Comparative Cold Tolerance and Climate Matching Of

Biological Control 58 (2011) 96–102

Contents lists available at ScienceDirect

Biological Control

journal homepage: www.elsevier.com/locate/ybcon

Comparative cold tolerance and climate matching of coastal and inland Laricobius nigrinus (Coleoptera: Derodontidae), a biological control agent of hemlock woolly adelgid ⇑ D.L. Mausel , R.G. Van Driesche, J.S. Elkinton

University of Massachusetts, Department of Plant, Soil and Insect Science, 102 Fernald Hall, Amherst, MA 01003, United States article info abstract

Article history: Laricobius nigrinus (Coleoptera: Derodontidae) was first collected near the coastal city of Victoria, British Received 1 June 2010 Columbia, Canada for release as a biological control agent to suppress tree-killing densities of hemlock Accepted 8 April 2011 woolly adelgid, Adelges tsugae (Hemiptera: Adelgidae), in the eastern United States. Beetles established Available online 20 April 2011 in warm areas of the invaded range of A. tsugae, but had a low probability of establishment in cold areas. With the goal of locating beetles with greater cold-tolerance, we collected hundreds of adults in the Keywords: northern Rocky Mountains. To support planned releases of these inland L. nigrinus, the cold tolerance Tsuga canadensis of field-collected coastal (Seattle, WA) and inland (Coeur d’Alene and Moscow, ID) adults and climate Adelges tsugae match index scores (CLIMEX v.2) of these collection areas and parts of the eastern United States were Laricobius nigrinus Classical biological control compared. We found that individuals of inland L. nigrinus were more cold tolerant than those of coastal Intraspecific variation L. nigrinus, based on higher survival in a winter field cage study in Massachusetts and a lower superco- Cold tolerance oling point in a laboratory assay. Inland L. nigrinus also had higher survival than coastal L. nigrinus after Climate matching an 18 h exposure to sub-zero temperatures in another laboratory assay. Areas of the eastern United States infested with A. tsugae and receiving L. nigrinus releases matched the climate of Seattle and Coeur d’Alene reasonably well, but Coeur d’Alene had higher index values over a considerably larger area than Seattle. Accordingly, inland L. nigrinus appears preferable for release in the colder portions of A. tsugae’s invaded range in the eastern United States. Published by Elsevier Inc.

1. Introduction impact on A. tsugae populations (Mausel et al., 2008). This ‘‘coastal’’ L. nigrinus, however, had a low probability of establishment in Adelges tsugae Annand (Hemiptera: Adelgidae), native to Asia colder areas, either in more northerly locations or at high eleva- and western North America, was accidentally introduced to east- tions, likely due to winter low temperatures (Mausel et al., 2010) ern North America from Japan (Havill et al., 2006) and is recog- that exceeded the cold tolerance limits of the beetle (Humble nized as the most important pest of eastern hemlock, Tsuga and Mavin, 2005). canadensis (L.) Carrière, and Carolina hemlock, Tsuga caroliniana If available, release of a L. nigrinus ‘‘biotype’’ with greater cold Engelmann (Anonymous, 2005). Laricobius nigrinus Fender (Cole- tolerance could improve both establishment probability and im- optera: Derodontidae), a predator released against A. tsugae in pact on A. tsugae because L. nigrinus is most active during the cold the eastern United States, is native to the northwestern United seasons (Zilahi-Balogh et al., 2003a,b) and release areas in the east- States and western Canada (Fender, 1945). The original releases ern United States are colder than the Puget Trough (Humble and in the eastern United States were made with beetles collected from Mavin, 2005). L. nigrinus has been recorded from the Rocky Moun- the Puget Trough area (e.g., Victoria, British Columbia and Seattle, tains (e.g., Coeur d’Alene and Moscow, Idaho in this study) of the Washington), which has a mild maritime climate. After their re- northwestern United States and Canada (Fender, 1945; Lawrence, lease, beetles from Victoria established in the mid- and southern 1989; Bright, 1991), an area with a continental climate. In this ‘‘in- Appalachian Mountains (Lamb et al., 2006; Mausel et al., 2010), land’’ area, winters are colder and snowier than in the coastal areas where they proved to be well synchronized and demonstrated an from which the original L. nigrinus beetles were collected for release. Additionally, inland summers are warmer and precipitation occurs year-round, while on the coast summers are cooler and ⇑ Corresponding author. Fax: +1 413 545 5858. drier (Humble and Mavin, 2005). The similarity of seasonal tem- E-mail address: [email protected] (D.L. Mausel). perature and precipitation patterns between collection and release

