The brown marmorated stink bug, Halyomorpha halys (Stål) (Heteroptera: Pentatomidae), is an exotic species (Hoebeke & Carter 2003) that is rapidly expanding its range from its original point of establishment in Allentown, Pennsylvania (Hamilton 2009). Collection records for H. halys date from 1996, although it was not realized that the insect was foreign until 2001 after the population had exploded. Halyomorpha halys is native to northeast Asia and, in Japan, has become a serious fruit pest since post-world war II reforestation of old-growth deciduous forest with Japanese cedar, Crytomeria japonica, and false cypress, Chamaecyparis obtusa (Kiritani 2007). According to Kiritani (2007), H. halys and 2 other fruit-feeding stink bugs are dependent on cones to complete their life cycle. The bug also has an extraordinary propensity to form aggregations (Toyama et al. 2006), often in houses and other buildings where they are a nuisance similar to boxelder bugs (Tinker 1952). In fact, it has been suggested that an aggregation of H. halys entered the U.S. in a cargo container (Hoebeke & Carter 2003) and, subsequently, an aggregation was apparently transported via some type of vehicle to Oregon where H. halys also is established (Anonymous 2007). More recently, H. halys has been found established in the Zurich region of Switzerland (Wermelinger et al. 2008).

Pheromone-baited traps offer a convenient and potentially powerful means to monitor insect populations, and some stink bug pheromones are known (Millar 2005), but no pheromone is known for H. halys. However, in Japan H. halys is crossattracted (Tada et al. 2001a,2001b; Funayama 2008) to the pheromone of another pentatomid, Plautia stali Scott, whose males produce methyl (E,E,Z)-2,4,6-decatrienoate (Sugie et al. 1996). In tests begun in 2003 at Allentown (Khrimian 2005; Khrimian et al. 2008), we verified that H. halys adults and nymphs are, indeed, attracted to methyl (E,E,Z)-2,4,6-decatrienoate and related geometric isomers. Surprisingly, methyl 2,4,6-decatrienoates also attracted significant numbers of the native green stink bug, Acrosternum hilare (Say) (Aldrich et al. 2007), even though methyl 2,4,6-decatrienoates have not been found from A. hilare and the reported pheromone components of this bug (trans-/cis-(Z)-α-bisabolene epoxides) (Aldrich et al. 1989; Aldrich et al. 1993; McBrien et al. 2001) are chemically unlike methyl 2,4,6-decatrienoates.

Here we report the results of our efforts to monitor H. halys and A. hilare at the Beltsville Agricultural Research Center from the 2004 through 2008 growing seasons using traps baited with methyl 2,4,6-decatrienoates and trans-/cis(Z)-α-bisabolene epoxides, the reported pheromones of Plautia stali (Sugie et al. 1996) and A. hilare (Aldrich et al. 1989; McBrien et al. 2001), respectively. Over this time period, the population of H. halys in Maryland increased from being undetectable in 2004 to becoming much more abundantly trapped than A. hilare in 2007 and 2008. Previously reported data for 2004 and 2005 (included herein, Aldrich et al. 2007) showed that A. hilare is particularly attracted to methyl (E,E,Z)-2,4,6-decatrienoate.

MATERIALS AND METHODS

Field Trapping

Field experiments were carried out at BARC, Prince George's County, Maryland, from 2004 through 2008 beginning in late spring or early summer through mid-Oct. A summary of the field experiments is presented in Table 1.

From 2004 through 2007, four replicates per treatment were tested with traps hung in the same location each year about 1.8m above ground from tree branches (≥20 m apart) in patches of mixed deciduous forest bordering a field consisting of alternating strips of corn and soybean (see supplemental material). Replicates were grouped into sets of traps, with (≥100 m between sets), and traps were rotated one position within a set every 2–3 weeks. In 2004 and 2005, container traps (30 cm tall transparent plastic, 10 cm removable bottom, and two 9.5 inwardly projecting mesh funnels see supplemental material) were fabricated by Sterling International, Inc. (Spokane, WA). In 2006, a new kind of baffle trap (see supplemental material) was tested side-by-side with the container traps deployed for the first 2 years (4 replicates/treatment/trap type) (Table 1). Baffle traps were constructed from 0.25 inch corrugated plastic (Coroplast®, Dallas, TX) after the design of Mizell (1996) (61 cm tall and 30.5 cm at the base), except that the collector consisted of the top of a reusable yellowjacket trap (Sterling International, Inc.) held in place with wire pins. In 2007 and 2008, the baffle trap was used exclusively except when traps were baited with H. halys males as described below (Table 1).

