Reproductive and Developmental Biology of Acroclisoides sinicus, a Hyperparasitoid of Scelionid Parasitoids

Simple Summary Acroclisoides sinicus is a pteromalid of Asian origin that has recently been detected in Europe and North America. It has been frequently found in association with several scelionid and eupelmid primary parasitoids of hemipteran eggs, including the brown marmorated stink bug Halyomorpha halys, leading us to suspect that A. sinicus is an obligate or facultative hyperparasitoid. Laboratory tests were conducted using pentatomid egg masses (H. halys, Acrosternum heegeri, and Dolycoris baccarum) parasitized by primary parasitoids, including Trissolcus japonicus, Trissolcus mitsukurii, Telenomus sp., Anastatus bifasciatus, and unparasitized H. halys egg masses, to test this hypothesis. Our studies confirmed that A. sinicus is an obligate hyperparasitoid of the pupal stage of scelionid primary parasitoids but not of eupelmid ones, at least under the tested conditions. Abstract Acroclisoides sinicus (Hymenoptera: Pteromalidae) was described in 1988 from China, but recent findings in Europe and North America within the framework of Halyomorpha halys (Hemiptera: Pentatomidae) biological control indicate a Holarctic distribution. The few records and fragmented information on A. sinicus are derived from generic observations of other species belonging to the same genus, and its biological and ethological traits are still completely unexplored. It was suspected to be a facultative or obligate hyperparasitoid of many egg parasitoid species (e.g., Scelionidae and Eupelmidae), especially those parasitizing Pentatomidae eggs. Laboratory colonies of A. sinicus were established from specimens collected in the field in Europe and the USA, which allowed us to investigate for the first time the life traits of this somewhat enigmatic species. Our studies confirmed the obligate hyperparasitoid hypothesis for species of Scelionidae but not of Eupelmidae. Laboratory studies revealed that A. sinicus is extremely selective in its host recognition as only the pupal stage of its host species is exploited for parasitization. Taking into consideration its hyperparasitoid habit, the adventive A. sinicus populations in Europe and North America may potentially be severe threats to pentatomid natural control as new components in the trophic chain of pentatomids and their parasitoid guilds.


Introduction
Many members of the hymenopteran family Pteromalidae, one of the principal taxa of the superfamily Chalcidoidea, are difficult to identify taxonomically, and their biological and ethological traits are often poorly known [1]. The members of this family are distributed worldwide and most species are primary or secondary parasitoids of a great variety of insect species, and therefore play an important role in the natural control of pests [2][3][4]. In particular, species of the genus Acroclisoides Girault and Dodd, which are distributed

Origin of the Insects for Studies in Europe
Acroclisoides sinicus were reared from H. halys egg masses collected during the summer of 2019 at infested sites in northeastern Italy [8,13]. Field-collected egg masses were reared in glass tubes (15 cm long and 2 cm wide, closed on both ends with a plastic net of 250-µm mesh) and maintained in the climatic chambers at standard conditions of 26 • C, 60% RH, and 16:8 L:D until egg hatch or parasitoid emergence. Identification of adult parasitoids was based on recent taxonomic literature [8,17]. The adults of A. sinicus were separated and placed in glass tubes, and reared in the climatic chambers under standard conditions using pure honey droplets as food. Voucher specimens of A. sinicus were deposited in CREA laboratory collections and are available upon request.
Trissolcus mitsukurii and A. bifasciatus were obtained from laboratory colonies at CREA that originated from H. halys egg masses from infested field sites, while the T. japonicus colony was established with specimens supplied by the USDA-ARS Beneficial Insects Introduction Research Unit, in Newark, Delaware, USA (strain collected in Beijing, China) and imported into Italy under quarantine conditions and authorization number for laboratory studies DG/DISR/DISR05/0013647-19/04/2018. All colonies of the egg parasitoids were maintained on fresh H. halys egg masses (age < 24 h). Adult parasitoids were fed with pure honey droplets replenished twice a week. Adults and parasitized host eggs were reared in glass tubes in climatic chambers under the standard conditions mentioned previously.
Telenomus sp. (identified as Telenomus sp. 1, Francesco Tortorici collection, University of Turin, Italy) were obtained from egg masses of A. heegeri collected occasionally in 2020 during entomological field trips. A Telenomus sp. colony was reared and maintained under the same conditions as the other egg parasitoids but using its natural host, A. heegeri.
The H. halys colony originated from individuals collected at infested sites in northern Italy in 2019 and 2020. Specimens were maintained in rearing cages (BugDorm ® 4F4545 insect rearing cage, MegaView Science, Taiwan) using fresh fruits (apples, kiwi), vegetables (carrots, green beans), potted plants (soybean), and seeds (peanuts and soybean) as food; water was provided with wet cotton. The colony of A. heegeri and D. baccarum was established by collecting adults in the field during the 2019 and 2020 growing season, and reared in the lab under the same conditions as H. halys. Fresh pentatomid egg masses were collected daily from the colonies for the tests.