1049-9644/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.biocontrol.2011.04.004 D.L. Mausel et al. / Biological Control 58 (2011) 96–102 97 sites (i.e., climate matching) is fundamental to natural enemy D. Don, infested with Pineus similis (Gillette) (Hemiptera: Adelgi- introductions, and mismatched natural enemies can fail to estab- dae) at the University of Idaho Arboretum, Moscow, Latah Co., ID lish or spread throughout a pest’s invaded range (Hopper et al., (46.7 N, 117.0 W; elevation 805 m). Beetles were shipped over- 1993). night on excelsior in escape-proof containers to the University of Another principle of classical biological control is to collect Massachusetts, Amherst, and were placed in environmental cham- large numbers of a natural enemy over a wide variety of environ- bers at 5 °C day and 2 °C night, 75% relative humidity (RH), season- ments in its native range to capture population variation in traits ally adjusted photoperiod, where they were fed locally collected that may influence effectiveness (Hopper et al., 1993). Coastal A. tsugae ad libitum before use in experiments. and inland L. nigrinus are disjunct, being isolated by a lack of suit- Adult beetles from both Coeur d’Alene and Moscow, ID (col- able host plants for adelgids in the arid Columbia Plateau area lected from different adelgid species) were confirmed morpholog- (Burns and Honkala, 1990), which divides the Cascade Mountains ically to be L. nigrinus by both Zilahi-Balogh (Canadian Food from the Rocky Mountains. Small but consistent differences in Inspection Agency) and Vandenberg (US Department of Agricul- COI gene sequences between coastal and inland L. nigrinus (0.5% se- ture, Agriculture Research Service). Furthermore, DNA sequences quence divergence) suggest that these beetles represent distinct from the mitochondrial COI gene from Coeur d’Alene and Moscow populations but that they are the same species (Havill, Personal beetles were indistinguishable from each other but were distin- Communication). Despite a lack of extensive data on the actual guishable from coastal beetles (Havill, Personal Communication). boundaries of L. nigrinus populations in a genetic context, re- The adelgid on western white pine was identified as P. similis by stricted gene flow between coastal and inland sources implies a comparing its COI sequence with other specimens determined by potential for long-term genetic differentiation in important life- morphology (Havill, Personal Communication). history features such as host-range and cold tolerance. Laricobius spp. are considered adelgid specialists (Lawrence and Hlavac, 2.2. Experiments 1979) and host-range testing of inland L. nigrinus suggests low risk to non-target organisms (Mausel et al., unpublished data) as deter- 2.2.1. Survival in field cages mined previously for coastal L. nigrinus (Zilahi-Balogh et al., 2002). To test for differences in survival of coastal (Seattle) and inland The existence of intraspecific variation in beetle cold tolerance has (Coeur d’Alene and Moscow) L. nigrinus adults under harsh winter been demonstrated previously in the Scolytidae (Bentz and Mul- conditions, we conducted a field cage study from 20 December lins, 1999; Lombardero et al., 2000; Khåi et al., 2007). 