Parasitoids of the eucalyptus gall wasp Leptocybe invasa (Hymenoptera: Eulophidae) in China

Leptocybe invasa Fisher & La Salle (Hymenoptera, Eulophidae) is an invasive pest in Eucalyptus plantations throughout the world. Potential biological control agents for L. invasa were investigated in the Fujian, Guangdong, Hainan, Guangxi, Jiangxi, and Sichuan provinces of China, where Eucalyptus spp. have been severely damaged by the eucalyptus gall wasp. Three hymenopteran parasitoids of L. invasa were identified: Quadrastichus mendeli Kim & La Salle (Eulophidae), Aprostocetus causalis La Salle & Wu (Eulophidae), and Megastigmus viggianii Narendran & Sureshan (Torymidae); M. viggianii is newly recorded in China. The percentages of parasitization by Q. mendeli, A. causalis, and M. viggianii were 2.96%–19.53%, 2.30%–26.38%, and 24.93%, respectively. The longevity and body length of females were significantly greater than for males in A. causalis and M. viggianii. No males of Q. mendeli were found in China. These parasitoids could be used as biological agents for L. invasa in China.


Introduction
Eucalyptus is one of the three major fast-growing tree species worldwide, which plays important roles in reforestation and the production of timber, pulp, potential bioenergy feedstock, and other forest products [25]. In China, the cultivated area under eucalyptus covers more than 3.68 billion hectares and produces a direct economic income that exceeds 100 billion Renminbi (RMB) [33]. However, the decline of ecosystem biodiversity is very obvious with the increased cultivated area of eucalyptus and results in a sharp rise of eucalyptus insect pests [22]. The number of eucalyptus insect pest species in China has increased from 53 in 1980 to 319 in 2011 and causes direct economic losses exceeding RMB 1.125 billion annually [23].
The eucalyptus gall wasp, Leptocybe invasa Fisher & La Salle (Hymenoptera: Eulophidae), originating from Australia, is a global pest in Eucalyptus plantations [18]. A recent study based on molecular and phylogenetic analyses suggested the occurrence of geographical variability in L. invasa populations and the existence of different putative species, among them a ''Chinese lineage'' [20]. The wasp populations investigated in our study were not characterized from a phylogenetic point of view; we therefore cannot indicate their exact taxonomical position. Moreover, as the lineage is not a taxonomic category coded by the International Code of Zoological Nomenclature, in the present paper the wasp is cited as L. invasa. Leptocybe invasa has expanded to more than 29 countries in Asia, Europe, Africa, and the Americas [18,35]. In China, L. invasa was first found in the Guangxi Zhuang Autonomous Region in April 2007 [30]. Subsequently, the pest has spread to Guangdong, Fujian, Hainan, Jiangxi, and Sichuan provinces [3,29,31,34]. Various management strategies have been explored to control L. invasa, including chemical control [13], breeding and the selection of resistant planting stock [5,34], and biological control [15,16]. However, chemical control is not widely accepted due to its varying success, negative effects on biodiversity, and environmental pollution. Sylvicultural control is largely ad hoc and is unlikely to represent a viable long-term solution against an increasing number and diverse range of damaging invasive pests. Biological control is considered an attractive alternative to other control methods due to its ecological and economic benefits [4].
In Australia, parasitoids play a very important role in limiting the populations of L. invasa [8,15]. The introduction of natural enemies from Australia has been considered an optimal way to control the eucalyptus gall wasp in epidemic areas [15,28]. However, only a few countries have adopted this method in view of increasing evidence of attacks against non-target hosts and the resulting threat to native biodiversity [15,28]. Recently, several L. invasa parasitoids have been found in the invaded regions, e.g., India, Israel, Turkey, Italy, Sri Lanka, Thailand, Argentina, and South Africa [7-9, 12-15, 21]. However, as mentioned above, the parasitic capacities of these parasitoids for L. invasa are different in these regions. Thus far, only Aprostocetus causalis La Salle & Wu (Hymenoptera: Eulophidae) and Quadrastichus mendeli Kim & La Salle (Hymenoptera: Eulophidae) have been reported to parasitize L. invasa in the Guangxi and Hainan provinces of China [10,17,32]. The number of parasitoid species of L. invasa and their parasitic capacities in the field are unknown. Therefore, it is necessary to widely investigate biological control agents for L. invasa in China.
The purpose of this study was to identify possible biological control agents for L. invasa occurring on Eucalyptus spp. In this study, we investigated the species of parasitoids present in some Chinese regions.

