J Ornithol (2006) 147:553-564 DOI 10.1007/sl0336-006-0079-z
ORIGINAL ARTICLE
Breeding ecology and nest site selection in allopatric mainland Citril Finches Carduelis[citrinella] citrinella and insular Corsican Finches Carduelis[citrinella] corsicanus
Marc I. Forschler • Elisabeth K. V. Kalko
Received: 11 October 2005 / Revised: 3 January 2006 / Accepted: 3 January 2006 / Published online: 12 July 2006 © Dt. Ornithologen-Gesellschaft e.V. 2006
Abstract The breeding ecology and nest site selection species is that Citril Finches breed mainly in half-open of mainland Citril Finches and insular Corsican Fin- conifer forests (especially pine forests), while Corsican ches have been studied throughout their limited range. Finches breed in the more open scrubby mountains of For many years both endemic forms were considered the Mediterranean islands dominated mainly by the to be two sub-species; however, based on evidence Tree Heath as an adaptation to the different land- from more recent molecular studies they have been scapes on the islands. In contrast to Citril Finches, this split into two species. This study provides data on the preference of the Corsican Finches for Tree Heath as variations in breeding ecology and nest site selection in nesting plants - even if suitable pines are available - is the different sub-populations of these little studied typical of the species. These behavioural changes result species. A secondary aim was to search out evidence of in a niche expansion into open habitats at lower alti- ecological differentiation between mainland Citril tudes. We suggest that the observed niche expansion Finches and insular Corsican Finches. We found that and behavioural variations are not suitable criteria for the studied sub-populations of both species largely taxonomic status, a proposal in contrast to that of overlapped in breeding ecology. Our data confirms the Sangster [Ibis 142:487-490 (2000)]. We further suggest close similarity of Citril Finches and Corsican Finches, that the few ecological differences found in this study both which are, similar to mountain birds, well adapted between the two (sub-)species are the result of the so- to the local habitat conditions of their different called insular syndrome, which includes changes in life mountain systems. Several differences were identified history traits such as morphology, demography and within the studied sub-populations of the two (sub-) behaviour. species with respect to nest site selection, probably caused by environmental conditions and local preda- Keywords Allopatry • Breeding ecology • tors. One of the main differences between the two Citril Finch • Corsican Finch • Insular syndrome • Nest site selection • Niche expansion
Communicated by F. Bairlein
M. I. Forschler (El) Introduction Vogelwarte Radolfzell, Max Planck Institute for Ornithology, 78315 Radolfzell, Germany The Citril Finch Carduelis [citrinella] is one of the few e-mail: [email protected] bird species restricted to European mountains (Voous M. I. Forschler 1960). Currently, two allopatric forms are distin- Natural History Museum, Pg. Picasso s/n, guished. The Corsican Citril Finch Carduelis [citrinel- Pare Ciutadella, 08003 Barcelona, Spain la] corsicanus lives exclusively on Corsica, Sardinia and several Tuscany islands (Capraia, Elba, Gorgona) E. K. V. Kalko • M. I. Forschler Department of Experimental Ecology, (Whitehead 1885; Jourdain 1911; Armitage 1937; Thi- University of Ulm, 89069 Ulm, Germany bault 1983; Moltoni 1975; Arcamone 1993; Cramp and
4y Springer 554 J Ornithol (2006) 147:553-564
Perrins 1994; Baccetti and Marki 1997; Thibault and In the study reported here, we analysed breeding Bonaccorsi 1999). The nominate form, the Citril Finch phenology as well as nest site selection of Citril Finches Carduelis [citrinella] citrinella, occurs at higher eleva- and Corsican Finches at selected sites considered to be tions in the mountain ranges of central and south- representative of their total range. Specifically, we fo- western Europe (Alps, Black Forest, Vosges, Jura, cussed on differences between sub-populations of Massif Central, Cevennes, Pyrenees, Cantabrians and Corsican Finches on Corsica, Sardinia and Capraia and the Sierras of Central Spain), generally above those of Citril Finches in the Pre-Pyrenees and the 900 m a.s.l. (Voous 1960; Bezzel 1993; Cramp and Black Forest. Perrins 1994; Bauer and Berthold 1996; Baccetti and Marki 1997; Glutz von Blotzheim and Bauer 1997; Holzinger 1997). Populations of nominate Citril Fin- Materials and methods ches in the northern part of their range, such as the mountain ranges of the Black Forest, have to adapt to Study areas wetter and colder conditions (Forschler 2001b) than birds in the southern mountain ranges, such as the We selected study sites throughout the total range of Catalonian Pre-Pyrenees where the weather is very hot the species. Breeding sites of Citril Finches were and dry (Gutierrez 1991; Borras et al. 2003). Similarly, investigated in the northern part of its distribution at Corsican Finches in the high central mountains of Mount Schliffkopf in the German Northern Black Corsica are exposed to colder conditions than are birds Forest (April-July 1999, April-June 2000) and in the in the dry and hot mountains of Sardinia. southern part of its distribution at Port del Comte in There is an ongoing debate about the taxonomic the Catalonian Pre-Pyrenees (April-July 2002) (Ta- status of nominate Citril Finches and Corsican Finches ble 1). Populations of Corsican Finches were studied in Europe. Despite recently detected genetic differ- on Corsica (April-June 2001, May-June 2003), mainly ences (mitochondrial DNA differs by 2.7%), some in the high valley of Niolo (Haute-Corse) and the authors still regard the two forms as conspecihcs mountain range of the Massif de l'Ospedale (Corse-du- (Pasquet and Thibault 1997; Thibault and Bonaccorsi Sud), at Monte Limbara (Gallura) and Monte Discudu 1999), whereas others assign species status to the (Gennargentu) in Sardinia (April-June 2003) and on Corsican Finch Serinus corsicanus, referring to the the island of Capraia (March-May 2003). For further same genetic results (Sangster 2000). Based on the information on the study areas see Table 1. recommendations of Sangster (2000) in combination with differences in morphology (Cramp and Perrins Searching for nests 1994; Pasquet 1994) and vocalization (Chappuis 1976; Cramp and Perrins 1994), the Association of the We conducted systematic searches for nests in the European Rarities Committees now treats the two study areas during the nest-building period (Forschler forms as full species (Sangster et al. 2002). 