Parasitoids of the Sawfly, Arge pullata, in the Shennongjia National Nature Reserve

Larvae of the argid sawfly, Arge pullata (Zaddach) (Hymenoptera: Argidae), feeds on leaves of birch (Betula spp.) in China, Europe, Siberia, and Japan. Parasitoids of A. pullata were studied in Shennongjia National Nature Reserve, Hubei Province, China, in 2009 and 2010. Five parasitoid species were found: Pleolophus suigensis (Uchida), Mastrus nigrus Sheng, Endasys parviventris nipponicus (Uchida) (Hymenoptera: Ichneumonidae), Vibrissina turrita (Meigen) (Diptera: Tachinidae) and Conura xanthostigma (Dalman) (Hymenoptera: Chalcididae). The average parasitism rate of A. pullata by parasitoids was as high as 11.0%. V. turrita was the dominant species, attacking 10.0% of the A. pullata cocoons. The emergence peak of V. turrita was from late May to early June. Three hyperparasitoids of V. turrita emerged from cocoons of A. pullata: Mesochorus ichneutese Uchida (Hymenoptera: Ichneumonidae), Pediobius sp. (Hymenoptera: Eulophidae), and Taeniogonalos maga (Teranishi) (Hymenoptera: Trigonalidae). Hyperparasitism rates were about 1.0% to 3.0%, with an average rate of 1.7%.

The biology, occurrence, and integrated management of A. pullata have been studied less in China (Li and Yuan 1993;Qi 2000) than elsewhere. In Japan, this sawfly was first recorded in Sugadaira, Nagano Prefecture, central Honshu, where it had a univoltine life cycle (Takizawa 1962). Hara and Shinohara (2008) reviewed the taxonomy, distribution, life history, and economic importance of A. pullata in Sapporo, Hokkaido, where A. pullata probably had two generations a year. In China, A. pullata had a univoltine life cycle (Li and Yuan 1993;Qi 2000).
High population levels of sawflies are not only affected by natural collapse, but are regulated by natural enemies (Luo et al. 2005). In China, most studies on parasitoids of sawflies have involved Diprionidae, Pamphiliidae, and some Tenthredinidae that damage forest trees (Wang et al. 1996;Zhang and Zhou 1996;Wang et al. 2000;Sheng and Chen 2001;Sheng et al. 2002;Chen and Sheng 2007). Worldwide, little is known about parasitism of Argidae, although they support a parasitoid complex that is highly distinctive (Pschorn-Walcher and Kriegl 1965).
Parasitoids play an important role in biological control of agriculture and forest pests. Up until now, only two parasitoids, Endasys parviventiris nipponicus (Uchida) and Mastrus nigrus Sheng (Hymenoptera: Ichneumonidae), have been reported on A. pullata (Min et al. 2010;Sheng and Zeng 2010). The data on additional parasitoids presented here are important for understanding the natural control of A. pullata, and future biological control effect against this economically important sawfly.

Materials and Methods
Overwintering cocoons of A. pullata under the bark of Betula spp. were collected on four occasions at the same site in Shennongjia National Nature Reserve, Hubei Province, in October 2009 and April 2010. The first sample (N = 2971) was collected on 8 October 2009, whereas the additional three samples (N = 4497, 12082 and 7075) were taken on 26 April 2010. The cocoons were reared in the laboratory at room temperature, and misted with distilled water one to two times per week in order to prevent desiccation. After one week, cocoons were stored individually in glass tubes (100 mm long and 15 mm in diameter) with a piece of filter paper dipped in distilled water (to prevent desiccation), and plugged with absorbent cotton.
All cocoons were checked daily for sawfly and parasitoid emergence until late autumn.
Emerged parasitoid larvae and pupae were kept in glass tubes at room temperature until adult emergence. After emergence of sawflies and parasitoids was complete, all remaining cocoons were dissected, and their condition (i.e., status of argid sawfly, and parasitism) was recorded. Parasitism rates data (p) were transformed by arcsin(p) 1/2 in order to better fit the assumptions of normality and homogeneity of variances for ANOVA. The means were analyzed by one-way ANOVA, followed by the Ryan-Einot-Gabriel-Welsh (REGW) multiple Q test (SPSS 13.0 for Windows) at α = 0.05. (Dalman) (Hymenoptera: Chalcididae). The overall parasitism rates of A. pullata larvae by these parasitoids ranged from 8.1% to 13.8%, with an average rate of 11.0% (Table 1).

