sign in sign up
<<
Home , Alpine chough, Great bustard, Great rosefinch, Little bustard, Mammoth steppe, Ring ouzel

T y r b e r g : Glacial Relicts in© Ornithologischethe West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 29

Om. Verh. 25, 1991: 29-49

Arctic, Montane and as Glacial relicts in the West Palearctic

By Tommy Tyrberg

1. Introduction

The theory that speciation of forest birds putative refuge areas, especially Southern has taken place in isolated forest refugia , the Southern and Asia Mi­ during glacial periods is well established nor. both in the Palearctic and elsewhere (e. g. It is however not generally appreciated H a ffer 1974, H a r r iso n 1982, M o r e a u 1966). that a similar process of ränge fragmenta- While this concept is very likely correct in tion and creation of relict populations by its essentials it is not verifiable by fossil evi- the glacial/interglacial climatic cycle also dence even in the West Palearctic (which affects arctic, montane and steppe birds, has by far the richest avian fos­ and that at least in the West Palearctic the sil record in the World), due to the virtual fossil record of this process is fairly good. absence of Pleistocene avifaunas from the

2. The “ steppe” avifauna

It has long been known that steppe condi- they contain f orms which today are found in tions have prevailed over large parts of Eu- quite different (an extreme ex- rope during glacial periods. This steppe ample is MuskoxOvibos moschatus occur- which at times extended essentially unbro- ing together with Lion Felis leo). This also ken from the Atlantic to the Mackenzie Ba­ applies to the avifauna. Thus among the gal- sin of northwestern was characteri- linaceous birds (which were apparently zed by a cold, dry Continental climate, a V e­ abundant on the mammoth steppe) the nor- getation usually dominated by Artem isia mally dominant Ptarmigan Lagopus mutus and a spectacular mammalian megafauna. and Grouse Lagopus lagopus are This which has no exact counterpart regularily found together with Partridge today is known by a variety of names: Perdix perdix, Quail Coturnix coturnix and “steppe ”, “arctic steppe”, “perigla- Black Grouse Tetrao tetrix. cial steppe”, “loess steppe”, “arctic grass­ Other characteristic members of the land” and “mammoth steppe” mammoth steppe avifauna are raptors The avifauna of the mammoth steppe has (especially Golden Aquila chrysaetos, received much less attention than the mam- K estrel Falco tinnunculus, and Snowy Owl mals but at least in Europe it is quite well Nyctea scandiaca), Larks (usually Skylark known from several hundred sites. Alauda arvensis though other also It has often been noted that glacial mam­ occur) and corvids (particularily Raven malian faunas are “disharmonious” i. e. corax, Red-billed Pyrrho- 30 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991 corax pyrrhocorax and Alpine Chough Melanocorypha and Calandrella larks and Pyrrhocorax graculus). In addition to these Rosy Starling Sturnus roseus. largely arctic-alpine forms a number of There is a clear climatic gradient in the “steppe” and “mediterranean” species are composition of the faunas, the “mediterra­ also regularily found, examples being nean” species being more common in Sou­ Long-legged Buzzard Buteo rufinus, Pallid thern Europe, but there is an extensive over- H arri er Circus macrourus, Red-footed Fal- lap of northern and Southern forms with con Falco vespertinus, Alectoris partridges, birds like Snowy Owl and Alectoris sp. be­ ing regularily found in the same deposits.

3. Relict Distribution Patterns

Among the biogeographically significant W ater PipitAnthus spinoletta/petrosus, Al­ characteristics of the mammoth steppe avi- AccentorPrunella collaris, Citril fauna is the considerable fossil evidence Serinus citrinella, Snow Finch M ontifrin- that the present discontinuous ranges of a gilla nivalis, Red-billed Chough Pyrrhoco­ number of species in the West Palearc- rax pyrrhocorax and Alpine Chough Pyrr­ tic may plausibly be regarded as relicts of a hocorax graculus. continuous ränge during the last (Würmian) The Ring Ouzel Turdus torquatus also be- . These birds can be divided longs to this group but the fossil record of into three groups based on their present dis- this species is too questionable to warrant tribution: analysis since the european Turdus species 1. Predominantlyarcticbirdswithiso- are hardly osteologically distinguishable lated Southern montane populations: except by size. Ptarm igan Lagopus mutus, Willow Grouse 3. Eastern steppe species with an iso­ Lagopus lagopus, D otterelEudromias mori- lated population in the nellus and Shore Lark Eremophila alpe- and/or Maghreb: Long-legged Buzzard B u­ stris. teo rufinus, Imperial Eagle Aquila heliaca, 2. M ontane species with a number of Demoiselle crane Anthropoides virgo, Pin- isolated populations in different mountain tailed Sandgrouse Pterocles alchata, and ranges and in some cases also a Coastal po- Black-bellied Sandgrouse Pterocles orien- pulation along the Atlantic coast: Rock/ talis.