Establishment of Acroclisoides sinicus Laboratory Colonies in Europe
To test whether A. sinicus was a hyperparasitoid, we established lab-reared colonies of the primary parasitoids T. japonicus, T. mitsukurii, and A. bifasciatus. Fresh H. halys egg masses were offered to females of the three parasitoids for primary parasitization. Considering the limited information available on A. sinicus reproductive biology, H. halys egg masses (n = 15 per species) were first parasitized by the primary parasitoids and after 24 h offered to females of A. sinicus that had emerged from the egg masses collected in the field. To test the combination of A. sinicus with the two Trissolcus species, primaryparasitized egg masses were exposed to groups of three females of A. sinicus for 9 days (since under test conditions Trissolcus adults begin emerging from H. halys host eggs after 10 days of juvenile development in males and 11 days in females). To obtain an A. sinicus colony on A. bifasciatus, primary-parasitized egg masses were exposed for 20 days to females of A. sinicus (Anastatus adults begin emerging from H. halys eggs 22 days after oviposition under test conditions). In all combinations, after exposure to Acroclisoides, the H. halys egg masses were maintained in climatic chambers under standard rearing conditions and checked daily for parasitoid emergence. Any newly emerged adults were removed as soon as possible to prevent intra-and interspecific interference. These F 1 generations of A. sinicus adults were used to produce the next five generations in order to obtain large rearing colonies maintained on the different primary parasitoids.

Rearing Acroclisoides sinicus in the USA under Laboratory and Field Conditions
Acroclisoides sinicus was reared on H. halys egg masses parasitized by T. euschisti. Each of these species originated from individuals collected from infested sites in the southeastern USA in 2019 and 2020 (parasitoids only). The colony of H. halys was maintained in rearing cages (27.9 cm long × 26.7 cm wide × 20.3 cm tall) on whole bean pods, apple slices, and raw peanuts at the USDA, ARS Southeastern Watershed Unit in Tifton, GA; knit cloth (97% cotton, 3% spandex) (Jo-Ann Stores, LLC, Hudson, OH, USA) was used as a substrate for oviposition [18]. A T. euschisti colony was maintained in individual 25 × 150-mm culture tubes (Kimble ® , Rockwood, TN) containing honey and raisins. Refrigerated egg masses (≤12 h old when placed in a refrigerator with a temperature range of 2.8-3.3 • C for 24 h) were used for hosts. Eggs were refrigerated to eliminate emergence of first-instar H. halys, thus avoiding possible feeding by them on parasitized eggs. Individual female T. euschisti were provided with 2-3 H. halys egg masses for oviposition. After exposure to females, egg masses were held in a walk-in environmental chamber at 25 • C ± 2.0 • C, 70 ± 10%RH, and 16:8 L:D until adult emergence. The first step in the process towards rearing A. sinicus was to offer a H. halys egg mass to two T. euschisti females for 2 days in a culture tube. In 2019, 2-, 3-, 4-, 6-, and 8-day-old T. euschisti parasitized egg masses (n = 10 for 2, 3, and 4 days; n = 25 for 6 and 8 days) were exposed to a single A. sinicus female for 72 h in a culture tube. Only one female was placed in a culture tube because females killed additional females. To start a new colony in 2020, individual 5-day-old T. euschisti-parasitized egg masses (n = 30) were offered single A. sinicus females for 72 h in a culture tube. After exposure to females, egg masses were held in a walk-in environmental chamber at the same settings above until adult emergence. These A. sinicus adults were used to produce four generations.
In a preliminary timed exposure test in the field in 2020, lab-reared refrigerated H. halys egg masses were used as sentinels to provide hosts for the detection of resident parasitoids in sassafras (Sassafras albidum (Nutt.) Nees) for 3, 4, 5, and 6 days. In the laboratory, a portion of the knit cloth with an egg mass was cut from the cloth. Any damaged eggs were gently removed from an egg mass when eggs were counted. Each sentinel egg mass was placed on the underside of a sassafras leaf using a small hair clip. Three egg masses were deployed per treatment on 5 Sept. in Prattville, AL. Retrieved egg masses were held in a walk-in environmental chamber at the same settings above until adult emergence. Voucher specimens of A. sinicus were deposited in the Florida State Collection of Arthropods, Gainesville, FL, USA.