2007 to 17 January 2008 in Leverett, MA (42.4 N, 72.5 W) using That such natural enemy variation matters is shown in a 10 year eastern hemlock at a site with 183 m elevation and an eastern as- field experiment in New Zealand, in which release of Microctonus pect with 20% slope. Ten A. tsugae-infested eastern hemlock trees, hyperodae Loan (Hymenoptera: Braconidae) sourced from east of separated by 20 m along a transect, were selected for use, with the Andes outperformed releases sourced from west of the Andes treatments arranged in a randomized complete block design in a successful biological control program against the South Amer- (n = 10). Cages were bags made of sewn polyester fabric ican weevil Listronotus bonariensis (Kuschel) (Coleoptera: Curculi- (0.25 mm mesh size), 0.6 m wide, and 0.9 m long (National Filter onidae) (Phillips et al., 2008). Knowledge of L. nigrinus variation Media Corp., Wallingford, CT). Three branches (>2 m height) with in cold tolerance could be used to design an efficient collection high adelgid densities (i.e., nearly every needle base having an and release program. Here we report results of studies on the rel- adelgid) were tagged per tree, and each caged branch was ran- ative degree of cold tolerance of coastal and inland populations of domly assigned to a beetle treatment (Seattle, Coeur d’Alene, or L. nigrinus. Adult survival during winter was compared in field Moscow). Ten adult beetles were placed inside each cage. The cage cages on hemlock branches in western Massachusetts. In the labo- was pulled over the branch, and tightly closed with a cable tie ratory, the supercooling point of adults and the survival of adults around the bag and stem. One temperature logger (iButton, Maxim after an 18 h exposure to sub-zero temperatures (i.e., 1.5 °C above Integrated Products Inc., Sunnyvale, CA) was deployed per tree in a the mean supercooling point of coastal L. nigrinus) were compared. cage and data were recorded at 1-h intervals. After 1 month, The supercooling point is the temperature at which hemolymph branches with cages were cut and returned to the laboratory and freezes and freeze-intolerant insects die. Also, exposure to long kept at À3 °C until inspection (within 3 h). Cages were opened in- periods of sub-zero temperatures that are above the supercooling side a Bug Dorm (Bioquip, Rancho Dominguez, CA) in a cool room point may cause cold stress and mortality (Block, 1990). Last, we (15 °C) to recover live and dead L. nigrinus. Despite repeated used climate-modeling software (CLIMEX v.2) to determine which searching, four Seattle (4% of total), three Coeur d’ Alene (3%), coastal and inland L. nigrinus collection areas were climatically and two Moscow (2%) adults were not recovered and these individ- most similar to eastern North America, and thus a potentially bet- uals were excluded in the calculation of percent survival. Two-way ter source for biological control agents for release against A. tsugae ANOVA was used to determine whether a significant portion of the in that part of the country. variation in mean percent survival was attributable to the L. nigrinus source, followed by a Tukey’s HSD test to separate significantly different means (p = 0.05) using SPSS 16.0 (SPSS Inc., 2007). 2. Materials and methods Proportion data were arcsine square root transformed before analysis to account for unequal variance. Averages are presented as 2.1. Insect populations means ± standard error (SE).