4. The Würmian fossil record

There is some confusion about the terms on data from some 850 sites with Würmian used to designate the and avifaunas. The quality of the fossil record is the period to which they apply. In the con- very uneven geographically, in particular text of this paper “Würmian” is regarded as with regard to . The best-studied a to “Weichselian”, “Devensian” area by far is Southern , other areas and “Valdai” and is considered to begin with fairly good fossil records include Swit- with Isotope Stage 5d about 115 000 BP and zerland, , and parts of Yugo- end with the slavia, and . Many rieh Plei- c. 10 000 BP. stocene avifaunas are also known from The fossil records in fig. 1 — 14 are based , and Czechoslova- T y r b e r g : Glacial Relicts in© Ornithologische the West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 31

kia but these were mostly studied long ago It must be emphasized that the Würmian and badly need restudy. In the Soviet Union lasted for approximately 100 000 years and the Pleistocene avifaunas of Caucasus, Cri- encompassed several colder and milder cli- mea and parts of are reasonably matic phases and that the maps include re­ well known while there are few records cords spanning the whole of this time inter- from other areas. Pleistocene avifaunas are val. The records therefore indicate the “en- also scarce or unknown from , velope” of a species’ Würmian ränge rather most of the N orth E uropean Plain, than the ränge at any one time. Most of the and Southern Spain, the Southern Balkans, avifaunas are however late glacial Asia Minor, North Africa and the Near East. (c. 15000-10000 BP).

5. Arctic/Alpine Species

5.1 P ta r m ig a n learctic; in the , the , the Scot- tish Highlands, the Scandinavian moun- The Ptarmigan Lagopus mutus is the clas­ tains, the northern Urals, Iceland and sic example of an arctic species with isola- Spitzbergen. These populations, with the ted Southern montane populations. There is possible exception of the last two, are com- at present seven populations in the West Pa­ pletely isolated and have presumably been

Figure 1: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Ptarmigan.Gegen­ — wärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) des Alpenschneehuhns. © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at 32 Orn. Verh. 25, 1991 so since the end of the Würmian. Each popu- spite the general scarcity of fossil birds from lation is usually regarded as a separate sub- these areas the absence of Ptarmigan is species. probably real since there are several records The map (Fig. 1) shows that Ptarmigan of Willow Grouse (Fig. 2) and Black Grouse were quite common throughout the uplands from both the Russian and the North Euro­ of central and during the pean plains. This probably means that the Würmian. In the southernmost part of the European Ptarmigan were more or less iso- glacial ränge it was apparently restricted to lated from the populations in the Urals and the highest mountain ranges (the Pyrenees, further east even at the glacial maximum. the Cantabrian mountains, the Appennines This accords well with the fact that the Py- and the Stara Planina). The species seems to renean, Alpine, Scottish and Scandinavian have avoided plains. The Ptarmigan was ap­ Ptarmigan form the greyish “mutus” group parently quite scarce in the plain of Aqui­ of while the birds in the Urals, taine in southwest France in contrast ot the Spitzbergen and Iceland belong to the Willow Grouse (M o u r e r -C h a u v ir ö 1979). In browner “rupestris” group. Iceland has it is only found in the foothills of the presumably been colonized from Green- Carpathians, in the mountains of Crimea land, and Spitzbergen either from Green- and in the Urals and there is virtually no land or from Siberia by way of Franz Josefs’ records from the North European plain. De- Land (Ptarmigan from Greenland appa-

Figure 2: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Willow Grouse. - Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) des Moorschnee­ huhns. T y r b e r g : Glacial Relicts in© Ornithologische the West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 33

rently still occasionally reach Iceland (G u d - Tetrao mlokosiewiczi and the Caucasian m u n d s s o n 1972)). W hether the Ptarm igan co- Snowcock Tetraogallus caucasicus which lonized the Scandinavian mountains at the have been present in the Caucasus at least end of the Würmian from since the Middle Pleistocene (B a r y s h n ik o v & across the North European plain or from C h e r e p a n o v 1985, B u r c h a k -A b r a m o v ic h 1975). Britain across the North Sea is uncertain. The end of the Würmian apparently cau- The distance across the then partly dry sed a straightforward withdrawal of the North Sea from to Norway is no Willow Grouse northward across the North longer than from Greenland to Iceland and European plain where (in contrast to the the North Sea was then presumably at least Ptarmigan) there are a few late glacial re­ as icebound as the Strait today. cords from northern Germany, Denmark Ptarmigan have recently been found in de- and Southern . During this process posits of Mid-Würmian interstadial age the population on the was cut (c. 30 000 BP) in Skjonghelleren Cave in W e­ off by the North Sea and adapted to the m a­ stern Norway ( L a r s e n et al. 1987) and the ritime conditions on the heather moorlands possibility that the species might have sur- of Great Britain and . That no relict vived the glacial maximum on some ungla- Willow Grouse survived on the European ciated refugium on the Norwegian coast mainland is presumably due to the absence should perhaps not be entirely discounted. of suitable lowland moor and scrub .

5.2 W illo w G ro u s e 5.3 D o tte r e l

The Willow Grouse Lagopus lagopus has The Southern European populations of only one isolated population in the West Pa­ the Dotterel Eudromias morinellus are only learctic, the scoticus subspecies in the Bri­ doubtfully glacial relicts. They are all small tish Isles. This subspecies occurs in several (Fig. 3) and subspecifically identical to the separate areas but these are probably not main North Eurasian population, and since isolated with the possible exception of the dottereis spend the winter in the subdesert Shetland birds. zone of North Africa and the Middle East The fossil record (Fig. 2) shows that the the montane population may have been Willow Grouse was quite common in central established during the by shorte- Europe during the Würmian. The Willow ned migration. Such a process would be fa- Grouse apparently did not ränge quite as f ar cilitated by the fact that Dottereis seems to south as the Ptarmigan, since there are no have relatively little ortstreue. records south of the Pyrenees or the Alps. The fossil record however indicates that On the other hand the species is found at se­ the Dotterel was widespread on the glacial veral sites in the Russian and North Euro­ of central Europe (Fig. 3) and it is pean plains. Both these differences in the perhaps equally likely that at least some of Würmian ränge compared to the Ptarmigan the Southern populations have persisted is presumably due to the Willow Grouse not since the Würmian. This at least seems to be being dependent on upland habitat. It is no­ the most likely origin of the rather larger table that neither the Willow Grouse, the populations in the mountains of Central Ptarmigan or the Black Grouse seem to have Asia. reached the Caucasus though all three spe­ The two late glacial records from Apulia cies occurred in Crimea. Possibly they were are probably of wintering birds. This part of excluded by the Caucasian Black Grouse Italy was apparently very dry steppe or 34 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991