2000, 2002a). With this aim, we selected 100- to 180-ha The breeding ecology of the two forms has received sample plots in suitable areas [Schliffkopf (180 ha), little attention, mostly due to their limited distribution Port del Comte-Vansa (150 ha), Port del Comte-Prat and the methodological constraints involved in study- de Botons (150 ha), Capraia (100 ha), Niolo-Calasima ing them in their mountainous habitats (Forschler (100 ha), Niolo-Albertacce (100 ha), Massif de 2000). Generally, both forms prefer the transitional l'Ospedale (125 ha), Massif de l'Ospedale-Cartalvonu zone between woodland and open landscape as a (100 ha), Monte Limbara (100 ha), Monte Discudu breeding habitat. However, Citril Finches appear to be (100 ha)] in which we mapped all nesting pairs and more strongly associated with wooded areas in higher territories. The largest number of all nests was found elevations, whereas Corsican Finches may exhibit a by following females engaged in nest building. Another habitat-niche expansion into lower Mediterranean suitable period to search for nests is during egg incu- habitats (Thibault 1983; Blondel et al. 1988; Cramp and bation when females are fed by males at their nest. Perrins 1994; Thibault and Bonaccorsi 1999). Niche This is a particularly good time to search for nests as breadth expansion of insular populations in compari- during the first days of egg incubation, females call son to mainland populations has also been observed in softly from their nest, presumably to attract their mate. other birds of the Mediterranean islands settled by Finally, we found a few nests just before the fledgling Corsican Finches (Martin 1982, 1992; Blondel 1985; period, when young birds standardly utter soft calls Blondel et al. 1988, 1991). from the nest. In total, 164 nests were detected at all
4y Springer J Ornithol (2006) 147:553-564 555
CD study sites: 118 nests of the Citril Finch and 46 of the Corsican Finch. >> -a c 3 ■ JS -3 .5/ -a ^^ Nest parameters o 5 a *3 5 .S cd : g > s S..S « g ■ % ^ « e ego, 4= £ After a nest was found, we noted the following char- 3 .^-§ u •Ola -j 3 TO - -a ft, OJ 2 e R acteristics of the tree or shrub in which the nest was A cd $■ a £ cd CD _ 2 Si- T3 cd ■aR °^ placed: plant species, height of plant and nest and o E-g cd ^ CD R J2 3 CD A 1> ■a o distance of nest to main trunk of tree and to the end of K B £ s § E R -C .be, a,>> c 3 ""R 5 ^R branches. To characterize the surroundings of the eg O g ft, '3 -R % a cd s 2 nesting sites, we recorded the minimum distances of _o a-a K TO ■R -R 8 .§ -3 bbJZ.S E eg J3 +- R 3 R _, -R cd CO 8*. the nests to an open landscape or forest patches as well -a ffi £ a as the distance to permanent bodies of water and cd >, nearest patches of important food plants with available 60 _o "o seeds, in particular pines (Pinus mugo, Pinus sylvestris, CD 60 o * c Pinus nigra), herbs and grasses {Taraxacum, Rosmari- >> R o . nus, Rumex, Anthoxanthum, Poa) (Forschler 2001a; U O "o E 60 £ 60 Borras et al. 2003; Forschler and Kalko 2006). a) 1> 0 PS a O O m 3 3 O We also determined clutch size for each nest, which M a) was checked every 3 days during the whole breeding D, U S^ o attempt (nest-building period: 5-10 days; hatching CD o " o _>,(N "33 °° u period: 14+2 days; nestling period: 18+3 days; see « "3 o Forschler 2002a). Finally, we calculated hatching, CD o u-, ^■s 1> breeding and nesting success. Hatching success was CD * B 60 O 28 o E o defined as the number of young in relation to the total 2 a o ^f. U number of eggs per nest; breeding success refers to the R o number of fledglings in relation to the total number of S o S b O ^H found a significant difference with respect to the time o o O N Q9 0 4y Springer 556 J Ornithol (2006) 147:553-564 no significant difference in nesting time between Citril cantly earlier than the start of the nesting period at Finches and Corsican Finches (Table 2). However, the higher altitudes on Sardinia (median: 26.5th pentad) studied Corsican Finches were nesting at a lower ele- and Corsica (median: 24.5th pentad) (Table 4). vation. In Corsica and Sardinia, most birds were building nests between the 21st and 35th pentad (mid- Nesting plants and nest position April to mid-June; median: 25th pentad). Corsican Finches at lower altitudes began to build nests signifi- We found 12 plant species that Citril and Corsican cantly earlier than those at higher elevations (linear Finches used for nesting. Citril Finches nested exclu- regression, #=0.722; p< 0.001; n=46). In the low sively in conifers (Fig. 1), whereas Corsican Finches breeding areas in Capraia, birds started nesting at the were mostly breeding in Tree Heath Erica arborea or end of March (median: 19th pentad). This was signifi- other small bushes (Fig. 2). Nest height between the Table 2 Comparison of nesting site parameters3 of Citril Finches Carduelis [citrinella] citrinella (Black Forest, Pre-Pyrenees) and Corsican Finches Carduelis [citrinella] corsicanus (Corsica, Sardinia, Capraia) Nesting site parameters Citril Finch Corsican Finch MWU test Significance Elevation of nest (m a.s.l.) 1580.3±456.6; «=118 898+375; n=46 7=2147; p< 0.001 *** Nest building (pentad) 26.3+3.8; n=118 24.7±4.6; n=46 