until the first complete moult in the winter quarters. After the postjuvenile moult, their plumage is composed of one or two feather generations, i.e. juvenile (upper bird) or juvenile and postjuvenile feathers (lower bird), which differ in age by several weeks only. After the postbreeding moult in the breeding area, the plumage of the adults is usually also composed of two feather generations, i.e. postbreeding and prebreeding feathers (lower bird), which differ in age by about six to nine months (Phylloscopus trochilus see text). Some adults do not moult before autumn migration and thus show a completely old plumage (upper bird). After the moults in the winter quarters, second-year birds and adults usually have the same plumage, i.e. composed of one feather generation, occasionally of two when having performed an additional partial prebreeding moult.' The complex moults of some species in the winter quarters are not shown (see text and Table 8).

CHAPTER 6

Species Accounts Presentation of the data and directions for use The species accounts present quantitative data on the extent of moult in 58 species from our own observations, as well as summarizing data from the literature. Ageing criteria based on moult are thus derived and supplemented with notes on other important ageing criteria. Colour photographs of extended wings illustrate the range of completed moult patterns and the corresponding plumage criteria which indicate a bird's age. Except for moults in tropical winter quarters, which may occur during early or late winter, we have not given data on the timing of moult. First, in most species the timing of the postbreeding/postjuvenile moult is broadly set by the timing of the breeding season and subsequent autumn migration (for more detailed information see Ginn & Melville 1983 as well as Glutz & Bauer 1985, 1988, 1991 and Cramp 1988, 1992, Cramp & Perrins 1993). Second, the timing of moult often varies considerably between individuals in a population and between populations. Consequently, general remarks on the timing of moult are of little help when ageing individual birds, since the state of the plumage must be determined by inspection for each individual (e.g. whether it has not yet moulted, is moulting or has completed moult).

Material Each species account presents a quantitative description of the extent of the postjuvenile, postbreeding and, if present, of the prebreeding moult. Only birds which had completed moult were included. Data were collected in summer/autumn (usually late July until the end of October) from live birds caught at several bird ringing stations in Switzerland: on the Alpine pass of Col de Bretolet (46°09'N/6°47'E) in 1980-1982 and 1988-1991 (and for a few species from additional years also); near Portalban (46°55'N/6°56'E) in 1987-1989; and near Grone (46°15'N/7°26'E) in 1984-1985. Some additional data were collected from other places in Switzerland and from Swiss birds in the collection of the Natural History Museum, Basel. Data from winter are mainly from the collection of the Natural History Museum, Basel. Spring data were collected at bird ringing stations near Iragna, Switzerland, (46°20'N/8°58'E) in 1990 and on Ventotene Island, Italy (40°48'N/13°25'E) in 1989 and 1991, as well as from the collection of the Natural History Museum, Basel. For a few species, skins from Europe in the British Museum (Natural History) were inspected (Anthus campestris, A. pratensis spring, M. cinerea spring, Luscinia s. $ve~ cica, Turdus t. torquatus, T. iliacus, Sylvia curruca spring, Musdcapa striata spring and adult autumn).

Relevance of the data As shown for many species (see section 4.4.3), the extent of a partial moult may vary according to a bird's geographical origin. In the case of predominantly sedentary species, the average extent of moult may be larger or smaller in areas other than in Switzerland. For migratory species with large sample sizes, our data probably describe the whole range of moult extent shown by birds from N Europe to south-central

Europe, since we mainly examined birds on migration. For migrants with small sample sizes, moult patterns which rarely occur in Switzerland may be missing. Since one might expect that most autumn migrants examined by us originate from central Europe, the quantitative data on moult extent are probably biased towards the average for that part of Europe. Generally, birds from N Europe can be expected to show a more limited average extent of moult than that found in our Swiss material, while birds in S Europe may show a more extensive moult and perhaps also as yet undocumented moult patterns. Data from the literature which give clues to any geographical variation are presented. However, since hatching date seems to be the predominant factor determining the extent of postjuvenile moult (see section 4 A3) considerable variation would be expected in all areas.

Presentation and analysis of the data For each species treated, the extent of moult is described by a series of summary statistics and, usually, a schematic wing diagram. For the analysis of the variation in moult extent, we took into account especially those factors which may be of interest to ringers. If sample sizes were large enough, the data have usually been analysed for differences between sites and the sexes and for seasonal trends. Differences between years could only be analysed with large data sets since there was usually an effect of season as well. However, differences between years were not usually detected; exceptions are given in the species texts. Summary statistics on the extent of moult: Generally, the proportion of birds which have moulted a given number of feathers of a tract after completion of moult is presented. If there is no mention of the bodyfeathers, they are usually all moulted. Within the wing and tail, only the feathers renewed are listed, those not mentioned have not apparently been replaced. For the number of greater coverts moulted, the range, arithmetic mean and mode (most frequent value) are given. Furthermore, the proportion of birds which have moulted none or all greater coverts is indicated, since this is of particular interest for ageing. The sample sizes are usually larger for greater coverts than for the other feather tracts, since we started our study on the moult of the greater coverts first and only included the other feather tracts of the wing in later years. However, data from any one feather tract was always collected over the entire migratory season, so avoiding any bias due to seasonal trends. Schematic wing diagrams: Hatchings indicate the proportion of individuals which have moulted a given feather or feather tract after completion of moult (for sample sizes see the summary statistics). This allows recognition of the variation in moult extent and the places where moult limits are to be expected. In the case of the marginal and median coverts, the entire feather tract is considered and the percentage given is reached by summing the percentage of birds which have moulted all the feathers of the tract plus half the percentage which have moulted part of the tract. The schematic wing diagrams refer to birds caught in Switzerland, thus exclude particular moult patterns occurring only in S Europe which are mentioned in the text.

62

Species Accounts

Graphs of the relationship between the extent of moult in different feather tracts: For those species with samples of sufficient size, the patterns of moult and their variability are shown by graphs showing the relationships between the moult extent of the various feather tracts of the wing. The number of greater coverts moulted is taken as a reference (x-axis) against which the percentage of individuals which have moulted one or more feathers of other tracts are shown. Only birds whose moult is completed and for which all feather tracts have been inspected are included. Graphs showing the seasonal variation of moult extent: In most species, there is a decrease in the extent of moult during autumn migration (see section 4.4.3). To save space, this trend is usually only shown for the extent of greater covert moult, although in many species it is also observed in other feather tracts. The data are usually grouped into fiveday periods (Table 9) or, in the case of small sample sizes, into ten-day periods or by month. Depending on the number of greater coverts moulted, the percentage of individuals which have moulted a given number of greater coverts or the mean number of greater coverts moulted after completion of moult is indicated. In the latter case, lines indicate the 10—90th percentiles and thus indicate 80% of the birds. If significant differences in moult extent between sexes have been observed, data are given for both sexes separately. Statistical tests: For all species with sufficient data, differences in the extent of moult between sites and changes with season have been tested for independence by chi-square tests. Usually, there was no site effect. In species whose sex can be determined, the extent of moult was analysed for the effects of sex and season by fitting a log-linear model to the three-way contingency table (sex, season, number of greater coverts moulted) to test the interactions for significance. An analogous procedure was used when testing for effects of site and season and when testing large data sets for differences between years, seasons and sex. Differences significant at the P = 0.05 level (two-tailed) have been called significant in the text.

Third, for birds before or during the postjuvenile/postbreeding moult the general criteria given in sections 5-3.1-5.3.4 can be used. For birds after the postjuvenile/postbreeding moult and after the prebreeding moult, those given in the species accounts apply. The sections 'Comments on ageing' give ageing criteria relative to moults, not seasons, since plumage criteria are dependent on moult and since the timing of moult is often variable. Use as many criteria as possible (for skull pneumatization see the appendix). The ageing criteria given apply in the first place to central European populations. According to the indications given on geographical variation in moult extent, the plumage criteria may be weighted and adapted to the birds to be examined. Care has been taken to explain the plumage ageing criteria for the whole range of moult extent, not just the most typical ones. However, be aware of the possibility that the extent of moult may be more or less than given in the species accounts, especially when working in Mediterranean or northern areas. Also be aware that due to early hatching dates, first-year birds may complete skull pneumatization in S Europe earlier than indicated by our Swiss material. Additional information may be obtained from part I of this book via the species index, which may be especially useful when examining a bird showing an unusual moult pattern. In this case, it may be useful to take a series of photographs, i.e. of the whole bird in the hand, of both entire wings and, if indicated, of parts of the wing (when spreading out the wing, place your fingers on the basal parts of the outer primaries, but not on the primary coverts; place all the feathers in their correct position; take the pictures against a background of medium shade, e.g. green vegetation). Table 9. Five-day periods according to Berthold (1973), grouped by decades (ten-day periods). Jan 1-5 Jan 6-10

May 1-5 May 6-10

Aug 29-Sep 2 Sep 3-7

Jan 11-1 5 Jan 16-20

May 11-1 5 May 16-20

Sep8-12 Sep 13-1 7

Jan 2 1-25 /an 26-30

May 21-25 May 26-30

Sep 18-22 Sep 23-27

Procedure of ageing

Jan31-Feb4 Feb 5-9

May 31-Jun4 Jun 5-9

Sep 28-Oct 2 Oct 3-7

This book is intended for users with a certain basic knowledge and as a complement to other ageing guides. For instance, the determination of the species is taken for granted. Ageing criteria other than plumage characters are not illustrated and the user is referred to other ageing guides, notably that by Svensson (1992). When ageing an individual of a given species, the following points should be borne in mind. First, when ageing on plumage characters, begin by understanding the moult cycle of the species concerned, which can be looked up in section 5.3. This provides the general plumage ageing criteria. We are convinced that it is vital to understand the moult cycle of a species when ageing on plumage characters and that the mere application of a set of rules may be misleading. Second, determine the state of the plumage of the bird, i.e. whether it is before, during or after the postjuvenile/postbreeding or the prebreeding moult. This may appear trivial in most cases, but is an important step which, in our experience of instructing ringers, may prevent some serious mistakes.

Feb 10-14 Feb 15-19

Jun 10-14 Jun 15-19

Oci 8-12 Oct 13-17

Feb 20-24 Feb25-Marl

Jun 20-24 Jun 25-29

Oct 18-22 Oct 23-27

Mar 2-6 Mar 7- 11

Jun 30-Jul 4 Jul 5-9

Oct 28-Nov 1 Nov 2-6

Mar 12-16 Mar 17-21

Jul 10-14 Jul 15-19

Nov 7-1 1 Nov 12-16

Mar 22-26 Mar 27-31

Jul 20-24 Jul 25-29

Nov 17-21 Nov 22-26

Apr 1-5 Apr 6-10

Jul30-Aug3 Aug4-8

Nov 27-Dec 1 Dec 2-6

Apr 11-1 5 Apr 16-20

Aug9-13 Aug 14-18

Dec 7- 11 Dec 12-1 6

Apr 21-25 Apr 26-30

Aug 19-23 Aug 24-28

Dec 17-21 Dec 22-26 Dec 27-31

Riparia

riparia

63

Riparia riparia Sand Martin Extent of postjuvenile moult The complete postjuv moult usually begins after arrival in the wintering area, but moult of body-feathers can start while still in Europe (Bub & Herroelen 1981, own obs.) and may continue during the first part of migration (own obs. on Col de Bretolet). Exceptionally, it includes some MaC, MeC and the innermost one or two GC. The only indications of ly starting primary moult in Europe are provided by Belman (in Mead 1980) from Spain and Greece. Moreover, four ly caught at a roost in N Italy during August/September showed renewed or growing P 1, P 1-2 or P 1-3 (F. Spina in lift.). Thus, central European ly generally migrate to Africa in complete or nearly complete juvenile plumage and the complete moult is interpreted as a complete postjuv moult in the winter quarters. Estimates of moult duration of ad and ly in Africa vary between 141 days in Ghana, 121 days in Uganda and Zambia (Mead 1980) and 135 days in Zambia (D.M. Francis in Ginn & Melville 1983). In Kenya and Uganda, moult lasts from about late October/November to mid March/mid April (Ginn & Melville 1983). At the end of February, Zambian birds had renewed 75% of the P, half of the S and 30% of the R. Only one bird out of 43 had already finished its flight feather moult (Loske&Ledererl988).

Fig. 55. ly in juv plumage, 19 September. Whole wing juv. T broadly fringed whitish. Smaller light fringes are present on GC, MeC and MaC.

Extent of postbreeding moult The complete postbr moult usually begins after arrival in the wintering area, but moult of body-feathers, wing-coverts, T, P, perhaps R and exceptionally S can start while still in Europe. In contrast to ly, a very small percentage of ad regularly starts P-moult in Europe. Such records are reported from NW Germany (two out of 450; Bub & Herreolen 1981), Britain (1.9%, N=3465, Mead 1980), Belgium (two records in Bub & Herroelen 1981), Sweden (one out often, Persson 1979), N Italy (14 ad and four unaged birds, F. Spina in lift,), Rumania (three out of 18, Csorgo 1992) and Switzerland (one record from Col de Bretolet). Those beginning P-moult in Britain (Mead 1980, N=79) usually had renewed P 1 or P 1—2 (85%), and the four most advanced birds P 1—4 and 1—5. These four birds and an ad with two renewed P had also started to replace S. About half of the birds with new P had also renewed some T, and one bird some R. The birds from Italy had mostly P 1—2, or at most P 1—3 growing or renewed. The high percentage (10%) of birds with new R in July in Bub & Herroelen (1981) probably includes accidental replacement and not regular moult. One bird from Col de Bretolet with renewed P 1—2 also had growing CC and A l l .

Fig. 56. Ad before postbr moult, 5 September. Innermost MaC and MeC, as well as GC 8 recently renewed. Rest of wing acquired during the last postbr moult in winter. T and all wing-coverts without light fringes.

