Characterization of Two Non-Native Invasive Bark Beetles, Scolytus Schevyrewi and Scolytus Multistriatus (Coleoptera: Curculionidae: Scolytinae) Patricia L

527 Characterization of two non-native invasive bark beetles, Scolytus schevyrewi and Scolytus multistriatus (Coleoptera: Curculionidae: Scolytinae) Patricia L. Johnson 1 United States Department of Agriculture, Forest Service, Wallowa Valley Office, Box A, 88401 Highway 82, Enterprise, Oregon 97828, United States of America Jane L. Hayes United States Department of Agriculture, Forest Service, Pacific Northwest Forestry and Range Sciences Laboratory, 1401 Gekeler Lane, La Grande, Oregon 97850, United States of America John Rinehart Eastern Oregon University, One University Boulevard, La Grande, Oregon 97850, United States of America Walter S. Sheppard Department of Entomology, FSHN Building 166, Washington State University, Pullman, Washington 99164-6382, United States of America Steven E.· Smith School of Natural Resources, University of Arizona, 325 Biological Sciences East, 1311 East 4th Street, Tucson, Arizona 85721, United States of America

Abstract-Scolytus schevyrewi Semenov, the banded elm bark beetle, and S. multistrlatus Marsham, the smaller European elm bark beetle, are morphologically similar. Reliance on adult external morphological characters for identification can be problematic because of wide within- species variability and the need for good-quality specimens. The inability to identify developmental stages can also hamper early-detection programs. Using two character identification systems, genitalic (aedeagus) morphology, and DNA markers (random amplified polymorphic DNA polymerase chain reaction (RAPD-PCR)) to distinguish S. schevyrewi from S. multistriatus, we examined specimens from geographically distinct populations of both species collected from infested host trees or semiochemical-baited funnel traps. We found that aedeagus morphology can be used to identify the two species, The use of two oligonucleotide primers in the RAPD-PCR analysis yielded distinct DNA banding patterns for the two species. Species identification using RAPD-PCR analysis was validated by a blind test and used to make species identitications of larval specimens. These tools improve the ability to differentiate between S. schevyrewi and S. multistriatus at immature and adult stages, and could be developed and used for other scolytines as well.

Resume-Scolytus schevyrewi Semenov, le scolyte asiatique de l'orme, et S. multistriatus Mars- ham, Ie petit scolyte europeen de l'orme, se ressemblent morphologiquement. L'utilisation des seuls caracteres externes des adultes pourI'Identification pose un probleme, car il y a un fort de- gre de variabilite au sein de chacune des especes et il est necessaire d' obtenir des specimens de bonne qualite. L'impossibilite d'identifier les stades immatures peut aussi nuire aux programmes de detection hative. En utilisant deux systemes de caracteres diagnostiques, la morphologie geni- tale (edeage) et des marqueurs ADN (ADN polymorphe amplifie aleatoirement en chaine par polymerase (RAPD-PCR)), pour distinguer S. schevyrewi et S. multistriatus, nous avons examine des specimens proven ant de populations geographiquemenr distinctes des deux especes et recol- tes sur des arbres hotes infestes ou dans des pieges appates de produits semiochimiques. Nous

Received 13 May 2007. Accepted 6 May 2008.

Corresponding author (e-mail: [email protected]).

trouvons que la morphologie de I'edcage pennet d'identifier les deux especes. L'utilisation de deux sondes doligonucleotides dans I'analyse RAPD-PCR produit des patrons de bandes ADN distincts chez les deux especes. Nous avons valide I'identification par l'analyse RAPD-PCR dans un test aveugle; nous I'avons aussi appliquee a I'identification specifique de larves. Ces outils permettent de mieux differencier S. schevyrewi de S. multistriatus aux stades immatures et adulte; il serait possible d'en mettre au point et d'en utiliser de semblables pour d'autres scolytines.

[Traduit par la Redaction]