Figure 3: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Dotterel.Gegen­ - wärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) des Mornells. semi-desert during at least part of the last cates a high degree of isolation between po­ glaciation as indicated by the presence of pulations. e. g. sandgrouse (C a s s o l i et al. 1979) and it The Shore Lark is not very common as a would presumably have been suitable win- fossil but there are enough records to show ter habitat for Dottereis. The record from that it was widely distributed in central

Binagady near Baku (B u r c h a k -A b r a m o v ic h Europe during the Würmian (Fig. 4). Consi- 1975) is of broadly last interglacial date and dering the wide glacial distribution it is not probably from a migrating bird. immediately obvious why there are no relict populations in western Europe. Possibly the 5.4 Shore Lark montane habitats there are too wet for Shore Larks. In the Scandinavian moun- All the West Palearctic populations of the tains it is normally confined to the drier Shore Larks Eremophila alpestris are sub- montane heath and the same seems to be specifically distinct with ssp. flava in the true in the Atlas Mountains (pers. obs.) north, balcanica in SE Europe, atlas in Mo- The only other m ember of the , Tem- rocco,albigula east of the Caspian, penicil- minck’s Horned LarkEremophila bilopha, lata in parts of Anatolia and andbicor- occurs in desert habitat from to nis in Southern Anatolia and . Kuwait. It is tempting to speculate that this Some 35 other subspecies occur in Asia and form is derived from a population isolated North and South America. This strong ten- either in Maghreb or the Near East during dency to form subspecies presumably indi- some earlier interglacial. T y r b e r g : Glacial Relicts in© Ornithologischethe West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 35

Figure 4: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Shore Lark.Ge­ — genwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) der Ohrenlerche.

6. Montane Species

6.1 Rock / W ater Pipit probably indicate that the ranges of the Rock and Water Pipits were contiguous The Rock/Water Pipit complex has during the W ürmian. recently been split into two species, the Coastal Rock Pipit and the montane Water 6.2 C i t r i l F i n c h Pipit (R is b e r g in press). The Rock Pipits of NW Europe are usually The Citril Finch Serinus citrinella has a subdivided into three subspecies: petrosus very limited ränge in the mountains of on the British Isles, kleinschmidtii on the southwestern Europe (Fig. 6). The popula- Faeroes and littoralis in Scandinavia while tions on and Sardinia belong to a se­ the Water Pipits in the mountains of South­ parate subspecies, corsicana. ern Europe all belong to the nominateAn- The few Würmian records (Fig. 6) neatly thus s. spinoletta. link the existing populations in Massif Cen­ The fossil record of Anthus spinoletta tral and the Pyrenees. There is also two sensu lato is rather exiguous, consisting of Middle Pleistocene records from Southern half a dozen records in France, three in Eng­ France (M o u r e r -C h a u v ir E 1975) but neither land and one in (Fig. 5). these nor the younger finds give any clue to These however mostly lie well outside the the origin of the uniquely circumscribed present breeding ränge of the species and ränge of this species. 36 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991

Figure 5: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Rock/Water Pipit. - Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) von Strand/Wasser­ pieper.

6.3 S n o w F in c h puzzling since there is a fair number of well- studied avifaunas from Hungary, Yugosla- The Snow Finch Montifringilla nivalis oc- via and Romania where the species might be curs in a number of isolated montane popu­ expected to occur. The subspecific Separa­ lations from the Cantabrian Mountains to tion between the european populations and Tibet. The European populations belong to the Snow in Caucasia — Montifringilla nivalis nivalis while the may however indicate a discontinuous dis- birds in Caucasus - Armenia belong to the tribution in southeastern Europe even un- subspecies alpicola. The Snow Finch is der glacial conditions. quite common in Würmian avifaunas of Southern France and the Alpine countries (Fig. 7). The French sites link the extant po­ 6.4 C h o u g h s pulations in the Pyrenees and the Alps. Du­ ring the late Glacial the species apparently Next to the Ptarmigan and the Willow ranged f arther north in central Europe than Grouse the two Pyrrhocorax species are today (the northernmost finds are from among the commonest and most characteris-

Oberfranken (B r u n n e r 1939, 1959)). The tic birds in european Pleistocene avifau­ scarcity of finds in eastern Europe — there is nas. Unfortunately the nomenclature of none between and Crimea — is these two species is extremely confused in T y r b e r g : Glacial Relicts in© Ornithologische the West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 37