Acroclisoides sinicus Reproductive Biology Studies in Europe
Fresh H. halys egg masses were offered for 24 h to females from laboratory colonies of the two primary egg parasitoids, T. japonicus and T. mitsukurii, to ensure complete parasitization. Anastatus bifasciatus was not tested further due to the failure to establish a colony (see the results). After primary parasitization, H. halys egg masses were reared in a climatic chamber under standard conditions to obtain a range of development stages of the primary egg parasitoid (see Table 1 for a schematic representation of the experimental designs). Preliminary studies of immature development of both Trissolcus species revealed that egg hatch and larval development from the first-instar larva to the mature third-instar larva occurred over 5 days (see the results, Tables 2 and 3). Morphological differences among these transitional stages are readily recognizable and correspond with the descriptions of four scelionid species, Telenomus remus (Nixon), Telenomus solitus Johnson, Typhodites gerriphagus (Marchal), and Trissolcus basalis (Woll.) (see also the results, Figure 1) [19][20][21][22]. The pupal stage occurs approximately from 7 to 9 days, at which time the pharate adult is recognizable (see the results, Tables 2 and 3). Our preliminary studies revealed that at around day 5 following primary parasitization, the meconium is expelled and is often visible through the cuticle of the host egg (in T. mitsukurii, an additional black ring appears below the operculum of the host egg) [8,23]. When this occurs, the immature parasitoid enters, transitionally, the pupal stage. Eggs rearing for additional 3 ("6d"), 6 ("9d"), and 9 days ("12d") before dissection Halyomorpha halys egg masses were first parasitized by Trissolcus spp. (T. japonicus/T. mitsukurii) and subsequently exposed to females of A. sinicus in different combinations of ageing and exposure intervals. As a control, unparasitized fresh (<24 h) egg masses of H. halys were offered to females up to egg hatching (not shown here).  Tests were conducted in a climatic chamber under standard conditions using single A. sinicus females 7-10 days old with no previous ovipositional experience. We assumed females to be mated if males were present in the rearing units prior to collecting females. Females were identified by the presence of a black spot on their fore wing [8]. The black spot is not immediately apparent at female emergence but is clearly visible after 48 h (GL and GSP, personal observations). Additionally, females have a physogastric abdomen enlarged with ovary growth. Egg masses containing primary parasitoids at various developmental stages (1, 3, 5, 7, and 9 days after parasitization) were offered to A. sinicus females. Acroclisoides sinicus females were exposed to each "age" class for 24 and 72 h, except for the 9-day group, which was exposed only for 24 h, and the 1-day group with an additional treatment of 9 consecutive days of exposure (216 h). Each treatment of parasitoid age/exposure time was replicated 5 times with a single female exposed to an egg mass in each replicate. Three weeks after exposure to A. sinicus females, H. halys eggs from which neither a nymph or parasitoid emerged, and recorded as pentatomid "dead eggs", were dissected to detect the cause (dead primary parasitoid, dead A. sinicus, dead hemipteran, or undetermined content). To clarify whether A. sinicus could develop within an egg mass as a primary parasitoid (rebutting in this case our hypothesis), fresh egg masses of H. halys (n = 20) were offered to single females in the same manner as described above and checked daily until eggs hatched or parasitoids emerged.
To determine whether A. sinicus exhibits typical haplodiploid sex determination, a common trait in hymenopteran parasitoids, a fresh egg mass (n = 28 eggs) of H. halys previously parasitized by T. mitsukurii was offered to an unmated female of A. sinicus for 9 consecutive days in the same manner as described above. Unmated A. sinicus females were obtained from rearing units from which males were removed shortly after emergence and prior to emergence of females, so that females that emerged had no opportunity to copulate.