Introduction in North America in 2003 even though it had been collected from Denver, Colorado in 1994 Scolytus schevyrewi Semenov (Coleoptera: (Negron et al. 2005) and again from New Mex- Curculionidae: Scolytinae), the banded elm ico in 1997 (LaBonte et al. 2003). By 2004, es- bark beetle, recently introduced into the United tablished populations of S. schevyrewi were States of America, is morphologically similar to found in 21 states across the United States of another non-native invasive species that is well America (Negron et al. 2005). Harris (2004) established in North America, Scolytus found that up to 96% of S.· schevyrewi collected multistriatus (Marsham), the smaller European from American elm bolts were infected with elm bark beetle (Fig. 1). The two species utilize Dutch elm disease. Scolytus schevyrewi is a po- the same preferred hosts (species of Ulmus L. tential vector of Dutch elm disease (Seybold and (Ulmaceae)) and have been found in the same Lee 2004). The similar morphology and preferred host tree (Negron et al. 2005; P.L. Johnson, un- hosts of S. multistriatus and S. schevyrewi are the published data). Scolytus schevyrewi has been most likely reasons why S. schevyrewi was not known to attack other hosts such as species of identified until several years after it became es- Salix L. and Populus L. (Salicaceae), Prunus L. tablished in the United States of America (Liu (Rosaceae), Elaeagnus L. (Elaeagnaceae), Alnus and Haack 2003). Mill. (Betulaceae), and Quercus L. (Fagaceae) The negative impacts of non-native invasive (Wang 1992; Allen and Humble 2002; North scolytines such as S. multistriatus and American Plant Protection Organization (July S. scolytus F. (Weber 1990; Haack and Cavey 2003)). Scolytus multistriatus has attacked or- 2000; Allen and Humble 2002) and the estab- namental cypresses (Cupressus L. (Cupres- lishment of S. schevyrewi years before its detec- saceae)) in Australia (Neuman 1987). The tion underscore the need to develop additional known morphological differences between tools to distinguish between these species in. S. schevyrewi and S. multistriatus include the particular and other morphologically similar general size, shape, and color of the adults, and non-native scolytines at all life stages (Haack the size, shape, and location of the apical spine 2006) in general. The purpose of this study was on the underside of the abdomen (LaBonte et al. to examine alternative techniques and develop 2003). tools to characterize S. schevyrewi that could be Native to western Europe, the Middle East, used to distinguish it more easily from and northern Africa (Bellows et al. 1998), S. multistriatus and other scolytines, particu- S. multistriatus was first recorded in North larly at early life stages and in different condi- America in Boston, Massachusetts, in 1909 tions. The study included morphometric (Chapman 1910) and subsequently spread west quantification of the male aedeagus (external across the United States of America and into genitalia) and examination of diagnostic molec- southern Canada (Baker 1972). Scolytus ular markers using random amplified polymor- multistriatus is a primary vector of the Dutch phic DNA polymerase chain reaction (RAPD- elm disease pathogen Ophiostoma novo-ulmi PCR). Male genitalic morphology has been Brasier in the United States of America and used qualitatively to identify bark beetle species Canada (Baker 1972; Furniss and Carolin (e.g., Sharp and Muir 1912; Cerezke 1964; Vite 1977). The native range of S. schevyrewi in- et al. 1974). Molecular genetic markers identi- cludes Asian Russia, Mongolia, Turkmenistan, fied through RAPD-PCR have also been used to Uzbekistan, Tajikistan, Kazakhstan, southern differentiate variation within and between spe- Kyrgyzstan, Korea, and most of China (Wang cies, including non-native, invasive insects. For 1992). Scolytus schevyrewi was first identified example, the genetic relatedness of populations

and Utah, and S. multistriatus from Kansas, Maryland, and Oregon. Specimens of each species were reared from American elm (Ulmus americana (L.) (Ulmaceae)) limbs or collected from semiochemical-baited Lindgren funnel traps. Identification of all specimens was confirmed by experts using current external morphological criteria (La Bonte et al. 2003). Samples used for morphometries were from all five states, with five populations (n = 10 per population) for each species. Samples of. S. schevyrewi were reared from elm limbs obtained from Lakewood (2004) and Golden (2005), Colorado (CO1 (39°42'N, 105°06'W) and C02 (39°78'N, 105°22'W)), and La Grande (2005), Oregon (OR3 (45°20'N, 118°06'W)), and trap-collected from Salt Lake City (2005) and Tremonton (2006), Utah (UT1 (40o45'N, 111o52'W) and UT2 (41°73'N, 112° 19'W)). Samples of S. multistriatus populations were obtained from trap collections in 2005 and 2006 from Crawford County, Kansas (KSI and KS2 (37°30'N, 94°51'W)); Prince George County (2006), Maryland (MD (38°84'N, 76°83'W)); and La Grande, Oregon (OR1), plus a popula- of the non-native bark beetle Tomicus piniperda tion reared from elm limbs from La Grande (L.) (Curculionidae: Scolytinae) was traced us- (2005), Oregon (OR3). Scolytus schevyrewi ing RAPD-PCR by Carter et al. (1996), who used for analysis of molecular genetics were determined that United States populations re- collected from Lakewood, Colorado (COl), La sulted from multiple introductions. Similarly, Grande, Oregon (OR3), and Salt Lake City, differences between Asian and European geno- Utah (UTI). Samples of S. multistriatus for types of gypsy moth, Lymantria dispar (L.) analysis of molecular genetics were collected (Lepidoptera: Lyrnantriidae), were character- from Crawford County, Kansas (KS2), La ized using RAPD-PCR (Garner and Slavecek Grande, Oregon (ORl), and Elgin, Oregon 1996; Schreiber et al .l997). Hidayat et al. (OR2 (45°34'N, 117°58'W)). Specimens from (1996) showed that two sibling species of grain all populations except those obtained from weevils, Sitophilus oryzae (L.) and S. zeamais Maryland and Kansas in 2005 (all specimens of (Motschulsky) (Curculionidae: Dryophthorinae), which were used in the experimental analysis) could be distinguished not only by variations in are presently stored at the Forestry and Range their genitalic morphology but also by RAPD- Sciences Laboratory, USDA Forest Service Pa- PCR genetic markers. Cane et al. (1990) and cific Northwest Research Station, La Grande, Stauffer and Zuber (1998) used external Oregon; upon completion of this work voucher morphometries and molecular genetics to dif- specimens will be submitted to Oregon State ferentiate between species of Ips De Geer Arthropod Collection at Oregon State Univer- (Curculionidae: Scolytinae). We expected to sity in Corvallis. . find, therefore, that S. schevyrewi and S. multistriatus each exhibit unique genitalic Genitalic morphology and molecular characters. Extraction and measurement of the aedeagus Materials and methods The aedeagus, the intromittent organ of the male insect, is a sclerotized tube at the distal Sample collections part of the phallus (Torre-Bueno 1962). Two to Samples for this study were collected from five aedeagi of each species were dissected using five states across the continental United States of a technique modified from Sharp and Muir (1912) America: S. schevyrewi from Colorado, Oregon, and examined for variation in morphology. A