7=3,359.5; p=0.111 ns Nest height (m) 7.5±7.3; n=lll 1.4+1.8; M=40 7=1024; p< 0.001 *** Tree height (m) 9.6+1.1; «=111 1.9±2.6; M=46 7=1288; p< 0.001 *** Trunk distance (m) 0.72±0.94; n=108 0.01±0.07; M=40 7=1623; p< 0.001 *** End of branch (m) 0.30±0.16; n=108 0.24±0.14; M=40 7=2535.5; p=0.055 ns Next track structures (m) 34.6±38.8; n=118 31.4+44.9; n=46 7=3710; p=0.757 ns Next open space (m) 11.4+21.3; M=118 0.7±4.4; M=46 7=2646.5; p< 0.001 *** Next forest (m) 41.9+74.4; n=118 837.2±1197.2; n=46 7=5145; p< 0.001 *** Next pines (m) 14.0+36.7; n=118 652.2±1160.1; n=46 7=5683; p< 0.001 *** Next feeding sites (m) 48.6±45.5; n=118 31.6+26.7; n=46 7=3255.5; p=0.048 * Next fresh water (m) 281.4+315.5; «=118 76.3+80.4; n=46 7=2460.5; p< 0.001 *** Clutch size (eggs) 4.1+0.8; «=61 3.6±0.8; n=19 7=582; p=0.034 * Hatching success (%) 61.9+41.9; M=61 55.3±49.7; n=19 7=751; p=0.839 ns Breeding success (%) 47.7±46.0; n=61 40.9±46.6; n=19 7=723.5; p=0.607 ns Nesting success (%) 46.9±50.2; n=81 33.3+48; H=27 7=132; p=0.293 ns *p<0.05; ***p<0.001; ns, not significant 3All of the nesting site parameters are given as the mean : the standard deviation Table 3 Comparison of nesting site parameters3 in Citril Finches Carduelis [citrinella] citrinella in the Black Forest and the Catalonian Pre-Pyrenees Nesting site parameters Black Forest Pre-Pyrenees MWU test Sign Elevation of nest (m a.s.l.) 1018.8±27.2; n=44 1914.1±175.2; H=74 7=1034; p< 0.001 *** Nest building (pentad) 27.8+3.9; n=44 25.4±3.5; M=74 7=3271; p< 0.001 *** Nest height (m) 12.3+10.4; %=38 5.2+3.0; M=73 7=2538.5; p=0.0ll ** Tree height (m) 13.8±10.8; M=37 7.6+2.6; M=74 7=2255; p=0.254 ns Trunk distance (m) 0.11+0.2; M=37 1.04±0.2; M=71 7=1254.5; p< 0.001 *** End of branch (m) 0.40±0.17; M=37 0.24±0.13; M=71 7=2721; p< 0.001 *** Next track structures (m) 37.2+40.9; M=44 33.2+37.8; «=74 7=2767.5; p=0.407 ns Next open space (m) 12.7±17.4; M=44 10.5+23.3; n=74 7=2980.5; p=0.044 * Next forest (m) 23.4+44.2; n=44 52.8±85.9; M=74 7=2196.5; p=0.019 * Next pines (m) 36.9±53.0; n=44 0.4±2.5; M=74 7=3370; p< 0.001 *** Next feeding sites (m) 61.1±56.8; M=44 41.1±35.5; M=74 7=2861.5; p=0.176 ns Next fresh water (m) 61.1+61.2; M=44 412.3±332.7; H=74 7=1207; p< 0.001 *** Clutch size (eggs) 3.7+1.1; n=ll 4.1±0.7; M=50 7=283; p=0.280 ns Hatching success (%) 67.7±38.6; n=ll 60.6±42.9; n=50 7=751; p=0.839 ns Breeding success (%) 48.3±43.0; n=ll 47.6+47.1; n=50 7=723.5; p=0.607 ns Nesting success (%) 35.5±48.6; n=31 54+50.3; n=50 7=1127; p=0.164 ns *p<0.05; ***p<0.001; ns, not significant 3All of the nesting site parameters are given as the mean : the standard deviation 4y Springer J Ornithol (2006) 147:553-564 557 m m m in in !S m )n in "Dm o o o o o o o o o d P o P o o o d d d d d d d d d Q o o ° d ° d V V V «, V «. v 2 « a CS td '3 60 00 a. a U X! c r-i ca o \o ■* r- ri -t 0 II N II ii II c CI ~ ~ ~. c c 3C -C _3 c n o p~" s" r; co d r-i d -r r- ci d o rN r-i CI r-l +1 +i +1 +1 +1 +1 +1 +1 +1 +1 +1 ON !-) o "+ f ■; r> ri ■fe c - o CS o\ CN r>%; ri s, U "gs S r-i U II 00 ry: r-l II M II DC II T3 II = II ~ t c r ~ r T3 o c II P~ CN = si C o ~ r-l d' d \C 00 CO d' o - C- d I : ■T) o r> m or T3 co o o n e-> m <-> P~ i-; T -f ~T c E CS +1 +i +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 re c '3 o o M ■JD O t- CO ro O C3 ■3 CO \r. c-, m r; oo 0 U o 'f r-l -*■ r-l Tf r-i CN r-i CN CN r CN r-l o II ST C S K II o C S ~ n s f» ■* \o a o -! '5 ca l> ci d oo o o \o rr, m ■* ■* a CS +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 a. ro o l> CM o\ O -* C~ 00 o o CN -& d cS o o o -* ■* CO O 00 d o d c VI 0 ft c 0 « 8 &^ d 00 cm c CL CS 00 U d E a, S.B 3 X u VI o eg1 dS ft «1) o d CO g U d 6 '•3 d 00 co T 00 o 3 ■<3£ d 60 m open sp track sti forest (r '5. 0> ,d M O 3 .3 H O o _ JZ 3 "° XXX x X u vi 5 C d 1) U o o u o ft < H z W H W Z Z Z Z Z u x m z 8 Springer 55S J Ornithol (2006) 147:553-564 uncinata were nesting lower than birds in the Black Forest, but the differences were not significant (Ta- ■ Pirm rwgowonala ble 3). Corsican Finches showed less variation in the ■ Pinus mugo rotundate height of nest and preferred significantly lower nesting □ Picea sties sites than Citril Finches (Table 2). Although Corsican s Pinus sylvestris Finches nested mainly in Tree Heath, even when other Q Pinus nigra suitable trees, such as pines, were present (Table 4), □ Aties alts there were differences in nest height between the studied populations of Corsican Finches. Birds on Ca- praia nested significantly higher than birds on Sardinia (Table 4). The height of the nesting plants correlated Fig. 1 Nesting trees used by Citril Finch Carduelis [citrinella] positively with nest height, with significant differences citrinella in the Black Forest (n=44) and the Pre-Pyrenees («=74) between Citril and Corsican Finches (Table 2). No significant difference was found in the height of nesting plants in the Pre-Pyrenees and the Black Forest (Ta- ble 3), but there were significant differences between the heights of nesting plants in the three Corsican D Erica arborea Finch populations. Birds on Sardinia nested in signifi- cantly lower bushes than those on Capraia and Corsica E Rubits ottofoHus 0 Genista salimartnii (Table 4). N Juniperus c. nana A comparison of the average distance of the nest ■ Pmusn&ra from the main trunk of a nesting tree revealed that QPirws pinaster birds from the Black Forest and the Pyrenees placed o Quercus ilex their nests differently. Nominate birds from the Black Forest nested significantly closer to the trunk than those from the Pyrenees (Table 3). In the Black Forest, most nests were situated in the crown, whereas finches Fig. 2 Nesting shrubs and trees used by Corsican Finch in the Pre-Pyrenees nested in denser parts of the outer Carduelis [citrinella] corsicanus on Corsica (n=24), Sardinia branches as well. On Corsica, Sardinia and Capraia, it (M=12) and Capraia (n=10) was not possible to measure the distances from the nests to a discernable trunk, as most nests were placed two species varied strongly (Fig. 3). In the Black For- in the shrub Tree Heath, which does not possess a est, there was a large variation in nest height that was a discernable, central trunk (Table 4). result of differences in the two main nesting trees: the Citril and Corsican Finches showed a nearly signif- spruce Picea abies and the Mountain Pine Pinus mugo icant difference with respect to the distance from the rotundatalpumilio (Table 2). Birds in the Pre-Pyrenees nest to the end of branches (Table 2), and Citril Fin- that nested mainly in Mountain Pine Pinus mugo ches showed a significant difference with respect to Fig. 3 Nest heights of the nominate Citril Finch Carduelis [citrinella] citrinella in the Black Forest (n=38) and in the Pre-Pyrenees (M=73) and of Corsican Finch Carduelis [citrinella] .