Extent of prebreeding moult Whether the renewal of body-feathers starting in March described by Pearson (1971) indicates an inconspicuous prebr moult is still uncertain. Comments on ageing in summer and autumn ly: T, GC, MeC and MaC broadly fringed light rufous or whitish. Whole plumage fresh. Ad: GC, MeC, MaC and T without or with only very narrow light fringes. Plumage generally worn. After the complete moult in the winter quarters, 2y and ad are indistinguishable on plumage characters.

Fig. 57. 2y/ad after complete postjuv/postbr moult in winter, 25 April. Whole wing postjuv/postbr. After the complete moult in the winter quarters, ageing by plumage is no longer possible.

64

Hirundo rustica

Hirundo rustica Swallow Extent of postjuvenile moult ly and ad usually perform a complete moult in the winter quarters. Whether some of the body-feathers are renewed twice, as in Delichon urbica^ is not known. This complete moult in the winter quarters is interpreted as a complete postjuv moult in ly and a complete postbr moult in ad. Some ly start renewing body-feathers in Europe and, according to our observations on Col de Bretolet, continue doing so during the first part of migration. Rarely, some innermost MaC and MeC are also moulted. The only examples of ly starting P-moult north of the Mediterranean are two birds with growing or new P 1 from England (Mead 1975) and one bird with growing P 1 from Scotland (Cameron & Lynch 1983). Furthermore, at a roost in N Italy, four ly out of 6187 birds (ly and ad) were found with renewed or growing P 1 or P 1—2 in August/September (F. Spina in lift.). Compared with other small passerines, moult duration of ly and ad is very long, estimations varying between 4.5—5 (Broekhuysen & Brown 1963, Stresemann & Stresemann 1968c) and 6-6.5 (de Bont 1962, Kasparek 1976, Francis 1980) months. In Zambia (Francis 1980) and Zaire (Herroelen 1960, de Bont 1962) ly started moulting four to six weeks later than ad, whereas in South Africa there was no significant difference (Broekhuysen & Brown 1963). In all trans-saharan areas studied, many Swallows, presumably mostly 2y, were still moulting in April when northward migration had already started (summarized in Glutz & Bauer 1985). Hence, it is not surprising that some spring birds in Europe have been found with still growing outer P and/or R (von VietinghofT-Riesch 1955> Verheyen in Herroelen 1960, Loske 1984). One bird from Switzerland even interrupted wing-moult at P 8, retaining P 9 on 18 June (skin in Natural Hist. Museum, Basel). Two migrants from Ventotene, Italy from early May showed still growing (score 4) P 9, S 6 and R 6 (F. Spina in litt.).

score 6, some score 4 and 5); ly in early October usually show scores 1-4, some 5-7 (Winkler 1979). Thus scores 1-3 can safely be attributed to ly, scores 4—7 could be either ly or ad. After the complete moult in winter, 2y and ad are indistinguishable on plumage characters.

Fig. 58. ly in juvenile plumage, 19 September. Whole wing juv. Only MaC with some blue gloss. Rest of wing dark brown.

Extent of postbreeding moult Usually a complete moult in the winter quarters (see above). Moult of body-feathers can start in Europe and continue during the first part of migration; it may also include some MaC and MeC and, mainly in birds with P-moult, T 8. In contrast to ly, a small number of ad regularly begin P-moult in Europe. The percentages reported are 2.8% for Switzerland (Winkler 1975), 2.6% for Belgium (M. de Haen & P. Herroelen in Glutz & Bauer 1985), around 10% for S Germany (estimate, Kasparek 1976) and 19% for Spain (Pimm 1970) as well as 56 ad and two unaged birds out of 6187 (1 y and ad) caught in August/September at a roost in N Italy (F. Spina in lift.). Most of the birds examined had only P 1 or P 1-2 growing or new, the most advanced, in Spain, showed renewed P 1-4. One bird from Switzerland with renewed P 1 —2 had also moulted S 1 on one wing.

Fig. 59. Ad 9 before postbr moult, 2 October. GC 5 recently replaced. Rest of wing acquired during the last postbr moult in winter. Although bearing about half a year's wear, the wing is distinctly more glossy than in ly. GC 8—10 seem to be newer than 1—4 and 6—7, but show the same degree of abrasion as the adjacent 6-7.

Comments on ageing in summer and autumn ly: Upperparts and wing-coverts dark brownish-blue, only faintly glossy. Forehead and throat orange. Length of outermost tail-feather 60-75 mm, no overlap with ad 9 . Ad: Upperparts and wing-coverts glossy metallic-blue with a violet tinge. Forehead and throat mahogany. Length of outermost tailfeather over 75 mm. Skull pneumatization is of limited value as an ageing criterion in autumn. Many ad (29%) do not complete skull pneumatization (mainly

Fig. 60. 2y/ad 9 after complete postjuv/postbr moult in winter, 16 April. Whole wing postjuv/postbr. After the complete moult in the winter quarters, ageing by plumage is no longer possible. Fresh wing-coverts with prominent metallic gloss.

Delichon urbica

Delichon urbica House Martin Extent of postjuvenile moult The complete moult usually starts in Europe with the replacement of body-feathers, and exceptionally P. Generally remiges are only renewed after arrival in the winter quarters. In Europe, moult of body-feathers is frequent and, according to our observations on Col de Bretolet, can continue during at least the first part of autumn migration. It does not apparently include wing-coverts and is best observed on the rump, where the white feathers of the juv plumage are replaced by light brown ones. Of 75 birds caught around mid September on Col de Bretolet, 12% had changed at least two thirds or all rump feathers, 53% showed some new rump feathers, 29% had moulted some feathers of other body tracts and 5% had not yet started postjuv moult. The only indications of ly starting P-moult in Europe come from Switzerland (five out of 3705 with new P 1, Winkler 1975), Portugal (two out of seven with growing P 1-2, Mead 1975) and S Spain (Hill 1992). The few moult data from Africa indicate a flight feather moult lasting from November to April (South Africa). In November about 60% (N=107, mostly ly) had started P-moult in Transvaal, the most advanced birds showing growing P 5—6 (Skead & Skead 1970). By mid April near Cape Town, 21 out of 52 birds (not aged) were in the last stages of P-moult, and 41 had still growing outer tail-feathers (Broekhuysenl953).

65

Comments on ageing in summer and autumn ly: T broadly tipped white. Crown without metallic gloss. Ad: T without or with only very faint white tips. Crown with metallic gloss. Skull pneumatization is of limited value as an ageing criterion in autumn. Many ad (43%) do not complete skull pneumatization (mainly score 6, some score 4 and 5); ly in early October usually show scores 1-3, some 4-7 (Winkler 1979). Thus scores 1-3 can safely be attributed to ly, scores 4—7 could be either ly or ad. After the complete moult, 2y and ad birds are indistinguishable on plumage characters.

Fig. 61. ly in juv plumage, 10 September. Whole wing juv. T with distinct white tips.

Extent of postbreeding moult Similar to ly, the complete moult usually starts in Europe with the replacement of body-feathers, rarely P and T and, exceptionally, S and GC. Remiges are generally only renewed after arrival in the winter quarters. The beginning of P and T-moult was recorded in 45 (5%) of 948 ad in Switzerland. Ten of these birds had only replaced T 8, the others had renewed or growing P 1 or P 1-2, occasionally together with T 8, and the most advanced bird had moulted P l^i, S 1 and all T (Winkler 1975). In Portugal three out of five ad House Martins were found with renewed P 1 or 1-2 (Mead 1975, Mead & Watmough 1976). From S Spain, Hill (1992) reported that probably all ad renewed P partly, rarely all (usually three to six, range one to nine), many the T and some S; usually only one P was growing at the time, rarely two. The majority was still in active moult of remiges when departing. A migrant from 16 September on Col de Bretolet with a new P 1, growing P 2, new T 8 and GC 7 and some growing body-feathers also demonstrates that moult can continue at least into the first part of autumn migration.

Fig. 62. Ad before postbr moult, 18 September. Whole wing acquired during the last postbr moult in winter. T without white tips.

Extent of prebreeding moult: 2y and ad The prebr moult seems to coincide with the last stages of the postbr/postjuv flight feather moult. This is suggested by the observations of Broekhuysen (1953) near Cape Town who found moult of body-feathers in 94% of 52 birds having finished or nearly finished their flight feather moult. In comparison, only 8.6% of the Swallows examined at the same time showed moult of body-feathers. In 43 of the 49 House Martins with growing body-feathers, moult extended over the whole body, indicating that the prebr moult may comprise all body-feathers. Whether it also includes other feather tracts is not known. The prebr moult results in a change of the rump colour from light brown to white.

Fig. 63. 2y/ad after complete postjuv/postbr moult, 22 April. Whole wing postjuv/postbr. After the complete moult m the winter quarters, ageing is no longer possible. Both age groups may show faint white tips on T. In this bird, P 1-2 and PC 1-2 are more bleached than the rest of the wing. These feathers have most probably been already renewed in Europe, before autumn migration.

66

Anthus campestris

Anthus campestris Tawny Pipit Extent of postjuvenile moult Body-feathers: only partially moulted (cf. Natorp 1925). Some ly migrate in almost complete juv plumage, others moult up to three quarters of the body-feathers. MaC: about half of the ly moult some, the rest moult none. MeC: about one third of the ly moult part, the others moult none, GC: usually not moulted. Only one out of 18 ly had GC 9 renewed. Extent of postbreeding moult The postbr moult was studied by Stresemann & Stresemann (1968a) and Kriiger (1989). Presumably, most northern and central European ad suspend moult of the remiges before autumn migration (Natorp 1925, Kriiger 1989) and resume it in the wintering area, sometimes as early as October/November (Stresemann & Stresemann 1968a). Other birds, however, complete the postbr moult in the breeding area (Stresemann & Stresemann 1968a). These are probably birds from more southern populations (Roselaar in Cramp 1988). There are also indications that some birds may moult all remiges in the wintering area (Kriiger 1989, Svensson 1992). All eight migrants studied in Switzerland showed suspended moult. The bird with the least extensive moult only renewed P 1, T 8 and R 1, the bird in the most advanced stage P 1—5, T 7—9 and R 1+6. One bird with four P and all T moulted had also renewed S 6. Otherwise, no S were moulted. All birds renewed the body-feathers completely, or almost so, and most of the MaC. Only one bird renewed all MeC, the others only a few. One bird renewed nine, another seven, the others only a few inner GC.

moult. Moult limits in spring birds (Fig. 71) show that MaC, MeC and GC are moulted at least partially during the prebr moult (N=4). In 2y, as in ad, it is often impossible to judge in spring whether feathers have been renewed during a possible resumption of the postjuv moult in the winter quarters or during the prebr moult. If all feathers moulted in the wintering area are taken together, a very extensive partial 'winter' moult results comprising all MaC, MeC and GC (only one 2y had retained juv GC 1, Fig. 66), sometimes CC and Al, always all T and R, mostly S 6 (one 2y S 1-2+5-6, Fig. 67) and in two birds eccentrically one or two P (one bird P 7, one P 5+7, Fig. 68 and 69) (N=6). Comments on ageing after prebreeding moult In spring, ageing is difficult and not always possible. Moult limits due to the prebr moult occur in 2y and ad. Often, 2y can be distinguished from ad by having more bleached remiges and by the shape and coloration of the PC. The inner PC of 2y typically are more bleached and pointed and have a distinct white terminal fringe (Fig. 66). Those of ad are darker, more rounded and have ill-defined brownish or buffish fringes (Fig. 70 and 71).

Comments on ageing after postjuvenile and postbreeding moult in the breeding area Best criteria: In late summer and autumn, ageing in the breeding area is easy, since ly always show many juv body-feathers and wing-coverts. ly: All ly retain juv body-feathers which give them a scaly appearance on the upper side. These juv body-feathers are fringed whitish; postjuv ones have no distinct fringes. Juv MaC, MeC, GC and T are fringed whitish (Fig. 64). The breast of ly is usually more prominently spotted dark-brown than in ad, but this is not an infallible ageing criterion.

Fig. 64. ly after partial postjuv moult, 31 August. Most proximal MaC partially postjuv. Rest of wing juv. The well defined whitish fringes on MeC and GC are typical of ly.

Ad: No body-feathers with distinct light fringes present on upperparts. Most northern and central European ad show suspended P-moult during autumn migration which is easily recognized (Fig. 65). They often show pronounced moult limits within T and GC. The GC not moulted during the postbr moult are much more worn than the juv GC (Fig. 65). After a complete postbr moult, ad are recognizable by the fresh plumage and by having MeC, GC and T broadly fringed brownish not narrowly whitish. Ad before the postbr moult and ad which have not started to moult remiges are easily recognized by very worn and bleached P, T and wing-coverts. Resumption of postbreeding moult and extent of prebreeding moult Ad resume a suspended moult of remiges after arrival in the winter quarters or perhaps at stopover sites en route, In spring, it is often impossible to distinguish between remiges renewed before autumn migration and remiges renewed in the wintering area; likewise, it is difficult to judge whether wing-coverts, T and R have been moulted when resuming postbr moult in the winter quarters or later during the prebr

Fig. 65. Ad in suspended postbr moult, 31 August. Before suspension, most

MaC, MeC 2+4-6+8, GC 8-10, CC, T 8-9, P 1-4 and PC 1-3 have been moulted. Easily recognized as ad by the suspended P-moult.

Anthus campestris

67

Fig. 66. 2y after partial prebr moult, 20 April. MaC and MeC prebr. GC 1 juv, 2-10 prebr. T prebr. Rest of wing juv. Juv GC 1, inner PC and CC with welldefined whitish terminal fringes, indicating 2y.