Acroclisoides sinicus Immature Development Study in Europe
Single egg masses (n = 15) of H. halys previously parasitized by T. mitsukurii and aged for 5 days to permit immatures of the primary parasitoid to reach the mature larval stage and enter into the pupal stage (see the results) were offered to groups of three A. sinicus females (7-10 days old). Three egg masses were exposed for 48 h ("2d") and three egg masses for 72 h ("3d") and then dissected. The rest of the egg masses were exposed for 72 h to A. sinicus females and then dissected after three ("6d", 6 days total time), six ("9d", 9 days total time), and nine ("12d", 12 days total time) days from the removal of females. Eggs were dissected under a stereomicroscope and the contents carefully inspected. Microslides were prepared when needed for examination of eggs or early larvae. Careful characterization of the three larval instars is critical for assessing the developmental age in species of Pteromalidae. Previous works [15,16] separated larval ages through measurements of morphological features (mandible length and head capsule width) obtained by scanning electron microscopy in conjunction with statistical analysis for Pteromalus cerealellae (Ashmead) (Hymenoptera: Pteromalidae) and Spalangia cameroni Perkins (Hymenoptera, Pteromalidae), respectively, and recognized three larval instars in both species. In P. cerealellae and S. cameroni, the second larval instar was approximately 2-3 times bigger in size than the first instar, and the third instar was twice as large as the second one. Being aware of the complexity of instar determination in pteromalids, a detailed description of larval instars of A. sinicus will be a forthcoming study, and in our preliminary work, we followed a descriptive and practical approach in referring to three larval instars of A. sinicus based on body growth following observations reported in [15,16] (see the results, Figure 1). The immature development study was conducted in a climatic chamber under standard conditions.

Development of Acroclisoides sinicus in Association with Native European Pentatomids
Fresh egg masses of A. heegeri (n = 3, 27 eggs in total) were exposed to females of Telenomus sp. for 24 h for parasitization. Shortly after primary parasitization, egg masses were exposed for nine days to groups of three female A. sinicus (7-10 days old) obtained from the colony reared on T. mitsukurii. After exposure, egg masses were held individually in climatic chambers under standard conditions until parasitoids emerged. The same experimental design was used to test the suitability of D. baccarum replacing H. halys as the primary host in the trophic chain. A fresh egg mass of D. baccarum (n = 1, 21 eggs) was exposed to females of T. mitsukurii for 24 h for parasitization. After primary parasitization, the egg mass was exposed for nine days to groups of three female A. sinicus (7-10 days old) obtained from the colony reared on T. mitsukurii. After exposure, the egg mass was held in a climatic chamber under standard conditions until parasitoids emerged.

Acroclisoides sinicus Longevity Tests in Europe
Newly emerged A. sinicus adults (males and females) reared on T. mitsukurii were isolated individually in glass tubes and provided either with droplets of pure honey or without honey (water was not provided in any case). Specimens were maintained under standard conditions in a climatic chamber without exposure to hosts and checked daily for mortality. Longevity was similarly assessed using A. sinicus from the Telenomus sp. colony fed with honey.

Polyphenism and Female Egg Load of Acroclisoides sinicus Study in Europe
To determine whether host size influenced the body size of adult A. sinicus, we measured the thorax width of A. sinicus emerging from A. heegeri eggs parasitized by Telenomus sp., and from T. mitsukurii-parasitized H. halys eggs from our laboratory colonies. Morphological characteristics of the pentatomid eggs, including egg height (H), egg width (R = D/2), and width of the operculum (r = d/2), were also measured (n = 12 eggs per host species). Egg volume (V) was estimated using the parabolic barrel formula V = πH(3r 2 + 4Rr + 8R 2 )/15 [24,25]. The measurements of A. sinicus body size and egg volume of pentatomid species were compared with Student's t-test.
To determine ovarial egg load, newly emerged A. sinicus females reared on T. mitsukurii from the H. halys laboratory colony were randomly chosen and placed individually in glass tubes provided with honey droplets and reared in a climatic chamber for 1, 7, 14, or 21 days. After each interval, five A. sinicus females were dissected to assess the egg load in their ovaries. Eggs were stained with 1% toluidine blue and counted under a stereomicroscope. Acroclisoides sinicus egg dimensions (length and width) were measured on 15 eggs randomly chosen from four females.