=■ 20 corsicanus from Corsica (M=21), Sardinia (n=12) and Capraia (n=l). In the Black z 10 Forest the larger variation in nest height is due to the different nesting trees - Mountain Pine {lower nests) and Spruce (higher nests) Black Forest P re.Pyrenees Corsica Sardinia Capraia Research area 4y Springer J Ornithol (2006) 147:553-564 559 Fig. 4 Distances of Citril Finch Carduelis [citrinella] citrinella nests to the nearest forest patches in the Black Forest (BC, M=44) and the Pre-Pyrenees (PC, n=74) and of Corsican Finch Carduelis [citrinella] corsicanus nests on Corsica, Sardinia and Capraia (CC, n=46) 500 1000 1500 2000 2500 3000 3500 D istan ce in m distances from the nest to the outer branches, with the Poa annua (Poaceae) and Briza maxima (Poaceae) distances being shorter for Citril Finches in the Pre- (Forschler and Kalko 2006). Corsican Finches nested Pyrenees than in the Black Forest (Table 3). The nests significantly closer to their feeding places than Citril of Corsican Finches on Sardinia were situated signifi- Finches (Table 2). No significant difference was found cantly closer to the end of branches than nests on between the sub-populations of the Black Forest and Corsica (Table 4). the Pre-Pyrenees (Table 3). However, Corsican Fin- ches nested significantly closer to food resources than Parameters of nest surroundings birds on Capraia and Sardinia (Table 4). The Citril Finches studied nested significantly fur- Corsican Finches nested significantly closer to open ther away from the next water source than Corsican landscape than Citril Finches. Accordingly, the dis- Finches (Table 2). Birds in the Black Forest nested tance to next forest patches was significantly higher for closer to water than birds in the Pre-Pyrenees (Ta- Corsican Finches than for Citril Finches (Table 2; ble 3). Nests in Corsica were situated closer to water Fig. 4). Furthermore, we found a significant difference sources than those on Capraia and Sardinia (Table 4). between sub-populations in the Black Forest and the Pre-Pyrenees, with birds in the Pre-Pyrenees nesting Clutch size and reproductive success closer to the open landscape and further away from the closed forest (Table 3). No significant difference was In total, we obtained data on clutch size, hatching and found between Corsican Finch sub-populations with breeding success for 80 nests of nominate Citril Finches respect to distance to open landscape. However, birds and Corsican Finches. An overall count of 318 eggs led on Capraia nested significantly further away from next to an average clutch size of 4.0 (±0.8) eggs per nest. forest patches than birds on Sardinia and Corsica From 318 eggs, 188 young hatched (59%) and 149 (Table 4). Citril Finches nested significantly closer to fledged (46%). There was a significant difference in pines, which provide one of their main food, than clutch size between Citril Finches and Corsican Fin- Corsican Finches (Table 2). Additionally, birds in the ches (Table 2). Citril Finches laid on average more Pre-Pyrenees nested closer to pines than birds in the eggs than Corsican Finches. Clutch size in the Black Black Forest (Table 3). Significant differences were Forest was not significantly different from that in the also found between the sub-populations of Citril Fin- Pre-Pyrenees (Table 3). In addition, no significant ches, with birds on Corsica and Sardinia nesting closer difference was found the three Corsican Finch sub- to pines than those on Capraia (Table 4), which occa- populations (Table 4). In total, hatching, breeding and sionally were found to nest at some distance from any nesting success were higher in nominate Citril Finches pine. than in Corsican Finches. However, the differences In all of the study areas, the birds foraged during were not significant, neither between the distinct sub- breeding season on seeds of grasses and herbs, in populations nor among the two species (Tables 2-4). A particular those of Taraxacum officinale (Asteraceae), pairwise %^-test revealed that there was a significant Rumex acetosa (Polygonaceae), Capsella rubella difference in nesting success between the birds of the (Brassicaceae), Anthoxanthum odoratum (Poaceae), Pre-Pyrenees and Corsica. 