Irina Kadis, Alexey Zinovjev

A while ago, as newcomers to New England, we were eager to familiarize ourselves with the eastern North American willows. The list of common willows that we expected to find easily included eight or nine species: black willow (S. nigra Marsh.)—the only native large tree willow; shining willow (S. lucida Muhl.)—the eastern Noth American counterpart of the European laurel willow (S. pentandra L.); heart-leaf willow (S. eriocephala Michaux); silky willow (S. sericea Marsh.); slender willow (S. petiolaris Smith); Bebb's willow (S. bebbiana Sarg.)—the only old acquaintance of ours, as it is widespread outside North America, all across Siberia; American pussy willow (S. discolor Muhl.); and upland willow (S. humilis Marsh.)—along with its dwarf variety (S. humilis var. microphylla Fern. or S. tristis Aiton).

To our amazement, we found native willows to be scarce in eastern Massachusetts—to the point that willows were often missing from appropriate habitats, such as grassy wetlands or roadside ditches. Locating all of the "common" species turned out to be a daunting task! Only three or four species on our list were relatively common and easy to find. Those were black (S. nigra), heart-leaf (S. eriocephala), slender (S. petiolaris), and pussy willow (S. discolor). Silky willow (S. sericea) was less common, and upland willow (S. humilis along with its dwarf variety) appeared to be extremely rare. As to shining (S. lucida) and Bebb's willow (S. bebbiana), we failed to find any, whatsoever! This was quite astonishing, as we had assumed that these willows were common Massachusetts shrubs.

Our confusion was deepened by the fact that the most common willow in the area, seen again and again in disturbed wet habitats—along ditches, small streams, and on banks of reservoirs—was not a native species. Even though often times it vaguely reminded of S. bebbiana, it was different from the S. bebbiana we had known. Furthermore, it did not look like any European willow familiar to us. Attending field trips lead by local botanists did not help, either. The strange willow, when it occurred along the route, was introduced either as Bebb's or pussy willow, which did not sound satisfactory. Our attempts to key out the unfamiliar willow yielded unstable results, at times bringing us to S. bebbiana, other times to S. atrocinerea Brotero, a West European willow we had never observed in the wild.

A few years passed before we before we clarified the identity of this species. The answer surfaced in 2005 from an unexpected source: the Metro! A short newspaper article reported on the findings of Tom Rawinski, a US Forest Service Officer in New Hampshire. Yes, said the article, there is indeed an alien willow in eastern New England, particularly on the coast. It has been spreading unnoticed, being taken for one or another native willow. The "new" species is the European ash willow, S. cinerea L.— already an infamous invasive in Australia and New Zealand.

We were excited, though could not quite agree with the identification (ash willow is another European willow well familiar to us). A fresh look at our problem species showed it to be rusty willow (S. atrocinerea). Rusty willow is closely related to ash willow, so that both are frequently treated as subspecies of a single species S. cinerea: ssp. cinerea and ssp. oleifolia Macreight. Apparently, Tom Rawinski understood ash willow in the broad sense or rather meant just the ssp. oleifolia.

Since 2005 (Rawinski 2005, Gould 2005), the presence of rusty willow all along the New England coast (Tucker 2006, 2007) and beyond, from eastern Canada to North Carolina, has been completely acknowledged. It turned out that rusty willow had been first reported from the United States in 1985, when George Argus identified herbarium specimens from North Carolina as S. atrocinerea. The oldest of those samples were collected in 1929 and 1939 (Argus 1986). Rusty willow was later reported from New York—found by G. Tucker and identified by G. Argus (Tucker 1996).

The exact distribution of rusty willow in eastern North America is still unknown. According to our own observations, it is likely to be present in most of eastern Massachusetts. We have also found it in central MA, but so far never in the west, though we don't travel much across the western counties.

With the confirmation of the identity of the strange willow, a major part of the mystery was resolved. However, there remained other questions. Why is the alien willow so variable, even among other willows, which are generally famous for their infraspecific variability? Why do some of rusty willow specimens tend to resemble native Massachusetts species, particularly Bebb's willow? Why do we find specimens with characters intermediate between rusty willow and one of the native willows? Perhaps the absence of some native species in eastern Massachusetts has to do with the presence of the alien willow? We started to notice more and more situations that could be interpreted as evidence of some connection between the presence of the alien and absence of a few natives.

Willows are famous for their ease of producing hybrids—be it under experimental conditions or naturally. Hybrids are often fertile, thus giving birth various forms through subsequent segregation. Indeed, a common opinion for more than a century has been that willows hybridize to such an extent that normally hybrids even prevail over the species.

A pioneer researcher of willow hybrids was M. Wichura (1865). He estimated that natural willows hybrids account for fewer than 1% from the occurrence amount of willow individuals in nature.