Fig. 69. 2y after eccentric partial prebr moult, 11 May. MaC partially prebr, partially postjuv. MeC and GC prebr. CC juv. Al prebr. T and S 6 prebr. P 5+7 prebr. Rest of wing juv. PC and CC bleached with well-defined whitish terminal fringes. T and GC look older than S 6 and P 5+7, but are not juv.

Fig. 67. 2y after partial prebr moult, 30 April. MaC mostly prebr, some postjuv in the undermost row. MeC and GC prebr. CC and Al prebr. T as well as S 1-2 and 5-6 prebr. PC 1 prebr. Rest of wing juv. Recognizable as 2y by the juv PC with well-defined whitish terminal fringes. T look older than the renewed S, but not as old as the juv S 3^; they may have been renewed early in winter, and are also more exposed. All P have the same pattern of growth bars, confirming 2y.

Fig. 70. Ad after partial prebr moult, 23 April. Postbr and prebr feathers are not distinguishable. Recognizable as ad by the ill-defined brownish terminal fringes on the inner PC. P 1—5 are probably older than P 6-r9 and may have been renewed before autumn migration (suspended P-moult). S 1—6 are darker than the innermost P and may have been renewed in the wintering area together with P6-9.

Fig. 68. 2y after eccentric partial prebr moult, 5 May. MaC, MeC and GC prebr. T and S 6 prebr. P 7 prebr. Rest of wing juv. The well-defined whitish terminal fringes of the bleached CC and inner PC are diagnostic of 2y. T look older than S 6 and P 7, but are not juv.

Fig. 71. Ad after partial prebr moult, 13 May. GC 3-4+6+9 probably postbr, rest prebr. The other feathers are difficult to assign to feather generations. Recognizable as ad by the ill-defined buffish terminal fringes on the inner PC.

68

Anthus trivialis

Anthus trivialis Tree Pipit Extent of postjuvenile moult MaC: 20% moult all MaC, 80% leave some, predominantly in the undermost row, unmoulted (N=60). MeC: one third moult part or all MeC, two thirds moult none (N-60). GC: range 0-4, mean 0.3, mode 0, no GC 81.4% (N=775). T: none 87.6%, one 5.7% (mostly T 8 or T 9), two 6.3% (T 8-9), three 0.2% (N=526) (see p. 33). R: none 96.7%, one 3.3% (R 1) (N=520). Birds which renew T and R, but do not moult GC are rare (Fig. 73). The extent of postjuv moult decreases as the autumn migratory season proceeds (Fig. 74). In NW Russia, MeC and the undermost row of MaC are usually only partly moulted. Only five out of 96 ly replaced one or two GC and one had renewed T (Rymkevich 1990). Extent of postbreeding moult Usually whole plumage. Occasionally, interruption of P-moult at an early stage (after P 1-2 have been moulted) occurs (van Hecke 1980, Fig. 21) and may be related to breeding (see section 3.4.1). These birds are likely to complete the postbr moult before migration. Exceptionally, some birds interrupt moult before migration. Such birds may show unmoulted Al 1, Al 1-2 (Fig, 80), PC 9 and/or P 10; one ad was found to have retained S 6 and Al 1, another S 3—6 and Al 1. This latter bird apparently moulted S 5—6 of the right wing later during the prebr moult, but not on the left wing (Fig. 84). One migrant on Fair Isle had unmoulted S 4-6 and Al 1-2 (Riddiford 1990) and two out of 33 in Sudan had partially unmoulted S (Nikolaus & Pearson 1991). Comments on ageing after postjuvenile and postbreeding moult Best criteria: Skull pneumatization until at least the end of October (p. 204). Moult limits within MaC, MeC, GC, T and R. Plumage characters sometimes difficult to apply. Fig. 73.Relationships between the number of postjuv GC and the percentage of individuals with renewed T and R in 1 y Anthus trivialis which have completed their postjuv moult.

6

5

4

3

2

1

1

2

3

4

5

6

7

0 10 30 60 90 100%

Fig. 72. Extent of postjuv moult on the wing and tail in ly Anthus trivialis.

ly: 80% of ly show no moult limit within GC. Most of them can be identified by moult limits within MaC or MeC. Postjuv MaC and MeC have darker fringes and feather centres and are less worn than juv ones (Fig. 76 and 77). Very fresh juv GC (usually before mid-August) are not yet worn and bleached and are virtually identical in colour to the postbr GC of ad. ly with a moult limit within GC or T are easily identifiable. Fringes and feather centres of renewed GC and T are darker and contrast with the lighter and already worn fringes of the juv GC and T (Fig. 77 and 78). R cannot be used for ageing unless showing a moult limit (central pair renewed, rest juv)* Ad: Whole plumage fresh, fringes of GC and T not worn, no moult limits within MaC, MeC and T. Note that the innermost three to four GC have somewhat darker fringes than the outer GC and may simulate a moult limit (Fig. 79).

Extent of prebreeding moult: 2y and ad MaC: 32% moult no MaC, 61% part and 7% all (N=137). MeC: 8.7% moult no MeC, 63.1% one to seven and 28.2% all (N=149). 2y moult significantly more MeC (mean 6.0, N=54) than ad (mean 5.2, N=35). GC: range 0-10, mean 4.8, mode 5, no GC 0.5%, all GC 1.6% (N=191). 2y moult significantly more GC (mean 5.0, N=83) than ad (mean 4.4, N=47). T: none 2.1%, one 3.2%, two 2.7%, three 92.0% (N=188). R: none 12.0%, one 58.7% (mostly R 1), two 17.9% (mostly R 1-2), three to six 11.4% (N=184) (see p. 16). P and S: one bird with all T moulted had renewed S 6, one bird S 5^6 (Fig. 84). Ludlow (1966) mentions one bird from Nigeria with P 8 growing on both wings (eccentric P-moult).

Comments on ageing after prebreeding moult Fig. 74. Percentage of ly Anthus trivialis with at least one postjuv GC during autumn (data grouped in five-day periods; the first value includes the period 9-18 August, the last 28 September-17 October).

Moult limits due to the prebr moult occur in ad and 2y. The prebr moult is much more extensive than the postjuv moult. Therefore, three feather generations within GC of 2y are unlikely. Coloration, wear and bleaching of the renewed and unmoulted feathers are sometimes very similar between 2y and ad. Thus, even with experience, ageing every bird in spring is impossible. Usually, P and S of 2y are more bleached and have a more brownish tinge than in ad. The fringes of juv GC are often more worn than those of the postbr GC of ad (cf. Fig. 82 and 83).

Anthus trivialis

69

Fig. 75. Extent of prebr moult on the wing and tail in ad and 2y Anthus trivialis. For differences between ad and 2y see text. Fig. 78. ly after partial postjuv moult, 9 September. MaC postjuv, MeC postjuv. GC 1—7+10 juv, 8—9 postjuv. T 7 juv, 8—9 postjuv. Rest of wing juv. The fringes of the renewed GC and T are less worn and darker than the juv feathers in the corresponding tract.

Fig, 76, 1 y after partial postjuv moult, 16 August. MaC postjuv except one in the undermost row. MeC 1-2 postjuv, 3-8 juv. GC, T and rest of wing juv. The moult limits within MaC and MeC are diagnostic of ly. All GC and T are slightly and uniformly worn.

Fig. 79. Ad after complete postbr moult, 7 September. Whole wing postbr. Whole wing uniformly fresh, MaC, MeC, GC and T almost intact. Note the change in colour within the inner GC which may simulate a moult limit.

Fig. 77. ly after partial postjuv moult, 26 August. MaC postjuv except four juv ones in the undermost row. MeC 4-5+7 juv, rest postjuv. GC 1-8+10 juv, 9 postjuv. T and rest of wing juv. The postjuv GC 9 is fresh, longer and has a darker feather centre than the juv GC. Moult limits within MaC and MeC recognizable by the bleached and worn juv feathers.

Fig, 80. Ad after complete postbr moult, 9 September. Whole wing postbr except Al 1-2. Exceptionally, Al 1—2 remain unmoulted and are heavily bleached. GC 8 is shorter and more worn than the other GC. It may have been lost accidentally and replaced before the ordinary, complete moult.

70

Anthus trivialis Fig. 83. Ad after partial prebr moult, 20 April. MaC partly prebr, partly postbr. MeC 2+8 postbr, rest prebr. GC 1-6+10 postbr, 7-9 prebr. T prebr. Rest of wing postbr. Recognizable as ad because the fringes of the postbr GC are less worn than those of the juv GC in Fig. 82.

Fig. 81. 2y after partial prebr moult, 22 April. MaC mostly postjuv. MeC 1 postjuv, 2-8 prebr. GC 1-4 juv, 5-10 prebr. T prebr. Rest of wing juv. P and S of 2y are usually slightly more bleached and browner than those of ad (cf. Fig. 83). Especially, S 6 contrasts more with the prebr T than in ad.

Fig. 82. 2y after partial prebr moult, 28 April. MaC mostly postjuv. MeC 1-2 postjuv, 3-8 prebr. GC 1-6+10 juv, 7-9 prebr. T prebr. Rest of wing juv. The extent of prebr moult of this 2y bird is similar to that of the ad bird in Fig. 83. Identifiable as 2y because the fringes of the juv GC are more worn than those of the postbr GC in Fig. 83.

Fig. 84. Ad after partial prebr moult, complete postbr moult interrupted, 19 April. MaC mostly prebr. MeC 1-2 prebr, 3-8 postbr. GC 1-7+10 postbr, 8-9 prebr. T 7-8 prebr, 9 postbr. Al 1 not moulted during the last postbr moult, Al 2-3 postbr. S 1-2 postbr, S 3-4 not moulted during the last postbr moult, S 5—6 prebr. This exceptional bird has neither moulted Al 1 nor S 3-6 during the last postbr moult. During the prebr moult, S 5-6 have been renewed on the right, but not on the left wing (not shown).

Anthus pratensis

71

Anthus pratensis Meadow Pipit Extent of postjuvenile moult MaC: 33% moult all MaC, 67% leave some MaC unmoulted, usually in the undermost row (N=52). MeC: 68% moult none, 24% part and 8% all MeC (N=78). GC: range 0-6, mean 0.5, mode 0, no GC 70.5% (N=373). T: none 71.0%, one 8.9%, two 3.8%, three 16.3% (N=338) (see p. 33). R: none 81.5%, one 16.4% (usually R 1), two 1.5% (usually R 1+6), six 0.6% (N=336) (see p. 34). If no GC are moulted, T and R are rarely renewed (Fig. 86). The extent of postjuv moult is significantly more extensive at the beginning than at the end of the migratory season: before mid-October, 61.3% (N=80) of ly have renewed at least one GC, 60.8% (N=74) at least one T and 39.7% (N=73) at least one R; after mid-October, the respective proportions are 20.7% (N=290), 20.3% (N=261) and 12.7% (N=260). In SW Niedersachsen, at least 52% of actively moulting ly renewed one or more T and 40% one or more R (N=151, Hotker in Glutz & Bauer 1985). This corresponds to the values of the first part of the migratory season in Switzerland. On the island of Mellum (North Sea), 18.3% of ly moult at least one R and only a few ly one or more T (N=60, Henle 1983). Svensson (1992) noted that early-hatched and southern populations are more likely to moult MeC, GC and T than late-hatched and northern populations.

Fig. 85. Extent of postjuv moult on the wing and tail in ly Anthus pratensis,

Extent of prebreeding moult: 2y and ad MaC: not or only partially moulted (N=36). MeC: 75% moult none, 25% one to all MeC (N=72). GC: range 0-5, mean 0.8, mode 0, no GC 49.3% (N=75). T: none 8.1%, one 20.3%, two 14.9%, three 56.8% (N=74) (see p. 15). R: none 32.9%, one 48.6% (usually Rl), two 5.7%, three 2.9%, four 2.9%, five 2.9%, six 4.3% (N=70) (see p. 16).

Extent of postbreeding moult Whole plumage. Rarely, individual PC and Al may be retained. Comments on ageing after postjuvenile and postbreeding moult Best criteria: Skull pneumatization until the end of September (p. 204). ly usually identifiable by moult limits within MaC, MeC and GC ly: ly without a moult limit within GC (68%) are usually recognizable by moult limits within MeC or MaC (Fig. 88 and 89). ly with moult limits within GC are easily identifiable. Postjuv GC contrast with juv GC in their dark, olive-brown and intact fringe (Fig. 89 and 90). Also check for moult limits within T and between postjuv T and juv S 6. Ad: All MaC, MeC, GC and T fresh. The fringes of GC have a more olive tinge and contrast more with the dark feather centre than in ly, although there is some individual variation. The difference in colour between T and S 6 is only slight.

Fig. 86. Relationships between the number of postjuv GC and the percentage of individuals with renewed T and R in ly Anthus pratensis which have completed their postjuv moult.

Fig. 87. Extent of prebr moult on the wing and tail in ad and 2y Anthus pratensis.

Comments on ageing after prebreeding moult Moult limits due to the prebr moult occur in ad and 2y. Since coloration and wear of the feather generations acquired before the prebr moult are very similar in 2y and ad, ageing in spring is only possible in a few cases and by an experienced observer. Usually, P, S and the fringes of unmoulted GC are more bleached in 2y than in ad.

72

Anthus pratensis

Fig. 88. ly after partial postjuv moult, 3 October. MaC partly juv, partly postjuv. MeC juv. GC, T and rest of wing juv. Recognizable as ly by the mixed MaC. The postjuv MaC have fresher fringes and darker feather centres than the juv MaC. GC and T slightly and uniformly worn. The fringes of juv GC are usually whitish and gradually fade into the dark colour of the feather centres.

Fig. 91. Ad after complete postbr moult, 20 October. Whole wing postbr. MaC, MeC, GC and T virtually intact. Fringes of GC have a more olive tinge and contrast slightly more with the dark feather centres than in juv GC.