Establishment of Laboratory-Reared Colonies in Europe
Starting from A. sinicus females obtained from field-collected H. halys egg masses, the establishment of lab-reared colonies was successful on H. halys eggs (Figure 2A) parasitized by T. japonicus and T. mitsukurii but not with those parasitized by A. bifasciatus. Despite several attempts to establish a colony on H. halys egg masses (n = 15) first parasitized by A. bifasciatus, in no case was an A. sinicus F 1 generation obtained under laboratory conditions. From H. halys egg masses offered to A. bifasciatus for primary parasitization and then to A. sinicus, a mean of 56.4 % ± 9.58 A. bifasciatus and 37.71 % ± 8.58 H. halys nymphs emerged while 5.88 % ± 1.86 of the eggs died. The contents of dissected dead eggs could not be determined and no A. sinicus mature larvae, pupae, or pharate adults were observed. Several A. sinicus females that emerged from field-collected H. halys egg masses and held in the confined space of glass tubes were frequently observed to display aggressive behaviors, such as biting the antennae, mouthparts, and wings of other females. Aggressivity among females was observed within the rearing tubes but was more frequent when primary parasitized egg masses were available, and host guarding behaviour was observed when potential competitors were present (see Supplementary Materials Video S1). Mating was observed while handling colonies: males emerged (generally one per egg mass) ca. one day before females and lingered on the egg mass waiting for female emergence ( Figure 2B); upon female emergence, mating occurred promptly and lasted just a few seconds.

Reproductive Biological Traits of Acroclisoides sinicus in Europe and the USA
Acroclisoides sinicus emergence was obtained when egg masses previously parasitized by the two Trissolcus species were exposed to females of A. sinicus for nine consecutive days after parasitization by the primary scelionid, and when exposing 5-and 7-dayold primary-parasitized egg masses for 72 h to the females (Tables 2 and 3). No other exposure combinations produced any A. sinicus progeny; from those others, only T. japonicus or T. mitsukurii adults emerged. The single exception occurred for the combination of 7 days × 24 h with T. mitsukurii as the primary parasitoid, from which only one male A. sinicus emerged (Table 3). In no combination did any H. halys eggs hatch. When fresh unparasitized egg masses of H. halys were exposed to females of A. sinicus, only nymphs of the pentatomid emerged (96.42% ± 0.87 hatched nymphs).
In tests with T. japonicus as the primary parasitoid, followed by dissection of dead pentatomid eggs from the 5-, 7-, and 9-day age combinations with the 72-h A. sinicus exposure, means (± s.e.) of 7.14 % ± 3.27, 3.77 % ± 1.47, and 8.08 % ± 1.83 per egg mass, respectively, contained pupae and/or adults of A. sinicus, while means of 17.85 % ± 4.00, 4.37 % ± 2.07, and 13.82 % ± 0.00 per egg mass, respectively, had undetermined contents. Only in the test with nine days of exposure (1 day × 216 h (9 days)) was it possible to detect dead mature larvae and dead pupae of T. japonicus (8.25 % ± 2.28) within dead pentatomid eggs. In all combinations with parasitoid emergence, both sexes of the primary parasitoid T. japonicus and A. sinicus were represented ( Table 2).
In tests with T. mitsukurii as the primary parasioid, dissection of the dead pentatomid eggs in the 5-and 7-day age combinations with the 72-h A. sinicus exposure revealed that mean values (± s.e.) of 6.23 % ± 1.95 and 1.62 % ± 1.00 per egg mass, respectively, contained pupae and adults of A. sinicus; a mean of 2.50 % ± 2.50 per mass had undetermined contents for the 5-day combination. In all combinations with parasitoid emergence, both sexes of the primary parasitoid T. mitsukurii and the mated hyperparasitoid were represented (Table 3). Unmated A. sinicus females that had been exposed to parasitized T. mituskurii immatures produced only male progeny.
In the USA, A. sinicus emergence in the laboratory in 2019 was obtained only when exposing 6-and 8-day-old T. euschisti-parasitized egg masses for 72 h to the females. For 6-day-old T. euschisti-parasitized egg masses, A. sinicus males and females emerged from 25% of the egg masses while for 8-day-old parasitized egg masses, A. sinicus males and females emerged from 33.3 % of the egg masses. In 2020, both sexes of A. sinicus emerged from 100% of the egg masses previously parasitized by T. euschisti aged for 5 days and exposed to hyperparasitism for 72 h.
In the preliminary timed exposure test on sassafras in 2020, parasitism of scelionids by A. sinicus was significantly influenced by the duration of exposure time of H. halys eggs masses used as sentinels on the leaves of this tree. Adult A. sinicus did not emerge from egg masses exposed in the field for 2 and 4 days. However, a single A. sinicus male emerged from an egg mass exposed for 5 days; 3 A. sinicus larvae and 14 larvae and one pupa of T. podisi died. Adult A. sinicus and T. podisi emerged from two egg masses exposed for 6 days. From one egg mass, 6 A. sinicus adults (5 females, one male) and 10 T. podisi adults (6 females, 4 males) emerged; 2 A. sinicus larvae and 4 larvae and one pupa of T. podisi died. From the second egg mass, four females of both A. sinicus and T. podisi emerged; two A. sinicus larvae and seven larvae and one pupa of T. podisi died; one overwintering third-instar of A. reduvii was detected. Because T. podisi adults emerged from egg masses deployed in the field for 6 days, it was likely the primary scelionid host species parasitized by A. sinicus. The results of the field exposure test agreed with those of the laboratory study.