TABLE 1.

FIELD EXPERIMENTS CONDUCTED AT BETLSVILLE, MD, 2004–2008.

From 9 Aug through 18 Oct 2007, 4 traps were added in which 2–3 live, laboratory-reared H. halys males were used to bait traps. Halyomorpha halys were reared as previously for other stink bugs (Aldrich et al. 1991) but with organically grown green beans (Phaseolus vulgaris) (My Organic Market, College Park, MD) and buckwheat (Fagopyrum esculentum) seeds in addition to sunflower seeds (Helianthus annuus). Adult males at least 2 weeks old were placed with a bean in transparent plastic half pint cups having several 3-4- mm diameter holes punched in the bottom halves of the cups, then the cups were inverted and inserted into the bottom of the container traps instead of the normal trap bottom such that volatiles from the male bugs would be released inside the trap. Another set of traps was added on 18 Sep prepared as described above except that instead of live H. halys males, specimens of methylene chloride-extracted males were pinned inside the plastic cups to test for possible visual response to chemically neutral H. halys males.

From 18 Apr through 31 Oct 2008, four baffle traps baited with methyl (E,E,Z)-2,4,6-decatrienoate and 4 unbaited control traps were deployed as described above except that 2 sets of traps were hung from deciduous trees around the perimeter of the agricultural field in the same location as in previous years, and 2 sets of traps were deployed in 2 patches of mature coniferous trees located approximately 500 m northeast of the deciduous/soybeancorn site (see supplemental material). One conifer patch consisted exclusively of loblolly pine (Pinaceae: Pinus taeda L.); the other patch consisted mainly of eastern white pine (Pinus strobus L.), plus some Austrian (Pinus nigra Arnold) and Virginia pines (Pinus virginiana).

Among phytophagous stink bug species in the subfamily Pentatominae (Heteroptera: Pentatomidae), sexual communication is a bimodal process whereby long-distance attraction is mediated by sex pheromones, and shorter-range mate location is mediated by substrate-borne vibrational songs (Bagwell et al. 2008). Thus, pentatomines attracted chemically from long-range may fail to enter traps in the absence of vibrational signals; therefore, bugs in and within 10 cm of traps were counted as being attracted.

RESULTS

Halyomorpha halys populations increased from being undetectable in 2004 to being abundantly trapped in 2007 and 2008 (Fig. 1). Adults of both sexes and nymphs (second through fifth instars) of each species were caught in roughly equal numbers (data for A. hilare not shown). Captures of A. hilare adults remained constant from 2006 through 2008, while captures of H. halys adults went from being less than those of A. hilare in 2006 to being about 4 times greater than A. hilare for 2007 and 2008 (Fig. 2). In tests conducted in 2008 deploying traps from deciduous or coniferous trees bordering agricultural fields (see online supplemental material), all A. hilare indi-viduals and all but 1 adult male H. halys were captured at the deciduous/soybean-corn site (P < 0.0001). Significant numbers of bugs were never caught in control traps (usually none were caught); therefore, data for control traps are omitted from Figs. 1–5.

Fig. 1.

The total number of Halyomorpha holys adults and nymphs caught inside or within 10 cm of 2 traps baited with isomers of methyl 2,4,6-decatrienoate and deployed at the same location in the field at Beltsville, Maryland, from 2004 through 2008 (2004: treatment = methyl (E,Z,Z)-2,4,6-decatrienoate^*, 4 mg active ingredient/rubber septum in container trap, rebaited biweekly); 2005: methyl (E,Z,Z)-2,4,6-decatrienoate, 4 mg a.i./rubber septum in container trap, rebaited biweekly; 2006–2008: methyl (E,E,Z)-2,4,6-decatrienoate, 2.5 mg a.i./rubber septum in baffle trap, rebaited weekly). [^*Note: Methyl (E,Z,Z)-2,4,6-decatrienoate readily isomerizes to methyl (E,E,Z)-2,4,6-decatrienoate and other isomers when exposed to light (Khrimian et al. 2008).]