Fig. 89. ly after partial postjuv moult, 19 October. MaC postjuv. MeC 1-4 juv, 5-8 postjuv. GC 1—8+10 juv, 9 postjuv. T 7—9 postjuv. Rest of wing juv. Easily recognizable as ly by the postjuv GC 9 which is darker, fringed more olive and less worn than the juv GC. Moult limit between T and S 6.

Fig. 92. 2y after partial prebr moult, 14 April. MaC partly juv, partly postjuv, partly prebr. MeC 1-4+8 juv or postjuv, 5-7 prebr. GC 1-7+10 juv, 8-9 prebr. T 7+9 prebr, 8 probably postjuv. Rest of wing juv. P, S and the fringes of the juv GC are more bleached than in ad.

Fig. 90. ly after partial postjuv moult, 6 October. MaC and MeC postjuv. GC 1-6 juv, 7-10 postjuv. T postjuv. Rest of wing juv. The postjuv GC are less worn and contrast with the juv GC in their broad olive-brown fringes. Postjuv T darker than juv S 6.

Fig. 93. Ad after partial prebr moult, 7 April. MaC mostly postbr. MeC postbr. GC 1—8+10 postbr, 9 prebr. T prebr. Rest of wing postbr. P and S slightly darker and fringes of postbr GC slightly less bleached than in 2y.

Anthus spinoletta spinoletta

73

Anthus spinoletta spinoletta Water Pipit Extent of postjuvenile moult MaC: about 50% moult all and 50% retain some juv MaC in the undermost row. MeC: 50% moult none, 45% part, 5% all MeC. GC: range 0-2, mean 0.5, mode 0, no GC 59.4% (N=160). T: none 71.3%, one 23.1%, two 3-5%, three 2.1% (N=143) (see p. 33). R: none 95.8%, one 4.2% (R1)(N=144). Most ly which renew T and all ly which renew R moult at least one GC (Fig. 95). During the first part of the autumn migration season (until the end of September), 50% of ly had renewed one to two GC (N=l 11) and 34% (N=97) one to three T; thereafter, the percentages were significantly lower (18% and 18%, respectively, N=49).

Fig. 94. Extent of postjuv moult on the wing and tail in ly Anthus s. spinoletta,

Whole plumage. Exceptionally, Al 1 may remain unmoulted (one out of 49 ad).

N=102) than ad (mean 1.7, N=59, difference not significant). T: one 1.3%, two 5.8%, three 92.9% (N=224). Significantly more 2y moult all T (96.0%, N=101) than ad (85.2%, N=54). R: none 4.5%, one 87.9% (R 1), two 7.6% (usually R 1+2) (N=224) (see p. 16). S: two birds with all T moulted had also renewed S 6 on one wing. This was also observed in one bird by Herremans (1987).

Comments on ageing after postjuvenile and postbreeding moult

The extent of the prebr moult of spring migrants in Belgium is very similar to our data (Herremans 1987).

Extent of postbreeding moult

Best criteria: Skull pneumatization until at least November (p. 204). Moult limits within MeC, GC and T. Plumage criteria sometimes difficult to apply and not always conclusive. ly: It is important to realize that the inner juv GC as well as the inner postbr GC of ad are differently coloured than the outer GC of the same feather generation and may simulate a moult limit (Fig. 97 and 100). Furthermore, very fresh juv GC (usually birds in juvenile plumage) have brownish fringes like the GC of ad after postbr moult (cf. Fig. 97 and 100). Juv GC, however, bleach more rapidly than ad GC. In order to recognize moult limits within GC (40% of ly), check especially for differences in colour of the feather centre and wear, ly without moult limits in GC often have a moult limit within MeC (Fig. 98). Ad: No moult limits within MeC, GC and T. Fringes of GC contrast more strongly with the dark feather centres than in ly. Fringes on outer GC are usually brownish or reddish-brown, not whitish, although individual differences and bleaching may give them a more juv appearance. Fig. 96. Extent of prebr moult on the wing and tail in ad and 2y Anthus s. spinoletta. For differences between ad and 2y see text.

Extent of prebreeding moult: 2y and ad MaC: not moulted (N=35). MeC: 55% moult one to six central MeC, 45% none (N=35). At least one MeC is significantly more often moulted by 2y (68%, N=19) than byad(27°/o,N=ll). GC: range 0—6, mean 1,8, mode 2, no GC 7.1%, GC 10 is moulted by 4.9% only (N=224). 2y birds moult slightly more GC (mean 1.9,

Fig. 95. Relationships between the number of postjuv GC and the percentage of individuals with renewed T and R in ly Anthus s, spinoletta which have completed their postjuv moult.

Comments on ageing after prebreeding moult Ageing in spring is very difficult, even by experienced observers. Moult limits due to the prebr moult occur in both 2y and ad. The prebr moult is much more extensive than the postjuv moult. Therefore, three feather generations within GC (Fig. 102) are very rare (0.5%). Usually, S and P of ad are slightly darker and contrast slightly less with the prebr T than in 2y. Postbr GC of ad are usually slightly less bleached than the juv GC of 2y. Only birds with two moult limits within GC (0.5%) are safely identifiable as 2y (Fig. 102).

74

Anthus spinoletta spinoletta

Fig. 97* ly in juvenile plumage, 1 August. Whole wing juv. Note that the fringes of the juv GC 8—9 are browner than those of the adjacent GC and may simulate a moult limit.

Fig. 100. Ad after complete postbr moult, 5 October. Whole wing postbr. Fringes of GC contrast strongly with the feather centres. Fringes of outer GC brownish, not whitish. Note that the four innermost GC have darker fringes than the others which may simulate a moult limit.

Fig. 98. ly after partial postjuv moult, 1 October. MaC postjuv, a few outermost juv. MeC 1-3+8 juv, 4-7 postjuv. GC 1—8+10 juv, 9 postjuv. T and rest of wing juv, The moult limits within MeC and GC are diagnostic of ly. The renewed MeC and GC 9 have darker black feather centres and totally intact, browner fringes than the juv GC and MeC which have lighter feather centres and worn margins.

Fig. 101. 2y after partial prebr moult, 7 April. MaC postjuv. MeC 3+8 prebr, 1-2+4-7 postjuv. GC 1-7+10 juv, 8-9 prebr. T prebr. Rest of wing juv. Compared with the corresponding postbr feathers of ad (cf. Fig. 103), the juv GC, S and P are more bleached, especially when comparing S 6 with the dark prebr T.

Fig. 99. ly after partial postjuv moult, 10 October. MaC and MeC postjuv. GC 1-7+10 juv, 8-9 postjuv. T 7-9 postjuv. Rest of wing juv. The renewed GC have darker feather centres than juv GC and clearly defined fringes. A conspicuous moult limit is visible between the postjuv T and the juv S.

Fig. 102. 2y after partial prebr moult, 14 April. MaC postjuv. MeC postjuv. GC 1—7 juv, 8 postjuv, 9 prebr, 10 juv or postjuv. Rest of wing juv. Identifiable as 2y by the three feather generations within the GC. The postjuv GC 8 is darker, especially near the tip, and less worn than the juv GC 7, but not as dark and intact as the prebr GC 9.

Anthus spinoletta spinoletta

75

Fig. 104. Ad after partial prebr moult, 6 April. MaC mostly postbr. MeC 6+8 prebr, rest postbr. GC 1—4 postbr, 5—10 prebr. T prebr. Rest of wing postbr. Rare case of an extensive prebr moult.

Fig. 103. Ad after partial prebr moult, 17 April. MaC postbr. MeC 5+6 prebr, rest postbr. GC 1—7+10 postbr, 8—9 prebr. T prebr. Rest of wing postbr. The postbr GC have darker feather centres than the juv GC of 2y birds (cf. Fig. 101 and 102). S and P are also darker than in 2y. S 6 contrasts less with the prebr T than in 2y.

76

Mo tacilia fla va

Motacillaflava Yellow Wagtail Extent of postjuvenile moult MaC and MeC: usually all. In about 10% (almost exclusively ly without renewed GC), individual MaC of the undermost row and MeC (rarely all) may remain unmoulted (N=79). GC: range 0-10, mean 1.9, mode 0, no GC 48.2%, all GC 0.7% (N=876). CC: 1.6%. Al: none 97.9%, one 2.1% (N=610). T: none 90.6%, one 4.2%, two 4.7%, three 0.5% (N=832) (see p. 33). R: none 90.1%, one 5.2% (usually R 1), two 2.9% (often R 1+6), three to five 1.8% (N=829) (seep. 34). The extent of postjuv moult is correlated among GC, CC, Al, T and R (Fig. 106). The extent of postjuv moult decreases as the autumn migratory season proceeds (Fig. 107). Scandinavian birds perform a considerably less extensive postjuv moult than that recorded in NW Russia, Switzerland and England: less than 5% moult at least one GC and only a small proportion moult MeC (Svensson 1992). In NW Russia (Rymkevich 1990), 50% moult MeC and 28% one to four GC (mean of all ly 0.5, N=133). In England (Hereward 1979), 52% of ly were recorded with at least one renewed GC and the frequency distribution of the number of moulted GC is similar to that in Switzerland, though the percentage of birds with all GC moulted (6.7%) is higher than in Switzerland (0.7%). In England, the number of GC moulted decreases during late summer, apparently due to the later appearance of late-hatched birds moulting fewer GC (Hereward 1979). In E Germany (Dittberner & Dittberner 1987), about 30% of ly were recorded as moulting a few or all R, some GC and some T; strangely, moult of PC was recorded (no details given) and one bird, apparently aged as ly, had P 1—2 growing.

Fig. 106. Relationships between the number of postjuv GC and the percentage of individuals with renewed CC, Al, T and R in ly Motacillaflava which have completed their postjuv moult.

Fig. 107. Mean number of postjuv GC during autumn of ly Motacilla flava which have completed the postjuv moult (data grouped in five-day periods; the first value includes the period 14-28 August, the last 3— 17 October).

Fig. 105. Extent of postjuv moult on the wing and tail in ly Motacillaflava,

Extent of postbreeding moult Whole plumage. Moult during postbreeding movements or the beginning of autumn migration was observed in Great Britain and Germany (Hereward 1979, Dittberner & Dittberner 1987). Exceptionally, moult interruption occurs. Among 98 ad caught on Col de Bretolet, one had S 6, Al 2-3, GC 10, MeC and some MaC (Fig. 118) and two birds Al 1 unmoulted (Fig. 117), the Al being among the last feathers on the wing to be renewed (Spina & Massi 1992). In NE Spain, an ad which retained S 5-6, Al 1, some MeC and MaC was observed (Aymf & Jaume 1992). Moult interruption during the first stages of P-mouIt may occur as well (Herroelen 1982a), but these birds are likely to complete moult before migration. Comments on ageing after postjuvenile and postbreeding moult Best criteria: Skull pneumatization until at least the end of October (p. 204). Most ly recognizable by moult limits within GC, MeC or between MeC and MaC. A few ly with all GC moulted are more difficult to separate from ad on plumage characters alone. ly: 51% show a moult limit within GC which is usually conspicuous (Fig. 111-115). The juv GC, especially the central ones (GC 4-7), have a well-defined white fringe. Postjuv GC have an ill-defined yellowish, greenish or light greyish-brown fringe, varying individually in colour. Note that juv, postjuv and postbr GC 8—10 have slightly differently coloured fringes than the other GC of the same feather generation (Fig. 109, 110 and 117) and may simulate a moult limit. Therefore, always check for differences in wear and also in the colour of the feather centres, ly with no GC moulted (48%) are often recognizable by having a moult limit within MeC (Fig. 109) or else by the white fringes on all GC and the contrast between postjuv MeC and juv GC (Fig. 110). ly with all GC moulted (0.7%) are difficult to distinguish from ad and may be recognized by moult limits within Al, T and R. Juv and postjuv T and R can be distinguished by differences in wear rather than by differences in colour. For instance, T 7 is always darker than T 8—9 of the same feather generation and is fringed white. Ad: Fringes of all GC without white, fading into the dark feather centres. No moult limits within T, R and Al (Fig. 116 and 117). The few birds retaining individual feathers during the postbr moult (Fig. 117 and 118) are easily identified as ad by the strongly worn and bleached feathers. Extent of prebreeding moult: 2y and ad In spring, the wing-coverts are often difficult to assign to the correct feather generation. Only a few birds migrate through Europe with fresh

Motacilla and intact wing-coverts, T and R (Fig. 122). Most birds show different stages of wear. Especially the exposed GC 8-9 and T 8-9 are often worn to such an extent that it is not possible to determine whether they have been renewed very early during the prebr moult or whether they are juv or postbr (cf. Roselaar in Cramp 1988). The feathers of ad are easier to assign since their plumage is composed of only two feather generations (postbr and prebr). In some ad, the outer GC are definitely postbr, the central GC prebr and the inner GC are already worn and also prebr, since they are not so bleached as the outer postbr GC (Fig. 123). In such birds, T 8—9 are usually more worn than T 7 and R. These observations suggest that the prebr moult of many Yellow Wagtails extends over a long period or is divided into two phases. Accordingly, the innermost GC, some MaC and MeC and T 8—9 are likely to be moulted during the first phase, and the central and outer GC, the remaining MaC and MeC, T 7 and all R during the second phase. A long-lasting prebr moult, lasting about 90 days, is confirmed from Nigeria (Wood 1976). In Kenya, the body-feathers are moulted in two distinct phases (Pearson & Backhurst 1973, DJ. Pearson in lift.). During a first phase in November—early December, many feathers of the head, throat and mantle are renewed. During a second phase, lasting about 60 days, in January—March the actual prebr moult takes place and the head feathers are renewed again. In Italy, 10.6% of the birds caught during spring migration (N=871) still showed growing feathers on head and neck (Serra 1992). For the reasons given above, the following data on the extent of prebr moult are subject to reservations. The percentage of inner GC, T and perhaps R moulted during the prebr moult might be lower than given here. Our data, however, agree well with those of Wood (1976) and Serra (1992), who found about the same number of GC moulted and also suggest a moult of all T and R. MaC: 47.7% moult all, 43.0% part and 9.3% no MaC (N=107). MeC: 52.5% moult all, 46.8% part of the MeC. Only one bird did not moult any MeC (N=l4l). Ad and 2y moult similar numbers of MeC. GC: range 4-10, mean 6.7, mode 6, all GC 6.2% (N=145). 2y birds moult significantly more GC (mean 7.1, N=56) than ad (mean 6.4, N=34). CC: 2.2%. Al: none 94.7%, one 5.3% (N= 94). T: one 6.9% (T 7), two 2.1% (T 7+9), three 91% (N= 145). R: probably all birds renew all R (N=131). S: one bird with all R, T and GC as well as Al 1 and CC moulted had also renewed S 6 on both wings (Fig. 121).

flava

77

Fig, 108. Extent of prebr moult on the wing and tail in ad and 2y Motacilla flava. For differences between ad and 2y see text.

