Present distribution and probable primary range
of brittle willow (Salix fragilis L.)

A. K. Skvortsov

Problemy biogeotsenologii, geobotaniki i botanicheskoy geografii. Leningrad, Nauka, 1973, pp. 263-280

Translation: Irina Kadis

Introduction

It was Andersson (1868: 209) who first came up with the idea that Salix fragilis L. was not truly wild and autochthonous in Europe and that it probably originated in southwest Asia. Andersson considered another willow that he had described from southwest Asia, S. australior Anderss., to be a wild ancestor race of S. fragilis. This provided grounds for his hypothesis about the Asiatic origin of S. fragilis. However, later on it became clear that Andersson was mistaken when he considered S. australior to be closely related to S. fragilis. S. australior (priority name S. excelsa Gmel.) rather belongs to the cycle of S. alba L. Therefore, the hypothesis about brittle willow's Asiatic origin was abandoned. No further studies were conducted in order to find out whether it was aboriginal in Europe.

While completing the taxonomical revision of the willows distributed within the USSR and surrounding countries, I, too, had to face the problem with S. fragilis. The analysis of all available material has brought me to the conclusion that S. fragilis is adventive in the USSR and Europe. Its primary range is within northern and northeastern Asia Minor. Though my reasoning behind the judgement on the motherland of S. fragilis is entirely different from Andersson's, I arrived to a very similar conclusion. Even though I have already presented it in the review of the willows of the Caucasus and Asia Minor and also in the review of the USSR willows (Skvortsov 1966: 116-117; Skvortsov 1968: 68, 111-113.), detailed arguments were beyond the scope of these works. The purpose of the article is to provide this argumentation.

I used the herbarium material at most major depositories in this country: herbaria of the botanical institutes in Leningrad, Kiev, Tbilisi, Baku, and Yerevan; Moscow and Tomsk university herbaria; and also the Herbarium of the Natural History Museum in Lvov. Of foreign depositories, I examined all the material at the Prague National Museum and collections from Asia Minor at the Royal Botanic Garden in Edinburgh. I express my cordial gratitude to all the colleagues who run collections at these institutions. I wish to thank Dr. E.L. Swann (Norfolk, England) and Dr. A.R. Pinto da Silva (Oeiras, Portugal) for some valuable material they mailed for the study. I am indebted to Dr. Neumann (Vienna) for his important comments on the distribution of S. fragilis in Western Europe. The late E.I. Steinberg provided translations of Estonian texts.

I have conducted field observations primarily in Moscow, Smolensk, Kaluga, and Tula oblast's. Later on I included the vicinity of Leningrad, the Carpathians, and Volgograd Oblast. Additionally, I conducted some observations at a lesser scale in Pskov, Novgorod, Ryazan, Voronezh, Saratov, Belgorod, and Kiev oblast's, and also in Latvia and central German Democratic Republic. Literature data were extensively used as well.

I. Brittle willow in central European Russia

Here is what Tsinger (1886: 393) said about S. fragilis in his Anthology of Knowledge:

"It is very common everywhere along rivers, around ponds, at dams, near dwellings, and at roadsides; for Tver Government it is listed only as cultivated. It probably becomes rare in the southeast, in the proximity of the Volga banks." This description was inherited by the early editions of Mayevskiy's Flora of Central Russia (1892: 413; 1895: 435; 1902: 469). The note regarding Tver Government found in Tsinger (1886) as well as the first three editions of Mayevskiy's Flora, had been originally made by Bakunin (1879: 348).

Petunnikov (1901: 37-39), citing mostly F.A. Teploukhov's opinion (Teploukhov apparently never published the results of his willow studies), stated that the majority of samples treated as S. fragilis within Moscow Government were actually the so-called S. viridis Fr., i.e., hybrids of S. fragilis and S. alba. At the same time, according to Petunnikov, the true, non-hybrid S. fragilis was rather rare in Moscow Government. Remarkably, all of non-hybrid specimens of S. fragilis that Petunnikov was aware of turned out to be staminate. Siuzev (1901) was another author citing Teploukhov's opinion on the hybrid nature of most S. fragilis in this country. He provided quotations from a letter by Teploukhov, which even indicate that Teploukhov probably had doubts about the aboriginal nature of S. fragilis. At the same time, Siuzev argued that indeed he himself had found pistillate specimens of S. fragilis on the Oka River in Moscow Government, and therefore Teploukhov finally agreed that S. fragilis was a native species in the flora of this country.

Syreshchikov (1907: 15, 19; 1927: 98, 99) noted in support of Petunnikov's view that S. viridis was common in Moscow Government, while S. fragilis was rare. The Flora of Central Russia by Mayevskiy, starting from edition 4 (1912: 500), also says that "crosses with S. alba occur more often than the typical" S. fragilis; while the note about Tver Government vanishes from the text. Starting from that time, the suggestion that S. fragilis might be not native in central European Russia is not repeated in the literature anymore. It is not found in the regional "floras" of Vladimir and Kaluga (Flerov 1902, 1912), Tula (Rozen 1916), or Kalinin (Nevskiy 1952).

My own observations in central European Russia completely support the conclusions of the earlier authors about domination of hybrids S. fragilis x S. alba over "pure" S. fragilis. Though the two species belong to the same section, they have a number of quite distinct differences. Therefore, recognition of hybrids should not be just an intuitive procedure based solely on inspiration, as—alas!—it often happens. Instead it can be based on distinct morphological criteria.

The diagnostic characters of the two species are superposed here below.