The ‘‘coastal’’ L. nigrinus adults used in the following experi- 2.2.2. Supercooling points ments were collected in early November 2007 and 2009 from A. Supercooling points of coastal (Seattle) and inland (Coeur tsugae-infested western hemlock, Tsuga heterophylla (Raf.) Sarg., d’Alene and Moscow) L. nigrinus adults were tested on 14–16 at the Washington Park Arboretum, Seattle, King Co., WA (47.6 N, January 2008 (n= 21). Beetles were placed on excelsior in plastic 122.3 W; elevation 24 m) using canvas beat sheets (Bioquip, Ran- containers without food or water and kept at 0 °C for 48 h before cho Dominguez, CA). ‘‘Inland’’ adults were collected during the the test. The supercooling point testing equipment and methods same periods either from A. tsugae-infested western hemlock in used were the same as described in (Bentz and Mullins, 1999), with Coeur d’Alene, Kootenai Co., ID (47.7 N, 116.7 W; elevation the exception of a new circulator bath (Lauda K-2/R, Lauda- 658 m) or from western white pine, Pinus monticola Douglas ex Brinkmann, Delran, NJ), data acquisition module (DaqTemp, iotech 98 D.L. Mausel et al. / Biological Control 58 (2011) 96–102

Inc., Cleveland, OH), and software (Daqview v.7.13.14, iotech Inc., 3. Results Cleveland, OH). Four adults were tested at a time and each adult was placed between two thermocouples (accurate to 0.1 °C) in a 3.1. Survival in field cages drop of zinc oxide thermal grease. Adults were cooled at a rate of 2 °C/minute from 24 to 5 °C, 1.0 °C/min from 5 to À10 °C, and Field survival of coastal and inland L. nigrinus adults in cages 0.4 °C/min from À10 °C until the adult froze. As ice formed in tis- differed significantly (F = 34.7; df = 2, 18; p < 0.0001). Mean sur- sues, a temperature spike occurred due to the release of latent heat vival of coastal adults (Seattle = 49% ± 0.07) was lower than that of fusion. The supercooling point was recorded as the mean tem- of inland adults (Coeur d’Alene = 90% ± 0.02 and Mos- perature of the thermocouples on the dorsal and ventral side of cow = 90% ± 0.05), despite the fact that one cage from the Seattle each adult, one second before the spike. One-way ANOVA was used treatment became buried in snow during the experiment and to determine whether a significant portion of the variation in mean had 90% survival (Fig. 1). Average temperature recorded in the supercooling point was attributable to the L. nigrinus source area, cages during the 1 month period was À1.1 ± 0.04 °C, average min- followed by a Tukey’s HSD test to separate significantly different imum temperature was À18.1 ± 0.2 °C during a bout of arctic air on means, as described previously. 2–3 January 2008, and average maximum temperature was 12.1 ± 0.6 °C. Sub-zero temperatures above the coastal L. nigrinus mean supercooling point (i.e., 0 to À16.9 °C, see experiment below) 2.2.3. Sub-zero temperature exposure Survival of coastal (Seattle) and inland (Coeur d’Alene) L. nigrinus adults after a long exposure to temperatures 1.5 °C above the mean supercooling point of coastal beetles were compared on 15–16 April 2010 (n= 75). Individual beetles were placed into ven- 7 tilated plastic containers and held at 5 °C, 75% RH, and 12:12 pho- Seattle, WA Coeur d'Alene, ID toperiod on excelsior (without food or water) for 24 h before the 6 Moscow, ID test. In an environmental chamber (Percival Scientific, Inc. Model LT36VL, Perry, IA) the mean temperature of the beetles during 5 the experiment was À0.6 ± 0.03 °C for 2 h, À8.0 ± 0.03 °C for 2 h, and then À15.4 ± 0.02 °C for 18 h. The mean (±SE) temperatures inside the chamber was calculated from data logger (iButton) read- 4 ings at 2-min intervals. At the end of the test, beetles were warmed gradually over 2 h to 21.0 ± 0.02 °C and placed on A. tsu- 3 gae-infested eastern hemlock with moistened filter paper in plastic Number of cages containers for 4 days. At the end of 4 days, coastal and inland bee- 2 tle survival was recorded. Dead beetles were unresponsive and were monitored periodically over a day to account for death feign- ing. Live beetles exhibited normal walking and feeding behavior. 1 The binomial comparative trial (Zar, 1999) data were analyzed using the conservative Yates continuity correction in SPSS 16.0 0 (p = 0.05). 0 102030405060708090100 Survival (%)

2.2.4. Climate matching Fig. 1. Survival of coastal (Seattle, WA) and inland (Coeur d’Alene and Moscow, ID) The basic ‘match climates’ algorithm was used in CLIMEX v.2 L. nigrinus adults over a 1 month long ﬁeld cage exposure study on eastern hemlock (Sutherst et al., 2004; Sutherst, 2005) to create two maps compar- branches in Leverett, MA (20 December 2007–17 January 2008). ing the climate match index of the Seattle (47.8 N, 122.3 W) and Coeur d’Alene (47.8 N, 166.8 W) collection areas to parts of eastern North America. The climate match index is similar to a correlation coefﬁcient and values vary from zero (no match) to 1.0 (an exact 9 Seattle, WA match). Generally, an index >0.8 is a close match, 0.6 < x < 0.8 is Coeur d'Alene, ID reasonable, and <0.6 is poor (Sutherst et al., 2004). The climate 8 Moscow, ID match index we used was a composite of minimum temperature, 7 maximum temperature, total rainfall, and seasonal rainfall pattern estimated from the 0.5° grid of meteorological data (mean monthly 6 norms, 1961–1990) included in CLIMEX v.2 (New et al., 1999). Two other maps were produced comparing mean monthly minimum 5 temperature (from October to April) of the collection areas to eastern North America, as establishment was positively related to 4 minimum winter temperature in earlier studies (Mausel et al., 3

