Field-Cage Methodology for Evaluating Climatic Suitability for Introduced Wood-Borer Parasitoids: Preliminary Results from the Emerald Ash Borer System

Field-cage methods were developed to evaluate the abilities of Tetrastichus planipennisi Yang (Hymenoptera: Eulophidae) and Spathius agrili Yang (Hymenoptera: Braconidae), biocontrol agents of Agrilus planipennis Fairmaire (Coleoptera: Buprestidae), to parasitize, develop and overwinter following three late-season releases at both a northern (Michigan) and a southern (Maryland) location within the current North American range of A. planipennis. In August, September and October of 2009, five young green ash trees were selected at each location. Tetrastichus planipennisi and S. agrili were each randomly assigned to one of two cages attached to each tree, surrounding separate sections of trunk in which late-instar A. planipennis had been inserted. The following April, the caged trunk sections were dissected to determine the fate of each A. planipennis larva and the developmental stages of all recovered parasitoid progeny. At both locations, T. planipennisi and S. agrili were able to parasitize hosts and successfully overwinter (i.e., reach adulthood the following spring). For T. planipennisi, successful parasitism (i.e., parasitoid progeny reached adulthood) occurred for all caged releases in Maryland, but only for the August and September releases in Michigan. At both locations, percent parasitism by T. planipennisi was higher in August and September than in October. For S. agrili, successful parasitism occurred for all caged releases in Maryland, but only for the August release in Michigan. In Maryland, percent parasitism by S. agrili in August and September was higher than in October. The caging method described here should be useful in determining the climatic suitability of other regions before proceeding with large-scale releases of either species and may have utility in other wood-borer parasitoid systems as well.


Introduction
Although less than a decade has passed since the emerald ash borer, Agrilus planipennis Fairmaire (Coleoptera: Buprestidae), an invasive beetle from Asia, was first identified as the cause of ash tree (Fraxinus spp.) mortality in North America (Haack et al. 2002), the species has already become one of the most ecologically and economically significant invasive forest pests in North America (Gandhi and Herms 2009;Kovacs et al. 2010). In Michigan alone, A. planipennis has killed an estimated 40 million ash trees, with tens of millions more killed in other infested states and provinces (Emerald Ash Borer Information 2009). Efforts by regulatory agencies to eradicate A. planipennis were unsuccessful, and chemical control options are expensive (Sadof 2010), temporary, and not suited for widespread use (Poland and McCullough 2006;Herms et al. 2009). Although woodpeckers (Lindell et al. 2008) and some native parasitoid species (Cappaert and McCullough 2009) are important natural enemies of A. planipennis in North America, their overall impact is insufficient to reduce A. planipennis population densities below a lethal threshold for most ash trees (Bauer et al. 2004;Duan et al. 2009;Kula et al. 2010). A classical biological control program, involving three Hymenopteran parasitoids of A. planipennis in China (Liu et al. 2003Gould et al. 2005), is under way in several infested states. The two larval parasitoids, Tetrastichus planipennisi Yang (Eulophidae) and Spathius agrili Yang (Braconidae), have so far been released in Michigan, Maryland, Ohio, and Illinois.
Tetrastichus planipennisi is a gregarious koinobiont endoparasitoid that preferentially parasitizes large larvae (Liu et al. 2003;Yang et al. 2006;Ulyshen et al. 2010b;Ulyshen et al. in press). Adults range from 1.6 to 4.1 mm in length with a female to male sex ratio of ~ 2.5:1 (Yang et al. 2006). Agrilus planipennis larvae parasitized by T. planipennisi remain active and continue to feed for about a week. Between 4 and 172 offspring are produced per host (Ulyshen et al. 2010a). After consuming the host larva, parasitoid larvae exit from the integument and pupate within the A. planipennis gallery (Yang et al. 2006). In China, four or more generations are produced per year and levels of parasitism average 22.4% , ranging from 0% to 65% (Liu et al. 2003;). Although T. planipennisi overwinters as larvae or pupae in a dormant state (pers. obs.; , there is no evidence that the species enters diapause. Spathius agrili is a gregarious idiobiont ectoparasitoid that also prefers to parasitize older A. planipennis larvae (Yang et al. 2005). Adults range from 3.4 to 4.3 mm in length with a female to male sex ratio of ~3:1 (Yang et al. 2005). Females permanently paralyze their hosts by envenomation when ovipositing and produce 1-20 offspring per host (Yang et al. 2005(Yang et al. , 2010. After consuming the host larva, mature larvae spin individual cocoons within the gallery (Yang et al. 2005). Adults from a single brood exit from their cocoons and emerge from the tree over a period of 5 to 44 days (Yang et al. 2010). In China, the species completes up to four generations per year and levels of parasitism range from 30% to 90% (Yang et al. 2005(Yang et al. , 2010. Spathius agrili overwinters as prepupae inside reinforced cocoons (Yang et al. 2010) and, unlike T. planipennisi, apparently enters diapause (JG, unpublished data), presumably in response to changing photoperiod and/or temperature (Leather et al. 1993).