Figure 6: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Citril Finch.Ge­ — genwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) des Zitronengirlitz. the older literature and in some cases it is Chough then also occurred in several areas virtually impossible to decide which species where it is no longer found e. g. the Bale- the records refer to (for some details on the arics, northern Yugoslavia and the Crimea. nomenclatural problems seeJ ä n o s s y 1954). However in most areas the Red-billed Chough is rather less common as a fossil than the Alpine Chough and this is parti- 6.4.1 Red-billed Chough cularily true of the Balkans. Four subspecies of Red-billed Chough The three records from Great Britain Pyrrhocorax pyrrhocorax occur in the west (Fig. 8) are well separated from the other Palearctic, pyrrhocorax in Great Britain Würmian records. One record (Kirkdale and Ireland, erythrorhamphus in Western Cave in Yorkshire [H a r r is o n 1980, L y d e k k e r Europe, barbarus in Maghreb and docilis in 1891]) may in fact be from the Eem ian (Ips- eastern Europe and the Near East. wichian) Interglacial and in that case the During the Würmian the Red-billed presence of this typically montane bird in Chough occurred throughout Southern Euro­ the British Isles would be a very interesting pe (Fig. 8). The northern limit of the distri- parallel to the species’ ränge during the bution was apparently the foothills of Mas- present interglacial though the species does sif Central, the Alps and the Balkan moun­ not occur in Yorkshire today. tains south of the Danube. The Würmian The other two sites (Chudleigh and Pavi- records link virtually all extant West Pale­ land Cave [B e l l 1915, 1922, H a r r is o n 1980]) arctic populations and the Red-billed are either of late Glacial or (more likely) Ho- 38 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991

Figure 7: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Snow Finch.Ge­ — genwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) des Schneefinks. locene age. The Red-billed Chough presum­ Carpathians. Also in contrast toPyrrhoco­ ably recolonized the British Isles during the rax pyrrhocorax the Alpine Chough expan- Late Glacial by following the Atlantic coast ded northwards to the edge of the North north from Spain and Southern France since European plain during the late Glacial (the it seems unlikely that the species could have northernmost sites are in Luxemburg, Thu- survived the Glacial maximum in the Bri­ ringia and the Ojcow region of Southern Po- tish Isles. It was definitely present in land (B o c h e n s k i 1974, F e r r a n t & F r ia n t 1940, by the early Holocene when it is recorded M u s il 1980). The differences in the W ürmian from Port Eynon Point Cave (H a r r is o n ranges of the two Pyrrhocorax species may 1987). be due to a greater tolerance for cold Conti­ nental climates by the Alpine Chough. 6.4.2 Alpine Chough

All european populations of the Alpine 6.5 A lp in e A c c e n to r Chough Pyrrhocorax graculus belong to the nominate subspecies. The Alpine Chough is Three subspecies of The Alpine Accentor even more common than the Red-billed Prunella collaris occur in the West Palearc- Chough in Würmian avifaunas (Fig 9). How- tic. The nominate subspecies occurs in SW ever, in contrast to the latter species the Europe and Maghreb, subalpina in SE Euro­ Alpine Chough was also common in hilly pe, and W Anatolia and m ontana in areas in the Pannonian Basin and in the Caucasus and Iran. T y r b e r g : Glacial Relicts in© Ornithologischethe West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 39

Figure 8: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Red-billed Chough. - Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) der Alpenkrähe.

Most Würmian records are from SW Euro­ tions between eastern Anatolia and the lar- pe and link the populations in the Pyrenees, gely coastal subspecies in NW Europe and Massif C entral and The Alps (Fig. 10). There also in having a lowland subspecies on the is only two records from SE Europe, on Transcaspian steppes. Crete and Karpathos (W e e s ie 1984, 1987). The Crimson-winged Finch Rhodopechys The find from Karpathos (where the species sanguinea is a somewhat anomalous case. It no longer occurs) suggests that the Alpine occurs in the mountains of Eastern Anatolia Accentor colonized Crete from Anatolia. and reappears in the Atlas mountains. As pointed out by M o r e a u (1966) it is unlikely that this strictly montane bird could ever 6.6 O th e r s p e c ie s have crossed the lowland deserts of and Libya and it must thus be presumed to In addition to the species mentioned have colonized the Atlas around the north above some other species which lack a fossil side of the Mediterranean, though it has left record have similar distribution patterns: no trace either in the form of fossils or relict The Wallcreeper Tichodroma muraria and populations. The Rock PigeonColumba li- the Twite Acanthis flavirostris clearly be­ via may also originally have been a “type 2” long to the species with “type 2” disconti- species, but in this case the original ränge nuous montane ranges, though the Twite is has been blurred by the presence of feral po­ unusual in lacking any montane popula­ pulations. 40 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991

Figure 9: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Alpine Chough. - Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) der Alpendohle.

7. Steppe Birds

7.1 L o n g -le g g e d B u z z a rd which is also found at scattered Würmian sites across Europe but was apparently Two subspecies of the Long-legged Buz­ quite common in southwestern France du­ zard occur in the West Palearctic, Buteo ru- ring part of the late Würmian(D e l p e c h 1975, finus cirtensis in Maghreb and Libya and 1983). the nominate subspecies from the Balkans eastward. The Long-legged Buzzard is not common as a Würmian fossil (Fig. 11). 7.2 I m p e r ia l E a g le There are single records from , Hungary and Luxemburg and a group of six The spanish population of the Imperial sites in southwestern France. Eagle forms a well defined subspecies This distribution of finds probably indi- Aquila heliaca adalberti which may qualify cates that the Long-legged Buzzard occured as a separate species while the birds in the on the mammoth steppe across Europe, per- rest of the species ränge from Eastern Euro­ haps with a concentration in SW France. pe to Lake Baikal belong to the nominate The concentration of sites in the plain of subspecies. Aquitaine is parallelled in another steppe There are a few Würmian records of Im­ species, the Saiga tataricaperial Eagle well west of its present ränge in T y r b e r g : Glacial Relicts in© Ornithologische the West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 41