Fig. 109. ly after partial postjuv moult, 1 September. MaC postjuv, except some outermost juv. MeC 1-5 postjuv, 6-8 juv. GC, T and rest of wing juv. Recognizable as ly by the conspicuous moult limit within MeC. The juv GC have well-defined white fringes which are often narrow on GC 8-10 and may simulate a moult limit.

Comments on ageing after prebreeding moult Ageing in spring is difficult and not always possible for every bird. Moult limits due to the prebr moult occur in all 2y and ad. Since the prebr moult is more extensive than the postjuv moult, prebr GC usually directly adjoin the juv GC in 2y. Three feather generations within GC are exceptional (one bird out of 65, but 16.5% out of 85 in the sample of Serra 1992). If the GC which have not been moulted during the prebr moult are not heavily worn, well-defined white fringes (juv GC) are diagnositc of 2y (Fig. 119). Ill-defined greyish, yellowish or greenish fringes (postbr GC) indicate ad (Fig. 122 and 123). Many (but not all, see Fig. 120) 2y also have well-defined white fringes on Al 1 and CC (Fig. 119). Ad usually have ill-defined, not white, fringes. Advanced wear may prohibit use of these criteria from the end of April. Then, 2y may usually be distinguished from ad by having more heavily worn and bleached P, S and those GC not moulted during the prebr moult (Fig. 120). The same ageing criteria as given above have recently been proposed by Serra (1992). Fig. 110. ly after partial postjuv moult, 18 September. MaC and MeC postjuv, GC juv. T 7+9 juv, 8 postjuv. Rest of wing juv. Exceptional case of a ly which moulted a T without having renewed any GC. Recognizable as ly by the white fringes on all GC and the contrast between postjuv MeC and juv GC.

78

Motacilla flava Fig. 111. 1 y after partial postjuv moult, 8 September. MaC and MeC postjuv. GC 1-7+10 juv, 8-9 postjuv. T and rest of wingjuv. Conspicuous moult limit within GC. The renewed GC have illdefined greenish fringes.

Fig. 112. ly after partial postjuv moult, 7 September, MaC and MeC postjuv. GC 1-4+10 juv, 5-9 postjuv. T and rest of wing juv. Conspicuous moult limit within GC. GC 10 remained unmoulted although five GC have been renewed.

Fig. 115- ly after partial postjuv moult, 6 September. MaC and MeC postjuv. GC 1 juv, 2-10 postjuv. CC postjuv. Al 1 postjuv, 2-3 juv. T 7 juv, 8-9 postjuv. Rest of wing juv. GC 1, the only juv one, contrasts with the other GC in having a well-defined white fringe. The renewed T 8-9 are recognizable as postjuv by their illdefined greenish fringes. T 7 is always fringed white, but slightly more worn than the other T.

Fig. 113. 1 y after partial postjuv moult, 11 September. MaC and MeC postjuv. GC 1—7 juv, 8—10 postjuv. T 7+9 juv, 8 postjuv. Rest of wing juv. Conspicuous moult limit within GC. GC 10 is also moulted. Fig. 116. Ad $ after complete postbr moult, 18 September. Whole wing fresh. Fringes of all GC ill-defined and not white. No moult limit within T.

Fig. 114. 1 y after partial postjuv moult, 8 September. MaC and MeC postjuv. GC 1+5-8+10 juv, 2-4+9 postjuv. T and rest of wing juv. Irregularities within the sequence of GCmoult occur more frequently than in other species.

Fig. 117. Ad 9 after complete postbr moult, 4 October. Whole wing postbr, except Al 1 and MeCl. Exceptionally, Al 1 and MeC 1 remained unmoulted. Fringes of all GC ill-defined and not white. Note that the fringes of GC 8-10 are slightly darker than in the other GC and may simulate a moult limit.

Motacilla

Fig. 118. Ad 9 after postbr moult interrupted at S 6, 24 September. Most MaC, all MeC, GC 10, Al 2-3 and S 6 remained unmoulted. These feathers are heavily worn and bleached. Fig. 119. 2y 9 after partial prebr moult, 26 April. MaC mostly prebr. MeC 1 postjuv, 2-8 prebr. GC 1-4 juv, 5-10 prebr. T prebr. Rest of wing juv. Recognizable as 2y by the juv GC which are heavily bleached and, like A l l and CC, have well-defined white fringes.

flava

79

Fig. 121. Probably 2y 6 after partial prebr moult, 23 April. MaC, MeC, GC, CC, Al 1, T and S 6 prebr. Rest of wing probably juv. Exceptional case of an extensive prebr moult which included even S 6. The age cannot be told with certainty. The bleached S and P, the Al 2 fringed white and the worn outer PC suggest a 2y.

Fig. 122. Ad c3 after partial prebr moult, 2 May. MaC and MeC prebr. GC 1-2 postbr, 3-10 prebr. T prebr. Rest of wing postbr. GC 1-2 are moderately worn and have illdefined greyish, not white fringes. The same holds for the fringes of Al and CC. These characters indicate the ad. All prebr wing-coverts are similarly fresh.

Fig. 120. 2y 8 after partial prebr moult, 24 April. MaC and MeC prebr. GC 1-3 juv, 4-10 prebr. T prebr. Rest of wing juv. Juv GC heavily worn and their white fringes almost completely worn off. Juv CC and Al 1 without white fringes.

Fig. 123. Ad 8 after partial prebr moult, 22 April. MaC mostly prebr. MeC 1 postbr, 2-8 prebr. GC 1-4 postbr, 5-10 prebr. T prebr. Rest of wing postbr. P 6 has probably been renewed following an accident (right wing only). The postbr GC 1-4, CC and Al still show ill-defined greyish fringes, which distinguish them from the corresponding juv feathers and indicate the ad. GC 7-9 and T 8-9 are more heavily worn than the other prebr GC 5-6 and T 7; they have probably been renewed during the first part of the prebr moult and considerably earlier than GC 5-6 and T 8-9.

80

Motacilla cinerea

Motacilla cinerea Grey Wagtail Extent of postjuvenile moult MaC and MeC: usually all. 1 y with no or few moulted GC may retain some juv MeC. GC: range 0-10, mean 3.8, mode 0, no GC 18.7.%, all GC 1.1% (N=91). CC: 17.4%. Al: none 71.0%, one 24.6%, two 4.3% (N=69). T: none 36.5%, one 10.8%, two 16.2%, three 36.5% (N=74) (see p. 33). R: none 42.7%, one 10.7% (usually R 1), two 13.3% (usually R 1+6), three to five 17.3%, six 16.0% (N=75) (see p. 34). CC and Al 1 may be renewed when at least three, T and R when at least two GC are moulted, ly with at least seven GC moulted have all T and usually three to six R renewed. In Switzerland, the extent of postjuv moult decreases as the autumn migratory season proceeds: before the end of September, 4.6 GC are moulted on average (10 and 90 percentiles 1-8, N=56), and after that only 2.4 (0-6, N=29). In Belgium, the extent of postjuv moult is very similar and also decreases during autumn migration; four out of 308 ly were found to have renewed PC 1 (Herremans 1988a). Extent of postbreeding moult Whole plumage.

Fig. 124. Extent of postjuv moult on the wing and tail in ly Motacilla cinerea.

Extent of prebreeding moult: 2y and ad MaC: three quarters of 2y moult none, one quarter moult the proximal MaC only (N=21). MeC: usually not moulted. One 2y had a single new MeC (N=32). GC: range 0-2, mean 0.5, mode 0, no GC 68.2% (N=44). GC are often moulted out of the normal sequence. T: none 22.7%, one 20.5%, two 25.0%, three 31.8% (N=44) (see p. 15). R: none 57.5%, one 15.0% (usually R 1), two 17.5% (usually R 1+6), three to five 7.5%, six 2.5% (N=40) (see p. 16).

Comments on ageing after postjuvenile and postbreeding moult Best criteria: Skull pneumatization until at least mid-winter (p. 204). Moult limit within GC diagnostic of most ly, otherwise check for moult limits within Al, T or between T and S. ly: 80% show a moult limit within GC which is easily recognizable by a difference in colour of the fringes if situated within the central GC. Juv GC 4—8 have a whitish, postjuv GC 4—8 a greyish fringe (Fig. 128), and are often longer (Fig. 129). The feather centres of juv GC are less dark than in postjuv GC (Fig. 128). This is important for recognizing moult limits within outer GC where the colour of the fringes is less decisive. Note that juv, postjuv and postbr GC 9 and 10 are slightly greyer than the other GC of the same feather generation and may simulate a moult limit (cf. Fig. 126, 127 and 130). Also check for moult limits within T. 1 y with no GC moulted (19%) are recognizable by the slight difference in colour between the renewed MeC and the juv GC (Fig. 127). ly with all GC renewed (1.1%) are difficult to separate from ad. Often, a renewed CC, a moult limit within Al or T or between T and S 6 is diagnostic.

Fig. 125- Extent of prebr moult on the wing and tail in ad and 2y Motacilla cinerea.

Ad: No moult limits within GC, T and Al. S 6 and T 7 have the same black colour (Fig. 130). Note that GC 9 and 10 are slightly greyer than the adjacent GC. Comments on ageing after prebreeding moult Ageing spring birds is difficult. Moult limits due to the prebr moult occur in 2y and ad. Since the prebr moult is usually less extensive than the postjuv moult (68% without any prebr GC), 2y are often recognizable, as in autumn, by moult limits within GC due to the postjuv moult (Fig. 132 and 133). Some 2y have three feather generations within GC, thus showing two moult limits (Fig. 131). Birds showing no moult limit or only a moult limit due to the prebr moult within GC (all ad, some 2y) are difficult to age. In ad, P and S are usually slightly darker than in 2y and contrast less with T (Fig. 134).

Motacilla cinerea

Fig. 126. ly in juv plumage, 3 August. One central MaC postjuv. Rest of wing juv. GC 1-8 fringed whitish. GC 9 and 10 are greyer than the others and may simulate a moult limit. No marked contrast between MeC and GC.

Fig. 127. ly after partial postjuv moult, 21 September. MaC and MeC postjuv. GC juv. T and rest of wing juv. Recognizable as ly by the postjuv MeC being darker blackish-grey than the juv GC.

Fig. 128. ly after partial postjuv moult, 17 September. MaC and MeC postjuv. GC 1—6 juv, 7—10 postjuv. T postjuv. Rest of wing juv. Conspicuous moult limit within GC. The juv GC 4—6 (but not 1—3) have whitish fringes and brownish-grey feather centres, the inner postjuv GC greyish fringes and blackish feather centres. The renewed T contrast with the juv S in their black colour.

81

Fig. 129. ly after partial postjuv moult, 19 September. MaC and MeC postjuv. GC 1-7+10 juv, 8-9 postjuv. T 7+9 juv, 8 postjuv. Rest of wing juv. The postjuv GC 8-9 contrast with the adjacent juv GC in their darker feather centres and their length. Note that the slight difference in colour among the two renewed GC corresponds to that in ad (cf. Fig. 130).

Fig. 130. Ad after complete postbr moult, 29 September. Whole wing postbr. S as dark as T. Difference in colour between MeC and GC less than in those ly with no renewed GC (cf. Fig. 127). The two innermost GC are slightly greyer than the others and may simulate a moult limit.

82

Motacilla cinerea Fig. 133. 2y after partial prebr moult, 17 April, MaC mostly postjuv. MeC postjuv. GC 1—7 juv, 8-10 postjuv. T 7+9 prebr, 8 juv or postjuv. Rest of wing juv. As in Fig. 132, this bird moulted only a few MaC during the prebr moult. Although the fringes are largely abraded, the moult limit due to the postjuv moult is still recognizable.

Fig. 131. 2y after partial prebr moult, 15 April. MaC partly prebr, partly postjuv. MeC postjuv. GC 1-6 juv, 7+9 prebr, 8+10 postjuv. T prebr. Rest of wing juv. Recognizable as 2y by the three feather generations within GC. The juv GC are brownish and the inner ones (GC 4-6) fringed whitish. The postjuv GC 8 shows a colour of the feather centre intermediate between that of the juv GC 1-6 and the prebr GC 7+9.

Fig, 132. 2y after partial prebr moult, 17 April. MaC mostly postjuv. MeC 5 prebr, rest postjuv. GC 1—3 juv, 4—10 postjuv. T prebr. Rest of wing juv, This bird moulted no wing-coverts during the prebr moult, except some inner MaC and MeC 5. Within the GC, which are all worn, the moult limit due to the postjuv moult is easily recognizable.