comparison of shape and size was made to the opening at the narrowest part of the aedeagus identify characteristics that differed between (Fig. 1). For both species, AW2 and AW3 were species. A number of measurements were made recorded at 38% and 69%, respectively, of the to quantify observed differences and establish a distance between the tip of the apex and the protocol for this diagnostic technique. Then, for basal edge of the membrane between the lobes each species, 50 aedeagus samples from popu- of the aedeagus (Fig. 2), Because the lobes var- lations from three states across the United States ied in length, they were not included when of America were analyzed (S. schevyrewi from measuring AW2 and AW3 for either species. COl, C02, OR1, UT1, and UT2; S. multistri- For both species, AW4 was recorded at the atus from KS1, KS2, MD, OR1, and OR3). basal edge of the membrane between the lobes Five measurements were taken: aedeagus length (Fig. 2). Seventy percent ethanol was used to (AL) and four width measurements at specific preserve and (or) moisten the aedeagi as they locations on the aedeagus (AWl, AW2, AW3, were being measured. All genitalic-morphology and AW4) (Fig. 2). AL was measured from the measurements were obtained at a magnification tip of the apex of the aedeagus to the base of its of 80x using a binocular dissecting scope and lobes. AWl was recorded at the apical end below ocular micrometer.

Data analysis ground. Specimens collected in propylene gly- col and then stored in ethanol were washed Aedeagus lengths and widths were not nor- three times with l x TE buffer before continuing mally distributed and could not be normalized with the above protocol. Specimens collected using transformation, therefore we chose to use and stored in isopropanol were placed in gradu- nonparametric procedures in all analyses of ated concentrations of ethanol until 95% was these data. We determined the statistical signifi- reached, then washed with insecticidal soap for cance of differences in mean length and widths 5 min or less, rinsed in sterile distilled water, between the species by means of Wilcoxon's and processed as described above. Dry beetles two-sample test using PROC NPAR WAY in 1 were ground with a mortar and pestle, then SAS/STAT software (SAS Institute Inc. 2002). washed with 150 uL of NLS and the solution We used a nonparametric nearest neighbor was transferred to a 1.5 mL microcentrifuge discriminant analysis procedure to evaluate the tube with 150 uL of NLS, using the same pi- usefulness of the aedeagus measurements in iden- pette tip. This process was repeated with tifying unclassified insects among the taxa stud- 100 uL of NLS and then 200 uL of NLS was ied. The discriminant model was constructed added to bring the total to 600 uL. The samples using all five aedeagus traits with equal weighting were then processed as described above. Bee- in all cases and PROC DISCRIM (METHOD = tles stored in 95% ethanol were dried for 24 h NPAR, k = 5; SAS Institute Inc. 2002). We eval- before being processed as described above. uated model accuracy using cross-validation. The model was initially constructed and ana- RAPD-PCR lyzed using data from all 100 samples. Then, to validate its use for identification, the model was The PCR reaction for RAPD-PCR had a final reconstructed sequentially with data that ex- volume of 50 uL with 25 uL of Taq PCR reac- cluded 1 of the 10 populations from the analy- tion mix containing 1.5 mmol MgCl2 (Sigma, sis, i.e., each new model was based on the St. Louis, Missouri), 2 uL, of DNA (16 ng/uL), remaining 90 insects. Using this model, the 10 2 uL of primer (20-25 ng/uL), and 21 uL of individuals from the excluded population were distilled water. The PCR was carried out in a then identified to species and the results were PTC-100 Programmable Thermal Controller compared with their previously determined (MJ Research, Inc., Waltham, Massachusetts) identification. This was done for all 10 popula- with the following program: initial denaturing step at 95°C for 5 min, then 45 cycles of the tions. For all significance tests, P = 0.05. following steps - denaturing at 95 °C for Molecular genetics 1 min, annealing 34°C for 1 min, and exten- sion at 72 °C for 2 min, followed by a final ex- DNA extraction tension step at 72 °C for 7 min. A final holding Total genomic DNA was extracted using the temperature of 4 °C was maintained until the Wizard Genomic DNA Kit (Promega, Madison, PCR reaction product was visualized on a 1% Wisconsin) and