Figure 10: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Alpine Accentor. - Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) der Alpenbraunelle. the Balkans. The westernmost are Saccopas- 7.3 D e m o is e lle C ra n e tore near Rome (C a s s o l i 1978) and Coten- cher in (J ä n o ssy 1980) (Fig. 12). The Demoiselle Crane Anthropoides It is interesting to note that of the six virgo breeds mainly in the Würmian records four (Saccopastore, Co- zone from the Ukraine to . There tencher, Kaiman Lambrecht Cave in Hun- is also a small population in Eastern Anato­ gary and Ripa in Romania) are of early Wür­ lia and another, very isolated and now al­ mian age (>60 000 years BP), one (Grotta most extinct, in Maghreb. No subspecies Romanelli in Apulia) is from the late Glacial have been recorded. (10 000 —12 000 BP) while one (Teufelsluk- There are only four Würmian records of ken in Austria) cannot be closely dated. Ad- this species: Binagady near Baku, Karmalki mittedly the evidence is slight, but it seems on the Upper Volga, Pin Hole Cave in north- likely from these dates that the main Wür­ ern England and Gorham’s Cave in Gi­ mian expansion of the Imperial ’ braltar(B urchaic -A bramovich 1975, Cowles ränge took place early during the glacial 1981, Eastham 1968). Unfortunately none of cycle. This fits in well with the considerable the finds are well dated, but the one from differentiation of the adalberti subspecies Pin Hole Cave may be of Mid-Würmian age which argues that this population has been (Cowles 1981) and that from is isolated for a long time. probably rather older(Eastham 1968, Gam­ ble 1986). Despite the sparseness of the fossil record a wide Würmian ränge linking 42 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991

Figure 11: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Long-legged Buz- zard. — Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) des Adler­ bussards.

the isolated population in Maghreb with the while the other two, Grotta della Madonna main ränge of the species on the Eurasian near Torre a Nave and Grotta Romanelli in steppe seems likely. Apulia, are from the late Glacial.

7.4 Black-bellied Sandgrouse 7.5 P in - T a ile d S a n d g r o u s e

The nominate subspecies Pterocles o. The subspeciation pattern in this species orientalis occurs in the Iberian Peninsula, is different from that in the Black-bellied Maghreb, and Anatolia and is repla- Sandgrouse. The nominate Pterocles a. al- ced by ssp. arenarius from Iran and the Cas- chata occurs in Southern France and the pian eastwards. There are four Würmian re­ Iberian penninsula while ssp. caudacutus is cords of this species (Fig. 13). Of these the found in Maghreb, Libya and south-west three finds from Southern Italy are the most Asia. significant from a biogeographic viewpoint There is only one Würmian record of this since they are approximately halfway bet­ species, from the late Glacial of Grotta Ro­ ween the extant populations in Anatolia manelli in Apulia (C a s s o l i 1972). and . Of these three finds one, Torre The palaeoecology of the Grotta Roma­ a Nave in Calabria, is of early Würmian age nelli avifauna form Dryas 3, the final “cold T y r b e r g : Glacial Relicts in© Ornithologische the West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 43

Figure 12: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Imperial Eagle. — Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise (Punkte) des Kaiseradlers.

snap” of the Würmian has been studied by spread of sandgrouse in the central Medi- C a s s o l i ( C a s s o l i et al. 1 9 7 9 ) .The sandgrouse terranean area, both by narrowing water- apparently lived on the semi-desert littoral gaps and by providing suitable habitat since sandplains uncovered by the low glacial both Black-bellied and Pin-tailed Sand­ sea-level. The existance of this habitat may grouse generally avoid rocky and broken have been important in facilitating the country.

8. Discussion

8.1 Dispersal On the other hand there are no fossil re­ cords of any montane species which occured It is notable that most arctic or steppe in Europe during the Würmian and which species which had a glacial ränge extending has failed to survive to the present. The only over much of Europe have left no relict po- possible exception would beMelanocory- pulations. Some such species are e. g. Gyr- pha m axim which a is only found on the Ti- falcon Falco rusticolus, Pallid Harrier Cir­ betan high plateau today, but has been re- cus macrourus, Snowy Owl Nyctea scan- ported from a site in Liguria by C a s s o l i diaca and Snow Bunting Plectrophenax ni­(1980), but the identification is uncertain. It valis. would seem — not surprisingly — that the 44 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991

Figure 13: Present breeding ränge (horizontal hatching) and Würmian records (black dots) of Black-bellied Sand­ grouse. — Gegenwärtiges Vorkommen (Schraffur) und würmeiszeitliche Nachweise des Sandflughuhns.

montane species are better adapted to sur- Maghreb and the Pyrenean Peninsula either vive for long periods in isolated montane by way of the glacial steppes north of the “islands” than the arctic or steppe birds. Mediterranean or along the North African The only real barrier to dispersal of the littoral. The northern route would have “type 2” montane birds between Himalaya been open during most of the Würmian and the Atlantic Ocean seems to have been while the hyperarid desert of Egypt and Li- between Asia Minor and Balkan. Several bya would have been most easily traversed montane birds have ranges that extend from during periods of increased precipitation to the mountains of Asia Minor either during the mid Würmian interstadial or Caucasus (SnowcocksTetraogallus sp., complex or during the early Holocene cli- White-throated Robin Irania gutturalis, matic optimum. Güldenstedts Redstart Phoenicurus ery- A third alternative is a “diagonal” route throgaster, Red-fronted Serin Serinus pu- from the Balkan to Tunisia by way of South­ sillus, Raddes Dunnock Prunella ocularis ern Italy-Sicily. Under glacial conditions and Great Carpodacus rubicilla) dispersal by this route would have been fa- while every montane bird species which oc- cilitated by greatly narrowed water-gaps curs on both sides of the Aegean-Black Sea and the emergent areas would probably barrier has a ränge that extends all the way have been suitable habitat for steppe birds. to the Atlantic. The virtual absence of Pleistocene avifau­ The “type 3” birds could have colonized nas from North Africa makes it impossible T y r b e r g : Glacial Relicts in© Ornithologische the West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 45