Parasitoids
Only seven adults of the three parasitic ichneumonids, P. suigensis, M. nigrus, and E. parviventris nipponicus (Hymenoptera: Ichneumonidae), emerged from the cocoons of A. pullata, and 0, 28, 26 and 25 larvae were discovered from dissecting host larvae of the samples 1-4, respectively. The larval ichneumonids were indistinguishable from each other, so their parasitism rates were pooled and discussed as a group. Parasitism rates of A. pullata cocoons by ichneumonids were from 0.1% to 0.6%, with an average rate of 0.3%. Ichneumonids pupated inside the sawfly cocoons after feeding on the sawfly larva. Adults made an irregular emergence hole in the sawfly cocoons when emerging.
V. turrita larvae developed in the body of final instar A. pullata larvae, and killed them after they emerged from hosts. The tachinid final instar larvae pupated inside or outside the sawfly cocoon until adult emergence. Emergence holes were observed on the sawfly cocoons. One dead tachinid pupa was found inside the final instar larva of A. pullata when dissected. There were 401, 245, 694, and 535 adults tachinids that emerged from the cocoons of A. pullata in each sample, respectively. Parasitism rates of A. pullata larvae by V. turrita were from 7.0% to 13.5%, with an average rate of 10.0%.
C. xanthostigma larvae developed in the body of final instar A. pullata larvae, killed them after they emerged from hosts, and pupated inside the sawfly cocoon until adult emergence, leaving their puparium in the sawfly cocoon. Parasitism rates of A. pullata larvae by C. xanthostigma were from 0.1% to 1.6%, with an average rate of 0.6%.   There were significant differences in the parasitism rates among the three groups of parasitoids (F = 54.97; df = 2, 9; p < 0.001). V. turrita was the dominant parasitoid, with its parasitism rates being significantly higher than those of both ichneumonids and C. xanthostigma (p < 0.001). There were no significant differences in the parasitism rates between the latter two parasitoid groups (p > 0.05) (Figure 1).
The seasonal occurrence of V. turrita adults emerging from final instar larvae of A. pullata was recorded in 2010 ( Figure 2). The emergence of V. turrita had two peaks. The first peak occurred from 22-28 May 2010, with 114 flies emerging on 28 May, 2010. The second peak was on 1 June 2010, with 111 flies emerging. The number of V. turrita emerging dropped gradually from 3 June 2010, and ended on 15 June 2010. Overall, the emergence peaks of V. turrita were from late May to early June.

Hyperparasitoids
Three species of hyperparasitoids, Mesochorus ichneutese Uchida (Hymenoptera: Ichneumonidae), a species of Pediobius sp. (Hymenoptera: Eulophidae), and Taeniogonalos maga (Hymenoptera: Trigonalidae), were reared from V. turrita. Hyperparasitism rates were from 1.0% to 2.8%, with the average of 1.7%. M. ichneutese was only found in the 4th sample of A. pullata cocoons, with a hyperparasitism rate of 1.1%. One to two Pediobius sp. emergence holes were found on the puparium of V. turrita, and its hyperparasitism rates ranged from 0.2% to 0.7%, with an average rate of 0.5%. An irregular emergence hole on the pupae of V. turrita was characteristic of T. maga. Its hyperparasitism rates were from 0.2% to 1.6%, with an average rate of 1.0% (Table 2).

Discussion
Under the experimental conditions, the average parasitism rate of A. pullata by parasitoids was as high as 11.0%. There may be a difference in parasitism rate between our experiment and field conditions. Natural enemies can continuously parasitize sawflies, so the parasitism rate for A. pullata sawflies in nature may be higher than it was in experimental conditions. Parasitoids play an important control effect on A. pullata populations in nature.
Hyperparasitism rates of Compsilura concinnata (Meigen) by trigonalid wasps were from 16.0% to 47.0% (Kellogg et al. 2003). Hyperparasitism rates of V. turrita by T. maga were from 0.2% to 1.6%. The biology and host relationships of T. maga need more investigation.
The populations and effectiveness in biological control of parasitoids are sometimes restrained by hyperparasitoids. Marietta carnesi (Howard) and Ablerus perspeciosus Girault are hyperparasitoids of Coccobius azumai Tachikawa, which was introduced from Japan to China (Huang 1994). Hyperparasitism can be as high as 16.0% in nature (Huang et al. 2005). The parasitoid Aphidius gifuensis Ashmead is a very important agent in biological control efforts against of Myzus persicae (Sulzer). However, hyperparasitoids have one of the main restriction factors in M. persicae control (Ren et al. 2000).
In biological control, introduced natural enemies may sometimes compete with native natural enemies, and weaken their overall control effect on target pests. Thus, future studies on host-parasite-hyperparasite interactions, and relations between foreign and native natural enemies, would provide valuable information needed for a successful biological control program against target pest insects, including the argid sawfly, A. pullata.