Alexey Zinovjev and Irina Kadis

Willows are usually considered a difficult group, many species treated differently by different authors. This is true for the Atlantic West-European species Salix atrocinerea Brotero, or rusty willow, naturally growing in Britain, France, Belgium, Spain, Portugal, on some Mediterranean islands, and in North Africa. It has been considered a subspecies of the European/West-Siberian gray willow: S. cinerea ssp. oleifolia (Smith) Macreight (Meikle 1984, Berg 2000, Haines 2008, and others). However, many contemporary authors, including the world authorities on the genus Salix, treat rusty and gray willow as two different species (Argus 1997, 2005, 2007; Chmelar, Meusel 1986; Rechinger 1993; Skvortsov 1968, 1999; etc.). Remarkably, phenetic analysis of a Spanish population of S. atrocinerea demonstrated that it clusters with S. aurita L., S. starkeana Willd., and S. bebbiana Sarg.—and not with S. cinerea L. This supported the species rank for rusty willow (Argus 1997: 76, footnote).

The fact that rusty willow was included with gray willow has produced confusion. While gray willow is well known to have been introduced from Europe to many countries, rusty willow has not been distinguished in the majority of cases. It is highly probable that some records of gray willow actually refer to rusty willow. For example, in Australia and New Zealand gray willow is considered to be a highly invasive plant. The presence of rusty willow in Australia was directly confirmed only recently by Cremer (2003), who listed both subspecies of S. cinerea.

While S. cinerea had been long known as an introduced species in North America, S. atrocinerea was first identified in 1985 (Argus 1985, 1986) as herbarium specimens from North Carolina, the oldest of which had been collected in 1929 and 1939. Recently rusty willow was found in New York, most of New England, and southern Canada (Tucker 1996, 2006, 2007; Rawinski 2005a, 2005b; Argus 2005, 2007). In fact, we have found this willow to be the most conspicuous in eastern Massachusetts and particularly around Boston (Rawinski 2005a, 2005b; Kadis 2007; Zinovjev, Kadis 2008). We found rusty willow common in nearly all of eastern Massachusetts counties: Essex, Middlesex, Norfolk, Suffolk, Bristol, Plymouth, and Barnstable—i.e., everywhere where we made observations. We also observed it in central Massachusetts (Worcester County), where it appears to be more rare. In addition to examining willows in nature, we had an opportunity to study some recent willow collections from Worcester County provided by Robert Bertin as well as eastern Massachusetts samples in the Harvard University Herbaria.

Our observations of S. atrocinerea have been restricted to only New England populations. Apparently, the variability range in the introduced S. atrocinerea cannot be the same as in the area of its natural distribution in Atlantic Europe. Yet the only specimen of S. atrocinerea we observed in Europe looked quite similar to American plants. (It was a young cultivated shrub in a private willow collection north of its natural range, in southern Finland.)

Despite its prominence, the alien willow has been overlooked for decades, taken for either some native willows, mostly S. bebbiana or S. discolor Muhl., or else for other alien willows (S. caprea L. or S. cinerea). Apparently, we still don't have knowledge of the entire area rusty willow has acquired in eastern North America. Most likely, it is still being overlooked in some parts of the Atlantic Coast, particularly between New York and North Carolina.

S. atrocinerea is rather closely related to those American and European willows with which it has been confused, nearly all of them belonging to the same large group: sect. Cinerella (Argus 1997, 2007) or sect. Vetrix, subsect. Laeves (Skvortsov 1968, 1999). The only exception is S. bebbiana, which is related somewhat more distantly. It has been placed in sect. Fulvae (Argus 1997) or in a different subsection (subsect. Substriatae) of section Vetrix (Skvortsov 1968, 1999). One important morphological difference between S. bebbiana and the rest of the willows in question is the bud gradation type. Skvortsov (1968, 1999) distinguished three generic types of bud gradation along the shoot in willows and named them alba-, arctica-, and caprea-type (fig. 1-2). While willows belonging to Cinerella are all characterized by the caprea-type of bud gradation (fig. 2), in S. bebbiana buds exhibit a different pattern and rather approach the alba-type (fig. 1).