2010). In addition, we compared the climate and minimum Number of beetles dying temperature match indices between the collection areas and re- 2 cent coastal and inland L. nigrinus release areas in ﬁve northeastern States with paired t-tests (n = 5). These recent L. nigrinus release 1 areas were near Caywood, NY on the Finger Lakes National Forest (42.8 N, 76.8 W); Slate Run, PA on the Tiadaghton State 0 -12 -13 -14 -15 -16 -17 -18 -19 -20 -21 -22 -23 -24 -25 -26 -27 Forest (41.3 N, 77.3 W); Wendell, MA in the Wendell State Forest Temperature (ºC) (42.8 N, 72.3 W); Brattleboro, VT on Cersosimo Lumber Co. land (42.8 N, 72.8 W); and Kittery, ME on non-industrial private forest- Fig. 2. Supercooling points of adult L. nigrinus from coastal (Seattle, WA) and inland land (43.3 N, 70.8 W). (Coeur d’Alene and Moscow, ID) collection areas. D.L. Mausel et al. / Biological Control 58 (2011) 96–102 99 occurred for 374.4 ± 3.2 h and below the supercooling point from Moscow (À18.6 ± 0.6 °C) was not signiﬁcantly different from (i.e.,<À16.9 °C) for 13.5 ± 1.9 h. those of Seattle or Coeur d’Alene.

3.2. Supercooling points 3.3. Sub-zero temperature exposure Supercooling points of coastal and inland L. nigrinus adults differed signiﬁcantly (F = 4.7; df = 2, 60; p = 0.01) (Fig. 2). The mean Survival of coastal and inland L. nigrinus after an 18 h exposure 2 supercooling point of coastal adults (Seattle = À16.9 ± 0.3 °C) was to À15.4 °C differed signiﬁcantly (vy = 9.6, df = 1, p = 0.002). Only 2.3 °C higher than that of inland adults from Coeur d’Alene 36% of coastal (Seattle) adults survived while 63% of inland (Coeur (À19.2 ± 0.7 °C). However, the supercooling point of inland adults d’Alene) adults survived.

Fig. 3. Climate match index maps produced by CLIMEX v.2 to predict areas of reasonable-to-good climatic similarity (0.6–1.0 indices) in the eastern United States for L. nigrinus collected from source areas in (A) Seattle, WA and (B) Coeur d’Alene, ID. The index is a composite of mean monthly minimum temperature, maximum temperature, total rainfall, and seasonal rainfall pattern (1961–1990), with 1 being an exact match. 100 D.L. Mausel et al. / Biological Control 58 (2011) 96–102

3.4. Climate matching States, and some areas of southern Ontario, Quebec, and the Atlan- tic Provinces. The areas of the eastern United States currently infested with A. The eastern United States areas that matched the mean tsugae matched the climates of both Seattle and Coeur d’Alene rea- monthly minimum temperature of Seattle reasonably well (i.e., sonably well (i.e., >0.6) (Fig. 3). However, Coeur d’Alene had higher >0.6) were the southeastern States and mid-Atlantic coast climate index values over a considerably larger area than Seattle. (Fig. 4). The eastern areas that matched the minimum temperature The eastern areas that best match Seattle (i.e., 0.68–0.76) were in of Coeur d’Alene reasonably well were the entire Appalachian the mid- to southern Appalachian Mountains, coastal mid-Atlantic Mountain chain, mid-Atlantic States, southern and coastal New States, coastal southern New England, some areas of the midwest- England, the midwestern States, southeast Ontario, and southern ern States, and southeastern Ontario. Coeur d’Alene’s best matches Nova Scotia. (i.e., 0.68–0.80) include the entire Appalachian Mountain chain, At ﬁve L. nigrinus release areas in the northeastern United mid- to southern New England, coastal New England, midwestern States, signiﬁcantly better climate matching resulted from the

Fig. 4. Minimum temperature match index maps produced by CLIMEX v.2 to predict areas of similarity in the eastern United States for L. nigrinus collected from source areas in (A) Seattle, WA and (B) Coeur d’Alene, ID. The index is based on mean monthly minimum temperature only (1961–1990), with 1 being an exact match. D.L. Mausel et al. / Biological Control 58 (2011) 96–102 101

Table 1 Climate and minimum temperature match indices of coastal (Seattle, WA) and inland (Coeur d’Alene, ID) L. nigrinus collection areas with experimental release areas in ﬁve northeastern United States using CLIMEX (v.2).