Fig. 134. Ad after partial prebr moult, 5 April. MaC partly prebr, partly postbr. MeC, GC, T and rest of wing postbr. Birds having no moult limit, or one due only to the prebr moult, are difficult to age. The postbr GC of ad are darker and fringed greyish, not whitish, P and S are darker than in 2y and hardly contrast with the renewed T.

Motacilla alba alba

83

Motacilla alba alba White Wagtail Extent of postjuvenile moult MaC and MeC: usually all. ly with no renewed GC may retain some juv MeC. GC: range 0-10, mean 5.3, mode 7, no GC 10.6%, all GC 4.5% (N=358). CC: 9.7%. Al: none 93.5%, one 6.5% (N=339). T: none 36.9%, one 15.6%, two 17.3%, three 30.3% (N=347) (see p. 33). R: none 48.2%, one 15,2% (usually R 1), two 18.4% (usually R 1+6), three 7.3%, four 4.4%, five 2.0%, six 4.4% (N=342) (see p. 34). P and S: one bird from S Italy (on both wings P 8—9 postjuv, PC juv, GC 2-10, T and R postjuv; Winkler & Jenni 1987) and one bird from Switzerland (P 6 postjuv on one wing only, PC juv, GC 1—10, Tand R postjuv) showed eccentric P-moult. Another bird from Switzerland had renewed S 5—6 on the left wing and S 6 on the right wing (GC 1—10, T and R postjuv on both wings). The extent of postjuv moult is correlated among GC, CC, Al, T and R (Fig. 136). CC and Al 1 may be renewed when at least six GC are moulted. When more than six GC are moulted, over 50% of ly moult at least one T and R. The extent of postjuv moult decreases as the autumn migratory season proceeds (Fig. 137). As observations from the Netherlands (Jukema & Rijpma 1984), NW Russia (Rymkevich 1990) and England (M. a. yarrellii, Baggott 1970, Broom et ai 1976) show, the postjuv moult is faster and less extensive in late-hatched than in early-hatched ly. In England, most ly renew seven to eight GC and all T and R, but late-hatched ly less than seven GC or none and only some or no T and R (Baggott 1970, Broom et aL 1976); furthermore, $ replace more GC than 9 (Baggott 1970). In NW Russia, usually all MeC, four to ten GC (mean 5.6, all GC 9-1%, N=166), rarely T 8 or T 8-9 and one R are moulted, but never CC and Al. Thus, the moult of T, R, CC and Al in NW Russia is less Fig. 136. Relationships between the number of postjuv GC and the percentage of individuals with renewed CC, Al, T and R in ly Motacilla alba which have completed their postjuv moult.

Fig. 137. Mean number of postjuv GC during autumn of ly Motacilla alba which have completed the postjuv moult (data grouped in five-day periods; the first value includes the period 8—27 September).

Fig. 135- Extent of postjuv moult on the wing and tail in ly Motacilla alba. extensive than in Switzerland, the Netherlands and England, and even considerably less in Scandinavia where most ly renew only one or a few GC(Svensson 1992).

Extent of postbreeding moult Whole plumage. Exceptionally Al 1 may remain unmoulted (one bird). Comments on ageing after postjuvenile and postbreeding moult Best criteria: Skull pneumatization until mid-October (p. 204). Moult limits within GC, between GC and MeC and between T and S 6. ly: 85% show a moult limit within GC which is most conspicuous if situated in the area of GC 4—8 (Fig. 142) and usually more distinct in 6 than in ? . Compared with juv GC, postjuv GC have darker feather centres, are less worn, and have better defined fringes which form a more distinct step between the inner and outer web, especially on GC 4—8. ly with no GC moulted (11%) are recognizable by the difference in colour between the juv GC and the postjuv MeC (Fig. 139 and 141). ly with all GC renewed (5%) usually moult all T and show a contrast between the black base of the postjuv T 7 and the more greyish base of the juv S 6 (Fig. 145). Since such birds usually also moult CC and Al 1, moult limits within Al and between CC and PC may also be helpful. As ly may renew all R, they can be used for ageing only if they show a distinct regular moult limit. Ad: No moult limits between GC and MeC, within GC, Al and between CC and PC. Difference in colour between the bases of T 7 and S 6 only slight.

Fig. 138. Extent of prebr moult on the wing and tail in ad and 2y Motacilla alba. For differences between ad and 2v see text.

84

Motadlla alba alba

Extent of prebreeding moult: 2y and ad MaC: not or only partially moulted (N=63). MeC: 6% moult all, 81% part (mostly three to four central) and 13% moult no MeC (N=63). 2y and ad moult similar numbers of MeC. GC: range 0-7, mean 3.4, mode 3, no GC 6% (N=124). 2y moult slightly more GC (mean 3.5, N=81) than ad (mean 3.0, N=42, difference not significant). T: none 12%, one 38% (usually T 8), two 17%, three 33% (N=123) (seep. 15). 2y moult slightly more T (mean 1.8, N=81) than ad (mean 1.5, N=42, difference not significant). R: none 49%, one 24% (usually R 1), two 24% (usually R 1+6), three to four 3% (N= 117) (see p. 16). Comments on ageing after the prebreeding moult in spring, ageing is difficult ana in some cases impossible. Moult limits due to the prebr moult occur in 2y and ad. During the prebr moult, more T, but fewer R and considerably fewer GC are moulted than during the postjuv moult. Therefore, moult limits between juv and postjuv GC are still retained in many 2y. Such birds show three feather generations within GC, i.e. from outside to inside juv, postjuv and prebr GC. Only 2y showing three feather generations within GC (54.5%) are safely determinable (Fig. 148 and 149). 2y with only one moult limit within GC, as most ad have, are difficult to age; these are either birds which did not renew any GC during the prebr moult (3.9%) or birds which renewed more GC during the prebr moult than during the postjuv moult (41.6%, Fig. 150). They may often be recognized as 2y by the juv GC being more bleached than the postjuv GC of ad. 2y which moulted all GC during the postjuv moult (Fig. 151) are even more difficult to distinguish from ad. With much experience, they may be recognized by their more heavily bleached and browner P and S. In ad, P and S are usually more blackish and PC and Al have often more distinct white fringes (cf. Fig. 149 and 152). Fig. 139. ly (3 after partial postjuv moult, 13 October. MaC and MeC postjuv. GC, T and rest of wing juv. A S with broader fringed GC than in 9 (cf.Fig. 141). All GC similarly worn with illdefined fringes merging gradually into the greyish-brown feather centres. Again, the contrast between postjuv MeC and juv GC is diagnostic of ly.

Fig. 140. ly after partial postjuv moult, 13 October. MaC and MeC postjuv. GC 1—8+10 juv, 9 postjuv. T and rest of wing juv. The postjuv GC 9 contrasts with the juv GC. It has a darker feather centre and is less worn.

Fig. 141. ly after partial postjuv moult, 13 October. MaC and MeC postjuv. GC, T and rest of wing juv. GC all similarly worn with greyish-brown feather centres. Since the fringes of the GC are narrow, this bird is probably a 9. Recognizable as ly by the contrast between postjuv MeC and juv GC which is slightly more pronounced than in ad (cf. Fig. 147).

Fig. 142. ly after partial postjuv moult, 4 October. MaC and MeC postjuv. GC 1-6+10 juv, 7-9 postjuv, T and rest of wing juv. Recognizable as lyby the conspicuous moult limit within GC. Moult limits within GC 4—8 are most easily recognizable.

Fig. 143- ly after partial postjuv moult, 20 October. MaC and MeC postjuv. GC 1-5 juv, 6-10 postjuv. T7 juv, 8-9 postjuv. Rest of wing juv. Distinct moult limit within GC. The juv T 7 is only slightly worn, but recognizable as juv by its greyish-brown, not black base which does not contrast with the base of S 6.

Motacilla alba alba

Fig. 144. ly after partial postjuv moult, 4 October. MaC and MeC postjuv. GC 1—3 juv, 4—10 postjuv, T postjuv. Rest of wing juv. Recognizable as ly by the conspicuous moult limit within GC. The postjuv GC have black, the juv GC greyish-brown feather centres. The postjuv T 7 has a black base which contrasts with the juv S 6. Fig. 145. ly after partial postjuv moult, 4 October. MaC and MeC postjuv. GC postjuv. T postjuv. CC postjuv. Al 1 postjuv, 2-3 juv. Rest of wing juv. All GC are similarly fresh and intact and have black feather centres. Such birds are not always easy to distinguish from ad, but can be recognized as ly by the contrast between the postjuv T and juv S, especially between the bases of T 7 and S 6. Futhermore, the renewed CC and Al 1 are slightly darker than the juv PC and Al 2-3.

Fig. 146. ly after partial postjuv moult, 5 October. MaC and MeC postjuv. GC postjuv. T and S 6 postjuv. CC, Al and rest of wing juv. Exceptional ly which renewed S 6 together with all T.

85

Fig. 147. Ad after complete postbr moult, 20 October. Whole wing postbr. Compared with ly with no GC moulted, the contrast between GC and MeC is less distinct (cf. Fig. 139 and 141). Compared with ly with all GC and T moulted, the difference in colour of the bases of T 7 and S 6 is less distinct (cf.Fig. 144 and 145).

Fig. 148. 2y S after partial prebr moult, 11 April. MaC postjuv. MeC 1-3+8 postjuv, 4-7 prebr. GC 1-2 juv, 3—5+10 postjuv, 6—9 prebr. T prebr. Rest of wing juv. Typical 2y with three generations of GC. The juv GC 1-2 are lighter and contrast with the adjacent postjuv GC 3-5 which have darker feather centres. The prebr GC 6—9 show the darkest feather centres and the broadest, most intact fringes. Since the juv GC 1-2 are not exposed on the closed wing, they are not much more worn than the postjuv GC 3—5.

Fig. 149. 2y 9 after partial prebr moult, 15 April. MaC and MeC postjuv. GC 1—2 juv, 3-6+10 postjuv, 7-9 prebr. T 7+9 postjuv, 8 prebr. Rest of wing juv. 2y with three generations of GC. Since this bird is a 9, the colour differences between the three generations of GC are less distinct. Nevertheless, GC 1-2 are recognizable as juv by the browner tinge compared with the adjacent postjuv GC 3-

86

Motacilla alba alba

Fig. 150. 2y after partial prebr moult, 29 April. MaC postjuv. MeC probably all postjuv. GC 1-5 juv, 6-10 prebr. T prebr. Rest of wing juv. In this bird, more GC were moulted during the prebr moult than during the postjuv moult and, hence, only two generations of GC axe present. The juv GC adjoin the prebr GC without being separated by postjuv GC as in Fig. 148 and 149. Such 2y are difficult to distinguish from ad which also show only two generations of GC. They can usually be recognized as 2y by their GC being more bleached than the postbr GC of ad.

Fig. 151. 2y after partial prebr moult, 11 April. MaC postjuv. MeC 1—4+7—8 postjuv, 5-6 prebr. GC 1-6+10 postjuv, 7-9 prebr. CC postjuv. Al 1 postjuv, 2—3 juv. T 7+9 postjuv, 8 prebr. Rest of wing juv. This 2y moulted all GC, CC and Al 1 during an extensive postjuv moult. Thus, there are only two generations of GC present in spring, consisting of postjuv and prebr GC. Such 2y are very difficult to distingush from ad. This bird can be recognized as ly by the postjuv CC and Al 1 contrasting with the juv PC and Al 2—3 in their darker colour.

Fig. 152. Ad after partial prebr moult, 7 May. MaC postbr. MeC 1—2+4—7 prebr, 3+8 postbr. GC 1-6+10 postbr, 7-9 prebr. T 7+9 postbr, 8 prebr. Rest of wing postbr. Recognizable as ad by dark PC edged white (not distinct in all ad), by the broad white fringe of Al 2 and by the dark, only moderately bleached P and S. 2y usually have slightly browner, more bleached P and S (cf. Fig. 149).

Troglodytes troglodytes

87

Troglodytes troglodytes Wren Extent of postjuvenile moult

MaC and MeC: all. GC: range 0-8, mean 4.5, mode 6, no GC 2.8% (N=109). CC: 2.4%. Al: none 84.3%, one 157% (N=83). T: none 60.8%, one 27.5%, two 10.8%, three 1.0% (N=102) (see p. 33). The extent of post] uv moult is correlated among GC, Al and T (Fig. 154). T may be moulted when at least three GC are renewed, Al 1 when at least five GC renewed. The extent of postjuv moult decreases as the autumn migratory season proceeds and is lowest during winter. In spring, birds with a more extensive postjuv moult appear (Fig. 155). This seasonal variation in the extent of postjuv moult is negatively correlated with wing-length (see p. 42 and Jenni & Winkler 1983). In England, ly starting postjuv moult late (probably late-hatched birds) moult no T and fewer GC and R than ly with an early start of postjuv moult (Hawthorn 1971, 1974) Extent of postbreeding moult Whole plumage. Prebreeding moult While Witherby etal. (1943) mentioned no prebr moult, other authors found indications of a limited prebr moult including body-feathers and exceptionally MaC, MeC, shoulder-feathers and underwing-coverts (Vaurie 1951, Glutz & Bauer 1985). Comments on ageing Best criteria: Skull pneumatization until the beginning of December (p. 204). Moult limit within GC and difference in colour between GC and MeC and MaC diagnostic of ly/2y, although sometimes difficult to recognize.