It was some ten years ago that we tentatively named a willow from the Boston area S. atrocinerea for the first time. We had not been familiar with S. atrocinerea in Europe; besides, the overall morphological variability of willow populations around Boston was so high that segregating willows into species appeared not as simple as in European Russia or Finland. Separating S. atrocinerea from the native S. bebbiana was particularly difficult. The latter is supposed to be one of the most common all across Massachusetts; however, it turned up to be impossible to find a typical Bebb willow in the Boston area. Those specimens we originally named S. bebbiana never looked exactly as the S. bebbiana we had observed in upstate New York, Canada, and different parts of Eurasia. It was much later that we found typical S. bebbiana in western and central Massachusetts. Yet we never identified a single typical S. bebbiana in the east, though plenty of old samples from eastern Massachusetts exist in herbarium collections. The presence of strange specimens reminding of Bebb willow along with complete absence of typical S. bebbiana can be explained if we assume that this willow has been eliminated around Boston due to hybridization with an alien species.

Willows are known to easily produce hybrids, though frequency of hybrids is often over-estimated. For example, many authors claim frequent hybrids between the common European willows S. caprea, S. cinerea, and S. aurita. However, according to Buser (1940: 653-655) and Skvortsov (1968, 1999: 72), hybrids between these species either do not exist or at least are extremely rare. Our own observations in northern Europe fully support this opinion. Abundant 'hybrids' found in nature often are nothing but misunderstood normal species. Skvortsov (1968, 1999), agreeing with Wichura (1865) and Buser (1887, 1940), stated that hybridization becomes significant only in specific habitats—either naturally unstable or disturbed by humans; yet even in those areas that are comparatively hybrid rich, hybrids usually don't dominate over parent species. However, when alien willows are introduced outside their native range due to human activities, the geographic isolation of species ceases to exist. Thus previously isolated species come into direct contact without having developed isolation mechanisms that prevent hybridization in nature. Most of the time adventive species occupy disturbed habitats known to constitute hybrid-rich zones. An exceptionally high rate of hybrid formation in these areas would not be normally possible in nature. The situation with S. atrocinerea in New England appears to present one of such cases of abnormal prolific hybrid formation. We are going to address the problem of hybrid formation with alien species in a different publication. Here we are going to demonstrate the existence of hybrids between S. atrocinerea and one more native willow from sect. Cinerella, S. humilis Marsh., by observing the seedlings originating from wild-collected S. humilis seeds.

Willows from the sect. Cinerella present a particular challenge when it comes to identification of dry samples; it is much more effective to look at live specimens. Winter is the best time to observe such important diagnostic characters as bud shape and position. Bark and wood surface characteristics are also available in winter (fig. 3-30). Observations of aments in spring are important by all means; yet the best time for willow identification is late summer and fall—the time when both leaves and buds are fully developed. Here is a brief account of some important characters in rusty willow, as compared to those of closely related species.

Rusty willow, if it is not damaged, grows as a tree to 10-12 m tall (fig. 53, 56), even though much more often we have seen it growing as a multi-stem tree/shrub (fig. 52) or even just a shrub (fig. 51). Of New England related willows, only S. bebbiana can grow as a tree. (The rest of tree willows in New England—S. nigra Marshall, S. alba L., S. fragilis L., and hybrids of S. babylonica L.—do not belong to the same subgenus and look quite different.) Tree growth is also characteristic of S. caprea; however, the latter has hardly ever become naturalized in New England. All the three species, differ from each other by the bark (fig. 21-30).

Prominent, dense and long longitudinal ridges (striae) are formed on S. atrocinerea wood underneath the bark (fig. 17). The wood ridges correspond to longitudinal depressions on the bark surface. The smooth, light gray bark on mature trunks and limbs is thus characteristically 'wavy' (fig. 21-22). Ridged wood is a character shared with S. cinerea, but that species does not grow as a tree. Besides, in European S. cinerea we never observed such prominently wavy bark as in populations of S. atrocinerea in Massachusetts. In S. caprea and S. bebbiana, wood ridges are either completely absent or short and sparse; the corresponding 'waves' on the bark are never formed. The bark on limbs and younger trunks of S. caprea is smooth, frequently quite distinctly greenish; its surface not wavy (fig. 23-25). The bark of mature S. bebbiana is very peculiar (fig. 28-30): with a beautiful 'braided' pattern, so that mature specimens of S. bebbiana can hardly be confused with other willows.