Climate match indexa Minimum temperature match indexb Release area Seattle, WA Coeur d’Alene, ID Seattle, WA Coeur d’Alene, ID Caywood, NY 0.671 0.770 0.321 0.689 Slate Run, PA 0.696 0.760 0.338 0.736 Wendell, MA 0.691 0.740 0.266 0.665 Brattleboro, VT 0.661 0.726 0.226 0.610 Kittery, ME 0.711 0.753 0.277 0.683 Average ± SE 0.686 ± 0.009 0.750 ± 0.008 0.286 ± 0.02 0.677 ± 0.02

a Mean monthly minimum temperature, maximum temperature, rainfall total, and rainfall pattern (1961–1990). b Mean monthly minimum temperature from October to April (1961–1990).

Coeur d’Alene collection area versus Seattle in terms of the climate nal movement but in a manner that did not allow then to walk, match index (t4 = 6.5; p = 0.003) and minimum temperature match articulate the head, or feed, and they soon died. index (t4 = 57.8; p < 0.0001) (Table 1). The mean climate match in- Climate matching of collection and release areas is considered a dex improved 6% points and the minimum temperature index im- fundamental requirement in choosing agents for establishment proved 39% points when comparing Seattle to Coeur d’Alene and impact in classical biological control projects (Debach and Ro- collecting areas. sen, 1991). Climatic matching provided further insights into areas at which inland L. nigrinus might have a survival advantage over coastal beetles. In terms of climatic similarity of collection areas 4. Discussion to intended release areas, we found that coastal beetles were best matched with warm areas as compared to colder areas, a conclu- Because L. nigrinus is active and is released for potential biolog- sion borne out by empirical results of actual patterns of establish- ical control of A. tsugae during fall through early spring, winter sur- ment (Mausel et al., 2010). Based on the high match indices vival is critical for population establishment. Results from our between the inland collection area and colder areas of the A. tsu- studies suggest that inland L. nigrinus adults (e.g., northern Idaho) gae-invaded range, we anticipate better establishment rates of in- are more cold tolerant and more likely to survive the winter condi- land L. nigrinus in these areas, where the coastal L. nigrinus has not tions found in cold areas of the eastern United States than are those yet demonstrated establishment. Furthermore, A. tsugae has not of the coastal population (e.g., Seattle, WA). Inland beetles may occupied its entire potential range in the northeastern States, therefore have a higher probability of establishment and impact. mid-western States, and Canada and may become a pest in these However, our field cage study could have artificially restricted cold areas due to continued genetic selection on the pest for in- some of the beetle’s normal behaviors. Surviving L. nigrinus adults creased winter cold tolerance (Butin et al., 2005) and climate were typically clustered where the cage was cinched to the branch change (Paradis et al., 2008). stem suggesting they were trying to move off the branch towards A difference between L. nigrinus populations in cold tolerance in the tree bole and perhaps the forest floor, which would likely have our common environment studies suggests a genetic basis for the increased survival. Adult aggregations were also found in branch trait, but maternal effects were not controlled. Crossing or half- axes inside the cage. Such sheltering behavior is known in Ips sib analysis should be done to verify and investigate if the pheno- grandicollis (Eichoff) (Coleoptera: Scolytidae) adults, which in Wis- typic variation we observed has a genetic basis (Hopper et al., consin migrated to the soil litter layer and survived À27 °C(Lom- 1993). Because L. nigrinus is collected yearly from the western Uni- bardero et al., 2000). Mean supercooling point temperatures ted States for direct biological control release in the eastern United were lower for L. nigrinus adults from Coeur d’Alene compared to States and to supply mass-rearing laboratories, maternal effects on Seattle, but by only a couple of degrees and the importance of this cold tolerance are potentially important for establishment and difference might be reduced if beetles sought shelter in thermally should not be ignored. To test the prediction of improved estab- buffered microsites. lishment and impact by inland L. nigrinus when released directly Freeze avoidance is the most common adaptation in arthropods in the field, we have initiated paired coastal and inland beetle re- to avoid lethal freezing in cold environments and the supercooling leases and control sites for long-term experimental assessment of point represents the lower lethal temperature (Block, 1990). Previ- establishment, population growth, and impact on A. tsugae and ous work has shown that coastal L. nigrinus adult survival declines hemlock health. Previously, large coastal L. nigrinus release sizes with increasing time at temperatures just above the mean superco- were recommended to establish populations in the climatically oling point (Humble and Mavin, 2005) and this aspect of cold tol- unfavorable cold areas of the eastern United States (Mausel et al., erance must be understood to fully assess relative cold tolerance 2010). It is possible that the biological control of A. tsugae with L. (Block, 1990; McDonald et al., 2000). For example, in a comparative nigrinus could be more efficient by allocating coastal and inland cold tolerance study of two populations of Delia radicum (L.) (Dip- beetles to the areas of greatest climate match. tera: Anthomyiidae), temperature changes and duration of sub- zero temperatures were more biologically important than the Acknowledgments supercooling point (Turnock et al., 1990). Similarly, the large population differences in L. nigrinus survival in our cage study com- We are indebted to J. Fidgen, S. Cook, L. Pederson, B. Onken, B. pared to the small differences in supercooling points (or no Thompson, S. Lyons, R. Kimmich, and R. McDonald for assistance difference in Moscow beetles) is probably explained by mortality collecting L. nigrinus; G. Zilahi-Balogh, N. Vandenberg, and N. Hav- due to the nearly 400 h of sub-zero temperatures above the super- ill, for morphological and DNA identification of species used this cooling point, as suggested by our sub-zero temperature exposure study; L. Humble for helpful discussion of methods and results; experiment. Many coastal and some inland beetles that were alive D. Mullins for use of his supercooling point equipment; Cowls immediately after the exposure experiment exhibited leg or anten- Lumber Co. for access to the field research site; Climate Research 102 D.L. Mausel et al. / Biological Control 58 (2011) 96–102