Acroclisoides sinicus Immature Development Studies in Europe
Using H. halys egg masses previously parasitized by T. mitsukurii and aged for 5 days (permitting T. mitsukurii specimens to reach the transition stage from mature larva to pupa), considerably less than 40% of H. halys eggs contained hyperparasitoid eggs after a 48-h initial exposure to A. sinicus females ("2d"), and the rest contained immatures of the first larval instar (LI) (Figures 3 and 4A,B). At the end of the 72-h exposure to A. sinicus females ("3d"), most of the immatures were in the first larval stage (LI). After a 72-h exposure to A. sinicus females and successive rearing of the hyperparasitized H. halys egg masses, the following development stages were observed: after three days of rearing ("6d") most of the immatures were pupae with a few remaining as second and third instars (LII and LIII, respectively) (Figures 3 and 4C-E). After 6 days of rearing ("9d"), most of the specimens were in transition between dark pupae and pharate adults; after 9 days of rearing ("12d"), most of the adults had already emerged. A schematic representation of the immature stages of parasitoids is shown in Figure 1. Hyperparasitized H. halys eggs were dissected after the following ageing periods: "2d" = after 48 h of continuous exposure to females; "3d" = after 72 h of continuous exposure to females; "6d" = 72 h exposure + 3 days ageing; "9d" = 72 h exposure + 6 days ageing; "12d" = 72 h exposure + 9 days ageing.
Eggs of A. sinicus are typically hymenopteriform (spindle-shaped) with a smooth surface, while those of Trissolcus species are stalked (with a tube-like extension at one end) [26]. All larval instars of A. sinicus are typically hymenopteriform (spindle-shaped, body segmented) with two spur-like mandibles, whereas in scelionids, the first-instar larva is teleaform (body segmentation not visible, cephalothorax separated from the abdomen by a constriction, strong mandibles very large and curved, sclerotized caudal process) and the older instars are hymenopteriform [26] (Figure 1). Dissections of T. mitsukurii hyperparasitized by A. sinicus in H. halys eggs revealed that this hyperparasitoid is an ectoparasitoid, with the first larval stage starting to feed on the body of its host and eventually completely consuming it ( Figure 5A,B). Eventually the A. sinicus mature larvae (and successively the pupae) completely occupies the space within the H. halys egg ( Figure 5C-E), which would have been occupied otherwise by the pupae of its Trissolcus host ( Figure 5F). Traces of meconium of A. sinicus expelled by the mature larvae before pupating are clearly visible close to the meconium of its host (Figures 4E and 5D). and (E) pupa; arrow a = large, semicircular meconium of the host species (Trissolcus sp.), arrow b = meconium of A. sinicus appear as dark-gray grains above the creamy-white meconium of its host.