Figure 14: Present breeding ranges and Würmian records of Gyrfalcon (cross-hatching/black dots) and Saker (ho­ rizontal hatching/triangles). —Gegenwärtiges Vorkommen und würmeiszeitliche Nachweise von Ger­ falke (Kreuzschraffur/Punkte) und Würgfalke (Horizontalschraffur/Dreiecke).

to decide the issue one way or the other but 8.2 S p e c ia tio n at least in the cases of the Imperial Eagle and the Demoiselle Crane the northern It would seem that the approximately route would seem to be more likely. 10 000 years since the Würmian have in ge­ It is possible that several more species had neral been insufficient for speciation to oc­ “type 3” ranges during the early Holocene cur though Anthus spinoletta/petrosus is but that the distribution patterns have been certainly a borderline case and some other blurred through adaption to the artificial subspecies e. g. Lagopus lagopus scoticus “steppe” habitat created in Europe by agri­ and Serinus citrinella corsicana may also be cultural activities from the Mid-Holocene close to species status. onwards. This in particular applies to the On the whole the diversity is low among White Stork Ciconia ciconia, Great arctic and steppe birds which argues that Otis tarda and Tetrax tetrax. speciation due to interglacial isolation is These three species were fairly common on not a common event in these groups. This is the Glacial steppe and today have strong- probably due to the relative shortness holds in the E urasian steppe zone and in the (10000-20000 years) and low frequency Iberian peninsula, though they also occur (about every 100 000 years) of interglacials. more or less widely on färmland in central There is however a few likely cases of such Europe. speciation in addition toEremophila alpe- 46 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991 stris/bilopha mentioned earlier. One is tral Asia with e. g. some ten species each of Ptarmigan/Willow Grouse. The Willow Prunella, Montifringilla and Carpodacus. Grouse has occured in Europe since the be- Exactly why there has been so much specia­ ginning of the Pleistocene (J a n o s s y 1974) tion in this area is unclear, but the large and while the P tarm igan shows up only tow ards geographically complex mountain system the end of the Middle Pleistocene (J a n o ssy and the frequently extreme aridity of the in­ 1976, 1980, 1984,M o u r e r -C h a u v ir £: 1975). It termontane basins have probably facili- seems reasonable that the “new” species tated isolation. The montane avifauna of the evolved from some isolated population, but West Palearctic on the other hand seems to there is of course no indication where this consist largely of “immigrant” species happened. which have evolved in Central Asia. Only Another possible case ist the Gyrfalcon the Citril Finch seems likely to have evolved Falco rusticolus and the Saker Falco cher- in Europe. Interestingly speciation has ap­ rug. A probable ancestor of these closely re­ parently been rather more common in the lated species Falco antiquus is known from Middle East where two montane endemics deposits of probably penultimate glacial with Palearctic affiliations occur; Syrian age in France and Hungary (J a n o ssy 1977, Serin Serinus syriacus and Arabian Accen- MouRER-CHAuvmfi 1975) while during the tor Prunella fagani. Würmian both the Gyrfalcon and the Saker An intriguing parallell to the Willow occurred in Europe and apparently behaved Grouse/Ptarmigan species pair is found as parapatric species (Fig. 14). This makes it among the montane birds in the two Pyrrho­ likely the speciation occurred during the corax species. In the Pyrrhocorax case it is Eemian interglacial, probably with the an- Pyrrhocorax graculus which has occurred cestral Gyrfalcons in the Arctic and the an- in the West Palearctic from the Early Plei­ cestral Sakers in the Eurasian steppe zone stocene while Pyrrhocorax pyrrhocorax and the two populations isolated from each only shows up tow ards the end of the Middle other by the beit. Pleistocene. However in this case too, there Among montane birds there is a distinct is no clue where speciation took place. center of diversity in the mountains of Cen­

Summary

Arctic, Montane and Steppe birds as Glacial relicts in the West Palearctic

A number of birds with predominantly arctic, dence for a continuous lowland distribution du­ alpine or steppe distributions have discontinuous ring the Würmian for Snow Finch, Red-billed ranges in the West Palearctic. Chough, Alpine Chough and Alpine Accentor Fossil evidence shows that in several cases the (Fig. 7 -1 0 ) and some data indicating the sam e for modern ränge is relict and a remnant of a conti- Rock/Water Pipit and Citril Finch (Fig. 5 — 6). A nuous ränge on the periglacial steppes of the latest sim ilar history is likely for a num ber of other spe­ (Würmian) glaciation. cies with similar ranges but with no fossil record, Among arctic species this is definitely so in the e. g. Wallcreeper, Twite and Crimson-winged cases of Ptarmigan and Willow Grouse (Fig. 1 — 2)Finch. and probable for Shore Lark and Dotterel There is also a number of steppe birds with iso­ (Fig. 3-4). lated populations in SW Europe and/or Maghreb Among montane birds with more or less isola­ i. e. Long-legged Buzzard, Imperial Eagle, Demoi- ted populations in the mountains of Central andselle Crane, Black-bellied Sandgrouse and Pin- Southern Europe there is extensive fossil evi­ tailed Sandgrouse. For these species there is also T y r b e r g : Glacial Relicts in© Ornithologische the West Palearctic Gesellschaft Bayern, download unter www.biologiezentrum.at 47