Fig. 153. Extent of postjuv moult on the wing and tail in ly/2y Troglodytes troglodytes.

ly/2y: 97% show a moult limit within GC, which is distinct when among the outer GC (Fig. 159-161), but sometimes difficult to detect when among the inner GC (Fig. 158). Moult limits within GC recognizable by an abrupt change in the individually very variable colour pattern and by differences in length and colour between juv and postjuv GC. Postjuv GC, T and Al have a more bronze tinge, while juv are more reddish-brown. Postjuv GC are often tipped white (Fig. 159 and 161). The CC may, or may not, show a white tip in the juv plumage (cf. Fig. 156 and 157). To recognize ly/2y without any or with few renewed GC, the difference in colour between MaC and MeC, which are always postjuv, and the juv GC is diagnostic. The banding pattern on P 7 and 8 as well as the colour pattern of Al 3 (Drost 1932, Hawthorn 1971) are not a reliable ageing criteria (Jenni & Winkler 1983). Ad: No moult limit within GC. No difference in colour between MaC, MeC and GC. Many, but not all ad show white tips on GC (cf. Fig. 162 and 163).

Fig. 154. Relationships between the number of postjuv GC and the percentage of individuals with renewed Al and T in ly/2y Troglodytes troglodytes which have completed their postjuv moult.

Fig. 155. Mean number of postjuv GC in the course of the year of ly/2y Troglodytes troglodytes which have completed the postjuv moult (the two values for October refer to the first and second half of the month).

Fig. 156. ly in juv plumage, 11 August. Whole wing juv. All GC have the same reddish-brown colour. Note the loosely textured MaC and MeC. CC tipped white.

88

Troglodytes troglodytes

Fig. 157. ly in juv plumage, 14 August. Whole wing juv. MaC and MeC loosely textured, CC without white tip.

Fig. 158. ly after partial postjuv moult, 25 September. MaC and MeC postjuv. GC 1—7 juv, 8-10 postjuv. Rest of wing juv. Moult limit among inner GC not conspicuous. The juv GC are more reddishbrown than the bronze tinged postjuv GC. The bronze tinged postjuv MaC and MeC contrast with the reddish-brown juv GC.

Fig. 161. ly after partial postjuv moult, 18 September. MaC and MeC postjuv. GC 1—3 juv, 4—10 postjuv. CC postjuv. Al 1 postjuv, 2—3 juv. T 7—8 juv, 9 postjuv. Rest of wing juv. Moult limit within GC best recognized by the difference in colour between juv and postjuv GC. Postjuv GC 4-6 tipped white.

Fig. 159. ly after partial postjuv moult, 14 October. MaC and MeC postjuv. GC 1—4 juv, 5—10 postjuv. Rest of wing juv. Juv GC reddish-brown, less spotted, and without white tips. Postjuv GC with prominent dark spots, tinged bronze and with white tips on GC 5-7. Fig. 162. Ad after complete postbr moult, 3 October. Whole wing postbr. No moult limit within GC. All GC have a similar colour pattern without abrupt change. Almost no difference in colour between MaC, MeC and GC. This bird has no white tips on GC.

Fig. 163. Ad after complete postbr moult, 29 October. Whole wing postbr. No moult limit within GC which all have a similar colour pattern. In this bird, the outer GC are tipped white.

Fig. 160. ly after partial postjuv moult, 18 September. MaC and MeC postjuv. GC 1—5 juv, 6—10 postjuv. Al 1 postjuv, 2—3 juv. T 7—8 juv, 9 postjuv. Rest of wing juv. Recognizable as ly by moult limits within GC and T. Postjuv GC and T tinged bronze and with distinct dark marks, resulting in an abrupt change of the colour pattern within the row of GC, No white tips on postjuv GC.

Prunella modularis

89

Prunella modularis Dunnock Extent of postjuvenile moult MaC and MeC: usually all. MeC may remain partially unmoulted. GC: range 0-10, mean 1.4, mode 0, no GC 60.0%, all GC 0.7% (N=1806). CC: 3.4%. Al: none 89.7%, one 9.5%, two 0.8% (N=1357). T: none 96.9%, one 2.2%, two 0.6%, three 0.3% (N=1355) (see p. 33). R: we could detect regular moult of R, as cited in the literature (e.g. Cramp 1988), in only one bird (Fig. 173) which had all GC, CC,, Al 1, all T and all R renewed. The extent of postjuv moult is correlated among GC, CC, Al and T (Fig. 165). The extent of postjuv moult decreases as the autumn migratory season proceeds (Fig. 166). In NW Russia, the postjuv moult is slightly less extensive than in Switzerland (Rymkevich 1990): all ly moult the MeC, 30% at least one GC (usually up to three, occasionally up to seven), 4% renew Al, 2% CC and one bird T. Extent of postbreeding moult Whole plumage. Comments on ageing Best criteria: Skull pneumatization until at least the beginning of November (p. 204). At least until the end of October, the colour of the iris is diagnostic in most birds. Usually, ad have a reddish-brown and ly a greyish-brown or greyish-olive iris, while birds with an intermediate iris colour may be ad or ly. Among 788 ly and 160 ad studied between July and October, 10% of ly and 9% of adults had an intermediate colour and only 0.4% of ly had already acquired the adulttype iris colour. There was no apparent seasonal trend. According to Spencer & Mead (1978a), the iris colour can be used up to at least

Fig. 165. Relationships between the number of postjuv GC and the percentage of individuals with renewed CC, Al and T in lyI'2y Prunella modularis which have completed their postjuv moult.

Fig, 166. Percentage of ly Prunella modularis with at least one postjuv GC during autumn (data grouped in fiveday periods; the first value includes the period 9-28 August, the last 23 October-6 November).

Fig. 164. Extent of postjuv moult on the wing and tail in ly/2y Prunella modularis. December. Ageing based exclusively on plumage characters is often difficult and should only be done with experience acquired by using other characters in addition. ly/2y: A few birds retain juv MeC, recognizable by the yellow, not rusty-brown tips (Fig. 168). 39% of ly/2y show a moult limit within GC which is usually recognizable by differences in colour and shape of the light tips between juv and postjuv GC. However, due to individual variability in these tips, moult limits are difficult to detect in some birds. Conspicuous moult limits as in Fig. 173 are rare. On juv GC, tips are yellowish, conspicuous on GC 2—9, and present on both webs of the inner juv GC; the black field around the shaft is well-defined, especially in the terminal part which contrasts well with the light tips. Postjuv GC (typical in Fig. 174) have ill-defined whitish tips which fade into the terminal black field and may be virtually absent on the innermost GC (GC 9 in Fig. 169, GC9-10 in Fig. 170). Outer renewed GC often have a greyish tinge (Fig. 172 and 175). Most ly/2y have no moult limit within GC. Those with well-defined tips on outer and inner webs of GC 8 and 9 can be aged ly/2y up to spring. Ad: Unless skull pneumatization and iris colour can be used, identification of ad is difficult in some cases. Usually, ad have whitish tips on the outer GC and inconspicuous or missing tips on the inner GC. Light tips, black field around the shaft and brownish fringes are ill-defined and fade into each other.

Fig. 167. ly at the beginning of partial postjuv moult, 9 August. Whole wing juv, except some growing MaC. MeC and inner GC with yellowish tips on outer and inner web.

90

Prunella modularts Fig. 168. \j after partial postjuv moult, 13 October. MaC postjuv, MeC 1-5 juv, 6-8 postjuv. GC and rest of wing juv. Juv MeC tipped yellowish. Tips on inner GC yellowish and well-defined.

Fig. 169. ly after partial postjuv moult, 24 September. MaC and MeC postjuv. GC 1—8+10 juv, 9 postjuv. Rest of wing juv. Renewed GC 9 without distinct light tips, black field around the shaft less conspicuous than on the adjacent GC.

Fig. 173. ly after partial postjuv moult, 29 September. Exceptionally, this bird has seven S (but three T) and 11 GC. MaC and MeC postjuv. GC 1-3 juv, 4-11 postjuv. CC postjuv, Al 1 juv or postjuv, 2-3 juv. Innermost T postjuv. Rest of wing juv. Exceptionally conspicuous moult limit within GC. GC 1—3 with yellowish, well-defined tips, renewed GC with whitish and ill-defined tips.

Fig. 170. 2y after partial postjuv moult, 17 April. MaC and MeC postjuv. GC 1-8 juv, 9-10 postjuv. Rest of wing juv. Renewed GC 9 and 10 virtually without light tips, less worn and more firmly textured than juv GC.

Fig. 171. ly after partial postjuv moult, 9 Sept. MaC and MeC postjuv. GC 1-4, 6-7+10 juv, 5+8-9 postjuv. Rest of wing juv. GC 5 with only a faint and ill-defined light tip. GC 8 and 9 without light tips. All juv GC have a more conspicuous black area around the shaft than the postjuv GC.

Fig. 174. ly after partial postjuv moult, 3 October. MaC and MeC postjuv. GC postjuv. CC and Al 1 postjuv, 2-3 juv. T 7-9 postjuv. PC 1-3+5 postjuv. Rest of wing juv. Exceptional case of a very extensive postjuv moult. Recognizable as ly by the moult limit between postjuv T 7 and juv S 6.

Fig. 172. Ad after complete postbr moult, 13 April. Whole wing postbr. Tips already worn, whitish and ill-defined.

Fig. 175. Ad after complete postbr moult, 29 September. Whole wing postbr. Inner GC with ill-defined light tips, fading into the black field around the shaft. Outer GC tipped whitish.

Erithacus rubecula

91

Erithacus rubecula Robin Extent of postjuvenile moult MaC and MeC: usually all. Exceptionally, individual juv MaC and MeC may be retained by ly moulting no or few GC. GC: range 0-10, mean 4.7, mode 5, no GC 0.03%, all GC 0.05%

(N=ll,108). CC: 11.7%. Al: none 75.0%, one 24,9%, two 0.1% (N=8701). T: none 99.8%, one 0.2% (N=8499). R: none 99.9%, one (R 1) 0.1% (N=8490). The extent of postjuv moult is correlated among GC, CC and Al (Fig. 177). CC and Al may be renewed when at least three GC are moulted. The extent of postjuv moult decreases as the autumn migratory season proceeds. It is lowest in early winter and increases again during late winter and spring (Fig. 178). During April, there is no trend in the number of GC moulted. Spring migrants (March, April) on Ventotene Island, Italy (not included in Fig. 178) have slightly more GC moulted than spring birds in Switzerland (Italy: mean 4.5, N=492; Switzerland: mean 4.3, N=1252, difference not significant). In NW Russia, the postjuv moult includes all MaC and MeC, one to six GC, none to three Al, never CC, and T 9 in one bird only. The extent of postjuv moult was found to depend on hatching date: while early-hatched Russian birds renew usually five to six GC and Al, latehatched birds moult only two to three GC and no Al (Rymkevich 1990). In ly birds caught in Sweden at Falsterbo and Ottenby, the postjuv moult is slightly less extensive than in birds caught in Switzerland (mean 4.4—4.5 renewed GC, re-calculated on the basis of ten GC; Karlsson et al. 1986a, Pettersson etai 1990). At Falsterbo, no bird was found having all or fewer than three GC moulted. At both sites, the number of GC moulted decreases during autumn migration (at Falsterbo from about 4.6 at the beginning of September to about 3.5 at the end of October). During spring migration at Falsterbo no seasonal trend is observed (Karlsson etaL 1986a). Based on a small data

Fig. 177. Relationships between the number of postjuv GC and the percentage of individuals with renewed CC and Al in ly/2y Erithacus rubecula which have completed their postjuv moult.

Fig. 176. Extent of postjuv moult on the wing and tail in ly/2y Erithacus rubecula.

set, Pettersson et al. (1990) concluded that birds wintering in the E Mediterranean (Greece and Cyprus) have fewer postjuv GC than birds wintering in the W Mediterranean (S France and Spain). There is probably no difference in the extent of postjuv moult between the sexes (Karlsson etal. 1986a). Extent of postbreeding moult Whole plumage. Suspension of primary moult in one bird after renewal of P 1-3 was due to remating and a second breeding attempt (Harper 1984). One ad retained PC 5 (Fig. 191). Comments on ageing Best criteria: Moult limit within GC diagnostic in almost all ly/2y. Skull pneumatization until mid-September, but many ly have incomplete pneumatization until the beginning of November (p. 204). Supplementary ageing criteria which leave some birds undetermined are given below. ly/2y: 99.9% show a moult limit within GC which is usually easily recognized. Juv GC can always be distinguished from postjuv GC by differences in the colour of the feather centre and outer fringe, irrespective of the distinctness of the light tips. Juv GC are brown with a yellow-brownish tinged fringe, postjuv GC have an olive tinge on the fringe and feather centre. The light tips of GC vary individually, but, when present, may be used to detect the moult limit. In many ly/2y birds, light tips are present on juv GC, but not on postjuv GC (Fig. 182). Birds having light tips on all GC show an abrupt change in shape and colour of the light tips at the moult limit (Fig. 183), This sudden change is not present in ad showing light tips on GC (Fig. 189 and 190). ly/2y with small (Fig. 185), worn off (Fig. 184) or missing (Fig. 186) light tips on GC are recognizable by differences in the colour of the feather centres between juv and postjuv GC. The very few ly/2y with all GC moulted (0.05%) show a contrast between the olive tinged CC and Al 1, which are always renewed in these birds, and the browner juv Al 2-3 and PC (Fig. 187). Ad: No moult limit within GC and no contrast between CC, Al and PC. Fringes of PC tinged olive, similar to GC. The light tips on GC are generally smaller and often absent (Fig. 188). Ad with light tips on GC show a gradual decrease in size from outside to inside and gradual change in shape (Fig. 189 and 190).

Fig. 178. Mean number of postjuv GC in the course of the year ofly/2y Erithacus rubecula which have completed the postjuv moult (from August until October: five-day periods; the first value includes the period 30 July-18 August).