S. atrocinerea produces buds of two contrasting types, groups of large floriferous buds found in-between small vegetative buds. Skvortsov (1968, 1999) called this bud pattern caprea-type (fig. 2). It is typical for most species of the sect. Cinerella. This character helps separate S. atrocinerea from more distantly related S. bebbiana, whose floriferous buds are hardly distinguishable from vegetative (alba-type of bud gradation, fig. 1).

Floriferous buds in rusty willow are usually positioned at an angle to the branchlet, as the adaxial bud surface is convex; the lateral carina typically forms a sigmoidal curve when the bud is viewed from the side (fig. 7-9). This bud shape makes S. atrocinerea similar to the European related species (fig. 3-5) but different from all the related willows native to New England (fig. 12-14).

Branchlets in S. atrocinerea are typically slender, glabrous, reddish, though these characters are considerably variable. In this character, S. atrocinerea resembles the native S. bebbiana as well as European S. aurita; this should help distinguish S. atrocinerea from the native S. discolor and introduced S. cinerea with much thicker branchlets and branches. In S. cinerea branchlets also differ in color—they are entirely gray, dull 'ash' (hence the name gray willow or ash willow for the latter species).

Aments (fig. 37-38) in S. atrocinerea are short-stalked, densely flowered, their rachises never exposed. Bracts are black. The aments can hardly be confused with those of S. bebbiana, which are borne on foliated stalks, become very loose when ripe, and have light-colored bracts. (fig. 39).

In S. atrocinerea leaf venation (up to the third order) is prominent on the lower leaf surface forming reticulum (fig. 35) and a corresponding faint rugose pattern on the upper surface. This character is shared by nearly all of the discussed species including S. bebbiana. However, this makes S. atrocinerea different from S. discolor, which does not have reticulum on the lower leaf surface (fig. 36).

Rusty willow has received its common name for the ferrugineous trichomes that are often present on its lower (and sometimes upper) leaf surface together with grayish trichomes. While the presence of these rusty trichomes is diagnostic, their absence does not necessarily exclude S. atrocinerea. Ferrugineous trichomes may be more or less abundant or completely absent. Ferrugineous trichomes may be also found in S. discolor, but not in S. bebbiana, neither S. cinerea.

In S. atrocinerea the upper leaf surface is bright green and shining (fig. 31). S. cinerea leaves (as well as its gray branchlets), in contrast, are characteristically dull (fig. 32).

Stipules in S. atrocinerea are typical for most of the willows from Cinerella: broad, distinctly inequilateral, which helps distinguish S. atrocinerea from S. humilis.

As to the leaf shape, in rusty willow it is highly variable. Even though for a trained eye the variability range is somewhat different from that of other species, it is difficult to describe the difference concisely.

S. atrocinerea has Mediterranean and South-European Atlantic distribution type, while S. cinerea is distributed in Eurasia, east and somewhat north of S. atrocinerea range. Therefore, it is only natural that S. atrocinerea is much more common in eastern Massachusetts, on the Atlantic Coast than the continental S. cinerea. In eastern Massachusetts we saw just some few willows that could be reliably named S. cinerea. Even though it is sometimes difficult to separate dry herbarium samples, in nature the two willows typically look rather different from each other.

S. cinerea can be normally distinguished from S. atrocinerea as follows. This is a multi-stem shrub to 5-6(7) m, never growing as a tree; its branchlets stout; branches, branchlets, and buds grayish, with dense persistent pubescence (fig. 10-11); leaves with maximal width distinctly above the middle, gradually attenuate toward base; upper leaf surface rather dull, lower surface with grayish trichomes; ferrugineous trichomes completely absent.

Bebb willow, S. bebbiana Sargent (S. starkeana ssp. cinerascens Hultén) is nearly circumboreal. It is distributed across the entire North America and all of Siberia, reaching west to Scandinavia. On a large part of the European territory S. bebbiana is replaced by vicarious S. starkeana Willd. (Both can be as well treated as subspecies of a single species.) Within New England, we saw the typical S. bebbiana only on territories less infested by S. atrocinerea, such as western Massachusetts or New Hampshire; however, we have recently examined it in Colorado and earlier in New York State, Canada, different parts of Siberia, and northern Finland.