The 2006 data for the 2 different types of traps showed that baffle traps worked as well as or better than container traps in capturing H. halys and A. hilare adults (Fig. 3, and supplemental online material). The distribution of insects and location (i.e., inside versus on traps) were significantly different for container- and baffle-type traps (χ² = 10.5, P < 0.0147, df = 3). Halyomorpha halys adults and nymphs were more apt to enter baffle traps than container traps (P < 0.0001); therefore, in 2007 baffle traps were used for all treatments except when container traps were baited with H. halys males.

Fig. 2.

The total number of Halyomorpha halys and Acrosternum hilare adults caught inside or within 10 cm of 2 traps baited with methyl (E,E,Z)-2,4,6-decatrienoate and deployed at the same location in the field at Beltsville, Maryland, from 2006 through 2008 (2.5 mg active ingredient/rubber septum in baffle trap, rebaited weekly). Asterisks over bars indicate significant differences between captures for each species within a year (P < 0.0001).

Fig. 3.

Comparison of the efficiency of container-type and baffle-type traps for the total number of adults of Halyomorpha halys and Acrosternum hilare during 2006 (“in” = inside the trap; “on” = standing on the trap or within 10 cm of the trap; treatment = methyl (E,E,Z)-2,4,6-decatrienoate, 4 baffle traps, 2.5 mg active ingredient/rubber septum, rebaited weekly). Asterisks over bars indicate significant differences between captures for the trap types (P < 0.0001).

The combined captures (bugs inside and within approximately 10 cm of traps) of H. halys and A. hilare for 2006 and 2007 by date are shown in Figs. 4 and 5, respectively. For H. halys, nymphs were the first individuals to be caught in mid-summer, followed by the appearance of adults from late Aug to early Oct. For A. hilare, a few adults were caught in May and Jun, and then nymphs appeared in mid-summer, followed by adults from Aug to early Oct. Some of the adults for both H. halys and A. hilare that were caught in the latter part of the season had soft, incompletely sclerotized cuticle, indicating that they were newly molted adults.

Fig. 4.

The total number Halyomorpha halys adults and nymphs caught inside or within 10 cm of traps baited with methyl (E,E,Z)-2,4,6-decatrienoate by date during 2006 and 2007 (4 baffle traps, 2.5 mg active ingredient/rubber septum, rebaited weekly).

Fig. 5.

The total number Acrosternum hilare adults and nymphs caught inside or within 10 cm of traps baited with methyl (E,E,Z)-2,4,6-decatrienoate by date during 2006 and 2007 (4 baffle traps, 2.5 mg active ingredient/rubber septum, rebaited weekly).

The attraction H. halys and A. hilare adults in 2006 to traps baited with methyl (E,E,Z)-2,4,6-aecatrienoate (EEZ), a 5:95 ratio of trans-/cis-(Z)-αbisabolene epoxides (BX), and the combination EEZ+BX, versus unbaited control traps is shown in Fig. 6. The distribution of H. halys and A. hilare was significantly different for these treatments for each species (χ2 = 93.82 and 83.07, respectively, P < 0.0001, df = 3). H. halys and A. hilare adults of both sexes were attracted to traps baited with methyl (E,E,Z)-2,4,6-decatrienoate, but neither species was attracted to traps baited with the bisabolene epoxide blend. Combining methyl (E,E,Z)-2,4,6-decatrienoate with the bisabolene epoxide blend did not significantly increase captures of either bug (P = 0.1548 and P = 0.063 for A. hilare and H. halys, respectively). The 2007 test results involving EEZ and BX treatments corroborated the 2006 results showing attraction of both species to EEZ-baited but not to BX-baited traps, and baiting traps with a 5 times greater dose of the bisabolene epoxide blend did not result in increased attraction of either species (Fig. 7). In addition, the following tachinid fly parasitoids were caught in 2006 and 2007 in or near traps baited with methyl (E,E,Z)-2,4,6-decatrienoate: Gymnosoma par (Walker) (11 females), Euthera tentatrix Loew (4 females), Euclytia flava (Townsend) (6 females and 3 males); or traps baited with the blend of bisabolene epoxides: Trichopoda pennipes (F.) (9 females and 3 males), E. flava (3 females and 2 males).