4y Springer 560 J Ornithol (2006) 147:553-564 Discussion Nesting plants and nest position Breeding phenology and altitude The selection of plants for nesting differed consider- ably between Corsican Finches and nominate Citril We did not find significant differences between the Finches. The preference of Citril Finches in the Black mainland Citril Finches and insular Corsican Finches Forest and the Pre-Pyrenees for conifers (Pinaceae: P. with respect to the start of nest building. This result mugo, P. abies) has also been observed in the Alps is in contrast to those obtained in studies comparing (Glutz von Blotzheim 1962; Maestri et al. 1989; Kaniss insular and mainland populations of other species, and Pfeiffer 1994; Glutz von Blotzheim and Bauer among which the Blue Tits Parus caeruleus (Blondel 1997). In contrast, we found that Corsican Finches et al. 1991). Both insular and mainland finches mainly prefer to nest in shrubby habitats, usually Tree Heath, nested between the 21st (11-15 April) and 35th (20- broom or bramble, rather than in pine species that are 24 June) pentad. However, the Corsican Finches of also in the area. This confirms the little data available our study clearly positioned their nests at lower ele- on nesting site selection of Corsican Finches (Armitage vations (Table 2), which translates into the relative 1937; Marzocchi 1990; Cramp and Perrins 1994; Thi- date of breeding being later than for those on the bault and Bonaccorsi 1999). Consequently, the nests of mainland; this result fits with other components of Corsican Finches were also situated much lower in the insular patterns (Blondel et al. 1991, 1999; Blondel vegetation than those of Citril Finches. However, some 2000). Only finches on Capraia appear to start with differences were found between the two studied sub- earlier broods in March due to their breeding habitat populations of Citril Finches, as birds in the northern at lower elevations. As most Citril and Corsican Black Forest tended to utilize the crown of higher trees Finches live in mountainous areas, the start of nest and locate their nest close to the trunk, whereas Pre- building depends strongly on the climate (Glutz von Pyrenean finches often preferred the ends of the lower Blotzheim and Bauer 1997; Holzinger 1997) and on lateral branches. food availability (Forschler and Kalko 2006). A small These adaptations in nesting behaviour may be portion of both forms starts building nests as early caused by local predators. Avian predators, such as opportunistic breeders at lower elevations under good corvids, are supposed to predate elevated nests more food conditions, including those in the Black Pine P. easily, whereas mammalian predators prefer lower nigra forest on Catalunya (Borras and Senar 1991) nests (Best and Stauffer 1980; Sockmann 1997; Li- and in low coastal Macchia on Corsica and Capraia ebezeit and George 2002; Schafer and Barkow (Jourdain 1911; Thibault and Bonaccorsi 1999; per- unpublished). Consequently, finches nest higher in sonal observation). Similarly, a few early broods have areas inhabited by mammalian predators, like the been recorded in the Alps (Jouard 1930; Glutz von Red Squirrel Sciurus vulgaris and Dormouse Gliridae, Blotzheim 1962; Glutz von Blotzheim and Bauer while in areas with more avian predators, birds are 1997). supposed to choose lower nesting sites. These effects Corsican Finches are reported to be one of the few could lead to the differences in nest height and po- forest species of the Mediterranean islands that exhibit sition among the various study populations. Accord- an exceptionally large habitat-niche expansion (Blon- ingly, the placement of nests in the crowns of trees of del et al. 1988). Within the context of habitat descrip- the northern Black Forest is perhaps associated with tions, Corsican Finches are characterized, contrary to the high pressure exerted by mammalian predators, nominate Citril Finches, as nesting in all types of low as the Red Squirrel is particularly abundant in that vegetation, from coastal areas up to locations area (personal observation), whereas selection of 2250 m a.s.l. (Thibault 1983; Cramp and Perrins 1994; lower heights for nests on lateral branches in the Pre- Pasquet 1994; Thibault and Bonaccorsi 1999). Al- Pyrenees might be associated with the higher local though we found a general trend supporting this abundance of avian predators, such as the Eurasian observation, most of the Corsican Finches on Corsica Jay Garrulus glandarius (Forschler 2002b). Addi- and Sardinia also nest in mountainous habitats domi- tionally, even ants may play a role in predation and nated by pines and Tree Heath between 800 and could force Citril Finches to nest on outer branches 1500 m a.s.l., which is more typical of Citril Finch because nests near the trunk and on large lateral habitats (personal observation). Only a small portion branches are more vulnerable to ant runs (Forschler of Corsican Finches select lower areas for breeding, et al. 2001). probably due to spill-over (Blondel et al. 1988) from As mammalian predators prefer lower nests and high-density populations in the mountains. avian predators prefer higher nests, the low nests on 4y Springer J Ornithol (2006) 147:553-564 561 Corsica and Sardinia could also be due to the scarcity have to fly large distances to reach fresh water of mammalian predators on these islands (no Red sources due to the general scarcity of water bodies. Squirrels), whereas the Eurasian Jay is abundant in the woodlands of Corsica and Sardinia (personal observa- tion). The higher placement of nests on Capraia may Clutch size and reproductive success be explained by the lack of jays and the abundant presence of snakes. The Dark Green Snake Coluber Overall, the average clutch sizes of four eggs per nest viridiflavus is particularly abundant on Capraia (Lam- (range: 2-5 eggs) found in this study confirm the little bertini 2002; personal observation) and is known to be data available for Corsican (Armitage 1937; Thibault an important predator of eggs and nestlings of pas- and Bonaccorsi 1999) and Citril Finches (Glutz von serine birds (Delaugerre and Cheylan 1992), especially Blotzheim 1962; Maestri et al. 1989; Kaniss and Pfeiffer on Montechristo (Bruno 1975), the neighbouring island 1994; Glutz von Blotzheim and Bauer 1997; Forschler closest to Capraia. 