Bebb willow is either a shrub or a tree up to 10 m. Pistillate specimens of S. bebbiana can be readily separated from related species in New England at the end of the flowering season, when their characteristic aments with light-colored bracts become particularly loose due to growth of ovary stipes.

Bebb willow may be also easily distinguished from rusty willow by the bud shape and pattern: its floriferous buds aren't very different from vegetative (cf. fig. 6). Argus (2005) has classified S. bebbiana bud pattern with the alba-type (fig. 1). The buds are much more flat and elongate than those of S. discolor, S. atrocinerea, S. cinerea, S. caprea, or S. aurita, looking similar to those of vicarious European S. starkeana (fig. 6).

Mature Bebb willows can be distinguished from all other willows by their peculiar bark with characteristic 'braided' pattern (fig. 28-30). ^[1] A somewhat similar though less regular pattern can be observed on old stems of S. discolor (fig. 26-27). The bark pattern seems to be a reliable character—at least in those populations we observed in the US. Bark in specimens from central Massachusetts, upstate New York, and even Colorado looked quite similar. The ridges under the bark are either inconspicuous or completely absent.

The specimens we treat as hybrids between the two willows can be distinguished even during the winter, as their buds are more or less similar to those of S. bebbiana (fig. 49-50), while the bark looks much like that of S. atrocinerea: smooth and wavy, as the wood ridges are developed much better than in typical S. bebbiana (fig. 18). We have found the 'braided' bark character in S. bebbiana reliable enough for us to consider 'bebbiana-like' mature willows with smooth bark as hybrids, though further observations of putative hybrids, particularly during the time of flowering, are desirable. So far we only observed willows partially resembling S. bebbiana in their summer and autumn stage and never had a follow-up ament observation for any known putative S. atrocinerea x S. bebbiana hybrid.

Upland or prairie willow S. humilis Marsh. is an eastern North American shrub of small to medium size, not more than 3 m tall.

Though we have found S. humilis to be rather common in central Massachusetts, we located only a few scattered individuals in the Boston area: some few in the Middlesex Fells Reservation north of Boston and only three in the Stony Brook Reservation, southern Boston. All three solitary individuals in Stony Brook are rather low and delicate, even resembling dwarf upland willow S. occidentalis Walter (S. tristis Aiton).

Many authors consider the latter as just a variety of S. humilis: S. humilis var. tristis (Aiton) Criggs or S. humilis var. microphylla (Anderss.) Fern. Others treat it as a distinct species (Gleason, Cronquist 1963: 239, 1991: 173; Skvortsov 1971). S. occidentalis is a delicate shrub only up to 0.7 m with miniature leaves. It is impossible to confuse this willow with S. atrocinerea. So far we have found it only twice: in the Middlesex Fells Reservation and Myles Standish State Forest.

In addition to its habit, S. humilis can be distinguished from S. atrocinerea by the following characters. Floriferous buds in S. humilis are oblong-ovoid (fig. 14-15). Leaves are more elongate, with somewhat revolute margins, their maximum width well above the middle. Lower leaf surface is densely tomentose, at least in young leaves. Stipules, when present, are subsymmetrical, nearly lanceolate to linear (fig. 14). In the latter character S. humilis (together with S. occidentalis) is different from the rest of the willows discussed in this article.

S. atrocinerea is abundant in the Stony Brook Reservation. Due to the rareness of same-species 'partners' around, scattered individuals of S. humilis might be pollinated by the readily available pollen of rusty willow and thus produce hybrid seed. In order to test this hypothesis, we grew plants from seeds collected from one of S. humilis. The progeny grew into vigorous, fast-growing, coarse single-stem plants (fig. 44-45), which apparently have little to do with a delicate mother shrub (fig. 42-43). One of the two grown plants (fig. 46) has distinctly wavy bark characteristic of S. atrocinerea. Both willows would actually fit S. atrocinerea, if it were not their elongate leaves (fig. 48), somewhat decumbent shoots (as in S. humilis or S. occidentalis), atypical aments (fig. 47), and floriferous buds of an odd shape (fig. 16).