Unit, Univ. of East Anglia, for the 0.5° grid of meteorological data Lawrence, J.F., 1989. A catalog of the Coleoptera of America north of Mexico, Family: used in CLIMEX; and N. Havill, B. Onken and two anonymous Derodontidae. USDA Forest Service Publication Agriculture Handbook, pp. 529– 565. reviewers for their comments. Funding was provided by USDA For- Lawrence, J.F., Hlavac, T.F., 1979. Review of the Derodontidae (Coleoptera: est Service, Northeastern Area, Cooperative Agreement No. 05-CA- Polyphaga) with new species from North America and Chile. Coleopterists 11244225-009. This is manuscript No. 3445 of the Massachusetts Bulletin 33, 369–414. Lombardero, M.J., Ayers, M.P., Ayers, B.D., Reeve, J.D., 2000. Cold tolerance of four Agricultural Experiment Station. species of bark beetle (Coleoptera: Scolytidae) in North America. Environmental Entomology 29, 421–432. References Mausel, D.L., Salom, S.M., Kok, L.T., Fidgen, J.G., 2008. Propagation, synchrony, and impact of introduced and native Laricobius spp. (Coleoptera: Derodontidae) on hemlock woolly adelgid in Virginia. Environmental Entomology 37, 1498–1507. Anonymous, 2005. Hemlock Woolly Adelgid Pest Alert. USDA Forest Service Mausel, D.L., Salom, S.M., Kok, L.T., Davis, G., 2010. Establishment of the hemlock Publication NA-PR-09-05. woolly adelgid predator, Laricobius nigrinus (Coleoptera: Derodontidae), in the Bentz, B.J., Mullins, D.E., 1999. Ecology of mountain pine beetle (Coleoptera: eastern United States. Environmental Entomology 39, 440–448. Derodontidae) cold hardening in the intermountain west. Environmental McDonald, J.R., Head, J., Bale, J.S., Walters, K.F.A., 2000. Cold tolerance, Entomology 28, 577–587. overwintering and establishment potential of Thrips palmi. Physiological Block, W., 1990. Cold tolerance of insects and other arthropods. Philosophical Entomology 25, 159–166. Transactions of the Royal Society of London B 326, 613–633. New, M., Hulme, M., Jones, P., 1999. Representing twentieth-century space-time Bright, D.E., 1991. Superfamily Dermestoidea, family Derodontidae, tooth-necked climate variability. Part I. Development of a 1961–90 mean monthly terrestrial fungus beetles. In: Bousquet, Y. (Ed.), Checklist of Beetles of Canada and Alaska. climatology. Journal of Climate 12, 829–856. Agriculture Canada Publication 1861/E, pp. 195–196. Paradis, A., Elkinton, J., Hayhoe, K., Buonaccorsi, J., 2008. Role of winter temperature Burns, R.M., Honkala, B.H., 1990. Silvics of North America. USDA Forest Service and climate change on the survival and future range expansion of the hemlock Publication Agriculture Handbook, p. 654. woolly adelgid (Adelges tsugae) in eastern North America. Mitigation and Butin, E., Porter, A.H., Elkinton, J., 2005. Adaptation during biological invasions and Adaptation Strategies for Global Change 13, 541–554. the case of Adelges tsugae. Evolutionary Ecology Research 7, 887–900. Phillips, C.B., Baird, D.B., Iline, I.I., McNeill, M.R., Proffitt, J.R., Goldson, S.L., Kean, J.M., Debach, P., Rosen, D., 