S. fragilis	S. alba
Trees developing in the open produce low, nearly spherical crowns; large oblique limbs originate at a fairly low level	Crowns of trees developing in the open mostly elongate; trunks upright, tall; limbs relatively short
Within old part of crown, 1-3-year-old branches never pendulous, spreading at 60-90° angles, breaking off easily when gently pressed	Within old part of crown, 1-3-year-old branches often pendulous, growing at angles less than 60°, not as brittle
It is only current-year branchlets, particularly suckers, that often become pigmented red or rich brown; older branches within old crown uniformly yellowish-gray or nearly ivory, occasionally with some orange tint or reddish spots	Current-year branchlets mostly green; older branches becoming rather bright reddish or brownish, pigmentation more pronounced on the lighted side
Current-year branchlets may be puberulent; older branches always completely glabrous	Current-year branchlets always pubescent; older branches mostly pubescent, at least in distal parts
Buds during late fall and winter colored same as branches; apices blackening due to bud scale dieback. Buds ovoid to lanceoloid, more or less acutish, abaxially convex or more or less carinate; slightly convex or sometimes also carinate on adaxial side	Buds during the fall and winter colored same as branches or brighter, not blackening, oblong lanceoloid, mostly blunt, abaxially somewhat convex; mostly completely flat adaxially
Outermost leaf primordium in bud mostly pale or grayish brown, broad, nearly round when spread, with several parallel veins originating from base. Catkin primordia in floriferous buds densely sericeous	Outermost leaf primordium mostly reddish, oval, its venation pinnate. Catkin primordia in floriferous buds puberulent, greenish; imbricate bracts form distinct pattern visible through pubescence
Young leaves in spring light yellowish-green; mature foliage dark green, without bluish tint	Young leaves more or less deep bluish-green; mature leaves pale bluish-green or sericeous
Stipules obliquely semi-cordate, distinctly inequilateral	Stipules narrowly lanceolate, linear-subulate, subequilateral, rarely somewhat falcate
Petioles 7-15 mm long; petiolar glands pronounced in all leaves on the branch, except for proximal ones	Petioles 2-8 mm long; petiolar glands pronounced mostly only in distal leaves and on suckering shoots
Young leaves mostly completely glabrous, occasionally sparsely hairy; mature leaves glabrous	Young leaves delicately sericeous; mature leaves also more or less silvery pubescent
Leaf blade broadly cuneate or rounded at base; reticulation network very conspicuous on underside. Margin dentation rather coarse, denticles mostly bent or rounded. Stomata on upper leaf surface mostly sparse or wanting	Leaf blade mostly cuneate at base; only major lateral veins are visible on underside, reticulation wanting; margin minutely serrulate, denticles mostly acute. Stomata dense on upper leaf surface
Catkins spread nearly at right angle from branch. All flowers open simultaneously in staminate catkins	Catkins positioned at acute angle on branch. Proximal flowers in staminate catkins open earlier than distal
Bracts 0.8-1.5 x 0.5-0.8 mm, covered with straight trichomes that extend by 0.8-2.0 mm beyond bract margin. Bracts remain pale on drying	Bracts 1.2-2.8 x 0.7-1.0 mm, puberulent; trichomes extend by 0.2-0.6 mm beyond bract margin. Bracts turn rusty-brown on drying
Filaments pubescent only at base (along ⅛ to ¼ of their length); dry anthers 0.5 (0.6) mm long	Filaments pubescent up to ¼-½ of their length; dry anthers (0.5) 0.6-0.7 mm long

Hybrids occupy and intermediate position, being closer to one or the other parent as regards characters expression. This variability is understandable, as we deal not only with F₁ generation, but various products of segregation formed as a result of reciprocal crosses with parental species as well as crosses within F₁. Heterosis is typical for the majority of hybrids. Usually hybrids grow faster and produce more robust branches than parents^[1].

There also exist hybrids of S. fragilis with S. pentandra L. and probably with S. triandra L.; however, those are rare in central European Russia and present no interest for this discussion.

Differently from Siuzev's experience, my search did not yield any clearly non-hybrid pistillate S. fragilis around the Oka River or elsewhere in central European Russia. Trees displaying typical species characteristics were invariably staminate. More than that, the herbarium search within material from this area yielded the same result: not a single pistillate specimen was found without hybrid traits. One could speculate that we are dealing here with a case of sexual dimorphism, i.e., pistillate plants are characteristically different from staminate ones in many respects. However, this assumption does not stand under scrutiny, as pistillate S. fragilis are found in other parts of the area, where they are similar to staminate plants in every respect (except for their sex). Siuzev's statement about presence of "pure" pistillate S. fragilis at the Oka River was apparently based on insufficient character analysis.

The predominance of hybrids along with complete absence of non-hybrid pistillate specimens in such a wide-ranging species as S. fragilis within the central European Russia presents a puzzle that needs explanation. It is Bakunin's opinion that comes to mind first. He believed that within Tver Government all of S. fragilis was just cultivated. However, brittle willow in Moscow, Kaluga, Smolensk, Tula, and Ryazan oblast's is commonly found in habitats were it sure could not have been planted. In the aftermath of a storm or during a windy day, multitudes of broken twigs are found around every old brittle willow. Many branchlets are carried away by wind or running water and root in 2-3 weeks, if they happen to be in favorable moisture situations. Thus S. fragilis is a spontaneous species undoubtedly belonging to the wild flora. However, does it also happen to be indigenous? In order to answer this question, it is necessary to consider habitats taken by the species and decide whether there exists a niche for it anywhere in undisturbed nature.

For solving a problem of this type, historic records also might be of some value; however, in this case chances of finding positive evidence are more than slim. Besides, historic evidence can be only relatively meaningful. If we cannot provide documentation describing the introduction of potatoes to Kamchatka, that does not necessarily mean that potatoes are native there. On the other hand, even though we know for sure that grapes have been cultivated in Tajikistan for ages and even millennia, we still cannot exclude a possibility for wild species to exist there. Whether or not we discover indications of oriental plane cultivation in the Varzob or Tsav river valleys, the outcome is not going to alter the fact that plane is a member of the native flora in Armenia. In all such cases, we should base our conclusions first and foremost on biological observations.

Outside truly secondary, man-made habitats, such as roadsides, ditches, or residential lots, S. fragilis has been found along streams and rivers, in damp depressions, and in small spring-fed fens. The most common willows on river banks in this country are S. viminalis L., S. dasyclados Wimm., S. triandra L., and S. alba L. There is no doubt that these four are native to the area. Whenever there appears a freshly deposited alluvium patch in a river valley, it is promptly taken by these species. It is mostly S. viminalis that occupies sandy deposits right along the water margin; S. triandra and S. dasyclados both require silt deposits and thus settle particularly along slow creeks or surface run-off depressions; still more often all the three species form mixed thickets in various proportions.

As for S. alba, it favors sandy-silty deposits that are formed during large floods at high levels, which remain dry during most of the year. Depressions that remain moist for a long time after the flood are particularly favorable. Thus it forms pure or nearly pure stands in large river valleys, usually producing small and dense, even-aged groves extended along river beds. Most of these groves have been clear-cut; however, I have found some along the Oka and Protva that were quite intact. When S. alba happens to grow along small streams, it is usually only solitary and, as a rule, occurs not far from some large river valley.

S. alba, S. viminalis, S. dasyclados, and S. triandra are consistently found at suitable habitats in semicultural landscapes as well as in those only moderately affected by human activities. The less the landscape has been altered by humans, the more the original ecological preferences of each species are pronounced, and the more often we encounter good seed reproduction at appropriate habitats.

In contrast to the species named above, S. fragilis never forms continuous single-aged groves in large river valleys, particularly along the Oka. We always find only scattered solitary trees of different ages that randomly occur in different parts of river valleys, most often within the floodplain, around oxbow lakes, along small tributaries, or in erosional depressions formed by floods. These habitats are also preferred by S. alba, which naturally grows there. Healthy natural groves of S. alba never contain any S. fragilis, even as a minor admixture. In undisturbed situations, all habitats suitable for S. alba become colonized by S. alba only, even though there are fruiting hybrids S. alba x S. fragilis growing in close proximity. These hybrids along with pure S. fragilis are encountered only in habitats fairly disturbed by humans. Hence, when natural habitats favorable for S. alba are available, hybridization between S. alba and S. fragilis, however extensive it is, has no effect on the integrity of S. alba, its ecological and morphological stability as a species. It is only S. fragilis that is subject to disruption and erosion as a species.