a protocol optimized for bark agarose gel stained with ethidium bromide. Six- beetles (J.L. Hayes and J. Rinehart, unpublished. teen oligonucleotides (Synthegen, LLC, Hous- data). The extraction protocol was modified for ton, Texas) were screened to identify those each of the different types of collection and giving consistent results (Table 1). These 16 storage methods. With the exception of Utah, primers were chosen because they had been where only 10 adult specimens of S. schevyrewi used in other bark beetle studies with reliable were available, a sample comprising 20 adult results (Carter et al. 1996; Ruiz 2005). Five of individuals was taken from each population the 16 RAPD-PCR primers (OPB-0l, (S. schevyrewi from COl and OR3; S. multi- OPAM-07, OPAM-11, OPM-0I, and OPT-05) striatus from KS2, ORl, and OR2). A sample yielded clear banding patterns and were then comprising 20 larvae and 8 adults of a species used for comparing DNA banding patterns spe- of Scolytus Geoffroy (not identifiable to species cific to S. schevyrewi and S. multistriatus. Of using current taxonomic criteria) reared at the these last five RAPD-PCR primers, OPB-l and Forestry and Range Sciences Laboratory was OPAM-07 yielded distinct banding patterns for also analyzed. Specimens that had been stored each species and were used for comparing spe- in an ultra-cold freezer (-81°C) were placed in cies. a 1.5 mL microcentrifuge tube with 150 uL of DNA bands were scored by placing the gel Promega Nuclei Lysis Solution (NLS) and then photograph under a magnifying light and using a

despite some variation within species, consistent differences between species could be detected. The aedeagus of S. schevyrewi is parallel-sided, narrowing toward the apically subtruncated tip, whereas that of S. multistriatus is parallel-sided, narrowing and constricted before the tip, which straight edge to score bands. The data were anais somewhat spatulate apically and expanded at lyzed using the absence (0) or presence (1) of the basal end (Fig. 2). Differentiation was estab- each band scored, with analysis of molecular lished by quantifying the characteristics of the variance (AMOVA) in GenAlEx 6 for species aedeagi of both species. differentiation. (Peakall and Smouse 2005). Our discriminant model based on aedeagus length and width was able to correctly classify RAPD-PCR validation and application 99 of the 100 insects to the appropriate species A blind test was performed to validate species . in an initial analysis where all insects were in- identified by RAPD-PCR with the primers OPB- cluded. In 10 sequential analysis runs where in- 01 and OPAM-07. Five specimens of each species sects of one population were excluded from the were randomly placed in microcentrifuge tubes model, again only one insect from one popula- by an assistant before they were extracted and an- tion was misclassified (Table 2). alyzed. We then attempted to identify these specimens based on the results obtained as described Molecular genetics above. In addition, the RAPD-PCR technique using the primers OPB-0l and OPAM-07 was ap- RAPD markers plied to DNA from unidentified larvae (n = 20) Two primers (OPB-01 and OPAM-07) and specimens that could not be identified to spe- showed differences in DNA banding pattern be- cies using existing taxonomic criteria (n = 8). The tween the species. Markers resulting from latter specimens were collected from elm limbs RAPD-PCR for primers OPB-0l and OPAM-07 containing thousands of emerging adults, approxi- are in a representative gel shown in Figure 3. mately 5% of which were S. schevyrewi. We sub- The banding pattern for each species was dis- sequently attempted to identify each larva and tinct for both primers. OPB-01 yielded a total each unidentified adult specimen to the correct of 15 bands ranging from 400 to 2000 base species based on the banding patterns obtained pairs (bp) over all populations of S. schevyrewi, from known specimens. Whereas S. multistriatus had a total of two bands at 900 and 1000 bp over all populations. Results OPAM-07 produced eight bands for each species, ranging from 400 to 1100 bp for S. schev- Genitalic morphology yrewi and from 500 to 2000 bp for S. multistriatus. The results from these two primers were used Aedeagus morphometrics for the AMOVA to apportion the genetic vari- The aedeagi of S. schevyrewi and S. multi- ance between species and populations, and indi- striatus were generally similar in appearance and, viduals within species and populations (Table 3).