Whereas rearing methods have been developed for both species and releases began in 2007 (Bauer et al. 2008;Ulyshen et al. 2010a;Yang et al. 2010), information on where and when to release either species is lacking. This is problematic considering that, historically, most introduced biological control agents have failed to be established (Clausen 1956) and the inability to adapt to the new climate has been responsible for more of these failures than any other single factor (DeBach and Bartlett 1964). In addition, because natural enemies are usually restricted to narrower geographic or ecological ranges than their hosts (Messenger and van den Bosch 1971), it may not be realistic to expect T. planipennisi and S. agrili to establish and persist wherever A. planipennis occurs in North America. Although the distributions of T. planipennisi and S. agrili are known to overlap in China (Liu et al. 2003), the extent to which this is the case, and whether their ranges fully encompass the distribution of the emerald ash borer in Asia, remain unclear.
Clearly, large-scale releases of these species will be more effective where and when conditions have been shown to be suitable. The ability to successfully overwinter, for example, should be demonstrated prior to proceeding with any large-scale releases (Leather et al. 1993), but this ability has not been well-documented for either species anywhere in North America. In addition, the point at which it becomes too late in the season to expect successful parasitism and establishment following a parasitoid release is likely to vary between species and with latitude, but this too remains unexplored. A major obstacle in addressing these questions has been the lack of a reliable sampling method for evaluating the establishment of wood-borer parasitoids. ! In the present study, field-cage methods were developed to preliminarily test the abilities of T. planipennisi and S. agrili to parasitize, develop and overwinter following three lateseason releases at both a northern (Michigan) and a southern (Maryland) location within the current North American range of A. planipennis. In addition to assessing the utility of the caging method, the information gained from this study should help determine where and when these species should be released to best achieve establishment and, ultimately, control of A. planipennis.
Parasitoids. Laboratory colonies of T. planipennisi, originally collected in China (Liaoning province), and S. agrili, also originally collected in China (Tianjin city), were used in this study. Only naïve wasps that had not been presented with hosts were used in experiments. They were at least one week old at the time of use and were presumed to have mated because mating is known to occur almost immediately for both species following female emergence (personal observation). October), five young green ash trees 4-7 cm in diameter at 1.5 m were selected at each location (i.e., 15 trees per location, 30 trees in total). Two cages were attached to each tree, surrounding separate ~ 0.5 m sections of trunk, in which ten third-or fourth-instar, field-collected A. planipennis had been inserted. The heights of the cages varied somewhat among trees, but never exceeded 3 m. The larvae were inserted into narrow grooves chiseled beneath small flaps of bark. The larvae were inserted "head down" to encourage feeding and the flaps of bark were held closed over the larvae with thin strips of Parafilm. The ten larvae in each trunk section were arranged in two "rings" encircling the trunk, each with 5 larvae, with the rings being ~ 10 cm apart. All larvae were inserted 1-3 m above the ground. The cages were made of fine fabric screening held ~5-10 cm away from the trunk by metal wire frames attached to each tree. Water in a vial with a cotton wick and honey streaked on the inside surface of a Petri dish lid were suspended with string in each cage. A section of plastic sheeting was stapled to the tree above each cage to provide shelter during rainstorms. For pictures of similar cages, see Ulyshen et al. (2010a). The cages were constructed over a two-day period; A. planipennis larvae were inserted and the wire frames attached on the first day and water, honey, screening and wasps were added on the following day. Staples were used to connect the edges of the screen and rubber bands were used to secure the screen to the trunk at the top and bottom of the cage. For each tree, 10 and ~ 5 T. planipennisi were randomly assigned to one cage and 5 and ~ 5 S. agrili to the other.
The following April, the caged trunk sections were dissected to assign each A. planipennis larva to one of the five following fates: parasitized (by T. planipennisi or S. agrili), preyed upon (including parasitism by native parasitoids), alive, dead (not killed by a predator or parasitoid), or lost (larvae that could not be found at the time of sampling). At the time of sampling, T. planipennisi progeny were assigned to one of the following four categories: larva, pupa, emerged adult (still in the galleries) and exited adult (already emerged from the tree as evidenced by emergence holes through the bark). Because all S. agrili progeny at the time of sampling were inside cocoons, their precise stages of development could not be determined. All recovered parasitoid broods were individually placed in Falcon Petri dishes (50 9 mm with tight-fit lid) lined with moistened filter paper and reared in the laboratory to determine overwintering survival. Data Analysis. Because larvae were inserted into the ash trunks by different groups of people in Michigan and in Maryland, and ash tree health varied between locations (i.e., the trees selected in Michigan generally appeared healthier than those used in Maryland), data on parasitism and other fates of A. planipennis larvae for each parasitoid species at each location were analyzed separately with the likelihood Chi-square test (SAS Institute  2008). Percent parasitism calculations were based on all A. planipennis larvae recovered at the end of the experiment, including dead larvae as these were all healthy and actively feeding when inserted into the trees (Note: although some larvae may have died soon after insertion, making them essentially unavailable for parasitism, the extent to which this was the case cannot be determined). All of the A. planipennis larvae that were lost due to insufficient tightening of the bark flaps or destroyed during the sampling process, however, were excluded from the data set.