During my tours of the major rivers (Oka, Moskva, Protva, Ugra, Osetr, and Upper Dnieper), I have never encountered any large-scale reproduction of S. fragilis or its hybrids. Consequently, we should admit that within the major river valleys in this country there are no habitats where S. fragilis could exist prior to human disturbance.

Yet what about those small rivers, brooks, and spring-fed fens? Perhaps S. fragilis could be a natural part of plant communities or successions there, could not it? All of our native forest trees, including those that never dominate climax formations or major successional stages, are constantly found (even though they might be scarce and depressed) in the oldest, minimally disturbed forests. This is not the case with S. fragilis. It is not consistently present in any climax communities, even as a rare species, neither found regularly at any particular successional stage. Stream banks and moist bottoms of depressions in this country used to be naturally covered by black alder groves and stream-bank spruce associations. We cannot imagine a light-loving tree with a broad low crown, such as S. fragilis, growing amidst a dense spruce forest or black-alder grove. It could not even temporarily be present during the natural regeneration stages of these communities.

Openings in black-alder groves and spruce forests that occur due to wind or insect damage never affect the forest environment to such a significant extent that it ceases being a forest; and S. fragilis never trespasses forest habitats, as far as we know from observations. Whenever moist forest grounds start receiving more light, tall herbs would conquer the spot along with S. caprea L., S. aurita L., and sometimes S. cinerea L., yet never S. fragilis. It is only present in the situations when light has been available for a long time and forest character of the habitat has been long lost to a large extent. It is only then that S. fragilis may settle, self-dispersed not from seed, but rather from broken twigs brought in by wind or water.

We should conclude that within the natural vegetation cover of central European Russia there is no niche for S. fragilis to occupy. We thus have to acknowledge that brittle willow is an adventive plant in this area.

II. Brittle willow in other regions of the USSR (excluding the Caucasus)

Many "floras" and reviews (Wolf 1900: 21; Seemen 1908-1910: 70; Nazarov 1936: 203; Pravdin 1951: 168; Rechinger 1957: 67, and others) inform that S. fragilis is distributed across the entire European territory of the USSR, except for the extreme north, and that its range also extends to West Siberia and the Altai Mts.; some believe that this willow even reaches the Lake Baikal (Penkovskiy 1901: 62). However, herbarium data signify that S. fragilis (together with its hybrids) hardly crosses the Volga. East of the Volga, it is represented only by some solitary findings. I have seen five specimens from Siberia and Kazakhstan (Biysk, Alma-Ata, Akmolinsk, Karaganda, and Telmanovo in the vicinity of Tselinograd) and some very scarce from the southern Urals and Bashkiriya. A single concrete reference to S. fragilis from the Altai Mts. (Edzhigan Valley—Krylov) has turned out to be a misidentification (Skvortsov 1956).

The notion of predominance of hybrids S. fragilis x S. alba over "pure" S. fragilis was extrapolated over the entire range of this species within the USSR by Nazarov (1936) in the Flora of the USSR. However, a detailed herbarium study and differential mapping suggest a more intricate and interesting relationship between hybrid and non-hybrid S. fragilis. Remarkably, all of the named five specimens from Siberia and Kazakhstan happened to be hybrids. Specimens from Alma-Ata and Biysk are known to be collected from cultivated plants. The remaining three samples from Kazakhstan originate from regions where S. alba is not distributed, which makes the formation of autochthonous hybrids impossible. Apparently, these were also collected from cultivated plants.

Non-hybrid S. fragilis was never ever collected from the Urals or Upper Vyatka Basin, neither from Volga Uplands. Hybrids of S. fragilis are not infrequent in Saratov and Volgograd oblast's, even away from human dwellings, where they appear to be growing naturally. However, a more careful examination always shows they have been planted. Apparently, in the Volga Basin there is no current advancement of S. fragilis hybrids, neither by seed dissemination, nor even from vegetative parts.

Further herbarium study has shown that the entire periphery of brittle-willow range within the European USSR has been occupied by hybrid plants only. "Pure" S. fragilis occurs only eastward, i.e., east of the line connecting Vyborg, Volkhov, Uglich, Vladimir, Arzamas, Tambov, Voronezh, Kremenchug, and Beltsy, with an isolated locus on the Crimea Peninsula.

According to Meinshausen (1878: 314), around St. Petersburg, S. fragilis "occurs only in cultivation, though some plants are occasionally found here and there in the woods, apparently as escapees from cultivation, as they are always only staminate." Aun (1939: 43-47) remarks that in Estonia there is no wild S. fragilis, either. All the specimens that Aun had found blooming (in Pärnu, Värru, Tartu, and Pechory in the vicinity of Pskov) happened to be staminate. He was unable to locate any pistillate plants despite his deliberate effort. Even local forest rangers were unable to point to any pistillate brittle willows in response to official requests through the Forestry Department. On the other hand, hybrid specimens, though not too common, were encountered in Estonia, and among those were some pistillate plants. My own observations around Leningrad and the herbarium studies fully support the reports by Meinshausen and Aun. I conclude that in Estonia and around Leningrad, the distribution of brittle willow is similar to what is observed around Moscow. The same refers to Pskov and Novgorod vicinities.

I have not found opinions similar to Meinshausen's or Aun's in the literature referring to other regions. However, according to the herbarium analyses, the area that harbors staminate non-hybrid plants only and where pistillate trees are lacking is actually considerably wider. It extends westward about to the line connecting Ventspils, Vilnius, Mogilev, Kiev, Zvenigorodka, Vinnitsa, Chernovtsy, and Kluzh. Non-hybrid specimens of both sexes can be found only west of this line. Therefore, if at all there are any native, primary habitats of brittle willow in Europe, it makes sense to look for them only west of this line. Apparently this boundary reveals the central-European distributional pattern of S. fragilis, and since the flora of central Europe is best represented within the USSR in the Carpathians and Transcarpathia, first and foremost one should search for primary habitats there.

I made two attempts to find such habitats during my trips to the Carpathians in 1957 and 1968; however, both were unsuccessful. Brittle willow is very abundant in the Carpathians. Occasionally (e.g., along a small river named Khustinka), I encountered populations with considerable individual variability among plants that appeared to be the "pure" species, — an evidence of normal seed reproduction. S. fragilis even occurred continuously for a few kilometers along some small streams, such as the Lyuta River, the left tributary of the Uzh.

However, in intact habitats where the virgin vegetation was more or less preserved, there was evidently no room for S. fragilis. In narrow gorges, the tall forest usually descended right to the banks of small rivers, allowing only minimum growth of alder and some S. purpurea L. or tall herbs here and there. At the same time, intact habitats on the banks of a large river (Tissa R.) were taken by pure groves of S. alba, much like in the central European Russia. There is no doubt left for me that S. fragilis is an adventive tree in the Carpathians and Transcarpathia, though it now appears to be well naturalized at some places there .

III. Brittle willow in Europe

Due to lack of observations and herbarium material, my analysis of brittle willow distribution in Western Europe has to be based largely on literature sources.