Fig. 6.

The total number of Halyomorpha halys and Acrosternum hilare adults caught inside or within 10 cm of traps baited in 2006 with methyl (E,E,Z)-2,4,6-decatrienoate (EEZ), a 5:95 ratio of trans-/cis-(Z)-α-bisabolene epoxides (BX), or EEZ+BX, versus unbaited control traps (4 baffle traps/treatment, 2.5 mg/lure, rebaited weekly). Bars followed by different letters are significantly different (P < 0.0001).

In 2007, baiting traps with live, laboratory reared H. halys males significantly attracted conspecific adults of both sexes and nymphs (4 females, 11 males and 6 nymphs; 29 Aug to 1 Oct, 2007, P < 0.0001). Traps containing solvent extracted, dead adult H. halys pinned males did not attract any conspecific individuals from 18 Sep through 16 Oct, during which time traps baited with live males caught 5 male and 2 female H. halys adults.

Finally, in feeding choice experiments between sunflower seeds and seeds of Japanese cedar (Cryptomeria japonica) or false cypress (Chamecyparis obtusa), second instar H. halys showed an overwhelming preference for sunflower seeds. For the sunflower/Japanese cedar test, 78 nymphs were observed aggregated on sunflower seeds, and none were observed on Japanese cedar seeds. For the sunflower/ false cypress test, 78 nymphs were observed aggregated on sunflower seeds, and 2 nymphs were observed on false cypress seeds.

Fig. 7.

The total number of Halyomorpha halys and Acrosternum hilare adults caught inside or within 10 cm of traps baited in 2007 with methyl (E,E,Z)-2,4:,6-decatrienoate (EEZ), a 5:95 ratio of trans-/cis-(Z)-α-bisabolene epoxides (BX), or 5 septa loaded with BX, versus unbaited control traps (4 baffle traps/treatment, 2.5 mg/ lure, rebaited weekly). Bars followed by different letters are significantly different (P < 0.0001).

DISCUSSION

Previous trapping experiments at Allentown, Pennsylvania, showed that methyl (E,E,Z)-2,4,6-decatrienoate is apparently essential for attraction of H. halys, but methyl (E,Z,Z)- and (Z,E,Z)-2,4,6-decatrienoates can be used to attract H. halys because these isomers substantially isomerize to methyl (E,E,Z)-2,4,6-decatrienoate when exposed to sunlight (Khrimian et al. 2008). Therefore, although our 2004 and 2005 tests used methyl (E,Z,Z)- and (Z,E,Z)-2,4,6-decatrienoates, we are confident that the trap counts recorded at Beltsville, Maryland, from 2004 through 2008 are a true reflection of the range expansion of H. halys into central Maryland. From a practical standpoint it is fortunate that methyl (E,E,Z)-2,4,6-decatrienoate is most active because this is the most convenient of the methyl 2,4,6-decatrienoate isomers to synthesize (Khrimian 2005), and isomerization of methyl (E,E,Z)-2,4,6-aecatrienoate in sunlight may even increase attractiveness to H. halys (Khrimian et al. 2008).

Our data confirm that H. halys is truly an invasive species (Richardson et al. 2000; Colautti & Maclsaac 2004) because in central Maryland the population has risen from being undetectable in 2004 to being more abundantly trapped than the indigenous A. hilare. Halyomorpha halys expanded its range some 280 k to reach Beltsville, Maryland, where it was first detected in traps in 2005 (1 adult and 3 nymphs) (Aldrich et al. 2007). Thus, the lag time (Mooney & Cleland 2001) from establishment to population explosion is relatively short for H. halys. In North America, the brown marmorated stink bug now also has been collected in Ohio, Virginia and Massachusetts, and populations in Pennsylvania and New Jersey have reached damaging levels in some apple orchards (Hamilton 2009). The extent to which H. halys becomes an agricultural pest, in addition to a nuisance when overwintering in dwellings, remains to be seen. However, the insect is continuing to spread in North America, and its northern range may be gradually extended as a result of global warming (Kiritani 2006, 2007). The discovery of H. halys in Switzerland (Wermelinger et al. 2008) portends a similar situation in Europe.