Development of Acroclisoides sinicus in Combinations with European Species
From the three egg masses of A. heegeri previously parasitized by Telenomus sp. and then exposed to A. sinicus females, three male and seven female A. sinicus emerged from one egg mass ( Figure 6); from the other 2 egg masses, 21 specimens of Telenomus sp. emerged. From the single egg mass of D. baccarum previously parasitized by T. mitsukurii and then exposed to females of A. sinicus, 2 male and 10 female A. sinicus emerged, one pentatomid egg hatched, and 8 eggs were dead with no emergence. (B) third-instar larva (LIII) and residues of its host after complete feeding; (C) third-instar larva (LIII) adhering to meconium of its host Trissolcus mitsukurii; (D) meconium of the host (arrow c) and first traces of A. sinicus meconium (arrow b) (partially separated for viewing); (E) pupa of A. sinicus (arrow d) adhering to the meconium of its host (arrow e) and displacing its volume within the H. halys egg; (F) T. mitsukurii pupa (arrow f) and its meconium (arrow g) when not parasitized by A. sinicus.

Discussion
The life cycle of A. sinicus is poorly known from the literature, but our laboratory study provides important details about its biology that will be valuable for planning more detailed biological studies in the future, such as studies of trophic host-parasitoid chains. The frequent association with different species of field-collected pentatomid egg masses worldwide suggested that this species could be an obligate or facultative hyperparasitoid of primary egg parasitoid species belonging to the genera Trissolcus, Telenomus, and Anastatus [7,8]. Our results confirmed the hypothesis of obligate hyperparasitism by A. sinicus of several Trissolcus and Telenomus species present in Europe and the USA, and it also explains several of the species associations reported in the literature. However, we were not able to rear A. sinicus using A. bifasciatus as a primary parasitoid, suggesting that this eupelmid species may not be a suitable host, at least under laboratory conditions. Since both these parasitoid species have emerged from field-collected egg masses, their interactions necessitate further studies. The absence of parasitoid progeny after offering females unparasitized egg masses of H. halys confirmed that A. sinicus is unable to be a primary parasitoid of H. halys, at least under our test conditions. Along with several other parasitoid associations recently documented worldwide [8,12], and the results of our laboratory tests using different hosts with primary parasitoids, we conclude that A. sinicus is an obligate hyperparasitoid of scelionids but not of eupelmids. However, further tests are needed for a definitive assessment under a wider range of laboratory conditions.
There is a very precise parasitization window for A. sinicus females to deposit their own eggs in their hosts to ensure their development, i.e., the time frame represented by the transition from mature larva and meconium expelling to pupal stages, or perhaps only the pupal stage of its host. Production of A. sinicus offspring was observed only in primary parasitoid combinations with T. japonicus and T. mitsukurii pupal stages (H. halys egg dissections revealed A. sinicus eggs and early instar larvae only associated with the presence of scelionid pupae). This was also demonstrated by the co-presence of the meconium of the primary parasitoid and the hyperparasitoid within the H. halys egg; thus, the pentatomid egg shell provides a protected environment for developing A. sincus immatures, which feed on sceliond pupae. In all cases so far observed, the meconium of A. sinicus is deposited directly on the meconium of its host and not randomly within the pentatomid egg [23]. The olfactory cues in the meconium could assist female A. sinicus in their search for hosts and subsequent acceptance, as is known to be the case with many parasitoid and predator species [27]. Progeny production was not observed when larval instars or pharate adults of Trissolcus species were offered for parasitization.
Hyperparasitism can occur with different levels of specialization depending on the associations between different host parasitoid and hyperparasitoid species. For example, the eupelmid A. bifasciatus is known to be a broadly generalist primary parasitoid, able to exploit both pentatomid and lepidopteran eggs, but it is also a facultative hyperparasitoid of several primary parasitoids at different development stages, including T. japonicus [28][29][30]. Many pteromalids are pupal parasitoids as is A. sinicus.
It was not possible to clearly define the developmental time for the A. sinicus immature stages because progeny production was obtained only by exposing the host for 72 h to parental females. Progeny production was not successful with only a 24-h exposure time except in a single case. The parasitization process likely occurs between the second and third day of host exposure, thus the developmental time of A. sinicus should be between 10 and 11 days. The single exception of progeny production after a 24-h host exposure resulted in a 10-day developmental time and produced a male. A similar developmental time was reported for the pteromalid P. cerealellae [15]. Under the same rearing conditions, immature development of A. sinicus is similar to that of Trissolcus species attacking H. halys [31,32].