According to many respectful sources (Hjelt 1902: 86; Hultén 1950: 144; Lid 1963: 251; Hylander 1966: 288), S. fragilis is an introduced species found only in cultivation on most of Scandinavian territory, except for extreme southern Finland, southern Sweden, and Denmark, where it also known to escape from cultivation. In Scandinavia hybrids with S. alba are not infrequent (described as S. viridis Fries 1928: 283). These hybrids, along with S. alba itself, also are either cultivated or feral.

According to the author of a monograph on the British willows, Linton (1913: 14), S. fragilis is native in England, southern Scotland, and Wales, being introduced to northern Scotland. According to Praeger (1934: 26), it is probably introduced in Ireland. A more recent British source (Clapham et al. 1962: 590) depicts the same view. However, the actual situation with S. fragilis on the British Isles appears to be quite different. Descriptions provided for this species by all authoritative English sources (Loudon 1838: 1516; White 1890: 348; Elwes, Henry 1913: 1754; Linton 1913: 14; Moss 1914: 17; Clapham et al. 1962: 590) clearly demonstrate that all the authors actually present hybrids S. fragilis x S. alba under the name "S. fragilis." Characteristically, Linton (l.c.) provided a synonym name S. russeliana Sm. for S. fragilis. As a matter of fact, S. russeliana has been long recognized as a hybrid even in the western literature (Wimmer 1866: 133, and others). The same conclusion is to be reached when one considers identifications on herbarium specimens provided by the English.

Along with S. fragilis, the British authors also list S. decipiens Hoffm. Some (White 1890: 348) consider it to be a hybrid of S. fragilis and S. triandra; others (Linton 1913: 17) think of it as a subspecies of S. fragilis. Some refrain from interpreting this taxon (Clapham et al., l.c.). According to Linton and the latest Flora (Linton l.c., Clapham et al. l.c.), S. decipiens appears to be found only in cultivation, sporadically, and infrequently in Britain. Reliable references exist only for staminate specimens. The study of Hoffmann's authentics of S. decipiens ("S. decipiens mihi. Erlangae" LE!) as well as material identified by the British as S. decipiens shows that all of these are actually true S. fragilis. Therefore, the situation with S. fragilis on the British Isles is no different from what we find around Moscow and Leningrad.

Though S. fragilis and its hybrids are distributed all across the British Isles, there is a distinct increase in density in the southeastern England, as it is shown in the Atlas of the British Flora (1962: 188). This is the most continental part of the British Isles; in those areas where the Atlantic flora dominates, S. fragilis becomes scarcer.

According to Coutinho (1939: 189), S. fragilis is scattered all across Portugal, both cultivated and "nearly wild" ("subespontaneo"). However, the local Flora (Rozeira 1944) of Tras-os-Montes, one of the largest provinces, does not list this willow. Of those four specimens that I received from Portugal under the name "S. fragilis," two (M. da Silva 2724; Rainha 5706) appeared to be hybrids with S. alba, and the other two had no relationship to S. fragilis (Rainha 5930—probably a puberulent variation of S. alba; M. da Silva 2432—S. babylonica or its hybrid.)

In his monograph of Spanish willows, Vicioso (1951: 47), citing Laguna (1883), provides that S. fragilis is found both wild and cultivated on most of the territory, though it is hard to tell autochthonous specimens from introduced, as it is always the case with plants cultivated from olden times. Apparently the author is not at all convinced that this willow is native in Spain. Merino (1906: 616) states that S. fragilis is definitely only a cultivated plant in Galicia. The description and illustration in Vicioso's monograph do not depict all characteristic features of S. fragilis. Most probably, hybrids S. fragilis x S. alba have been actually described in Spain as S. fragilis. Remarkably, Görz (1929: 15) believed that S. fragilis does not grow on the Iberian Peninsula, but is substituted there by S. neotricha Goerz, which resembles S. australior Anderss. (= S. excelsa Gmel.) from the Caucasus. Obviously, S. neotricha is yet another designation for the same hybrid cycle S. fragilis x S. alba. In any case, there is no talk about autochthonous S. fragilis existing on the Iberian Peninsula.

Major sources report S. fragilis to be distributed across "nearly the entire" France (Camus 1904: 76; Coste 1906: 271; Rouy 1910: 193); however, according to Bonnier (s. a.; ca. 1930: 40), S. fragilis is not found in western, southwestern, and Mediterranean France. Neumann, a contemporary Austrian authority on willows, believes, mostly basing on own observations, that S. fragilis no longer grows around Paris (Neumann in litt. a. 1962). According to Chassagne (1956: 260), in the French Massif Central S. fragilis is often cultivated, and it is hard to tell whether it is growing there naturally. Chassagne believes that the species has an oriental origin and advances due to cultivation. Thus we can hardly suggest that S. fragilis is aboriginal in France. As for hybrids with S. alba, they are not infrequent. According to Chassagne (1956: 259), these hybrids are found even more often than the parental species.

Buser (1940: 629) considered S. fragilis "by far, the rarest of all willows" in Switzerland. According to Buser, "true S. fragilis ... is just as rare, as hybrids S. alba x S. fragilis are common in Swiss piedmont. So it is not surprising that the Swiss have been taking hybrids for the species or at least have not been distinguishing between the species and hybrids... Nearly all of "S. fragilis" in local Swiss "floras" are actually hybrids. This situation can only be explained if we admit that S. fragilis as well as its hybrids are not native, but rather introduced" (Buser 1940: 631). The recent Flora of Switzerland (Hess et al. 1967: 666), which includes Tirol, the French Jura Mts., and Savoy, treats S. fragilis as an introduced species within the entire area.

The Italian researcher of the 19th century Bertoloni (1854: 303) reported S. fragilis for Palermo only. Later authors believed it was scattered all across Italy (Parlatore 1867: 220; Borzi 1885: 138; Fiori 1923: 341). However, local "floras" portrait S. fragilis as a random alien species that shows no affinity to any particular region or native plant community. Thus in Verona this "rare, random" tree is known from a single location (Goiran 1898: 19). For Triest (Marchesetti 1896-97), Reggio-Emilia and Modena (Negodi 1944), as well as the Silagian Mts. (or La Sila) in Calabria (Sarfatti 1959) S. fragilis is not at all listed. It is known from just a single locality in the Roman Apennines (Zangheri 1966: 86) and a single place within Teramo Province (Zodda 1954: 66). For the Sicily (Lojacono 1904: 392), it is only the hybrid S. alba x S fragilis that is mentioned; besides, the vague description makes one suspect that forms of S. alba are described instead. There are more examples like this. Hence in Italy S. fragilis is not aboriginal, either.