2002a) as well as for other finches (genera: Carduelis, Climate factors also play a role in nest site selection Serinus, Loxia) that also show similar clutch sizes and breeding success (Ricklefs 1969; Frey 1989a, b; (Glutz von Blotzheim and Bauer 1997). During a long- Bairlein 1996; Rauter et al. 2002; Forschler et al. 2005). term study of the Serin Serinus serinus, an average Thus, the prevalence of strong winds in the Black clutch size of four eggs per nest (range: 2-5 eggs) was Forest may force Citril Finches to nest in denser and recorded by Gnielka (1978), and Khoury (1998, 2001) more stable parts of trees that are closer to the trunk. found similar clutch sizes for the Syrian Serin, with an In contrast, Corsican Finches may have chosen Tree average of 4.1 eggs per nest (range: 4-5 eggs). Heath thickets over pine trees because the interior of We found significant differences in clutch size among the thickets offer relatively calmer areas that are pro- the study populations. In total, mainland birds laid sig- tected against the wind while at the same time pro- nificantly more eggs (4.1+0.7) than those from island viding safety from predators. populations (3.6+0.8). Smaller clutch sizes have been observed for various bird populations on Corsica in comparison to mainland populations (Martin 1992) and Nest surrounding is generally explained by differences in habitat selection and food availability (Martin 1982,1992; Blondel et al. At all of the study sites, the nests of Corsican Finches 1991). were found in more open areas, away from closed Very little information has been published on forest structures, in comparison to those of Citril thebreeding and nesting success of Citril and Corsi- Finches. This corresponds with the more open land- can Finches (Maestri et al. 1989; Forschler 2002a). In scapes of Corsica, Sardinia and Capraia, which in our study, hatching and breeding success did not some cases also lack any forest (e.g. Capraia). Al- differ significantly between these two finches. though Corsican Finches use pines for nesting - as do Breeding success of all nests was about 46% (41% Citril Finches - it was evident that the dependence of for Corsican Finches; 48% for Citril Finches), which the former on the presence of pine trees is not as is similar to the 45% recorded by Maestri et al. high as that of the latter (Forschler 2002a; Borras (1989) in the Italian Alps. Nesting success followed et al. 2003). Corsican Finches can also be found breeding success: there was no significant difference nesting far away from any pine forest (e.g. Capraia). in nesting success, but it was clearly higher in the Pine seeds do not play as important role in the diet mainland Citril (47%) than in the insular Corsican of the Corsican Finch nutrition as they do in that of Finches (33%). However, these results were obtained the Citril Finches (Forschler and Kalko 2006). In for each area from one breeding season only, so they contrast, the flight distance to other important food must be treated cautiously since reproductive success items (herbs and grasses) was lower for Corsican may vary between years due to severe weather in the Finches than for Citril Finches. The fact that Corsi- mountains (Bezzel and Brandl 1988; personal obser- can Finches generally nest closer to water sources vation). Strong weather conditions during the breed- than Citril Finches may be linked to hotter conditions ing season may have considerable effects on in the breeding areas. However, the difference may reproductive success (Forschler et al. 2005). A good also be related to the - in general -lower frequency example of this is the spring of 2001 in Corsica, when of water bodies in the studied sites of the mainland hatching success seemed to be lower than normal, Citril Finches. In particular, birds in the Pre-Pyrenees probably as a result of an unusually low availability 4y Springer 562 J Ornithol (2006) 147:553-564 of food (Forschler and Kalko 2006) and severe populations are not much larger than those between weather during the hatching period. the two mainland ones. The splitting of completely allopatric (sub-)species, as in the case of the Citril and Corsican Finches, is therefore subjective and rather Conclusions arbitrary. Our data on the ecological differentiation of insular Citril Finches and Corsican Finches are currently Corsican Finches fits models of the so-called insular treated as two species with independent evolutionary syndrome, which predicts changes in morphology, histories (Sangster 2000; Sangster et al. 2002). How- demography and behaviour due to reduced dispersal ever, both forms show similar breeding ecology and (Blondel 2000). The shift from dispersal to sedentari- behaviour. The main ecological difference that we ness and habitat fidelity on islands are potential factors observed was that Corsican Finches have expanded for population differentiation at much smaller spatial their range from exclusively half-open forests (Citril scales than on mainland regions, which enhances Finch) to more open landscapes located as some dis- within-species diversity (Blondel et al. 1999; Blondel tance from forest patches. We suggest that the key 2000). In the light of this model the obvious differen- component in this scenario is most likely the extraor- tiation of Corsican Finches with respect to their dinarily high abundance of Tree Heath in the moun- mainland counterparts in morphology (distinct plum- tainous regions of all the Mediterranean islands age coloration, smaller size), demography (lower occupied by Corsican Finches. Corsican Finches use hatching and breeding success, smaller clutches, higher Tree Heath as a kind of 'pine substitute', which enables portion of non-breeding birds) and behaviour (in- them to nest safely even when the next forest patch is creased niche breadth, niche expansion into open and relatively far away. Consequently, Corsican Finches lower habitats, preference for Tree Heath) may be also exploit - in contrast to Citril Finches - patches easily explained. located some distance away from forested areas, pro- vided these offer abundant and nutritious food re- sources. This behavioural difference is reflected as well Zusammenfassung by the variation in the composition of the diet during breeding season (Forschler and Kalko 2006). While Brutokologie und Nistplatzwahl in allopatrischen Citril Finches in the Black Forest and the Pre-Pyrenees