The AMOVA showed that although small, the those of S. schevyrewi and S. multistrlatus genetic variance (7%) between species for (Fig. 3). When known S. schevyrewi and S. schevyrewi and S. multistriatus was signifi- S. muitistriatus samples were compared with cant (P < 0.001). the eight unidentifiable samples by AMOVA, the variance was 7% for S. multistriatus but no Validation and application of RAPD markers difference was found for S. schevyrewi (Ta- ble 3). Individual banding patterns from each of The banding patterns for specimens from the the eight unidentifiable samples were then com- blind validation test were consistent with the pared with the bands of the parental populations previous results obtained for both S. schevyrewi of S. schevyrewi and S. multistriatus from Ore- and S. multistriatus (Fig. 3) and we were able gon. Comparison of the banding patterns of to correctly identify the species of each speci- these eight unidentifiable samples with those of men. We applied the results of the RAPD-PCR confirmed S. schevyrewi and S. multistriatus re- to distinguish S. schevyrewi from S. multi- vealed that the unknown specimens yielded striatus in samples derived from 19 of 20 indi- markers (six bands) shared by the two species, vidual larvae; one sample failed to yield DNA. as well as eight species-specific markers: two Of these samples, 18 showed banding patterns found only in S. multistriatus samples and six consistent with S. multistriatus and 1 showed a found only in S. schevyrewi samples. In addi- pattern consistent with S. schevyrewi (Fig. 4). tion, six bands were found inconsistently Samples from the eight specimens not identifi- among the eight unidentifiable samples that had able to species using current taxonomic criteria not been observed in either S. schevyrewi or yielded banding patterns that appeared to combine S. multistriatus (Table 4).

Discussion Aedeagus morphometries As expected, the aedeagi of these two species Our results show that features of the are generally similar in appearance, but differ aedeagus and RAPD-PCR markers can be reli- sufficiently in detailed features to allow species ably used to differentiate between S. schevyrewi differentiation. Because verbal descriptions can and S. multistriatus, We address each of these be difficult to apply readily to a given specimen techniques and findings below. without reference to multiple specimens, and

(Swaine), which is similar in external morphology and keys out in the same couplet as S. multistriatus in Wood (1982; Scolytus couplet 7(6). The use of om technique is limited to the male adult stage, but can be used for specimens that have been stored dry or in alcohol for years because the aedeagus is sclerotized and does not disintegrate as soft tissue does. A comparative study of other native and non- native Scolytus species using our quantification technique could be useful for identifying non- native, invasive species of concern, such as S. scalytus.

RAPD markers RAPD-PCR showed a significant difference (P < 0.001) in banding patterns between the there is likely to be considerable variation within S. schevyrewi and S. multistriatus specimens species, we attempted to develop a means of with 2 of the 16 primers screened. The use of readily quantifying differences in features, Al- more than one primer to distinguish species is though our sample was' relatively small, the important to verify the identity of individual specimens. aedeagi of S. schevyrewi and S. multistriatus proved to be unique to each species across each In addition to the blind validation test, the geographically distinct population used in this use of both primers on larval specimens not study. Scolytus schevyrewi and S. multistriatus only confirmed the species identification of showed some variation in the size of the aedeagus each specimen but also demonstrated that this within populations. This may be in proportion technique could be used on immature and adult to the overall variation in specimen size, a fac- . stages. tor for which we did not account but that On the other hand, the use of both primers on should be included in future work. In studies of the. unidentifiable adult specimens yielded pat- species of Carabidae, Noctuidae, and terns containing both unique bands and bands Drosophila Fallen (Diptera: Drosophilidae) shared by S. schevyrewi and S. multistriatus. (e.g., Garnier et al. 2005; Mutanen 2005; Soto The AMOVA showed no variance between the 2005), the shape of the male genitalia was unidentifiable specimens and S. schevyrewi, but quantified by outlining the male genitalia. In there was significant variance between these studies by Cane et al. (1990) and Stauffer and and S. multistriatus. These results suggest that Zuber (1998), external morphology was quanti- the unidentifiable specimens were S. schevyrewi fied and molecular genetics used to differentiate and not S. multistriatus. However, the fact that various species of Ips. To our knowledge the band-by-band analysis showed that these speci- genitalic morphology of no other species of mens have some unique bands, share some Scolytus has been described and (or) quantified bands with both S. schevyrewi and S. multi- using morphometries within this genus. We an- striatus, and share some bands with either one ticipate that our quantification technique will or the other species suggests that the unidentifi- allow not only S. schevyrewi to be distin- able specimens could be hybrids or a distinct guished from S. tnultistriatus, but also either but currently unknown species. species from other Scolytus species, although Data were obtained using a variety of differ- this remains to be tested. We were unable to ent specimen-collection and -storage methods obtain specimens from the native ranges of and from specimens in varying condition. How- these species for comparison; however, this ever, the particular method used to preserve would be the logical next step in the validation specimens can cause difficulties for DNA ex- of this technique. If consistent variability were traction. Specimens preserved in ethanol pre- to be detected, this might provide insight into sented the most difficulty for DNA extraction, the origins of Nearctic and other populations. It whereas those frozen immediately after collec- would be particularly interesting to compare the tion yielded the best results and required the conifer-feeding spruce engraver, Scolytus piceae simplest extraction procedures. Dry samples that