Hybrids are often characterized by low vitality and fertility. As "experiments of nature," most hybrids are apparently less adapted to the environmental conditions than parental species and thus are short lived. This is one possible explanation of the rarity of willow hybrids in nature. According to Skvortsov (1999), the abundance of natural willow hybrids has been habitually overestimated. Abundant "hybrids" found in nature most of the time are nothing but misunderstood normal species. Our own observations fully support this opinion.

Another common misconception is that natural hybrids are commonly formed by closely related species (for example, S. caprea L. and S. aurita L. or S. caprea and S. cinerea, which belong to the same section). However, this may be unlikely because closely related species are bound to be isolated from one another—otherwise, they would have ceased to be distinct! In case closely related species with normal cross-pollination co-inhabit the same territory, they have to "protect their identity" through isolation mechanisms, which include differences in flowering time, genetic incompatibility, and low viability of hybrids.

Skvortsov (1999) pointed out that parents of natural hybrids in willows usually belong to different sections. Though at the first glance this statement seems paradoxical, it can be readily explained by the "necessity" for species to sustain themselves.

It is only in specific habitats—either naturally unstable or disturbed by humans—that hybridization becomes significant. Yet even in those areas that are comparatively hybrid rich, hybrids usually don't dominate over parental species. Skvortsov (1999) listed only a few known exceptions to this general rule—situations when hybridization in willows occurs en masse—and between c losely related species.

One of these hybrid-dominated areas is a rather broad zone of contact between a pair of vicarious willows (that is, species of common origin that now occupy adjacent geographic areas): Bebb's willow (S. bebbiana Sarg.) and Starke's willow (S. starkeana Willd.). The hybrid zone is the territory where the areas of these two species overlap. It is a stripe crossing the European Continent from northwest to southeast, that is, from Scandinavia all the way to the Southern Urals.

A second, similar case is one involving another pair of vicarious species: S. repens L. and S. rosmarinifolia L. The hybridogeneous overlap zone for this pair exists in Central Europe: southern Germany, Czechia, and Poland. Both examples are cases of temporary geographic isolation of taxa, where the counterparts became separated during the glacial period and thus evolved independently for a while. However, once both returned to their former areas upon the retreat of the glacier, they started to hybridize all along the newly formed contact zone. The current situation with both species pairs has been even more aptly described using the category of subspecies: both pair members may just as well be considered as subspecies of a single species. S. rosmarinifolia has been also called S. repens ssp. rosmarinifolia Koch; another name for S. bebbiana is S. starkeana ssp. cinerascens Hultén.

The above examples show that closely related species, which are normally geographically isolated from one another, may still hybridize if external forces bring them in contact. And all of us are well aware of the importance of humans as an external force affecting the natural world.

A case of willow hybridization that presents a special interest for our analysis is that of brittle willow (S. fragilis L.) and white willow (S. alba L.). These two trees belong to the same section. White willow is native from Europe to West Siberia. For a long time, brittle willow was also considered to be a native European willow. It was Skvortsov (1973, English transl. 2008), who untangled and told the amazing history of this tree. He showed that brittle willow is actually an endemic species of Asia Minor with a small geographic range there and that it was introduced to Europe only around the Medieval Ages, by humans. Once it found itself in Europe, brittle willow very quickly acquired an enormous secondary area. The prompt advancement was accomplished mostly by vegetative reproduction: branchlets of brittle willow easily break off (hence its common name) and readily root.

While conquering Europe, brittle willow apparently was invading the range of the closely related native European white willow. Both species had been well isolated by mountain ranges before the advancement of brittle willow and thus used to be distinct. As a result of the introduction of brittle willow to Europe, the two tree willows started to hybridize on such a grand scale that the hybrids promptly became much more common than either species in most of Western Europe! Meanwhile, due to the expansion of humankind, the native white willow was completely deprived of its natural habitats all across Western Europe. As a result of habitat loss and hybridization with the alien willow, it lost its identity as a distinct species. The hybrid progeny of white willow and the alien brittle willow dominated instead, quickly becoming the most common adventive willow in Western Europe.

In the 18th-20th century, the deceptively common, seemingly "native" trees actually representing the complex of adventive hybrids were mistakenly mentioned multiple times in all the European floras and botanical manuals instead of the true brittle willow! The hybrids, of course, back-crossed with parental species—a process called introgressive hybridization—forming an intricate multi-generational complex.

This complex of hybrids between brittle and white willow was inadverently introduced to North America along with the true white and brittle willow. Luckily, there is no North American tree willow closely related to white and brittle willow. This introduction did not compromise the identity of the North American tree willow S. nigra, which belongs to a different, distantly related section.