Festlandspopulationen des Zitronenzeisiges feed predominantly on pine seeds, birds on Corsica and Carduelis [citrinella] citrinella und Capraia consume mainly herb seeds (Shepherd's purse Inselpopulationen des Korsenzeisiges Carduelis C. rubella; Rosemary Rosmarinus officinalis). How- [citrinella] corsicanus ever, as the Corsican Finches in Sardinia have also been observed to feed predominantly on pine seeds, Wir untersuchten die Brutokologie und Nistplatzwahl this trait can not be considered to be a general von Festlandspopulationen des Zitronenzeisiges und behavioural pattern that separates the two forms. Inselpopulationen des Korsenzeisiges. Beide fiir We consider the observed niche expansion by Europa endemischen Formen wurden lange Zeit als Corsican Finches (Blondel et al. 1988) to be a result of Unterarten betrachtet. Aufgrund von genetischen variations in habitat conditions rather than result of Untersuchungen werden sie neuerdings jedoch als ei- fundamental behavioural differentiation between the genstandige Arten behandelt. Wir beschaftigten uns two (sub-)species. In our opinion, niche expansion sowohl mit Variationen in der Brutokologie und der should not be used as an argument for taxonomic sta- Nistplatzwahl in verschiedenen Lokalpopulationen tus (see Sangster 2000), since insular sub-populations beider Arten, als auch mit der okologischen Differen- of other bird species, such as the Coal Tit Parus ater zierung zwischen Zitronenzeisigen und Korsenzeisi- and the Eurasian Tree-Creeper Certhia familiaris, dif- gen. In der Brutbiologie fanden wir groBe fer as well comparably from their mainland counter- Ubereinstimmungen, die die Ahnlichkeit der Leben- parts in morphology, habitat selection and vocalization sweise beider Arten unterstreicht. Beide sind als (Chappuis 1976; Martin 1992; Pasquet and Thibault Bergvogel gut an die lokalen Bedingungen ihrer Ge- 1997). In fact, most insular forms differ in many traits birge angepasst. Einige wenige Unterschiede finden from their mainland conspecifics and are not consid- sich in der Nistplatzwahl der untersuchten Populatio- ered to be full species (Blondel et al. 1999; Blondel nen, was auf die jeweiligen lokalen Umweltbedingun- 2000). Based on the results of our study, the ecological gen zuruckgefuhrt werden kann. Im Gegensatz zum in differences between the mainland and insular sub- Koniferen brutenden Zitronenzeisig bevorzugt der 4y Springer J Ornithol (2006) 147:553-564 563 Korsenzeisig die Baumheide als Neststandort, auch Blondel J (2000) Evolution and ecology of birds on islands: wenn geniigend Kiefern in der Umgebung verfiigbar trends and prospects. Vie Milieu 50:205-220 Blondel J, Chessel D, Frochot B (1988) Bird species impover- sind. Dies fiihrt dazu, dass Korsenzeisige auch in of- ishment, niche expansion, and density inflation in Mediter- fenerem Gelande mit niedrigem Gebiisch und Macchia ranean island habitats. Ecology 69:1899-1917 vorkommen, wahrend Zitronenzeisige auf halboffene Blondel J, Dervieux A, Maistre AM, Perret P (1991) Feeding Nadelwalder (vor allem Kiefernwalder) beschrankt ecology and life history variation of the blue tit in Medi- terranean deciduous and sclerophyllous habitats. Oecologia sind. Diese Verhaltensanderung ermoglicht es dem 88:9-14 Korsenzeisig in einer Art Nischenerweiterung auch Blondel J, Dias PC, Perret P, Masitre M, Lambrechts MM (1999) tiefere Lagen mit entsprechender Gebiischvegetation zu Selection-based biodiversity at small scale in a low-dispers- besiedeln. Dabei ist diese Nischenerweiterung unserer ing insular bird. Science 285:1399-1402 Borras A, Junyent (1993) Vertebrats de la Catalunya Central. Meinung nach nicht als Argument fur die taxonomische Manresa, Barcelona Einordnung des Korsenzeisiges geeignet (vgl. Sangster Borras A, Senar JC (1991) Opportunistic breeding of the citril 2000). Die gefundenen okologischen Unterschiede der finch Serinus citrinella. J Ornithol 132:285-289 beiden (Unter-) Arten konnen vielmehr als das Ergebnis Borras A, Cabrera T, Cabrera J, Senar JC (2003) The food of the Citril Finch in the Pyrenees and the role of Pinus seeds as a des sogenannten 'Insel-Syndroms' betrachtet werden, key resource. J Ornithol 144:345-353 dass schnelle Anderungen in Morphologic, Demogra- Bruno S (1975) Note rassuntive sull' Erpetofauna dell' Isola di phic und Verhalten in Lokalpopulationen aufgrund Montechristo (Archipelago Toscana, Mare Tirreno). Lav reduzierter Dispersion vorhersagt. Soc Ital Biogeo 5:1-98 Bruno C, Dupre G, Giorgetti G, Giorgetti JP, Alesandri J (2001) Chi tempu face? Meteorologie, climat et microclimats de la Acknowledgements The field work was kindly supported by Corse. Centre Regional de Documentation Pedagogique de Antonio Borras, Toni Cabrera, Josep Cabrera and Juan Carlos Corse, Ajaccio Senar (Museu Ciencies Naturals, Barcelona, Spain), Philippe Chappuis C (1976) Origine et evolution des vocalisations de Ferret (Centre d' Ecology Fonctionelle et Evolutive, Montpel- certains oiseaux de Corse et des Baleares. Alauda 44:475- lier, France), Nicola Baccetti and Fernando Spina (Istituto naz- 495 ionale per la Fauna Selvatica, Bologna, Italy), Sergio Nissardi Cramp S, Perrins CM (1994) The birds of the western Palearctic, (Cagliari, Italy), Ulrich Dorka (Tubingen, Germany) and Jiirgen vol. 8. Oxford University Press, Oxford Klager (Baiersbronn, Germany). Peter Berthold (MPI Vogelw- Delaugerre M, Cheylan M (1992) Batraciens et reptiles de Corse. arte Radolfzell, Germany), Ulrich Dorka (Tubingen, Germany) Pare Naturel Regional de Corse. Pampelune, Spain and Thomas Schafer (MPI Vogelwarte Radolfzell, Germany) Forschler M (2000) Untersuchungen zur Brutphanologie, Nah- provided helpful comments on the manuscript. Alexander Tahori rungswahl und Habitatwahl des Zitronengirlitzes Serinus (Tel Aviv, Israel) kindly improved the English. We thank Franz citrinella im Nordschwarzwald. Diplomarbeit, Universitat Bairlein and Jacques Blondel for helpful