© 2008 Entomological Society of Canada 536 .Can. Entomol. Vol. 140, 2008 were ground before NLS was added produced S. zeamais, were separable not only by minor multiple bands, whereas grinding with NLS variations in their reproductive morphology but produced only a single band when the OPT-Os also by RAPD-PCR genetic markers. Because primer was used. The OPT-05 primer produced S. schevyrewi and S. multistriatus originate multiple bands with both S. multistriatus and from geographically different areas (Wang S. schevyrewi in DNA extracted from frozen 1992; Solomon 1995) and thus are subject to beetles. Carvalho and Vieira (2000) recom- different selection pressures, we expected to mended that DNA extraction be done within find, and did find, species-specific molecular 1 year of specimen collection to avoid degrada- genetic markers. The different histories of these tion of DNA. two species in North America may have con- tributed to the differences that we detected, but our samples from multiple sites across the Conclusion United States of America provided reliably de- tectable species-specific markers. For our pur- Scolytus schevyrewi and S<, multistriatus are examples of scaly tine species that are difficult poses, RAPD-PCR is a particularly useful tool to distinguish by traditional taxonomic means, for species differentiation because of its relative especially at immature stages, for which no tax- simplicity, availability, and low cost. onomic references appear to exist. Taken to- Given their differences in origin, and that gether, our results provide two additional S. schevyrewi and S. multistriatus can live and diagnostic tools for identifying the two non- reproduce in the same host tree, we hypo the- native species targeted in this study. The use of sized that quantifiable differences in reproduc- more than one tool, whether external or internal tive morphology exist between these two morphology or DNA markers, allows greater species and that these differences help to pre- confidence that scaly tine specimens will be ac- vent interbreeding. The need for a more thor- curately identified when collected. However, ough examination of the reproductive the use of more than one diagnostic tool may morphology, not just the aedeagus, of these two not be possible, given the life stage or condition species is indicated by our findings. The dis- of the specimens, and only one technique may covery of possible hybridization suggests that be employed when there is a need for quick the differences between the two species may identification. These tools have helped to show not strictly prohibit successful reproduction. that S. schevyrewi and S. multistriatus not only The existence of a third unique species is possi- could occur in the same tree/limb but may also ble, but is less likely than hybridization, be- hybridize. The fact that RAPD-PCR of the eight cause no other evidence of another species has unidentifiable adult specimens produced results been reported. Moreover, other scolytines that similar to the analysis, using the same methods, are known to infest elm do not resemble either of identified specimens from other locations S. multistriatus or S. schevyrewi. suggests that the RAPD-PCR methodology we With the increasing potential for international employed is reliable and repeatable. The mark- transportation of known and unknown insect ers did not differ substantially in band fre- species and the necessity for detection and quency or band size. If these eight specimens eradication or other management measures, the represent a different species, we would have ex- ability to distinguish both native and non-native pected to see more unique specimens among insects is essential. We have described two the thousands that were reared from elm limbs tools that, in combination with classic taxo- in the Forestry and Range Sciences Laboratory. nomic procedures, may enhance our ability to Though beyond the scope of this study, the identify immature and adult specimens of two ecological and morphological similarities be- non-native bark beetles, S. schevyrewi and tween S. schevyrewi and S. multistriatus raise S. multistriatus. Identification via genitalic mor- questions concerning the degree of relatedness phology is limited to adult males but can be of these two species. Not only do they utilize used on intact specimens regardless of time the same host plant species, their mating biol- spent in storage or method of preservation. ogy is similar (Wang 1992; Solomon 1995) and, RAPD-PCR can be used on specimens in any as we have shown, the features of one compo- life stage or condition and that have been pre- nent of their reproductive morphology are simi- served in a variety of ways. This set of tools lar. Hidayat et al. (1996) showed that two sibling has the potential for use with a wide array of species of grain weevils, Sitophilus oryzae and scaly tines , both native and non-native, to assist