Black willow (S. nigra) has not been affected by the presence of alien willows and remains a common pond and riverside tree around Boston. However, this might not be the case with those New-England willows that belong the section Vetrix/Cinerella—once they found themselves in close proximity to the introduced rusty willow belonging to the same section. So far we don't have direct genetic data in support of this hypothesis, though there is some anecdotal evidence.

The section Vetrix or Cinerella is one of the most difficult, species-rich sections of willows, known for extremely subtle morphological differences between its members. Just by glancing at a certain willow, it is usually quite easy to tell that it belongs to this section; however, it is not always a simple task to pin-point the species. Indeed, it is so hard that many authors writing on these willows never manage to discriminate between species often growing close together and thus arrive to the conclusion about the hybrid nature of the majority of plants. The fact that the alien willow also belongs to Vetrix/Cinerella explains why it has been overlooked for decades, if not centuries, taken for some native willow of the same section. The three common native Vetrix/Cinerella willows of eastern New England are supposed to be S. discolor, S. humilis, and S. tristis, the latter often considered a dwarf variety of S. humilis. Additionally, S. bebbiana is placed either in a separate subsection within the same section (Skvortsov 1999) or in a different, closely related section, which contains only a single species in North America (Argus 1997).

Considering the current situation with these species in eastern Massachusetts, let's start with the most puzzling observation of ours. As much as we searched for S. bebbiana around Boston and elsewhere in eastern Massachusetts, we never found a single typical individual. At the same time, we collected typical Bebb's willow in central and western Massachusetts (in Berkshire and Worcester counties)—either in the complete absence, or at least a m uch more modest presence of the alien willow. S. bebbiana appears also to occur north of Massachusetts—in northern non-coastal New England and Canada.

On the other hand, we have observed and collected numerous specimens of the exotic willow that more or less exhibit some of S. bebbiana characters. No wonder that in the absence of Bebb's willow the alien has been taken for it! The paradoxical absence of a common species makes one wonder if the alien willow has reduced Bebb's willow in the Boston area—at least partially—by hybridizing with it, to the point that Bebb's w illow totally lost its identity.

The case of Bebb's willow seems to be the extreme. The next native willow that might be badly affected appears to be upland willow (S. humilis). This delicate shrub, whose Latin name aptly describes its appearance, usually produces only solitary individuals of a modest size, never forming thickets in our area. As its populations are scattered so much, the species apparently is facing additional challenge, once it comes to the cross-pollination.

We found this willow so scarce around Boston that we celebrated every specimen, marked all we found, returned to them multiple times, and collected from them for the purpose of propagation. We located a few scattered individuals growing along the East Boundary of the Stony Brook Reservation—an urban reservation in Boston. The seedlings produced from one of them ended up on the Arnold Arboretum's grounds—planted right along the Meadow Road, the main avenue of the Arboretum. The label says: Salix humilis collected by Irina... Alas! These vigorous, fast-growing, coarse shrubs apparently have little to do with S. humilis! They might serve, however, a curious piece of evidence regarding the probable work in progress: elimination of a native willow in eastern Massachusetts by the alien species by way of drowning its gene pool within the alien's own gene pool. Apparently, due to the lack of same-species "partners" around, the scattered individuals are being pollinated by the readily available alien pollen and thus produce hybrid seed. As a consolation, we shall add that perfectly normal, non-hybrid upland willow becomes common as far west as central Massachusetts. Perhaps some still exist in eastern Massachusetts.

As to the dwarf upland willow, S. tristis, we have been able to locate only two tiny clones (each about 10 inches tall and perhaps one yard in diameter) in two different reservations of northeastern and southeastern Massachusetts. We have never observed dwarf upland willow in flower, let alone producing fruit. One of the two clones appears to be in rapid decline. It is hard to guess the immediate cause of this degradation.

Of the Vetrix/Cinerella willows, pussy willow (S. discolor) seems to be the least affected by the presence of the alien willow. Large pussies "in good standing" are relatively common in eastern Massachusetts, though our general feeling is that there could have been more of them. However, even within this comparatively intact species, some doubtful individuals exhibit characters intermediate between pussy and rusty willow.

Finally, a thought about another mysterious gap in eastern Massachusetts: the absence of S. lucida, a species abundantly represented in century-old herbarium collections. This willow could not be genetically compromised by the presence of S. atrocinerea as it belongs to the section Pentandrae/Salicaster, a rather remote one from Vetrix/Cinerella, attributed to a different subgenus. However, S. lucida is genetically close to another alien complex mentioned here above—that of S. alba/fragilis—indeed, closer than any other native willow in Massachusetts...