Our finding that traps baited with live, laboratory-reared H. halys males attracted conspecific females, males and nymphs suggests that H. halys males do produce a pheromone despite our inability thus far to isolate any male-specific compounds from them. Alternatively, the captured conspecifics could have been attracted by substrate-borne vibrations from caged males. If attraction was due to the presence of a pheromone, the apparent low level of pheromone production by laboratory-reared H. halys males may be caused by an inadequate diet (see Morishima et al. 2005). In this regard it could be significant that we must feed the insects organically grown green beans for them to survive yet the colony never flourishes (J. R. A., personal observation). Nevertheless, in our feeding tests of Japanese cedar and false cypress seeds versus sunflower seeds, second instars showed no propensity to feed on the conifer seeds notwithstanding Kiritani's (2007) claim that H. halys must feed on cones of these conifers to complete their life cycle. Funayama (2005) presented evidence that the mature seeds of Japanese cedar are suitable for the development of H. halys nymphs, but that indehiscent cones are not. He concluded that most H. halys in his region of study (Akita Prefecture, Japan) cannot develop into adults or lay eggs by feeding on Japanese cedar cones because cones do not dehisce until Oct. On the other hand, H. halys adults laid eggs on Japanese bird cherry trees, Prunus grayana (Rosaceae), in early Jun, and nymphs from these eggs completed a generation by mid-Sep feeding on bird cherry fruit (Funayama 2007). Funayama believes that H. halys females in his region develop their ovaries while feeding on early maturing wild cherries (P. apatala, P. verecunda, and P. sargentii), and then immigrate into the later maturing bird cherry (P. grayana) to lay eggs (personal communication; Fruit-tree Experiment Station, Akita Prefectural Agriculture, Forestry and Fisheries Research Center, Daigo, Hiraka, Yokote, Akita, 013-0102, Japan; funayamak@pref.akita.lg.jp). The observation of Funayama (2007) that Prunus spp. are excellent hosts for H. halys in Japan, raises the possibility that North American Prunus spp., such as black cherry, Prunus serotina, are early season hosts of H. halys in the U.S.

The near total absence of H. halys in the coniferous forest site at Beltsville is further evidence that the establishment of H. halys in North America is not associated with coniferous forests or ornamentals. Our results suggest that H. halys was primarily associated with soybean at our study site, and sweep-net sampling of the soybean strips at the Beltsville deciduous/soybean-corn site confirmed the presence of H. halys nymphs and adults (J. R. A., unpublished data). Indeed, in Oct 2008 we observed a spectacular invasion of a home near Hagerstown, Maryland, by thousands of H. halys adults that were apparently originating from a soybean field about 50 m downhill from the residence (J. R. A. and A. K., personal observation). The abundance of overwintering adults in dwellings where the H. halys population has exploded suggests that the failure to catch adults early in the season in traps baited with methyl (E,E,Z)-2,4,6-decatrienoate reflects a seasonal difference in the responsiveness of H. halys to this compound.