© 2008 Entomological Society of Canada Johnson et al. 537 in detection and eradication efforts aimed at Chapman, J.W. 1910. The introduction of a Euro- limiting the introduction or spread of invasive pean scolytine (the smaller elm bark-beetle, species. Scolytus multistriatus (Marsh.) into Massachu- setts. Psyche (Cambridge), 17: 63-68. Furniss,R.L., and Carol in, Y.M. 1977. Western for- Acknowledgments est insects. UnIted States Department Agriculture Forest Service Miscellaneous Publication Support for this work was provided by the No. 1339. USDA Forest Service Pacific Northwest Re- Garner, K.J., and Slavecek, J.M. 1996. Identification search Station and Pacific Northwest Region, and characterization of a RAPD-PCR marker for and Washington State University. We thank distinguishing Asian and North American gypsy James LaBonte, Robert Rabagalia, Glen moths. Insect Molecular Biology, 5: 81-91. Salsbury, Jeffery Witcosky, Jose Negron, and Garnier, S., Magniez-Jannis, F., Rasplus, J.-Y., and Clint Burfitt for providing and identifying sam- Alibert, P. 2005. When morphometry meets genet- ples, and Micheal Minthorn and Larry Bare for ics: inferring the phylogeography of Carabus laboratory assistance. Haruyo Matsuyama as- . solieri using analyses of pronotum and male geni- sisted with PCRexperiments and band analysis. talia. Journal of Evolutionary Biology, 18: 269- Carol Anelli, William Turner, Gary Piper, Lia 280. doi: 10.1111/j.1420-9101.2004.00854.x. Spiegel, arid Joshua Johnson provided com- Haack, R.A. 2006. Exotic bark- and wood-boring Coleoptera in the United States: recent establish- ments on earlier drafts of this paper. ments and interceptions. Canadian Journal of For- est Research, 36: 269-288. References Haack, R.A., and Cavey, J.F. 2000. Insects inter- cepted on solid wood packing materials at United Allen, E.A., and Humble, L.M. 2002. Nonindigenous States ports-of-entry: 1985-1998. In Quarantine species introductions: a threat to Canada's forests pest risk for the forestry sector and their effects and forest economy. Canadian Journal of Plant on foreign trade: Proceedings of Silvotecuna 14, Pathology, 24: 103-110. 27-28 June 2000, Concepcion, Chile [CD-ROM]. Baker,W.L. 1972. Eastern forest insects. United La Corporacion Chilena de Ja Madera (CORMA), States Department Agriculture Forest Service Concepcion, Chile. pp. 1-16. Miscellaneous Publication No. 1175. Harris, J.L. 2004. Forest insect and disease conditions Bellows, T.S., Meisenbacher, c., and Reardon,R.C. in the Rocky Mountain Region. USDA Forest 1998. European elm bark beetle biological con- Service, Rocky Mountain Region, Renewable trol. In Biological control of arthropod forest Resources, R2-05-09. Available from http:// pests of the western United States: a review and www.fs.fed.us/r2/fhm/reports/r2cond_r2-05-09.pdf recommendations. USDA Forest Service, FHTET- [accessed 15 May 2006]. 96-21,. The University of Georgia, and Southern Hidayat, P., Phillips, T.W., and French-Constant, Forest Insect Work Conference. Available from R.H. 1996. Molecular and morphological charac- http://www.barkbeetles.org/Biocontol/europeanel ters discriminate Sitophilus oryzae and S. zeamais mbarkbeetle.html [accessed 15 September 2006]. (Coleoptera: Curculionidae) and confirm repro- Cane, J.H., Stock, M.W., Wood, D.L., and Gast, S.l. ductive isolation. Annals of the Entomological 1990. Phylogenetic relationships of Ips bark bee- Society of America, 89: 645-652. tles (Coleoptera: Scolytidae): electrophoretic and LaBonte, J.R., Rabaglia, . and Hoebeke, E.R. morphometric analyses of the grandlcollis group. R J. Biochemical Systematics and Ecology, 18: 359- 2003. A screening aid for the identification of the 368. banded elm bark beetle, Scolytus schevyrewi Carter, C.M., Robertson, J.L., Haack, R.A., Law- Semenov. Available from http://ceris.purdue.edu/ rence, R.K., and Hayes, J.L. 1996. Genetic relat- napis/pests/barkb/schevy/schevyrewiIDnewlA.pdf edness of North American populations of Tomicus [accessed 19 October 2003]. piniperda (Coleoptera: Scolytidae). Journal of Liu, H.-P., and Haack, R.A. 2003. Scolytus Economic Entomology, 89: 1345-1353. schevyrewi: pest report from the exotic forest pest Carvalho, A.O.R., and Vieira, L.G.E. 2000. Compar- information system for North America. Available ison of preservation methods of Atta spp. from http://spfnic.fs.fed.us/exfor/data/pestreports. (Hymenoptera: Formicidae) for RAPD analysis. cfm?pestidval = 163&langdisplay = english [ac- Anais cia Sociedade Entomologica do Brasil, 29: cessed 19 July 2004]. 489-496. Mutanen, M. 2005. Delimitation difficulties in spe- Cerezke, H.P. 1964. The morphology and functions cies splits: a morphometries case study on the of the reproductive systems of Dendroctonus Euxoa tritici complex (Lepidoptera, Noctuidae). monticule Hopk. (Coleoptera: Scolytidae). The Systematic Entomology, 30: 632-643. doi: Canadian Entomologist, 96: 477-500. 10.1111/j.1365-3113.2005.00296.x.