Males always emerge a day before the females and mating occurs immediately on the egg mass, evidence that sib-mating occurs frequently, as observed for P. cerealellae [15]. Sibmating is also a common behavoir in egg parasitoids of pentatomids, such as T. japonicus and Gryon pennsylvanicum (Ashmead) (Hymenoptera: Scelionidae) [32,33].
As shown by our preliminary studies on immature development of T. japonicus and T. mitsukurii, the first larval stages are teleaform, which is typical for scelionids, and mandibles are abnormally developed compared to the last larval instar. Different types of mandibles may reflect different behaviors by first-instar larvae of these parasitoids. Trissolcus japonicus and T. mitsukurii may be adapted to develop in a more competitive environment, fighting for resources with competitors in the H. halys parasitoid guild. In contrast, A. sinicus larvae apparently develop in a less competitive environment, attacking a host that is generally less active, i.e., the pupa of its scelionid host. In all cases, the purpose of mandibles of such larvae may be to puncture the host integument and imbibe fluid food oozing from the host body, as described for S. cameroni [16].
When provided with carbohydrate in the form of honey, A. sinicus adults live a long time in the lab, similar to T. japonicus or A. bifasciatus [34,35]. Without food A. sinicus can survive for several days, much longer than Trissolcus [34]. Longevity seems to be host dependant, for A. sinicus females developing on the H. halys-T. mitsukurii combination lived twice as long as those developing on the A. heegeri-Telenomus combination. These differences could due to the quantity and/or quality of food resources of the different hosts: A. sinicus that developed on the A. heegeri-Telenomus sp. association are smaller than those that developed on H. halys-T. mitsukurii one, and A. heegeri eggs are smaller than those of H. halys. This was also seen in T. mitsukurii, whose adults live longer when they develop on H. halys eggs than on N. viridula eggs [36,37].
The Asian A. sinicus attacks not only several exotic Trissolcus species that have become invasive in Europe, but also several European native egg parasitoids, developing in different host associations, such as Telenomus sp. developing on A. heegeri or T. mitsukurii developing on D. baccarum. The emergence of A. sinicus males and females from fieldcollected egg masses strongly suggests that that this hyperparasitoid might also attack at least some of the native Trissolcus and Telenomus species parasitizing a range of pentatomid host eggs in the USA [8], and the development of A. sinicus in the laboratory confirms its hyperparasitism behavior of the North American scelionid T. euschisti. In addition to the results reported in this paper, rearing records from other field surveys reported the development of A. sinicus on (unidentified) parasitoids within the pentatomid genera Brochymena sp., C. hilaris, Euschistus sp., P. prasina, and H. halys ( [8] and KAH unpublished data).
Hyperparasitism by A. sinicus of scelionids, including T. japonicus as shown in our study, could have important consequences for classical biological control of H. halys by the release of exotic parasitoids (independently of T. japonicus or T. mitsukurii). A related question concerns the ecological impact of such hyperparasitoid on the global natural control provided by native egg parasitoid guilds and their role in the ecosystem. The recent discoveries of adventive A. sincus in newly invaded areas of Europe and North America suggest that it is likely to become a permanent part of the local fauna, but the impact of this species to the ecosystem of the North American and European fauna is a matter to be assessed in the future [38]. Biological traits presented in our work can explain many aspects of biological and ethological features under field conditions and can be helpful for future planning of research with such complicated parasitoid species, where highly specific optimal timing of host suitability within the entire trophic chain is fundamental for the success of the parasitoid's reproduction.

Conclusions
We conclude that A. sinicus is an obligate hyperparasitoid of the scelionids T. japonicus, T. mitsukurii, and Telenomus sp. but not of the eupelmid A. bifasciatus. Adult emergence of the hyperparasitoid was obtained upon exposing 5-and 7-day-old primary-parasitized egg masses for 72 h to A. sinicus females, suggesting that this pteromalid is a highly specific hyperparasitoid exploiting hosts at the pupal stage. The recent findings in Europe and North America of this Asian hyperparasitoid emerging from pentatomids are concerning in that this pteromalid may interfere in the trophic chains that regulate the population dynamics of stink bugs.