Perhaps the main reason why the population of H. halys is rapidly increasing is that native scelionid wasp (Arakawa et al. 2004) and tachinid fly parasitoids (Aldrich et al. 2006) have yet to fully exploit H. halys as a new host. The only tachinid to emerge from H. halys adults collected in Allentown (2 flies from 834 H. halys adults) was Trichopoda pennipes (F.) (Aldrich et al. 2006), which is the sole known tachinid parasitoid of A. hilare in North America (Arnaud, Jr. 1978). Tachinid flies use pheromones of heteropterans to home-in on potential hosts (Aldrich 1995), so parasitism of H. halys by T. pennipes may be a clue that the pheromone of H. halys is chemically similar to (Z)-α-bisabolene epoxide. Furthermore, we found that T. pennipes is attracted to the reported “pheromone” of A. hilare (Aldrich et al. 1989; Aldrich et al. 1993; McBrien et al. 2001) even though A. hilare itself is not attracted to the 5:95 blend of cis-and trans-(Z)-α-bisabolene epoxides. Combining (Z)-α-bisabolene epoxides with methyl (E,E,Z)-2,4,6-dectrienoate did not increase attraction of A. hilare, nor did increasing the dosage of (Z)-α-bisabolene epoxides affect A. hilare attraction. Thyanta spp. are the only New World bugs known to use a methyl 2,4,6-decatrienoate (the (E,Z,Z)-isomer) as a pheromone component (Millar 1997; McBrien et al. 2002; Moraes et al. 2005); therefore, A. hilare may be eavesdropping on pheromone calling Thyanta males. Interestingly, females of the solitary wasp, Astata occidentalis Cresson (Sphecidae) use methyl (E,Z,Z)-2,4,6-decatrienoate as a kairomone to find Thyanta adults with which to provision their nests (Millar et al. 2001; Aldrich et al. 2007). Evans (1957) also lists Hymenarcys nervosa (Say) and Banasa calva (Say) (Pentatomidae) as preferred prey for A. occidentalis, suggesting that these stink bugs may employ methyl 2,4,6-decatrienoates in their pheromones.

Recently, Funayama (2008) has provided the most direct evidence that H. halys in Japan is attracted to the pheromone of P. stali to find food. By calculating a nutritional level index [= live weight (mg)/pronotum width (mm)³] for adults collected in traps baited with methyl (E,E,Z)-2,4,6-dectrienoate versus adults hand-collected on food plants, he found that the semiochemically collected insects were nutritionally inferior to those hand-collected from host plants. Our current and past data (Aldrich et al. 2007) indicate that the cross-attraction propensity of H. halys for the P. stali pheromone has been maintained in the population of H. halys invading North America. The fact that pentatomid nymphs, even young nymphs, are often attracted to synthetic pheromones of conspecific males (even when no adult conspecifics are present in traps to potentially produce substrate vibrations) (Aldrich 1988, 1995) also supports the idea that pheromones may be associated with food because nymphs are obviously not seeking a mate. Funayama (personal communication) suspects that H. halys adults and nymphs may need to feed on a series of host plants for optimal development. Similarly, A. hilare has a distinct preference for certain native hosts, mainly wild bushes and trees, and a succession of hosts appears to be necessary to sustain a population (Schoene 1933).

Under field conditions in Georgia, Tillman et al. (2009) verified our finding that traps baited with a 5:95 trans- to cis-(Z)-α-bisabolene epoxide blend failed to attract A. hilare, whereas traps baited with methyl (E,E,Z)-2,4,6-dectrienoate significantly attracted A. hilare. Additionally, however, it was demonstrated for the first time in the field that the southern green stink bug, Nezara viridula (L.), can be trapped using its reported pheromone, a 3:1 trans- to ds-(Z)-a-bisabolene epoxide blend (Aldrich et al. 1987; Baker et al. 1987), but that N. viridula is not attracted to methyl (E,E,Z)-2,4,6-dectrienoate. Males of Acrosternum and Nezara spp. produce distinctive ratios of the same trans/cis-(Z)-α-bisabolene epoxides, but A. hilare produced only about 0.3 µg/male/day versus 9 µg/male/day for N. viridula and about 20 µg/male/day for A. pennsylvanicum (Aldrich et al. 1989). Although we now have some indication that H. halys males do produce an attractant pheromone, our difficulty in isolating the pheromone suggests that the amount of pheromone produced is naturally low, as is the case as well for A. hilare.

In conclusion, H. halys seems to have suppressed pheromonal communication while relying more on pheromone cross-attraction to find the array of host plants required for optimal development. The semiochemical situation for A. hilare may be similar since production of male-specific compounds is low, and the insect apparently requires a series of host plants to complete development as for H. halys. The semiochemical parallels between H. halys and A. hilare would be validated if and when native North American species of bugs are found that produce methyl (E,E,Z)-2,4,6-dectrienoate as a part of their aggregation pheromones.