The North American pussy willow S. discolor Muhl. is a boreal, mostly eastern species of wide distribution. This tall shrub can grow up to 6(8) m. Usually pussy willow can be distinguished from related species by its leaves, which are rather broad, lacking impressed veins above and conspicuous reticulation beneath (with mostly just first-order veins prominent). Decorticated wood is mostly smooth or with sparse and short, indistinct ridges (fig. 20). Bark on old stems may form ridges producing irregular 'braided' pattern (fig. 26-27). The floriferous buds are large, often with black scales during the winter time. The bud shape is drastically different from that in related European willows, including S. atrocinerea. The bud shape in S. discolor is indeed closer to that of willows from other sections, such as S. hastata L. or even S. phylicifolia L. or S. pulchra Scham. (Skvortsov 1955: fig. 7; Belyaeva et al. 2006: fig. 41, 68, 71). The aments are somewhat similar to those of S. atrocinerea, especially at the start of their development. Precocious, appearing very early in spring, they are very large and dense. However, later on pistillate aments become loose, though never as lax as those of S. bebbiana (fig. 39). Another diagnostic character of S. discolor is long, spreading stigmas (fig. 40).

Even though a possibility of its hybridization with S. atrocinerea cannot be excluded (see the notes under S. caprea), S. discolor remains relatively common around Boston, sometimes growing together with S. atrocinerea in wet habitats.

Goat willow, S. caprea L. is an Eurasian willow having a habit of a tree—sometimes a mighty tree (Lagerström, Uronen 2005: 118, photo); otherwise, with a transitional multi-stem shrub/tree habit. This is one of the most mesophilous and shade-tolerant species among the European willows, so that it is frequently found in relatively dry woods, far away from any water (while both S. cinerea and S. atrocinerea are typical wetland species).

The bark on old goat-willow trunks forms coarse longitudinal fissures (fig. 23), as in most tree willows. On younger trunks and branches the bark is smooth, like in S. atrocinerea, though without any trace of wavy surface due to complete absence of ridges on the wood underneath the bark. The bark on younger limbs is greenish-gray (fig. 24-25).

Under the name 'pussy willow,' it has been frequently recorded as 'cultivated' and 'naturalized' in the United States. However, this can hardly be true, since S. caprea is very difficult to propagate from cuttings. In most cases, these records are to be attributed to S. cinerea or S. atrocinerea (Argus 1986: 115, 120).

On the other hand, some cultivated 'pussy' willows we observed did not look like S. cinerea or S. atrocinerea. In this respect, a noteworthy remark by A. Skvortsov comes to mind. While reviewing yet another Old World species from the same group, S. aegyptiaca L., he described it as 'favored for cultivation, since it is easily propagated from cuttings, unlike S. caprea' (Skvortsov 1999: 178). One may assume that S. aegyptiaca might have been also overlooked in the United States—taken for S. caprea. As far as we know, there are no records of wild-collected S. aegyptiaca in North America.

One more possibility is that some hybrid willows have been taken for goat willow. S. caprea is first and foremost characterized by large leaves—mostly broader than in the majority of related species (fig. 33). A few times we have observed around Boston strange non-native willows with unusually broad, often suborbicular leaves. In other respects they did not look anything like S. caprea. We cannot yet tell if these fit within the variability range of S. atrocinerea or also represent hybrid specimens. In the latter case, the parent from which the broad leaves had been inherited could only be S. discolor, as this is the only native species in our area with a similar leaf length-to-width ratio. These plants of obscure origin might have been taken for S. caprea.

We haven't had an opportunity to study the S. caprea samples from the southeastern US cited by Argus (1986: 115). So far none of those specimens we observed in Massachusetts seemed to be true S. caprea. More evidence is needed in support of the opinion that S. caprea has been naturalized in eastern North America.

Eared willow, S. aurita L. is another European species considered to be introduced and even naturalized in two of the eastern states: Pennsylvania and Massachusetts. However, it is important to note that, similarly to S. caprea, this willow cannot be easily propagated from cuttings. According to Belyaeva et al. (2006), the outcome of propagation from cuttings in S. aurita is less than 5%—nearly as low as in S. caprea (0-2%).