comments on the Tubingen, Tubingen manuscript. The study was conducted with financial support from Forschler M (2001a) Brutzeitliche Nahrungswahl des Zitronen- the Max Planck Research Centre for Ornithology, Vogelwarte girlitzes Serinus citrinella im Nordschwarzwald. Vogelwelt Radolfzell and the Landesgraduiertenforderung Baden-Wiirt- 122:265-272 temberg, University of Ulm, to M. Forschler. Forschler M (2001b) Witterungsbedingte Ausweichbewegungen des Zitronengirlitzes Serinus citrinella im Nordschwarzwald. Ornithol Beob 98:209-214 Forschler M (2002a) Brutbiologie des Zitronengirlitzes Serinus References citrinella im Nordschwarzwald. Ornithol Beob 99:19-32 Forschler M (2002b) Predation strategy of the Eurasian Jay Arcamone E (1993) Venturone. In: Meschini E, Frugis S (eds) Garrulus glandarius and antipredator response by the Citril Atlante degli uccelli nidificanti in Italia. Instituto per la Finch Serinus citrinella. Rev Catalana Ornithol 19:41-43 Fauna Selvatica. Ozzano Emilia, Italy Forschler MI, Kalko EKV (2006) Macrogeographic variations in Armitage J (1937) File notes on the Corsican Citril Finch. Br food choice of mainland citril finches Carduelis [citrinella] Birds 31:98-100 citrinella versus insular Corsican (citril) finches Carduelis Baccetti N, Marki H (1997) Citril Finch. In: Hagemeijer WJM, [citrinella] corsicanus. J Ornithol 147:441-447 Blair MJ (eds) The EBCC atlas of European breeding birds. Forschler M, Cabrera J, Cabrera T, Borras A (2001) Loss of a T and AD. Poyser, London, p 711 Citril Finch Serinus citrinella nest possibly as a result of ant Bairlein F (1996) Okologie der Vogel. Fischer, Stuttgart predation. Butll GCA 18:43-44 Bauer HG, Berthold P (1996) Die Brutvogel Mitteleuropas. Forschler MI, Borras A, Cabrera J, Cabrera T, Senar JC (2005) Aula, Wiesbaden Inter-locality variation in reproductive success of the citril Best LB, Stauffer DF (1980) Factors affecting nesting success in finch Serinus citrinella. J Ornithol 146:137-140 riparian bird communities. Condor 82:149-158 Frey M (1989a) Nahrungsokologie und Raumnutzung einer Bezzel E (1993) Kompendium der Vogel Mitteleuropas. Aula, subalpinen Population des Hanflings Carduelis cannabina. Wiesbaden Orn Beob 86:291-305 Bezzel E, Brandl R (1988) Der Zitronengirlitz im Werdenfels- Frey M (1989b) Brutbiologie des Hanflings Carduelis cannabina ener Land, Oberbayern. Anz Orn Ges Bayern 27:45-65 unter den Einflussen des Gebirgsklimas. Orn Beob 86:265- Blondel J (1985) Habitat selection in island versus mainland 289 birds. In: Cody ML (eds) Habitat selection in birds. Aca- Glutz von Blotzheim UN (1962) Die Brutvogel der Schweiz. demic, London, pp 477-516 Schweizerische Vogelwarte, Sempach 4y Springer 564 J Ornithol (2006) 147:553-564 Glutz von Blotzheim UN, Bauer KM (1997) Handbuch der Martin JL (1992) Niche expansion in an insular bird community: Vogel Mitteleuropas. Band 14. Aula, Wiesbaden an autoecological perspective. J Biogeogr 19:375-381 Gnielka R (1978) Zur Okologie und Brutbiologie des Girlitzes Marzocchi JF (1990) Contribution a l'etude de l'avifaune du Cap Serinus serinus. Orn Mitt 30:81-90 Corse. Editions Marzocchi, Bastia Gutierrez E (1991) Climate tree-growth relationships for Pinus Moltoni E (1975) L'avifauna dell'isola di Capraia (Arcipelago uncinata in the Spanish pre-Pyrenees. Acta Oecol 12:213- toscano). Riv Ital Orn 45:97-217 225 Pasquet E (1994) Venturon montagnard. In: Yeatman-Bertholet Holzinger J (1997) Die Vogel Baden-Wiirttembergs. Band 3.2. D (ed) Nouvel atlas des oiseaux nicheurs de France 1985- Eugen Ulmer, Stuttgart 1989. Societe Ornithologique de France, Paris Jouard H (1930) Der Zitronenzeisig als Winterbriiter. Ornithol Pasquet E, Thibault JC (1997) Genetical differences among Monatsber 38:137-139 mainland and insular forms in the Citril Finch Serinus ci- Jourdain FCR (1911) Notes on the ornithology of corse. Ibis trinella. Ibis 139:679-684 53:189-208 Rauter CM, Reyer HU, Bollmann K (2002) Selection through Kaniss M, Pfeiffer R (1994) Notizen zur Brutbiologie des Zi- predation, snowfall and microclimate on nest-site prefer- tronengirlitzes Serinus citrinella. Orn Anz 33:63-65 ences in the Water Pipit Anthus spinoletta. Ibis 144:433^-44 Khoury F (1998) Habitatwahl und Nahrungsokologie des Zed- Ricklefs RE (1969) An analysis of nesting mortality in birds. erngirlitzes Serinus syriacus in Jordanien. PhD thesis, Rhe- Smithson Contrib Zool 9 inische Friedrich-Wilhelms-Universitat Bonn Sangster G (2000) Genetic distance as a test of species bound- Khoury F (2001) The breeding ecology of Syrian Serin Serinus aries in the Citril Finch Serinus citrinella: a critique and syriacus in Jordan. Sandgrouse 23:68-69 taxonomic reinterpretation. Ibis 142:487-490 Lambertini M (2000) Capraia. Terra Mare. Pacini Editore, Os- Sangster G, Knox AG, Helbig AJ, Parkin DT (2002) Taxonomic pedaletto, Pisa recommendations for European Birds. Ibis 144:156 Lambertini M (2002) The Tuscan Arcipelago and the National Sockmann KW (1997) Variation of life-history traits and nest- Park. Pacini Editore, Ospedaletto, Pisa site selection affects risk of nest predation in the California Liebezeit JR, George TL (2002) Nest predators, nest-site selec- Gnatcatcher. Auk 114:324-332 tion, and nesting success of the Dusky Flycatcher in a Thibault JC (1983) Les oiseaux de la Corse. Pare Naturel Re- managed ponderosa pine forest. Condor 104:507-517 gional de la Corse, Ajaccio Maestri F, Voltolini L, Lo Valvo F (1989) Biologia riproduttiva Thibault JC, Bonaccorsi G (1999) The birds of Corsica. B.O.U. di una communita di Fringillidi in un Mugeto delle Alpi Checklist No. 17. British Ornithologists' Union, Tring, Retiche (Sondrio). Riv Ital Orn 59:159-171 Herts, UK Martin JL (1982) L'infiltration des oiseaux forestieres dans les Voous KH (i960) Atlas of European birds. Nelson, London milieux buissonnants de Corse. Rev Ecol (Terre Vie) Whitehead J (1885) Ornithological notes from Corsica. Ibis 27:39 36:397-419 4y Springer