1991. Biological control by natural enemies. Cambridge 2008. East meets west: adaptive evolution of an insect introduced for biological University Press, Cambridge, United Kingdom. control. Journal of Applied Ecology 45, 948–956. Fender, K.M., 1945. A new Laricobius from Oregon. Pan-Pacific Entomologist 21, 152. SPSS Inc., 2007. SPSS Statistical Sofware: Release 16.0. SPSS Inc., Chicago, IL. Havill, N.P., Montgomery, M.E., Yu, G.Y., Shiyake, S., Caccone, A., 2006. Sutherst, R.W., 2005. Prediction of species geographical ranges. Journal of Mitochondrial DNA from hemlock woolly adelgid (Hemiptera: Adelgidae) Biogeography 30, 805–816. suggests cryptic speciation and pinpoints the source of the introduction to Sutherst, R.W., Maywald, G.F., Bottomley, W., Bourne, A., 2004. CLIMEX v.2 User’s eastern North America. Annals of the Entomological Society of America 99, 195– Guide. Hearne Scientific Software, Melbourne, Australia. 203. Turnock, W.J., Reader, P.M., Bracken, G.K., 1990. A comparison of the cold hardiness Hopper, K.R., Roush, R.T., Powell, W., 1993. Management of genetics of biological of populations of Delia radicum (L.) (Diptera: Anthomyiidae) from southern control introductions. Annual Review of Entomology 38, 27–51. England and the Canadian Prairies. Canadian Journal of Zoology 68, 830–835. Humble, L.M., Mavin, L., 2005. Preliminary assessment of the cold tolerance of Zar, J.H., 1999. Biostatistical Analysis. Prentice Hall, Englewood Cliffs, NJ. Laricobius nigrinus, a winter-active predator of the hemlock woolly adelgid from Zilahi-Balogh, G.M.G., Kok, L.T., Salom, S.M., 2002. Host specificity of Laricobius western Canada. In: Onken, B., Reardon, R.C. (Eds.), Third Symposium on nigrinus Fender (Coleoptera: Derodontidae), a potential biological control agent Hemlock Woolly Adelgid in the Eastern United States, 1-3 February 2005, of the hemlock woolly adelgid, Adelges tsugae Annand (Homoptera: Adelgidae). Asheville, North Carolina. USDA Forest Service Publication FHTET 2005-1, p. Biological control 24, 192–198. 304. Zilahi-Balogh, G.M.G., Humble, L.M., Lamb, A.B., Salom, S.M., Kok, L.T., 2003a. Khåi, J.K., Ylioja, T., Billings, R.F., Régnière, J., Ayers, M.P., 2007. Impact of minimum Seasonal abundance and synchrony between Laricobius nigrinus (Coleoptera: winter temperatures on the population dynamics of Dendroctonus frontalis. Derodontidae) and its prey, the hemlock woolly adelgid (Hemiptera: Ecological Applications 17, 882–899. Adelgidae). Canadian Entomologist 135, 103–115. Lamb, A.B., Salom, S.M., Kok, L.T., Mausel, D.L., 2006. Confined field release of Zilahi-Balogh, G.M.G., Salom, S.M., Kok, L.T., 2003b. Temperature-dependent Laricobius nigrinus (Coleoptera: Derodontidae), a predator of the hemlock development of the specialist predator Laricobius nigrinus (Coleoptera: woolly adelgid, Adelges tsugae (Hemiptera: Adelgidae), in Virginia. Canadian Derodontidae). Environmental Entomology 32, 1322–1328. Journal of Forest Research 36, 369–375.