Eared willow has received its name for the characteristic prominent and persistent stipules, which we have not observed in specimens identified as S. aurita from Massachusetts. In Europe S. aurita mostly grows as a comparatively low shrub only to 1-2 m in height—much lower than S. cinerea. Its tall-growing variety is known in Scandinavia (Berg 2000, Roininen et al. 2001, Vikberg pers. comm.), though even when it grows tall, its stem never reaches the diameter of S. atrocinerea. In comparison with S. atrocinerea, leaves of S. aurita are more dull, of somewhat different color, with even more pronounced reticulate pattern on both sides (fig. 34). Its floriferous buds are slightly different from those in the rest of related willows (fig. 5). The bracts in S. aurita pistillate aments are paler than in S. atrocinerea and S. cinerea.

So far we don't have evidence which would justify the inclusion of S. aurita in the flora of North America. A few herbarium samples from Massachusetts identified as S. aurita we have examined, rather represent S. atrocinerea or, even more likely, its hybrids with S. bebbiana. S. aurita really resembles S. bebbiana or S. atrocinerea, so that poorly collected specimens of S. atrocinerea can be easily mistaken for S. aurita.

Generally speaking, willow identification is not as difficult as it is usually considered. Though it may not be easy to name every specimen, particularly when hybrids are involved, it should not be difficult to discriminate between the New England native willows and non-native S. cinerea and S. atrocinerea together with their putative hybrids—at least in the field. The alien willows are readily identifiable even during the winter. The peculiar smooth, wavy bark of S. atrocinerea and its hybrids never occurs in any native willows in New England. The sigmoidal floriferous bud shape typical for European S. atrocinerea, S. cinerea, and S. caprea is different from the bud shape in any related New England willows. Even when S. atrocinerea buds lack the typical sigmoidal shape, they still look very much different from buds of any native willows.

Both S. atrocinerea and S. cinerea start flowering earlier than the majority of native willows. When a tall staminate S. atrocinerea goes into flower, it can be spotted from afar. The only native willow around Boston that may flower at the same time is S. discolor. Though S. discolor can grow nearly as tall as S. atrocinerea, it has a different habit, usually forming a spreading rounded shrub, while S. atrocinerea typically develops a narrow upright crown. Certainly, upon noticing a staminate S. atrocinerea from the distance, one needs to take a closer look to confirm the initial identification.

Like many other alien invasive plants, such as glossy buckthorn or Oriental bittersweet, S. atrocinerea takes advantage of a prolonged vegetation season, as compared with the growing season of the native willows. In other words, it once more becomes prominent in the landscape late in the fall, retaining its characteristic dirty-green foliage well after all the native willows and the majority of other native woody plants have dropped their leaves. The leaves often stay on branches until the first frost and then sometimes remain as marcescent, thus, again, making the alien willow noticeable at the beginning of winter (fig. 57).

S. atrocinerea and S. cinerea (either in the broad or narrow sense) have not yet been included on the Invasive Plant List of Massachusetts, despite the evidence provided by Rawinski (2005a, 2005b, etc.). S. atrocinerea frequents roadside ditches and other roadside wet habitats, sometimes penetrating into open wetlands (fig. 51-52). We have found S. atrocinerea and its hybrids on nearly every pond shore around Boston, though only rarely in shaded vernal pools (Stony Brook Reservation). According to Rawinski (Cutko, Rawinski 2007: 86), the invasive willows are not infrequent in vernal pools.

While S. atrocinerea is capable of vegetative reproduction, the major mechanism of expansion appears to be seed dispersal. Rusty-willow seedlings and saplings (fig. 54-55) occur around many ponds, reservoirs, and lakes including habitats of rare pondshore species, such as Plymouth gentian. Massive germination of rusty willow (on average, 15 seedlings per sq. m for the total plot area of 92 sq. m) was recorded in the Blue Hills Reservation, at the bottom of the recently drained Blue Hills Reservoir (2006, own unpublished data). Our impression is that alien willows are on the rise in eastern Massachusetts, becoming more and more prominent every year.