Results
At both locations, T. planipennisi and S. agrili were able to parasitize hosts and overwinter at advanced stages of development. Tetrastichus planipennisi overwintered as late-instars or pupae after exiting the host, with about 89% (N=614) and 98% (N=208) of the progeny recovered from Michigan and Maryland, respectively, reaching adulthood in the laboratory. Spathius agrili overwintered inside cocoons, with about 53% (N=12) and 50% (N= 36) of the progeny recovered from Michigan and Maryland, respectively, reaching adulthood in the laboratory. T. planipennisi progeny were more developed at the time of sampling in Maryland than in Michigan due to climatic differences between the two locations ( Figure 1). For example, all T. planipennisi broods from the first two caging periods in Maryland were adults at the time of sampling and half of the adults had already exited the trees. In Michigan, by contrast, all T. planipennisi broods from the first caging period consisted of pupae and those from the second caging period consisted of pupae as well as larvae. While S. agrili progeny were likely at more advanced stages of development in Maryland as well, all progeny recovered at both locations were within cocoons so their precise stages of development could not be determined. In both Maryland ( Figure 2) and Michigan (Figure 3), caging period significantly affected percent parasitism of A. planipennis hosts exposed to T. planipennisi (Maryland: N = 81, 2 = 45.147, df = 6, p < 0.0001; Michigan: N = 146, 2 = 28.759, df = 6, P<0.0001) or S. agrili (Maryland: N = 60; 2 = 27.643, df = 6, p < 0.0001; Michigan: N = 141, 2 = 53.124, df = 6, p < 0.0001), although this may be due in part to the fact that the mortality rate of A. planipennis larvae varied considerably among caging periods, particularly in Maryland (Figures 2 and 3). For T. planipennisi, successful parasitism occurred during all three caging periods in Maryland (Figure 2), but only during the August and September periods in Michigan ( Figure 3). Percent parasitism by T. planipennisi in August (18.0 and 35.7%) and September (14.6 and 41.4%) was higher than in October (0.0 and 8.3%) in both Michigan and Maryland, respectively. Similarly, in Maryland, percent parasitism by S. agrili in August (28.6%) and September (43.48%) was significantly higher than in October (4.4%). In Michigan, by contrast, only 6.25% parasitism by S. agrili was observed in August and no parasitism was observed in September or October.

Discussion
At both locations, caging period significantly affected the parasitism rate and stage of development of T. planipennisi and S. agrili. Parasitism was recorded for both species from each caging period in Maryland, but at considerably reduced rates in October. It should be noted, however, that, for reasons unknown, a large proportion of the A. planipennis larvae inserted in October in Maryland were dead at the time of sampling. Although the extent to which this was the case cannot be determined, it is possible that some mortality took place soon enough after insertion to affect observed parasitism rates. In Michigan, parasitism was not recorded for either species in October and was recorded only for T. planipennisi in September. Though not detectable at the time of sampling, it is possible that some parasitism occurred following these releases but the parasitoid progeny were unable to develop adequately within or upon the host before the onset of winter. Alternatively, the apparent lack of parasitism in October in Michigan may be attributable to cool temperatures on the morning the wasps were added to the cages given that current weather conditions, especially temperature, can determine whether a field release will succeed or fail (Stiling 1993;Etzel and Legner 1999). Regardless, the results from this study suggest late-season releases of either species should be avoided when possible.
The parasitism rates exhibited by T. planipennisi and S. agrili were similar in Maryland for the three caging periods. In Michigan, by contrast, a greater proportion of hosts were parasitized by T. planipennisi than S. agrili in August and parasitism was recorded only by T. planipennisi in September. That S. agrili parasitized at lower rates than T. planipennisi in Michigan may be due, in part, to the fact that greater proportions of A. planipennis larvae died in August and September in the S. agrili cages (Figure 3). It remains unknown, however, if this mortality occurred soon enough after insertion to affect observed parasitism rates. In a field-cage study carried out within the same Michigan forest but several months earlier (i.e., June and July 2009), the number of hosts parasitized by S. agrili was comparable to that parasitized by T. planipennisi (Ulyshen et al. 2010a). (It should be noted that in the previous study, three female S. agrili or 10 female T. planipennisi were added to each cage compared to the five or 10, respectively, used in the present study). These inconsistent findings may either suggest that A. planipennis mortality did occur early enough in the present study to affect observed parasitism rates or that the effectiveness of S. agrili declines more rapidly near the end of the season than does that of T. planipennisi. More research will be needed to resolve this issue.
The caging methods described in this study should prove useful in assessing the climatic suitability of other sites before proceeding with large-scale releases of either species and may have utility in other wood-borer parasitoid systems as well. It is important to note, however, that a single documentation of overwintering following a field release is not conclusive evidence of suitability as it is common in biological control to document parasitism at immediate release sites even when conditions there are not suitable to maintain permanent populations (DeBach and Bartlett 1964). Consequently, long-term monitoring of release sites will be needed in order to fully assess the potential ranges and impacts of T. planipennisi and S. agrili in North America.