some fossil data indicating a continuous Wür­ There is however a few species-pairs in the mian ränge (Fig. 11-14) though the evidence is West Palearctic where speciation due to intergla- weaker than for most of the arctic or montane cial isolation seems likely, i. e. Ptarmigan/Willow birds. Grouse, Gyrfalcon/Saker, Shore Lark/Tem- A similar pattern of ränge fragmentation andminck’s Horned Lark and Red-billed Chough/Al- isolation has presumably occurred during pre-pine Chough, though in most of these cases spe­ vious interglacials. Despite this the diversity of ciation may have occurred outside the West Pale­ the affected groups is low in the West Palearctic, arctic. There is also two borderline cases of “inci- indicating that such isolation has only exception- pient speciation” among the taxa treated in this ally led to speciation. This is probably due to thepaper Auila( heliaca/adalberti andAnthus spino- relative shortness (10 000 — 20 000 years) and low letta/petrosus). frequency of interglacials.

Zusammenfassung

Arktische, montane und Steppenvogelarten als Eiszeitrelikte in der W estpaläarktis

In der Westpaläarktis weist eine Anzahl Vogel­ nich, das Sandflughuhn und das Spießflughuhn, arten mit vorwiegend arktischer, alpiner oder belegen die Fossilfunde ein kontinuierliches Steppenverbreitung unzusammenhängende würmeiszeitliches Verbreitungsgebiet, wenn­ Areale auf. Die Fossilfunde beweisen, daß in einer gleich die Evidenz schwächer als im Fall der mon­ Reihe von Fällen das gegenwärtige Areal ein Re­ tanen Vogelarten ist. liktvorkommen darstellt, welches ursprünglich Ähnliche Muster zwischeneiszeitlicher Areal­ die Eisrandsteppen der letzten Vereisungsperiode zersplitterungen verbunden mit Isolation der Po­ (Würm-Eiszeit) umfaßte. Für Arten mit arkti­ pulationen, dürften während der früheren Verei­ schem Verbreitungstyp trifft dies sicher bei Al- sungsperioden und Zwischeneiszeiten entstan­ pen- und Moorschneehuhn, wahrscheinlich auch den sein. Aber trotz der Unterschiedlichkeit der bei Ohrenlerche und Mornell zu. betroffenen Vogelgruppen ist die davon abzulei­ Für montane Arten gibt es umfassendes fossiles tende Artaufspaltung und Diversitätszunahme Belegmaterial für ein zusammenhängendes Tief­ nur vergleichsweise geringfügig ausgefallen. landsverbreitungsgebiet im Falle von Schnee­ Möglicherweise hängt das mit der kurzen Dauer fink, Alpenkrähe und Alpendohle sowie für die der Zwischeneiszeiten von nur 10 000-20 000 Alpenbraunelle. Diese Arten haben gegenwärtig Jahren und ihrer geringen Frequenz des Auftre­ ein diskontinuierliches, montanes Areal in den tens zusammen. Für einige Artenpaare, wie Al- Gebirgen von Mittel- und Südeuropa. Auch für pen- und Moorschneehuhn, Ger- und Würgfalke, Strand/Wasserpieper und für den Zitronengirlitz Ohren- und Hornlerche sowie für Alpenkrähe zeichnen sich derartige Verhältnisse ab. Für Ar­ und Alpendohle, läßt sich die zwischeneiszeitli­ ten ohne fossile Belege, wie Mauerläufer, Berg­ che Arealaufspaltung mit Arttrennung anneh­ hänfling und Rotflügelgimpel, dürfte dieselbe Er­ men, obgleich in diesen Fällen auch Einflüsse von klärung zutreffen. außerhalb der Westpaläarktis in Betracht zu zie­ Auch für eine Reihe von Steppenvogelarten mit hen sind. Die Zwillingsarten Kaiseradler und isolierten Populationen in Südwesteuropa und/ Spanischer Kaiseradler sowie Strand-(Fels-) und oder im Maghreb, wie beispielsweise für den Ad­ Wasserpieper befinden sich gerade im Prozeß der lerbussard, den Kaiseradler, den Jungfernkra­ Arttrennung.

Literature

B a r y s h n ik o v , G. F & G. O. C h e r e p a n o v (1985): B e l l , A. (1915): Pleistocene and later bird faunas Ptitsi Bol’shchogo Kavkaza epochi paleolita of Great Britain and Ireland. The Zoologist i mezolita. Ornitologiya 20: 139-160. 19:401-412. 48 © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at Orn. Verh. 25, 1991