In Austria and Germany most authors interpret S. fragilis as a common species growing in natural habitats. However, in these countries, as well as in Britain, Switzerland, and others, hybrids of S. fragilis and S. alba have habitually been taken for S. fragilis. A series of hybrid forms have been described under binary names: S. rubens Schrank 1789: 226; S. fragilissima Host 1828: 6; S. palustris Host 1828: 7; S. excelsior Host 1828: 8. Even though Wimmer (1866: 133-134) had pointed to their hybrid nature, K. Koch, an authority in German dendrology, included them all in S. fragilis (1872: 514). The German-Swiss Synopsis (W. Koch 1907: 2300), similarly to the English sources, designated the true, non-hybrid S. fragilis as "var. decipiens," while noting that it was known perhaps only in cultivation. Hartig (1851: 563) remarked that S. russeliana Sm. (that is, a hybrid) was common in northern Germany, while non-hybrid S. fragilis was so rare there that he had only seen it in botanic gardens. According to Toepffer (1914: 191), in Bavaria, all records for S. fragilis needed revision due to constant mixing with hybrid plants. As per Toepffer, the true S. fragilis (which he as well named "var. decipiens") was rare in Bavaria. In Wurttemberg, according to Bertsch (1951: 46), only hybrids were encountered, all of them originating from cultivated specimens. In the Northern German Lowlands, non-hybrid S. fragilis was reported rare and even completely missing from many areas (Görz 1922: 31). Neumann (in litt.) also believes that S. fragilis is not present in the western part of Northern German Lowlands and beyond, in the Netherlands.

During the fall of 1964, I found some small scattered non-hybrid S. fragilis growing along small streams in the Gartz southern piedmont. These habitats appeared to be completely similar to those occupied by this willow within this country: in the Carpathians and even around Smolensk and Moscow. I also found hybrids in-between Halle and Jena. The herbarium material from Germany that I had at my disposal contained a number of non-hybrid as well as hybrid specimens of both sexes. I concluded that within the southern GDR and central BRG the situation with brittle willow is similar to that in the western regions of our country, that is, the tree has been nearly completely naturalized, so that occasionally it appears as a native. At the same time, in the northwest and southwest the status of brittle willow is closer to that in France and England.

In Lower Austria, S. fragilis used to be considered as "one of the most common willows, particularly in the Danube Valley" (Kerner 1860: 185). However, later on Rechinger (1957: 67) remarked that all of S. fragilis from the Danube were actually hybrids.

In Belgium, S. fragilis has been known as mostly cultivated, its hybrids encountered much more often than the species (Lawalrée 1952: 36).

In Poland, Czechoslovakia, Hungary, and Romania, S. fragilis appears to be frequent. It is usually mentioned in "floras" as a native tree. However, it is hardly possible that the origin of S. fragilis in Romania and Czechoslovakia could be drastically different from its origin in our Carpathians; or else that the situation in Poland fundamentally differs from what takes place in Germany or the Baltic States. Even though non-hybrid plants of both sexes are not infrequent, the majority of specimens are of hybrid nature, as per Wołoszczak (1889: 292), for Transylvania or Szafer (1921: 33), for Poland.

In Yugoslavia, S. fragilis has been believed to be widespread (Beck von Managetta 1909: 115; Adamoviħ 1909: 194; Rohlena 1942: 19; Španovi�? 1954; Jovanovi�? 1955; Em 1967; and others). However, a note by Hirc (1904: 158) remarkably testifies for absence of herbarium collections from Croatia. S. fragilis is also characteristically absent from remote ravines in Macedonia, where the flora should still be most primeval (Soška 1938-39). Those few specimens from southern Yugoslavia that I examined were all hybrids.

Velchev (1966: 56) reports S. fragilis being distributed in Bulgaria "across the entire country." However, the description that he provides rather fits hybrids. According to Stefanov (1943), S. fragilis in Bulgaria is a boreal element restricted to the northwest. According to Chernyavskiy et al. (1959: 94), stands of S. fragilis in Bulgaria are of "secondary origin." Regretfully, this statement is elaborated no further.

In Greece S. fragilis apparently is completely absent.

While summarizing data on S. fragilis range in Europe outside the Soviet Union, we can distinguish the same three zones as within the Soviet Union (Fig. 1):

* core zone where S. fragilis has been nearly entirely naturalized and is represented by non-hybrid trees of both sexes;

* transitional zone, where the species is still quite common, though among non-hybrid plants staminate ones overwhelmingly dominate;

* peripheral zone, where virtually only hybrids are represented, being scattered, sporadic, or cultivated only.

Of course the demarcation lines are at this point very much approximate. For the refinement, much more information is needed than I have had at my disposal.

Fig. 1. Distribution of Salix fragilis L.

1. primary range

2. area where S. fragilis is nearly naturalized, non-hybrid plants of both sexes encountered

3. distribution limits for non-hybrid staminate plants, their reproduction spontaneous, vegetative

4. boundary of area where nearly exclusively hybrid plants are encountered, originating from plantings

Literature sources also report a rather large-scale introduction of S. fragilis to North America, where it appears to be also nearly naturalized. All those specimens from North America that I have examined happened to be only hybrids.

IV. Brittle willow in the Near East, Asia Minor, and the Caucasus

In eastern Asia Minor, southern and eastern Transcaucasia, Iran, and the Soviet Central Asia there is a common, widely distributed willow, S. excelsa Gmel. This species is a race of S. alba, closely related to it, though distinct in a number of characters that resemble S. fragilis: non-weeping crown, stout, brittle branches, stout buds, less pubescent leaves and branchlets, and more pubescent bracts. It was first described from northern Iran in 1774 by Gmelin, the younger, but then forgotten and described for the second time by Andersson (1867: 43) as S. australior Anderss. or S. fragilis δ australis (Andersson 1868: 210). Andersson placed this tree in the S. fragilis cycle by mistake and then suggested that the European S. fragilis had originated from southwest Asia (Andersson 1867: 41).

Relying on Andersson's works, Boissier (1879: 1184) also listed S. fragilis for Syria, Asia Minor, and Transcaucasia. However, it soon became evident that S. australior Anderss. was not at all a form of S. fragilis, but rather belonged to S. alba cycle. Despite this clarification, due to the authority of Boissier, almost until recently S. fragilis has been considered an "oriental" species present in Syria and Lebanon (Bouloumoy 1930: 316; Post 1933: 530); Jordan (Nábélek 1929: 24); Iran (Parsa 1950: 1351); western and southern Transcaucasia (Medvedev 1883: 228, Nazarov 1936: 203); Dagestan (Lvov 1956: 77); the western and eastern Caucasus (Makhatadze 1961: 51); and the Russian Central Asia (Fedchenko 1915: 295, Protopopov 1953: 36). There is no doubt that all these listings were based on either erroneous identifications or uncritical borrowing.

Meanwhile, the true S. fragilis was actually discovered growing in the East. It was quite wild, neither adventive, nor deliberately introduced, its populations represented by both sexes. Besides, hybridization with S. alba was completely absent or at least very insignificant. It was first collected in 1892 by Sintenis; however, the first reliable record of S. fragilis in Asia Minor belongs to Görz (1930: 113). He discovered specimens collected by S. Turkevich in former Kars [Province of Turkey; Russian Turkestan at the time of collection]. Even though Görz considered this find "amazing, " he still under-evaluated its importance. It did not even occur to him to re-examine the collections he had at hand. Indeed he failed to correctly identify some true S. fragilis that he collected himself.

