The Banded Elm Bark Beetle, Scolytus schevyrewi Semenov (Coleoptera, Curculionidae, Scolytinae) in North America: a taxonomic review and modifications to the Wood (1982) key to the species of Scolytus Geoffroy in North and Central America

Abstract In 2003, an Asian bark beetle, Scolytus schevyrewi Semenov (Coleoptera: Curculionidae: Scolytinae), the banded elm bark beetle, was detected for the first time in North America. This paper modifies the Wood (1982) key to the species of Scolytus Geoffroy to enable identification of Scolytus schevyrewi in North and Central America. Variation of diagnostic characters in Scolytus schevyrewi is discussed.


Introduction
A growing number of exotic wood boring or wood associated beetles have recently been found to be established in North America (e.g., Hoebeke 1994;Hoebeke 1999;Maier and Lemmon 2000;Vandenberg et al. 2000;Mudge et al. 2001;CFIA 2002;McCullough and Roberts 2002;LaBonte et al. 2005;Haack 2006; Lee et al. 2008). In response to this trend, a multiagency pilot project to detect exotic Scolytinae throughout the United States was initiated in 2001. Th is program was initially designated the Exotic Forest Pest Early Detection and Rapid Response Program, but is now known as the Early Detection and Rapid Response program, or EDRR (Rabaglia et al. 2008). Since 2007, the EDRR program has been implemented on a national level under the auspices of the USDA Forest Service and, as of 2009, most states in the U.S. have been particpants.
As the cooperating taxonomist for the western states participating in the 2003 EDRR program, specimens from Lindgren funnel traps used in this survey were sent to me for identifi cation. Early samples from the Denver, Colorado, metropolitan area contained several specimens of a species of Scolytus Geoff roy unfamiliar to me and that I was unable to key to any species in Wood (1982). I consequently sent specimens to Dr. Stephen L. Wood (deceased), Dr. Donald E. Bright (emeritus, Colorado State University, Fort Collins, Colorado) and Dr. Natalia J. Vandenberg (U.S.D.A., Agricultural Research Service, Systematic Entomology Lab, Washington, DC). Th ese taxonomic authorities determined the specimens to be Scolytus schevyrewi Semenov, an Asian species previously unknown from North America. Shortly thereafter, I found specimens of S. schevyrewi in samples from Ogden, Utah.
Subsequent trapping found this species to be abundant and clearly established in Denver and Ogden. Th ese data stimulated extensive trapping surveys throughout Colorado and Utah, where it was found at most sites. Th e realization that S. schevyrewi was widely distributed in Colorado and Utah prompted several neighboring states to initiate surveys as well. At the end of 2003, S. schevyrewi had been detected in twelve additional states (Negron et al. 2005). By 2008, this supposedly new exotic species had been found from coast to coast in twenty-eight states (with the earliest records from 1994), as well as in southern Canada (Lee et al. 2009). As yet, there are no records of S. schevyrewi from Mexico.

Methods
Existing taxonomic treatments for North American Scolyus (Bright 1975;Wood 1982;Furniss and Johnson 2002) do not include S. schevyrewi as it was unknown from this continent when those were published. Shortly after the detection of S. schevyrewi in Colorado and Utah, an image-based aid to the identifi cation of this species was placed on the Purdue University NAPIS (National Agricultural Pest Information Site) website (LaBonte et al. 2003). Th is aid is more streamlined than the following modifi cations to the Wood (1982) key and emphasizes the diff erences between S. schevyrewi and S. multistriatus (Marsham) because the latter is much more apt to be encountered in surveys than is S. piceae (Swaine). However, this early treatment is incomplete as it does not include some diagnostic characters subsequently found nor was the range of variation of some characters recognized.
Th e diagnostic characters used to diff erentiate S. schevyrewi from other species of Scolytus are mainly based on specimens acquired via the EDRR project and a variety of wood boring insect surveillance programs, most funded via the USDA Cooperative Agricultural Pest Survey (CAPS) program. I have examined over 7,600 specimens of S. schevyrewi from these surveys.