Negron, J., Witcosky, J.J., Cain, R.J., LaBonte, J.R., 20Meeting/1010%20SeyboldCAPest3-Scolytus.pdf Duerr, D.A., II, McElwey, S.J., Lee, J.C., and [accessed 15 May 2006]. Seybold, S.J. 2005. The banded elm bark beetle: a Sharp, D., and Muir, E 1912. The comparative anat- new threat to elms in North America. American omy of the male genital tube in Coleoptera. Entomologist, 51: 84-94. Transactions of the Entomological Society of Neuman, EG. 1987. Introduced bark beetles on ex- London. Part III. pp. 477-642: otic trees in Australia with special reference to in- Solomon, J.D. 1995. Guide to insect borers in North festations of Ips granicollis in pine plantations. American broadleaf trees and shrubs. Agriculture Australia Forestry, 50: 166-178. Handbook No. 706, USDA Forest Service, Wash- North American Plant Protection Organization .. ington, D.C. pp. 509-514. 2003. Detection of Scolytus schevyrewi Semenov Soto, I. 2005. Use of elliptic Fourier descriptors for in Colorado and Utah: official pest reports for quantification of male genitalia morphology of United States, phytosanitary alert. Available from cactophilic Drosophila. Drosophila Information http://www.ceris.purdue.edu/napis/pests/barkb/sch Service, 88: 42--45. evy/030715-nappo.txt [accessed 9 October 2004]. Stauffer, C., and Zuber, M. 1998. Ips amitinus var. Peakall, R., and Smouse, P.E. 2005. GenAlEx 6: ge- montana (Coleoptera, Scolytidae) is synonymous netic analysis in Excel - population genetic soft- to Ips amitinus: a morphological, behavioral and ware for teaching and research [computer genetic re-examintion. Biochemical Systematics program]. Molecular Ecological Notes, 6: 288- and Ecology, 26: 171-183. 295. Torre-Bueno, J.R. de la. 1962. A glossary of entomol- Ruiz, E.A. 2005. Estructura genetica poblacional de ogy, with Supplement A by G.S. Tulloch. Brooklyn Dendroctonus pseudotsugae Hopkins (Coleoptera: Entomological Society, Brooklyn, New York. Curculionidae: Scolytinae) : mediante marcadores Vite, J.P., Islas, S., Renwick, J.A.A., Hughes, P.R., moleculares. M.Sc. thesis, Instituto Politecnico and Kliefoth, R.A. 1974. Biochemical and biolog- Nacional, Mexico D.E, Mexico. ical variation of southern pine beetle populations SAS Institute Inc. 2002. SAS/STAT® software. Ver- in North and Central America. Zeitscrift fur sion 9.1. SAS System for Windows®. SAS Insti- Angewandte Entomologie, 75: 422-435. tute Inc., Cary, North Carolina. Wang, Z. 1992. Scolytus schevyrewi Semenov. In Schreiber, D.E., Garner, K.J., and Slavicek, J.M. 1997. Forest insects of China. 2nd ed. Edited by Identification of three randomly amplified polymor- G. Xiao. China Forestry Publication House, phic DNA-polymerase chain reaction markers for Beijing, People's Republic of China. pp. 633-634. distinguishing Asian and North American gypsy Weber, J.W. 1990. Relative effectiveness of Scolytus moths (Lepidoptera: Lymantriidae). Annals of the scolytus, S. multistriatus and S. kirschi as vectors Entomological Society of America, 90: 667-674. of Dutch elm disease. European Journal of Forest Seybold, S., and Lee, J. 2004. The banded elm bark Pathology, 20: 184-192. beetle: update on an invasive pest of elms. Wood, S.L. 1982. The bark and ambrosia beetles of USDA Forest Service, Pacific Southwest Research North and Central America (Coleoptera: Station, Davis, California. Available from Scolytidae), a taxonomic monograph. Great Basin http://www.caforestpestcouncil.org/2004%20CFPC% Naturalist Memoirs No.6.