I examined the following specimens of S. fragilis from its eastern range:

1. Paphlagonia, wilajet Kastambuli, Tossia, Utsch-tschesme, 3.VI.1892, P. Sintenis No. 4089, fl. femin. et fol.;

2. districtus Gümüsch-hane, supra Bayburt, 1500 m, frutex inter stirpes typicas S. australioris idque prorsus sponte nascens, 8.VI.1931. R. Görz (Sal. Asiat. No. 37: "S. australior var. pseudofragilis Goerz var. nova"), fol.;

3. inter Gümüsch-hane er Bayburt, 1600 m, sponte nascens, 16.VI.1931. R. Görz (Sal. Asiat. No. 37-a: "S. australior ad var. pseudofragilem Goerz vergens"), fl. femin.;

4. Prov. Kastamonu, 10 km from Daday to Effani, 900-1000 m, stream-side in pinetum, 30.VII.1962. Coode et Yaltirik No. 38610, fol.;

5. Prov. Kastamonu, between Seydiler and Kure, 1200 m, salix scrub near stream, 30.VII.1962, P. Davis No. 38464, fol.;

6. Karsskaya obl., bliz st. Promezhutochnoi, bereg r. Kyaklik [Kars Oblast, close to Promezhutochnaya Station, bank of the Kyaklik River], 11.V.1914, S. Turkevich No. 245, fl. femin., masc.;

7. Karsskaya obl., mezhdu Sarykamyshem i Karakurtom v Tyaklitskoi balke [Kars Oblast, between Sarikamis and Karakurt in the Tiyaklit Depression], T. Roop, fol.;

8. Sanjak Erzerum, tugai r. Khnys-chai bliz Khnyskaly [Erzurum Province, Hinisçayi River forested floodplain near Hiniskaly], 14.VII.1916, V. Sapozhnikov and B. Shishkin, fol.;

9. bliz Akhaltsikha, r. Poshov-chai [near Akhaltsikhe, the Posof-çayi River], 9.VIII.1920, D. Sosnovskiy, fol.

The latter is the only sample known so far from the USSR territory.

There is no indication that any of the listed specimens originate from cultivated plants. Accompanying hybrid series, which are so common in European collections, are absent here. Indeed hybrids (T. Roop 1911, LE) were collected only around the Kars Fortress, from the landscape that had been greatly disturbed for a long time.

According to Grossheim (1945: 27), S. fragilis is commonly cultivated in the Northern Caucasus. If this is true (I did not see any samples), then S. fragilis (or, more likely, its hybrids) apparently had been brought to the Northern Caucasus and Siberia with migrations of the Russian and Ukranian population from the west and northwest rather than directly from Asia Minor.

Conclusions

Thus from northern and northwestern Asia Minor, where it had a small primary range in the mountains, brittle willow was somehow introduced to Europe, where it advanced primarily by means of vegetative reproduction and formed an enormous secondary range. The plant has been nearly completely naturalized in the central part of this range; its alien nature is more obvious in the peripheral parts. Across all of its secondary range, S. fragilis hybridizes with S. alba on a large scale; in significant parts of the range, particularly peripheral, non-hybrid specimens are completely absent.

When and where exactly in Europe did S. fragilis first appear? Is it still continuing to advance? One might assume that it appeared first within the center of its present range. However, the actual events could as well be quite different: it is also possible that the formation of the present locus has been influenced by climatic factors rather than historic. S. fragilis could have entered Europe somewhere at a peripheral part of its modern range, such as Italy or the Balkans. In any case, S. fragilis was likely to be promptly transported over a great distance. As for its arrival on the Russian territory, it is clear enough that it came from the west rather than across the Caucasus.

Quite a number of S. fragilis herbarium samples have been preserved for 100-150 years. They provide evidence for the conclusion that during the recent 100-150 years there has been no significant dynamics in the distribution of this species. For example, non-hybrid staminate and hybrid pistillate specimens were collected around Moscow, respectively, in 1837 and 1823; around St. Petersburg, respectively, in 1850 and 1827; non-hybrid specimens of both sexes were already present around Mogilev in 1862, and so on. Apparently the process of colonization in Europe has essentially been completed, so that S. fragilis has attained its climatically determined limits. Certainly this is as well indicated by the concentric zoning pattern described above. Had the process of colonization been continuing actively into the present time, such a structure could not yet have been established. Non-hybrid plants definitely have a greater capacity for self-dispersal than hybrids. Hybrids continue to expand their range only through new plantings. From the economic standpoint, hybrids are more valuable, as they grow better. Additionally, hybrids might be more tolerant to climatic conditions of Mediterranean Region as well as those in Siberia, since one of the parental species, S. alba, grows naturally in both these regions. The similarity of the secondary range of S. fragilis and natural ranges of a few European species (Fig. 2) also signifies that climatic limits might have been already reached.

Fig. 2. Areas of some European species that resemble
S. fragilis area (adapted from Meusel et al. 1965)

1. Carex brizoides Juslen. 2. Carex montana L. 3. Anemone nemorosa L.

From the time of the Turkish conquest of the eastern Byzantine provinces in the 11-12th centuries and to the mid-19th century there was literally no connection between Europe and eastern Asia Minor. Hence it is safe to speculate that S. fragilis was introduced to Europe during the Byzantine Empire rule or maybe even during the Roman epoch. In one of local British "floras" (Grose 1957: 508), I came across some curious evidence. In ancient Saxon land deeds of Wiltshire County (there are fifteen of them, the oldest dated 796), some willows ("withies") are mentioned as markers of property boundaries. S. fragilis presently grows at nearly all of the named locations. Boundary markers were of course preserved and restored; therefore, it is not improbable that the willows growing there today are direct descendants of trees that were planted more than a thousand years ago.

Classic examples of ginkgo, metasequoia, and franklinia show that a plant with an extremely narrow relict range can be very viable, highly adaptable, and tolerant to a wide range of conditions. Hence a narrow primary range of S. fragilis should not be considered an obstacle for this willow to attain a wide secondary range. Besides, the phytoclimate of the forested regions in eastern Asia Minor has much in common with the climate of central Europe.

Veronica filiformis Sm. provides another example of a species with the natural range in the Caucasus and Asia Minor extensively naturalized in Europe. Much has been said about it (Lenmann 1942; Thaler 1953; Kornaś, Kuc 1954; Bangerter, Kent 1965; Jehlik, Slavik 1967). This veronica was first found in the wild in 1893 in France. It has continued colonizing the European countries since then, mostly (or perhaps exclusively) vegetatively. Its present range embraces nearly all of central Europe (Fig. 3). Recently it has been found in Estonia (Eichwald 1960). Similarly to S. fragilis, it appears to advance to the Russian territory from the west rather than south, i.e., not directly from the Caucasus.