Taking into consideration a possibility of hybridization between the alien and native willows on a large scale, one may predict that the harm produced by the invasive species in New England is going to exceed that inflicted in Australia or New Zealand, where there are no native willows. In New England, a few native willows might become extinct by way of producing hybrids with rusty willow (Kadis, Zinovjev 2008).

There already exists a precedent in the human history when an introduced willow eliminated a native species in Europe by way of forming hybrids with it. Centuries ago, when S. fragilis L. was introduced from Asia Minor to Europe, it produced a huge secondary range there and started to hybridize with the native European willow S. alba L. on such a grand scale that both pure S. alba and pure S. fragilis have become rare within a major part of Europe (Skvortsov 1973). This hybrid complex has been introduced to the New World under the names of S. alba and S. fragilis.

Tom Rawinski's breakthrough was critical for our understanding of the alien willow. He was the first person with whom we could share our initial concerns. We are grateful to George Argus for his continuous support, confirmation of the identity of the alien willow, and also for completing a short survey of S. atrocinerea populations in the Boston area in 2006. We are indebted to Robert Bertin, Paul Somers, and Tom Palmer for their helpful comments on drafts of this article. We are grateful to Robert Bertin for showing S. bebbiana and other willows from Worcester County. Photography of European willows would not be possible without generous help of Veli Vikberg (Turenki, Finland). European rusty willow was observed in Tapani Uronen's private willow collection in Kasiniemi, Padasjoki (Finland).

Argus, G.W. 1985. Computerized catalogue of herbarium specimens of Salix in the southeastern United States. National Museum of Canada. Ottawa, ON, Canada.

Argus, G.W. 1986. The genus Salix (Salicaceae) in the southeastern United States. Syst. Bot. Monogr. 9. 170 pp.

Argus, G.W. 1997. Infrageneric classification of Salix (Salicaceae) in the New World. Syst. Bot. Monogr. 52. 121 pp.

Argus, G.W. 2005. Guide to the genus Salix (willow) in New England and New York. New York Flora Association. Salix identification workshop. June 10-12, 2005. Website http://www.nyflora.org/Workshops/Salix_Workshop/GuideSalixNE-NY4Mar05.pdf (accessed 16 January 2009).

Argus, G.W. 2007. Salix L. (Salicaceae) distribution maps and a synopsis of their classification in North America, north of Mexico. Harvard Papers in Botany 12: 335-368. Online version: http://aknhp.uaa.alaska.edu/willow/pdfs/Argus_2007_Synopsis_Maps.pdf (accessed 16 January 2008).

Belyaeva I.V., O.V. Epanchintseva, A.A. Shatalina, and L.A. Semkina. 2006. Willows of Ural: Atlas and identification key. UB RAS, Ekaterinburg. 176 pp.

Berg, C.C. 2000. Salix, pp. 131-133. In: B. Jonsell, Flora Nordica 1. The Bergius Foundation, The Royal Swedish Academy of Sciences, Stockholm.

Buser, R. 1887. Die Brüggerschen Weidenbastarde, pp.49-92. In: A. Gremli, Neue Beiträge zur Flora der Schweiz 4. Aarau (cited after Skvortsov 1999).

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

Chmelar, J., W. Meusel. 1986. Die Weiden Europas. Gattung Salix.

Cremer, K.W. 2003. Introduced willows can become invasive pests in Australia. Biodiversity 4(4): 17-24.

Cutko, A., T. Rawinski. 2007. Flora of northeastern vernal pools. In: A.J.K. Calhound, P.G. deMaynadier (eds). Science and conservation of vernal pools in northeastern North America.

Gleason, H.A., A. Cronquist. 1963. Manual of vascular plants of northeastern United States and adjacent Canada. D. Van Nostrand Co., New York. 810 pp.

Gleason, H.A., A. Cronquist. 1991. Manual of vascular plants of northeastern United States and adjacent Canada. 2nd ed. The New York Botanical Garden, New York. 910 pp.

Gould, L.L. 2005. The invasive beat: what's here and where is it? Rhode Island Naturalist 12(2): 13-14.

Haines, A. 2008. Synonymized checklist of New England tracheophytes. Website http://www.arthurhaines.com/checklist.htm (accessed 16 January 2009).