Elimination of native species by a closely related alien perhaps is not unique for willows. Another interesting example is a case of American bittersweet, Celastrus scandens L. According to the Massachusetts Checklist (Sorrie and Somers 1999), which is based on herbarium specimens, this native vine is distributed within Massachusetts all the way from the Berkshires to Nantucket. A unique document of the late 19th century, the flora of the major eastern-Massachusetts state reservations (Deane 1896), listed American bittersweet as "occasional" in the Blue Hills, frequent in Middlesex Fells, and spread at least "in the woods west of Turtle Pond" in Stony Brook Reservation. At the same time, Oriental bittersweet (Celastrus orbiculatus Thunb.), an alien invasive vine, now one of the commonest vines in eastern Massachusetts, was amazingly missing from Deane's list! We assume that this adventive vine then was either absent or very insignificant—and thus possibly overlooked by Deane's team; and that it has made its advancement at a breathtaking speed during the last century.

American bittersweet is readily distinquishable from Oriental bittersweet in its flowers and fruit positioned at tips of branchlets For a few years we have searched the three reservations surveyed by Deane and his colleagues and elswhere in eastern Massachusetts in hope to find the native vine. We have asked field botanists about it. Despite all efforts, we failed to locate a single plant—and no one so far has been able to help us. Our only photo of American bittersweet had to be taken on the Arnold Arboretum grounds. The Arboretum's specimen had been collected wild in 2002 in Indiana.

Our observations suggest that alien willows may pose an immediate threat to the very existence of the native American willows—not only by competing with them for space and resources, but also through "swallowing" genotypes of closely related species by producing interspecific hybrids—a much more scary scenario!

However, if we try to grasp a wider picture by looking at current occurrences of other native species within the state of Massachusetts, we inevitably notice a certain trend in their current distribution, which seems to be true for many species. It is not just willows that have recently disappeared from the east. Many of those plants that are supposed to be common all across the state according to the old herbarium collections, are now found primarily in the western and central MA, but not anymore in the east. For example, bloodroot (Sanguinaria canadensis) is supposed to be present all the way east including Norfolk County. Have you often seen it in the Blue Hills or anywhere around there? Painted and purple trillium (Trillium undulatum and T. erectum) are to be found as far east as Plymouth Co.; strawbell (Uvularia perfoliata) should be present as far east as Bristol Co.; trout lily (Erythronium americanum)—even farther east, down to Barnstable Co.; etc., etc. There are scores of examples like this, involving both herbaceous and woody species: for example, hobble-bush (Viburnum lantanoides) or red-berried elderberry (Sambucus racemosa). Even if these species are still present in the east, they have been increasingly scarce there and much more difficult to find than in the west, where they are still abundant. Perhaps there are multiple reasons for plants to behave this way, the major cause appearing to be a profound habitat disruption in the east. Considering this general background, the situation with willows may not be particularly exceptional.

Perhaps the right answer to the multiple-choice question regarding the lack of common species in eastern Massachusetts would be "all of the above." It is only by employment of molecular biology methods that we will be able to tell exactly whether the hypothesis about genetic swarms of the native and alien willows is truly grounded.

Tom Rawinski's observations played a major part in our understanding of the alien willow. The confirmation of the rusty willow identity in our collections and during the field trip in October 2005 by George Argus has been of great importance to us. We wish to thank Robert Bertin for showing a good Bebb's willow in West Boylston, the closest one to eastern Massachusetts we know as of today; and also for improving the style of this article.

Argus, G. W. 1986. The genus Salix (Salicaceae) in the Southeastern United States. Systematic Botany Monographs. Vol 9. 170 pp.

Argus, G. W. 1997. Infrageneric classification of Salix (Salicaceae) in the New World. Systematic Botany Monographs. Vol. 52. 121 pp.

Flora of the Blue Hills, Middlesex Fells, Stony Brook, and Beaver Brook reservations, of the Metropolitan Park Comission, Massachusetts. 1896. Edited by Walter Deane. Preliminary Edition. Boston. C.M. Barrows & Co.

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

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

Skvortsov, A. K. 1973. [Present distribution and probable primary range of brittle willow (Salix fragilis L.)]. Problemy biogeotsenologii, geobotaniki i botanicheskoy geografii. Leningrad, Nauka, pp. 263-280.

Skvortsov, A. K. 1999. Willows of Russia and adjacent countries. Taxonomical and geographical revision. University of Joensuu, Faculty of Mathematics and Natural Sciences, Report Series # 39. Joensuu. English translation. 307 pp.

Skvortsov, A. K. 2007. Present distribution and probable primary range of brittle willow (Salix fragilis L.) Problemy biogeotsenologii, geobotaniki i botanicheskoy geografii. Leningrad, Nauka, pp. 263-280. English translation: www.salicicola.com

Sorrie, B.A. and Somers, P. 1999. The vascular plants of Massachusetts: A county checklist. MA Division of Fisheries and Wildlife. Natural Heritage and Endangered Species Program. Westborough, MA

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.

---

21-31 July 2008