Fig. 3. Distribution of Veronica filiformis Sm. (generalized)

1. primary range 2. secondary range

There remains a single question concerning S. fragilis that yet awaits an answer. Why do staminate non-hybrid plants have a wider range than pistillate non-hybrids? Perhaps those were only staminate plants that were initially introduced. Pistillate plants then might have appeared only many generations later, as a result of hybrid segregation. Or else, staminate specimens might have been more capable of vegetative regeneration. It is quite possible that the right answer is none of the above. It has yet to be discovered.

References

Adamoviħ L. 1909. Flora Serbiae austro-orientalis. Salix. Rada Jugosl. Akad., 177: 194-195.

Andersson N. J. 1867. Monographia Salicum 1. Kongl. Svenska Vetens.-Akad. Handl. 6(1). Stockholm.

Andersson N. J. 1868. Salix. In A.P. De Candolle (ed.). Prodromus systematis naturalis regni vegetabilis 16 (2): 190-323. Genève.

Atlas of the British Flora. F.H. Perring, S.M. Walters (eds.). 1962. London—Edinburgh.

Aun K. 1939. Eesti pajud. Tallinn.

Bakunin A.A. 1879. Spisok tsvetkovykh rasteniy Tverskoi flory [Flowering Plants of the Tver Flora. A Checklist]. Tr. SPb obshch. yestestvoisp. 10.

Bangerter E.B., Kent H.D. 1965. Additional notes on Veronica filiformis. Proc. Bot. Soc. Brit. Isles 6.

Beck v. Mannagetta G. 1909. Flora von Bosnien ... Salicaceae. Wiss. Mitt. aus Bosnien und Hercegovina 9: 114-119. Wien.

Bertoloni A. 1854. Flora italica 10. Bononiae.

Bertsch K. 1951. Jahresh. Vereins Vaterländ. Naturk. Württemberg 106: 46 (fide Lawalrée 1952: 38; non vidi).

Boissier E. 1879. Flora Orientalis 4. Genève—Basel.

Bonnier G. (s.a.; ca. 1930). Flora complète illustrée 10. Paris.

Borzi A. 1885. Compendio della flora forestale italiana. Messina.

Bouloumoy L. 1930. Flore du Liban et de la Syrie. Paris.

Buser R. 1940. Kritische Beiträge zur Kenntnis der schweizarischen Weiden. Ber. Schweiz. Bot. Ges. 50: 567-788.

Camus A., Camus E.-G. 1904. Classification des saules d'Europe et Monographie des saules de France 1. Paris.

Chassagne M. 1956. Inventaire analytique de la flore d'Auvergne et contrées limitrophes 1. Paris.

Chernyavskiy P.I., Ploshchakova L. Nedyalkova S., Dimitrov I. 1959. Drveta i khrasti v gorite na Blgariya [Bäumer und Sträucher in der Wäldern Bulgariens]. Sofiya

Clapham A.K., Tutin T.G., Warburg E.F. 1962. Flora of the British Isles. 2nd ed. Cambridge.

Coste H. 1906. Flore descriptive et illustrée de la France 3. Paris.

Coutinho A.X.P. 1939. Flora de Portugal. 2nd ed. Lisbõa.

Eichwald K. 1960. Niitjas mailane—Eesti rohumaid ohustav umbrohutaim. Eesti Loodus 3.

Elwes H.J., Henry A. 1913. Trees and shrubs of Great Britain and Ireland. 7. Edinburgh.

Em H. 1967. Pregled na dendroflorata na Macedonija. Skopje.

Fedchenko B.A. 1915. Rastitelnost Turkestana [Vegetation of Turkestan.] Petrograd.

Fiori A. 1923. Nuova flora analitica d'Italia 1. Firenze.

Flerov A.F. 1902. Flora Vladimirskoy gub. [Flora of Vladimir Government]. Moscow.

Flerov A.F. 1912. Flora Kaluzhskoy gub. [Flora of Kaluga Government] Kaluga.

Fries E.M. 1828. Novitiae florae Suecicae. 2nd ed. Lund.

Görz R. 1922. Über norddeutsche Weiden. Feddes Repert. 13.

Görz R. 1929. Les saules de Catalogne. Cavanillesiae 2, 7-10. Barcelona.

Görz R, 1930. Beiträge zur flora Kleinasiens. Salix. Feddes Repert. 28.

Goiran A. 1898. Juglandaceae et Salicaceae veronenses. Bull. Soc. bot. Ital. 18-24.

Grose J.D. 1957. Flora of Wiltshire. Devizes.

Grossheim A.A. 1945. Flora Kavkaza [Flora of the Caucasus] 3. 2nd ed. Baku.

Hartig T. 1851. Vollständige Naturgeschichte der forstlichen Kulturpflanzen Deutschlands. Berlin.

Hess H.E., Landolt E., Hirzel R. 1967. Flora des Schweiz. 1. Basel—Stuttgart.

Hirc D. 1904. Revizija Hrvatske flore. Salix. Rada Jugoslav. Acad. 159: 157-162.

Hjelt H. 1902. Conspectus florae fennicae. 2(1). Helsingforsiae.

Hoffmann G.F. 1791. Historia Salicum iconibus illustrata 2, fasc. 1. Lipsiae.

Host N.T. 1828. Salix. Vindobonae.

Hultén E. 1950. Atlas över växternas utbredning i Norden. Stockholm.

Hylander N. 1966. Nordisk Kärlväxtflora 2. Stockholm.

Jehlik, N., Slavik B. 1967. Doplňky k rozšireni Veronica filiformis v Československu. Preslia 39.

Jovanovi�? B 1955. Sumske fitocenoze i stamista Suve Planine. Glasnik sumarskog fakult. 9. Beograd.

Kerner A. 1860. Niederösterreichische Weiden. Verhandl. zool.-bot. Ges. Wien 10.

Koch K. 1872. Dendrologie 2 (1). Erlangen.

Koch W.D.J. 1907. Synopsis der Deutschen und Schweizer Flora 3. Aufl. Hersg. E. Hallier u. A. Brand. Leipzig.

Kornaś J., Kuc M. 1954. Veronica filiformis Sm. — nowy we flore polskiej uciażliwy chwast lakowy. Fragm. flor. et geobot. 1(1).

Laguna M. 1883. Flora forestal Española 1. Madrid (n.v.)

Lawalrée A. 1952. Salix. In Flora générale de Belgique 1(1). Bruxelles.

Lehmann E. 1942. Die Einbürgerung von Veronica filiformis in Westeuropa. Gartenbauwissenschaft 16.

Lid J. 1963. Norsk og Swensk flora. Oslo.

Linton E.F. 1913. The British willows. Suppl. to J. Bot. 51.

Lojacono M. 1904. Flora Sicula 2(2). Palermo.

Loudon J.C. 1838. Arboretum et frutecetum Britannicum 3. London.

Lvov P.L. 1956. Opredelitel glavneyshikh derevyev i kustarnikov Dagestana [Guide to Most Important Trees and Shrubs of Dagestan.] Makhachkala.