Other criteria: A number of other ageing criteria have been described and tested on a large number of birds (Pettersson 1983, Karlsson et al. 1986a, Pettersson et al. 1990). They may be useful to some ringers as

92

Erithacus rubecula

supporting criteria, but leave some birds undetermined or assigned to the wrong age class. Usually, ly have a yellowish inside to the upper mandible in autumn and still some yellow left in spring, but in autumn and especially in spring, this feature may be unclear or misleading (Frelin 1971, Pettersson 1983, Karlsson et ai 1986a, Svensson 1992, own obs.). In ly/2y, the R and PC are usually, but not always (cf. Fig. 187), more pointed and the colour of the iris and the legs darker than in ad (Karlsson et aL 1986a, Pettersson etal. 1990).

Fig. 179. ly in juv plumage, 11 August. Whole wing juv. MaC, MeC and all GC tipped yellowish, typical of the juv plumage.

Fig. 180. 2y after partial postjuv moult, 23 April. MaC mostly postjuv, a few at the edge of the wing juv. MeC 1-3 juv, 4—8 postjuv. GC and rest of wing juv. Very rare case of a 2y without any GC moulted. Recognizable as 2y by the retained juv MaC and MeC.

Fig. 181. ly after partial postjuv moult, 14 October. MaC and MeC postjuv. GC 1—8+10 juv, 9 postjuv. Rest of wing juv. Recognizable as ly by the renewed GC 9 tinged olive and without light tip.

Fig. 182. ly after partial postjuv moult, 20 September. MaC and MeC postjuv. GC 1—5 juv, 6—10 postjuv. Rest of wing juv. Conspicuous and typical moult limit within GC. Postjuv GC tinged olive and without light tips, juv GC browner and with large tips.

Fig. 183. ly after partial postjuv moult, 14 September. MaC and MeC postjuv. GC 1—4 juv, 5—10 postjuv. CC postjuv. Al 1 postjuv, 2—3 juv. Rest of wing juv. Moult limit within GC recognizable by the difference in colour of the feather centres and by the abrupt change in size, shape and colour of the light tips. The renewed CC and Al 1 are tinged olive, but not the juv Al 2-3 and PC. Fig. 184. 2y after partial postjuv moult, 10 April. MaC and MeC postjuv. GC 1-2 juv, 3-10 postjuv. CC postjuv. Al 1 postjuv, 2—3 juv. Rest of wing juv. Although the GC are heavily worn, the moult limit within the GC is still easily recognized by the difference between the browner juv GC 1-2 and the greenish-grey postjuv GC 3—10. The postjuv CC and Al 1 contrast with the juv Al 2—3 and PC in their greenish-grey colour.

Fig. 185. ly after partial postjuv moult, 14 September. MaC and MeC postjuv. GC 1—6 postjuv, 7—10 juv, Al 1 postjuv, 2-3 juv. Rest of wing juv. Moult limit within GC recognizable by the difference in colour between the browner juv GC and the olive tinged postjuv GC, but hardly recognizable by the pattern of tips. Juv GC only faintly tipped.

Erithacus mbecula

93

Fig. 186. 2y after partial postjuv moult, 19 April. MaC and MeC postjuv. GC 1-7 juv, 8-10 postjuv. Rest of wing juv. Juv GC without light tips, recognizable by their browner colour and more worn fringes, postjuv GC by their olive tinge.

Fig. 189. Ad after complete postbr moult, 13 April. Whole wing postbr. As Fig. 188, but with small tips on GC. The tips decrease gradually in size from outside to inside without abrupt change.

Fig. 187. ly after partial postjuv moult, 22 September. MaC and MeC postjuv. GC all postjuv. CC postjuv. Al 1 postjuv, 2-3 juv. Rest of wing juv. Rare case of a ly with all GC moulted on the right wing (left wing: GC 1-2 juv). Recognizable as ly by the renewed Al 1 and CC which contrast with the juv Al 2—3 and PC in their olive tinge. GC more olive than the juv PC and S.

Fig. 190. Ad after complete postbr moult, 17 September. Whole wing postbr. Exceptional ad with very large tips on GC. Recognizable as ad by the gradual, not abrupt, change in size and shape of the light tips over the GC and by the similarly coloured fringes of GC and PC.

Fig. 191. Ad after complete postbr moult, 21 September. Whole wing postbr except PC 5. Typical ad with only faint tips on GC. All GC of the same colour. Exceptionally, PC 5 has been retained.

Fig. 188. Ad after complete postbr moult, 27 September. Whole wing postbr. No moult limit within GC and Al and between CC and PC. All GC uniformly tinged olive. Typical ad without light tips on GC.

94

Luscinia megarhynchos

Luscinia megarhynchos Nightingale Extent of postjuvenile moult MaC and MeC: all. GC: range 3-9, mean 5.3, mode 5 (N=120). CC: 3.9%. Al: none 97.1%, one 2.9% (when at least seven GC renewed) (N= 102). T: none 98.0%, one 2.0% (N=102).

Extent of postbreeding moult Whole plumage.

Fig. 192. Extent of postjuv moult on the wing and tail in ly/2y Luscinia megarhynchos.

Comments on ageing Best criteria: Skull pneumatization until at least the beginning of October (p. 204). Moult limit within GC diagnostic of ly/2y. ly/2y: Moult limits within GC are usually easily recognizable by the presence of buffish tips on juv GC (Fig. 195). Occasionally, these tips are not distinct (Fig. 193). Therefore, it is advisable to always check for differences in colour of the feather centres between juv and postjuv GC. Juv GC are slightly more rusty-brown, of looser texture and often shorter than postjuv GC. Moreover, PC of ly/2y are narrower and more pointed than in ad and often have inconspicuous pale tips. In spring, the plumage is often surprisingly little abraded, so that the pale tips on juv GC are still visible (Fig. 194 and 195). Ad: All GC of similar colour and texture and without pale tips (Fig. 196 and 197). Fig. 195. 2y after partial postjuv moult, 18 April. MaC and MeC postjuv. GC 1-6 juv, 7-10 postjuv. Rest of wing juv. Conspicuous moult limit within GC recognizable by the presence of distinct pale tips on juv GC. Although a spring bird, the plumage is hardly worn. Fig. 193. ly after partial postjuv moult, 25 August. MaC and MeC postjuv. GC 1—5 juv, 6-10 postjuv. Rest of wing juv. Juv GC without or with only small illdefined pale tips. Moult limit recognizable mainly by a slight difference in colour of the feather centres between juv and postjuv GC.

Fig. 194. 2y after partial postjuv moult, 4 May. MaC and MeC postjuv. GC 1-3 juv, 4-10 postjuv. Rest of wing juv. Juv GC with pale tips, slightly more rusty feather centres and looser textured than postjuv GC.

Fig. 196. Ad after complete postbr moult, 12 August. Whole wing postbr. GC without pale tips and of similar colour, texture and length. PC broader and more rounded than in ly/2y.

Fig. 197. Ad after complete postbr moult, 20 April. Whole wing postbr. GC all similar in colour, texture and length and without pale tips. Although a spring bird, all feathers are surprisingly fresh.

Luscinia svecica

95

Luscinia svecica Bluethroat Extent of postjuvenile moult

MaC and MeC: all. GC: range 1-8, mean 2.9, mode 2 (N=74). CC: 12.8%. Al: none 94.9%, one 5.1% (N=39). TandR: one bird had T 7, another T 7-9 and R 1-5 renewed (N=64). In NW Russia, the postjuv moult includes all MeC, usually one to three GC (depending on the site 10—30% retain all juv GC) and occasionally CC and one or two Al (Rymkevich 1990). Fig. 198. Extent of postjuv moult on the wing and tail in ly/2y Luscinia svecica.

Extent of postbreeding moult Whole plumage.

autumn migration scores 3-5 occur (N=46). Light tips on juv GC and moult limit within GC diagnostic of ly/2y.

Extent of prebreeding moult The prebr moult comprises only the feathers of chin, throat and sides of head or may be completely suppressed (Glutz & Bauer 1988, Roselaar in Cramp 1988). Comments on ageing Best criteria: Skull pneumatization until at least October. During

ly/2y: Juv GC with rusty-buff tips, which are often still present in spring, but may be abraded from about May onwards. Postjuv GC with more greyish feather centres than juv GC. Many ly/2y show more or less distinct light tips on CC, Al and PC. Ad: GC without rusty-buff tips, but only narrow and slightly lighter fringes. Exceptionally, ad may show light tips on GC (Svensson 1992). Thus, always check for uniformity of colour of feather centres.

Fig. 201. Ad after complete postbr moult, 19 September. Whole wing postbr. No light tips on GC. All GC show similarly coloured feather centres. Fig, 199, ly after partial postjuv moult, 12 September. MaC and MeC postjuv. GC 1—8 juv, 9—10 postjuv. Rest of wing juv. Distinct moult limit within GC. Juv GC with rusty tips, postjuv GC without tips and with more greyish feather centres. Fig. 200. ly after partial postjuv moult, 16 August. MaC and MeC postjuv. GC 1—5 juv, 6-10 postjuv. Rest of wing juv. Moult limit within GC recognizable by the presence of light tips on juv GC, but also by a difference in colour of the feather centres between juv and postjuv GC.

Fig. 202. Ad after complete postbr moult, 15 April. Whole wing postbr. No light tips on GC and uniformly coloured feather centres.

96

Phoenicurus ochruros

Phoenicurus ochruros Black Redstart Extent of postjuvenile moult MaC and MeC: all. GC: range 0-10, mean 3.5, mode 2, no GC 1.0%, all GC 0.2% (N=477). CC: 4.3%. Al: none 85.9%, one 14.1% (N=398). T: none 87.6%, one 5.7%, two 5-0%, three 1.7% (N=418) (see p. 33). R: none 99.2%, one 0.8% (N=394). S: two birds moulted S 6 on one wing.

The extent of postjuv moult is correlated among GC, CC, Al and T (Fig. 204). The extent of postjuv moult decreases slightly but significantly as the autumn migratory season proceeds (Fig. 205). Extent of postbreeding moult Whole plumage. Prebreeding moult The indication of a possible prebr moult found in some publications probably goes back to a footnote in Witherby et al. (1943) mentioning two birds with growing body-feathers in February and March, but qualifying them as 'abnormal'. Among 108 birds (ad and 2y) examined by us in Italy and Switzerland during March and April, three had a few growing body-feathers and none showed feathers renewed during winter or spring. Thus, there is no, or only a very restricted, prebr moult in this species. Comments on ageing Best criteria: Skull pneumatization until the end of August (p. 204). Most ly/2y recognizable by moult limits within GC and T. ly/2y: 98.8% show a moult limit within GC. In ly $ in the ad-d1 -like plumage (^paradoxus morph), moult limits are conspicuous; juv GC

Fig. 203. Extent of postjuv moult on the wing and tail in ly/2y Phoenicurus ochruros.

are brown with narrow bufFish fringes, postjuv GC are dark grey with broad grey fringes (Fig. 208, 209 and 212). In 6 in the 9 -like plumage ('cairet morph) and in ?, moult limits are more difficult to recognize, especially when within the inner GC (Fig. 207 and 210). The postjuv GC have darker feather centres tinged greyish and darker fringes than the juv GC. Moult limits within outer GC are more distinct (Fig. 211-213). Some ly/2y with more than four GC moulted and all ly/2y with more than eight GC moulted show a moult limit within T or between T and the browner juv S (Fig. 212 and 214). Moult limits are still easily recognized in spring, but are obscured by wear in summer. A supporting, though not infallible, criterion is the pattern of the outermost R. Juv R 6 generally has dark marks on the tip and a dark shaft extending more than 9 mm from the tip inwards; R 6 of ad shows less prominent dark markings extending less than 8 mm from the tip inwards or (more rarely) is completely red (Svensson 1992). Ad: cT are recognizable by their uniformly greyish-black wing and by having white outer fringes not only on T but also on inner S (Fig. 215 and 216). 9 are recognizable by the lack of a moult limit within GC and T, by darker fringes on GC resulting in less contrast between feather centre and fringe than in ly/2y and by a similar colour of T and S.

Fig. 204. Relationships between the number of postjuv GC and the percentage of individuals with renewed CC, Al and T in ly/2y Phoenicurus ochruros which have completed their postjuv moult.

Fig. 205. Mean number of postjuv GC during autumn (data grouped in ten-day periods; the first value includes the entire August) and in spring of ly/2y Phoenicurus ochruros which have completed the postjuv moult.

Fig. 206. ly S at the beginning of partial postjuv moult, 5 August. MaC partly postjuv. Rest of wing juv. Recognizable as S by the renewed MaC having black feather centres and grey fringes. Exceptionally, juv T 7 shows some white at the base of the outer web.

Phoenicurus ochruros

97

Fig. 207. ly after partial postjuv moult, 15 October. MaC and MeC postjuv. GC 1-8 juv, 9-10 postjuv. Rest ofwingjuv. Inconspicuous moult limit within GC of a 9 or a 3 in 9 -like plumage. The renewed GC are slightly darker with a greyish tinge and less worn than the juv GC.

Fig. 208. ly 6 after partial postjuv moult, 16 October. MaC and MeC postjuv. GC 1-7 juv, 8-10 postjuv. Rest ofwingjuv. Conspicuous moult limit within GC, typical of 6. The postjuv GC are grey, the juv GC brown.

Fig. 209. 2y 3 after partial postjuv moult, 6 April. MaC and MeC postjuv. GC 1—8 juv, 9-10 postjuv. Rest of wing juv. Conspicuous moult limit within GC, typical of 6.

Fig. 212. ly c5 after partial postjuv moult, 2 October. MaC and MeC postjuv. GC 1—3 juv, 4—10 postjuv. Al 1 postjuv, 2—3 juv. T 7—9 postjuv. Rest of wing juv. Distinct moult limit within GC. The postjuv T contrast with the juv S and have white outer fringes, typical of ly