Distinguishing Scolytus schevyrewi from other species of Scolytus in North and Central America
Th e following modifi cations to Wood's (1982) key to the species of Scolytus in North and Central America will serve to identify typical specimens of S. schevyrewi. Very little of the text in his key remains in the couplets below, other than the distributions, hosts, and size ranges for species dealt with therein. Several characters used in his key to discriminate among these species are not used because they are unnecessary or lead to an unduly cumbersome key, are too variable, or cannot be accurately assessed without reliably identifi ed reference specimens, a resource many users of this key will lack. Th ese characters can include the relative sizes of interstrial and strial punctures, whether the elytral interstriae are impressed, and subtle diff erences in size among abdominal sternal punctures.

Diagnostic summary and character variation
Typical specimens of S. schevyrewi cannot be easily confused with any other species of Scolytus known from North America. Th e shape and position of the spine on sternum 2 (Figs 7-8, 10, 17-21), especially in males, and typical coloration (Figs 24-26) are unique relative to all other North American species. However, this is a highly variable species and it is advisable to use a suite of characters for its identifi cation. Th e following elaboration on variation in S. schevyrewi is based upon the examination of thousands of specimens of that species and of S. multistriatus, the species most apt to be confused with S. schevyrewi. Th ere can be great variation in the shape and position of the spine on sternum 2. Males most often have the spine well developed, with a blunt apex that is broader than the base, appearing triangular in ventral view (Fig. 10). Especially with females, this feature can be variously reduced, becoming almost absent in the most extreme cases (Fig. 20). Reduction of the spine can lead to possible confusion with specimens of S. multistriatus that have malformed or broken spines, which are not uncommon in large series thereof. However, as indicated previously, in S. schervyrewi, the base of the spine on sternum 2 is remote from the anterior margin, almost at the center, whereas in S. multistriatus the base of the spine is in contact with the anterior margin. Some S. schevyrewi, especially those with larger spines, can have the base of the spine positioned anterior of the center of sternum 2, compounding the possibility of confusion with aberrant S. multistriatus. Inadequately cleaned S. multistriatus may have debris under the base of the spine, which can make its lateral appearance broader than is the case. An extreme example of abdominal spine variation in S. schevyrewi is exhibited in a male from California (courtesy of R.L. Penrose, California Department of Food and Agriculture) (Figs 22, 23). In this instance, a second, accessory, sharply conical spine is present on the third abdominal sternum. Th is spine is asymmetrically positioned (Fig.  23), leading to my conclusion that this specimen represents a developmental aberra- tion rather than a diff erent species. In all other respects, this specimen appears to be a typical S. schevyrewi.
Th e elytron of a normal S. schevyrewi is distinctively colored, with a variably developed median dark band and the base and apex distinctly pale (Figs 24,25). Th is character enables rapid identifi cation of this species as no other North American species has this color pattern. Th e dark median band is best observed in well dried specimens; it can be obscured in specimens still damp from collecting in liquid or storage in alcohol. However, there are occasional specimens with apparently unicolorous dark or pale elytra (Fig. 26). Some samples exhibited higher proportions of these unicolorous specimens, but probably less than 10%. Th e existence of individuals with concolorous elytra dictates caution in complete reliance upon color for identifi cation of S. schevyrewi, as both S. multistriatus and S. piceae normally have concolorous elytra. Some S. multistriatus also have elytra with dark apices and bases, with pale median areas.
Th e pronotum of a typically colored S. schevyrewi is also distinctively colored, with relatively extensive areas of pale coloration (Figs 24-26). Th e extent of the pale areas is highly variable. Th e most common pattern is a fairly broad pale band along the posterior margin with a somewhat narrower pale band along the anterior margin (Figs 24,25). Many specimens have the pale coloration extending from the posterior margin into the median area (Fig. 26). Th is pale coloration can sometimes cover almost the entire dorsum of the pronotum. On the other hand, in some specimens the pronotum is essentially completely dark, with only hints of anterior and posterior pale marginal banding. Th e extent of pale pronotal coloration appears independent of the size and extent of the median dark elytral band and the darkness of the ground color of the elytra. Several of the specimens of S. multistriatus I've examined have large, nebulously paler areas in the median area of the pronotum. As with elytral coloration, pronotal coloration should be used with some caution to distinguish S. schevyrewi from S. multistriatus and S. piceae.
As indicated in the key, specimens of S. schevyrewi average larger than either S. multistriatus or S. piceae. However, small S. schevyrewi fall within the size range of both of the other species. Furthermore, large S. multistriatus approach the size of average or even large S. schevyrewi.
Specimens of S. multistriatus have distinct, pointed tubercles or "teeth" on the posterior lateral margins of sterna 2-4 (Fig. 2). Specimens of S. piceae and S. schevyrewi lack this feature (Figs 6, 8). Th is character is occasionally obscured in specimens of S. multistriatus swollen with liquid preservatives, but it can normally be observed by examining the lateral margins of the sterna from an oblique perspective.
Th e posterior margin of sternum 2 is distinctly truncate or slightly emarginate in almost all S. schevyrewi examined (Fig. 10). On the other hand, in S. multistriatus (Fig. 3) and S. piceae (Fig. 9) the posterior margin of sternum 2 is normally slightly to pronouncedly convex and in S. multistriatus is often slightly produced at the base of the spine (Fig. 3).
Th e elytron of a S. schevyrewi specimen in good condition typically displays three rows of relatively stout, long, discal setae that are about twice as long as the width of the elytral intervals . Th e elytra of most S. multistriatus lack distinct rows of discal setae, although there are generally setae at the elytral apices and there may be discal setae (Fig. 4). If discal setae are present, they are generally scattered, are less stout than those of S. schevyrewi and are shorter, about as long as the width of the elytral intervals. Of the small series of S. piceae examined, most lacked discal setae (Fig. 14). A few specimens had a row of 3 or 4 discal setae on interval 7, but these setae were short and fi ne, similar to those of S. multistriatus.
In summary, a suite of characters is best used to reliably identify S. schevyrewi. Especiallly with regard to S. multistriatus, these include (more or less in order of reliability and ease of assessment) the shape and position of the spine on sternum 2, normal elytral and pronotal coloration, the absence of lateral teeth on sterna 2-4, average size, the truncate or slightly emarginate posterior margin of sternum 2, relatively abundant and large discal elytral setae, and short, recumbant setae on sterna 3-5.