Makhatadze L.B. 1961. Salix. In Dendroflora Kavkaza [Dendroflora of the Caucasus] 2. Tbilisi.

Marchesetti C. 1897. Flora di Trieste. Trieste.

Mayevskiy P.F. 1892. Flora Sredney Rossii [Flora of Central Russia], 1st ed. Moscow;

Mayevskiy P.F. 1895. Flora Sredney Rossii [Flora of Central Russia], 2nd ed., S.I. Korzhinskiy ed. Moscow.

Mayevskiy P.F. 1902. Flora Sredney Rossii [Flora of Central Russia], 3rd ed., B.A. Fedchenko ed. Moscow.

Mayevskiy P.F. 1912. Flora Sredney Rossii [Flora of Central Russia], 4th ed., D.I. Litvinov ed. Moscow.

Medvedev Ya.S. 1883. Derevya i kustarniki Kavkaza [Trees and Shrubs of the Caucasus]. Tiflis.

Meinshausen K.F. 1878. Flora Ingrica. St. Petersburg.

Merino B. 1906. Flora descriptiva e ilustrada de Galicia 2. Santiago.

Meusel H., Jäger E., Weinert E. 1965. Vergleichende Chorologie der zentraleuropäischen Flora 1. Jena

Moss C. E. 1914. The Cambridge British Flora 2. Cambridge.

Nábélek F. 1929. Iter turcico-persicum 4. Spisy prirodoved. fak. univ. Brno 105.

Nazarov [Nasarov, Nasaroff] M.I. 1936. Salix. In Flora SSSR [Flora of the USSR] 5. Moscow—Leningrad.

Nazarov [Nasarov, Nasaroff] M.I. 1949. Salix. In Flora Belorussii [Flora of Byelorussia] 2. Minsk.

Negodi G. 1944. Flora delle provincia di Modena e Reggio Emilia. Atti Soc. Natur. Moden. 75.

Nevskiy M.L. 1952. Flora Kalininskoy oblasti [Flora of Kalinin Oblast] 2. Kalinin.

Parlatore F. 1867. Flora Italiana 4. Firenze.

Parsa A. 1950. Flore de l'Iran 4. Teheran.

Penkovskiy V.M. 1901. Derevya i kustarniki, kak dikorastushchiye, tak i razvodimyye [Wild and Cultivated Trees and Shrubs.] Kherson.

Petunnikov A.N. 1901. Salix. In [Critical Review of Moscow Flora. III.] Tr. SPb obshch. yestestvoisp. 31(3): 18-40.

Post G.E. 1933. Flora of Syria, Palestina, and Sinai 2. 2nd ed. Beirut.

Praeger R.L. 1934. The botanist in Ireland. Dublin.

Pravdin L.F. 1951. [Willow.] In Derevya i kustarniki SSSR [Trees and Shrubs of the USSR] 2. Moscow—Leningrad.

Protopopov G.F. 1953. Salix. In Flora Kirgizskoy SSR [Flora of Kirghiz SSR] 4: 9-37. Frunze.

Rechinger K.H. 1957. Salix. In Hegi G. Illustrierte Flora von Mitteleuropa 3(1): 44-135. Aufl. 2. München.

Rohlena J. 1942. Conspectus florae Montenegrinae. Preslia 20-21.

Rouy G. 1910. Salix. In Flore de France 12: 189-252. Paris.

Rozeira A. 1944. A flora da provincia de Tras-os-Montes. Memor. Soc. Broter. 3.

Rozen V.V. 1916. [Checklist of Plants Found in Tula Government up until 1916.] Izv. Tulsk. obshch. lyubit. yestestvozn. 4.

Sarfatti G. 1959. Prodromo della flora della Sila. Webbia 15.

Schrank F.P. 1789. Baierische Flora 1. München.

Seemen O., v. 1908-1910. Salix. In Ascherson P., Graebner P. Synopsis der mitteleuropäischen Flora 4: 54-350. Leipzig.

Siuzev P.V. 1901. [On some new plants of Moscow flora.] Tr. bot. sada Yuryev. un-ta 2(1).

Skvortsov A.K. 1956. [Some additions and amendments to the willow flora of West Siberia.] Sistem. zametki gerbariya Tomsk. un-ta 79-80: 13-15.

Skvortsov A.K. 1966. [Review of willows from the Caucasus and Asia Minor.] Tr. Bot. in-ta AN ArmSSR 15: 91-141.

Skvortsov A.K. 1968. Ivy SSSR [Willows of the USSR.] Moscow—Leningrad.

Soška T. 1938-1939. Beiträge zur Kenntnis der Schluchtenfloren von Südserbien 1-7. Glasnik Skopskog nauchnog drushtva 18, 20.

Španovi�? T. 1954. Vrbe naših podunavskih ritova. Šumarski List 78: 506-521. Zagreb.

Stefanov B. 1943. Fitogeografski elementi v Blgaria. Sb. Blg. Akad. 39. Sofia.

Syreishchikov D.P. 1907. Illustrirovannaya flora Moskovskoy gub. [Illustrated Flora of Moscow Government] 2. Moscow.

Syreishchikov D.P. 1927. Opredelitel rasteniy Moskovskoy gub. [Guide to the Plants of Moscow Government.] Moscow.

Szafer W. 1921. Salicaceae. In Flora Polska 2: 24-47. Kraków

Thaler I. 1953. Die Ausbreitung von Veronica filiformis. Phyton 5.

Toeppfer A. 1914. Salicaceae. In Vollmann F. Flora von Bayern: 187-202. Stuttgart.

Tsinger V. Ya. 1886. Sbornik svedeniy o flore Srednei Rossii [Anthology of Knowledge about the Flora of Central Russia.] Moscow.

Vicioso C. 1951. Salicaceas de España. Madrid.

White F.B. 1890. A revision of the British willows. J. Linn. Soc. 27.

Wimmer F. 1866. Salices Europaeae. Vratislaviae.

Wołoszczak E. 1889. Kritische Bemerkungen über siebenbürgische Weiden. Österreich. bot. Z. 62: 291-295; 330-332.

Zangheri P. 1966. Flora e vegetazione del medio ed alto Appenino Romagnolo. Webbia 21.

Zodda G. 1954. La flora teramana. Webbia 10.

NOTES

^[1] Some authors, for example, Nazarov (1949: 54) have had an impression that pistillate plants of S. fragilis tend to grow faster than staminate. Such observations without doubt have to be attributed to comparisons of hybrid pistillate and non-hybrid staminate specimens.

Translation I.Kadis
1 Jan 2008


Salicicola	Last updated: webmaster

Present distribution and probable primary range of brittle willow (Salix fragilis L.)

A. K. Skvortsov

Problemy biogeotsenologii, geobotaniki i botanicheskoy geografii. Leningrad, Nauka, 1973, pp. 263-280

Translation: Irina Kadis

Present distribution and probable primary range
of brittle willow (Salix fragilis L.)