— (1922): Pleistocene and later bird faunas ofH a f f e r , J. (1974): Avian Speciation in Tropical Great Britain and Ireland. Naturalist (Hull) South America. Publ. of the Nuttall Ornith. 37: 2 5 1 -2 5 3 . Club No. 14. Cambridge, Mass. B ocheiviski , Z. (1974): Ptaki mlodszego czwartor- H a r r is o n , C. (1980): A Re-examination of British z§du Polski. Warszawa-Krakow. Dvensian and Earlier Holocene Bird Bones in B r u n n e r , G. (1939): Die Gaiskirche im oberen the British Museum (Natural History). Jour­ Püttlachtal (Oberfranken). N. Jahrbuch f. nal of Archaeological Science 7: 53 — 68. Mineralogie Beil.-Bd. 79 Abt. B: 243 — 273. — (1982): An Atlas of the Birds of the Western — (1959): Das Schmiedberg-Abri bei Hirsch­ Palearctic. London. bach (Oberpfalz). Paläon. Z. 33: 152 — 165. — (1987): Pleistocene and Prehistoric Birds of B u r c h a k -A b r a m o v ic h , N. I. (1975): Die pleisto- South-West Britain. Proc. Univ. Bristol Spe- zäne Vogelfauna der UdSSR. Quartärpa­ laeol. Soc. 18 (1): 8 1 -1 0 4 . läontologie 1: 87-105. JA n o s s y , D. (1954): Fossile Ornis aus der Höhle C a s s o l i, P F (1 9 7 2 ): Lo Pteroclide (Aves, Ptero- von Istalloskö. Aquila 55/58: 205 — 223. clidae) fossile nei livelli del Paleolitico supe- — (1974): Upper Pliocene and Lower Pleisto­ riore e medio nel Pleistocene dell’ Italia me- cene Bird Remains from Poland. Acta Zoolo- ridionale. Quaternaria16: 225-245. gica Cracoviensia 19: 531 — 563. — (1978): L’avifauna pre-würmiana di Torre in— (1976): Plio-Pleistocene Bird Remains from Pietra. Quaternaria 20: 429-440. the Carpathian Basin I. . 1 Te- — (1980): L’avifauna del Pleistocene superiore traonidae. Aquila 82:13-34. delle Arene Candide (Liguria). Mem. Istit. — (1977): Plio-Pleistocene Bird Remains from Ital. Pal. Uman. N. S. 3: 155-234. the Carpathian Basin III. Strigiform es, Fal- C a s s o l i, P. F., A. G. S e g r e & E. S e g r e (1979): Evo­ coniformes, Caprimulgiformes, Apodifor- lution morphologique et ecologique de la mes. Aquila 84: 9 -3 6 . cöte de Castro (Pouilles) dans le Pleistocene — (1980): Plio-Pleistocene Bird Remains from final. In: D e S o n n e v il l e -B o r d e s , D.: La Fin the Carpathian Basin V Podicipediformes, des temps glaciaires en Europe: 325-332. Ciconiiformes, Otidiformes, Columbiformes, C o w l e s , G. S. (1981): The first evidence of D em oi­ Piciformes. Aquila 86: 19 — 32. selle Crane Anthropoides virgo and Pygmy — (1984): Die Jungmittelpleistozäne Vogel­ CormorantPhalacrocorax pygmaeus in Bri­ fauna von Hunas (Hartmannshof). H Ine l ­ tain. Bull. Br. Orn. CI. 101 (4): 3 8 3 -3 8 1 . l e r , F et al. Die Höhlenruine Hunas bei D e l p e c h , F (1975): Les faunes du paleolithique Hartmannshof (Landkreis Nürnberger superieur dans le sud-ouest de la France. Land). Quartär-Bibliothek 4. Bonn. These de doctorat a l’universite de Bor­ L a r s e n , E. et al. (1987): Cave StratigraphyW in e ­ deaux I. stern Norway; multiple Weichselian glacia- — (1983): Les faunes du paleolithique superieur tions and interstadial fauna. Bo­ dans le sud-ouest de la France. Cahiers du reas 16: 26 7 -2 9 2 . Quaternaire N:o 6. L y d e k k e r , R. (1891): Catalogue of the fossil birds E a s t h a m, A. (1968): The Avifauna of Gorham’s in the British Museum (Natural History). Cave, Gibraltar. Institute of Archaeology London. (London), Bulletin 8: 37-42. M o r e a u , R. E. (1966): The Bird Faunas of Africa F e r r a n t , V & M. F r ia n t (1940): La Faune Pleisto­ and its islands. New York — London. cene d’Oetrange (grand Duche de Luxem­ M o u r e r -C h a u v ir i S, C. (1975): Les Oiseaux du Plei­ bourg). Bull. Soc. Natur. Luxembourgeois stocene moyen et superieur de France. These 34: 1 8 5 -2 2 0 . Universite Claude Bernard Lyon. G a m b l e, C. (1 9 8 6 ): The Palaeolithic Settlement of — (1979): Les Oiseaux de la fin des temps gla­ Europe. Cambridge. ciaires en France. La disparition des especes G u d m u n d s s o n , F (1972): Grit as an indicator of froides. In: De S o n n e v il l e -B o r d e s , D.: La the overseas origin of certain Birds occuring Fin des temps glaciaires en Europe: in Iceland. Ibis 114: 582. 10 5-111. © Ornithologische Gesellschaft Bayern, download unter www.biologiezentrum.at T y r b e r g : Glacial Relicts in the West Palearctic 49

Musil , R. (1980): Die Großsäuger und Vögel der W e e s i e , P D. M. (1984): On some Pleistocene Bird Teufelsbrücke. In Feustel , F.: M agdalenien- Fossils from the South Aegean Island of Kar- station Teufelsbrücke II: Paläontologischer pathos. Geobios (Lyon) 17 (6): 8 4 5 -8 4 9 . Teil. Weimarer Monographien zur Ur- und — (1987): The Quaternary avifauna of Crete, Frühgeschichte 3: 5 — 27. . Ph. D. Thesis. University of Utrecht. R is b e r g , L. (in press): Sveriges Fäglar. Second Edition. Authors’ Adress: Tommy Tyrberg Kimstadsv. 37 S - 610 20 Kimstad Sw eden