Discussion
Th e U.S. specimens collected prior to 2003, along with the extensive U.S. distribution of S. schevyrewi and its great abundance in many areas, provide ample evidence that this exotic species is not new to the U.S. but is instead a legacy species that has been present for decades. Such legacy species, e.g., Xyleborinus alni (Niisima) (Mudge et al. 2001;LaBonte et al. 2005;Hoebeke and Rabaglia 2007), are probably more often the norm for newly detected exotic wood boring insects than otherwise. Th is is a consequence of the current surveillance technology, the limited survey eff orts of the past, and the limited taxonomic expertise available to deal with the many thousands of specimens generated by current surveys.
Th ere are profound taxonomic challenges presented by the remaining pool of undetected legacy species, truly newly introduced exotics, and the onslaught of continued new introductions as a consequence of global trade. Almost all existing taxonomic works for scolytines in North America, let alone other taxa of wood boring or wood associated insects, quite reasonably treat only those species previously known from this continent. Th e taxonomic infrastructure available to support surveillance for a wide spectrum of exotic wood borers has been eroding for decades and may have declined below critical and self sustaining levels. New technologies, such as extended depth of fi eld macroscopy and LUCID™ go far to bridge this taxonomic impediment, e.g., a recent guide to the North American Siricidae (Schiff et al. 2006). Nonetheless, exotic wood boring insects, such as S. s. schevyrewi, will continue to evade recognition and detection unless substantial funds and resources are devoted to expanding our taxonomic base.