Kadis, I. 2007. Unexpected guests become unwelcome owners. Friends of the Blue Hills Newsletter. 32(1): 1, 5. Online version: http://www.friendsofthebluehills.org/PDFs/JanMar07.pdf (accessed 16 January 2009).

Kadis, I., A. Zinovjev. 2008. The alien rusty willow Salix atrocinerea: A possibility of hybridization with American willows. Website http://www.salicicola.com/articles/atrocinerea2.

Lagerström, M., T. Uronen. 2005. Pajut puutarhassa. Küstannusosakeyhtiö Tammi. Helsinki. 216 pp.

Meikle, R.D. 1984. Willows and poplars of Great Britain and Ireland. Botanical Society of the British Isles Handbook 4.

Rawinski, T. 2005a. European gray willow (Salix cinerea): A message from Tom Rawinski, USFS. Invasive Plants Newsbriefs 14, New England Invasive Plant Group [NIPGro], 25 Aug 2005 (cited after Gould 2005).

Rawinski, T. 2005b. In: Rosenholm, G. Botanist says stealthy invasive willow putting local rare plants at risk. USDA Forest Service Northeastern Area News Release 14-05. Website http://www.na.fs.fed.us/nanews/archives/2005/oct05/oct14_05.htm (accessed 16 January 2009).

Rechinger, H.K. (revised by J.R. Akeroyd). 1993. Salix. In: T.G. Tutin et al. Flora Europaea, Vol. 1. ed. 2. (Salicaceae edited by J.R. Edmonson).

Roininen H., J. Tahvanainen, V. Vikberg, and A. Zinovjev. 2001. Salix aurita—the correct food plant for Euura cinereae Kopelke 1996 (Hymenoptera,Tenthredinidae). Entomologica Fennica 12: 129-130.

Skvortsov, A. K. 1955. The willows of Central European Russia and their identification during the wintertime. Bull. MOIP, otd. biol., LX (3): 115-127. In Russian; English translation at website http://www.salicicola.com/translations/Skvortsov1955.html.

Skvortsov, A. K. 1968. Ivy SSSR. Sistematicheskiy i geograficheskiy obzor [Willows of the USSR. Taxonomic and geographic revision.] Nauka, Moscow. 262 pp.

Skvortsov A. K. 1971. Euro-amerikanische verwandtschaftliche Beziehungen zweier mediterraner Salix-Arten. Feddes Repertorium 82(6): 407-420. English translation at website http://www.salicicola.com/translations/Skv1971EA.html.

Skvortsov, A.K. 1973. [Present distribution and probable primary range of brittle willow (Salix fragilis L.)], pp. 263-280. In: Problemy biogeotsenologii, geobotaniki i botanicheskoy geografii. Nauka, Leningrad. English translation at website http://www.salicicola.com/translations/Skv1973SF.html.

Skvortsov, A.K. 1999. Willows of Russia and adjacent countries. Taxonomical and geographical revision. Univ. Joensuu. Faculty of Mathematics and Natural Sciences Report Ser. 39. Joensuu, Finland. English translation of Skvortsov 1968. Ivy SSSR. Sistematicheskiy i geograficheskiy obzor [Willows of the USSR. Taxonomic and geographic revision.] Nauka, Moscow. 262 pp. Online version: http://www.salicicola.com/announcements/skv/skvortsov.pdf.

Tucker, G.C. 1996. Salix atrocinerea—an overlooked willow in New York State. New York Flora Association Newsletter 7(2): 2.

Tucker, G.C. 2006. Additions to the flora of Rhode Island. Rhodora 108: 65-71.

Tucker, G.C. 2007. Additions to the flora of Connecticut. Rhodora 109: 459-463.

Wichura, M. 1865. Die Bastardbefruchtung im Pflanzenreich, erläutert an den Bastarden der Weiden. Breslau (cited after Skvortsov 1999).

Zinovjev, A., I. Kadis. 2006. Willows of New England. Comparison of introduced species Salix atrocinerea and S. cinerea. Website http://www.salicicola.com/notes/atrocinerea_cinerea.

^[1]   A somewhat similar though less regular pattern can be observed on old stems of S. discolor (fig. 26-27).

---

26 Jan 2009