Materials and methods
Eucalyptus gall wasps were searched for by the typical bump-shaped galls they form on leaf midribs, petioles, and stems ( Fig. 1). Branches of DH 201-2 (Eucalyptus grandis · E. tereticornis) (Myrtales: Myrtaceae), E. tereticornis Smith, E. Exserta L., and E. grandis · E. urophylla damaged by L. invasa were collected from Fujian, Guangdong, Hainan, Guangxi, Jiangxi, and Sichuan provinces from 2015 to 2016. The sampling sites and sampling times for each province are shown in Table 1. Branches were placed in a glass container filled with water to retain freshness and transferred to a sealed net cage (40 cm · 40 cm · 80 cm) at 27 ± 1°C (the average air temperature of the sampling sites during the period of collection) with an L16:D8 photoperiod and 70-80% relative humidity to prevent the adults from escaping. The water in the glass container was replaced daily until the emergence of L. invasa and their parasitoids.
The emerged L. invasa adults ( Fig. 2) and their parasitoids were collected daily using 50 mL plastic tubes. The percentage of parasitization for each parasitoid collected from different geographical populations was calculated as the number of emerged parasitoids (EP) divided by the sum of the total numbers of emerged gall-formers and emerged parasitoids (EGP) [15,16].
A cotton ball soaked in a 10% sucrose solution and galled eucalyptus branches were supplied to allow oviposition by the parasitoid adults. Newly emerged female wasps were reared one by one for uniparental species and in pairs for biparental species. The honey-water and galled branches were renewed daily. The rearing conditions of these parasitoids were similar to those of the host insects mentioned above. The mortality of the male and female specimens was recorded daily to evaluate the longevity of the parasitoids. The body lengths of these dead specimens were subsequently measured using image-measuring software (Leica Application Suite version 4.6.0, Leica Microsystems, Germany). Images of the parasitoid adults were taken with a Sony digital camera (DSC-HX60, Sony, Kyoto, Japan). Identification of the parasitoids was performed with keys [7,15,32] and confirmed by Prof. Chao-Dong Zhu and Dr. Huan-Xi Cao (Institute of Zoology, Chinese Academy of Sciences, China).
Statistical analysis was performed using SPSS 16.0 (SPSS, Chicago, IL, USA). Adult longevities and body lengths were compared using the nonparametric Mann-Whitney U test. The results were considered significant at p 0.05.

Discussion
Quadrastichus mendeli is one of the indigenous parasitoids parasitizing L. invasa in Australia [15], with the percentage of parasitization varying from 7.9% to 95.6%. Recently, Q. mendeli was introduced from Australia to Israel and India as a biological control agent to limit the severity of damage caused by L. invasa. The parasitoid is now successfully established in Israel and India, and the percentage of parasitization was 73% in Israel and 81.74-94.03% in India [15,28]. However, initial efforts to establish the Q. mendeli population in quarantine facilities in South Africa and Kenya have failed [6]. Interestingly, Q. mendeli was collected from L. invasa galls on E. Camaldulensis Dehnh. in Italy beginning in 2013, although this parasitoid was never officially released [21]. Field data showed that mean parasitization percentage by Q. mendeli varied from 30.2% to 50.5% in Italy [21]. Here, we report Q. mendeli parasitizing L. invasa in China under natural conditions, although this parasitoid was never officially released in China. Although Q. mendeli is a widely distributed L. invasa parasitoid in China, the percentage of parasitization differed in these geographical populations (Table 1).
Aprostocetus causalis, a parasitoid of L. invasa in China, was described as a new species in 2014 [32]. However, the longevity of both sexes of this parasitoid in the laboratory was lower than the longevity found in a previous study carried out in Thailand [27], probably due to the different rearing method. Adults in the laboratory in Thailand were fed with a honey solution, whereas both a honey solution and galled foliage were provided in this study. Previous studies suggested that parasitization involves a high physiological cost for the parasitoid and that longevity is lower in ovipositing females [24,26].
Megastigmus viggianii was recorded as parasitizing the bud galls of Calycopteris floribunda Lamark (Myrtales: Combretaceae) in India [19]. In 2008, this parasitoid was found parasitizing L. invasa larvae for the first time in India [11], and the percentage of parasitization varied from 14.29% to 31.82% [28]. However, M. viggianii was only found in the Sichuan site; thus, we cannot ascertain whether this parasitoid is native to China. Furthermore, M. viggianii has a multi-host-range (i.e., C. floribunda and L. invasa) compared to Q. mendeli and A. causalis. Other M. viggianii hosts have not been found in China and need further investigation.
In the field, two or three parasitoids attacked L. invasa in China with possible competition for the monopolization of host resources. The percentage of parasitization by Q. mendeli was always higher than that by A. causalis when the two parasitoids attacked the eucalyptus gall wasp in Guangdong, Hainan, and Guangxi provinces. However, the percentage of parasitization by Q. mendeli was lower than those for A. causalis and M. viggianii when the three parasitoids attacked eucalyptus gall wasps in Sichuan province. We presumed that this could be attributed to the population dynamics of Q. mendeli in Sichuan province, but this needs to be confirmed. Previous studies showed that Q. mendeli is a solitary idiobiont ectoparasitoid [15] while A. causalis and M. viggianii are solitary koinobiont endoparasitoids [32]. The competitive dynamics and mechanisms of both the ectoparasitoid (Q. mendeli) and endoparasitoids (A. causalis and/or M. viggianii) in L. invasa need to be studied.
In general, parasitoids have host-tracked their hosts either simultaneously or following invasion by the pest. For example, Ophelimus maskelli Ashmead (Hymenoptera: Eulophidae) is a leaf-gall-inducing invasive pest of several Eucalyptus species in California, while its parasitoid, Closterocerus chamaeleon Girault (Hymenoptera: Eulophidae), was also found less than 1 year after Q. maskelli [1,2]. As Q. mendeli and M. viggianii were never released in China, their record leads us to suppose that they could shortly reach other neighboring countries around China.

Conclusion
In this study, we found that Q. mendeli, A. causalis, and M. viggianii could be used as biological control agents for L. invasa in China. However, further studies are needed on the biological characteristics, mass production and release, competitive dynamics and mechanisms of these parasitoids.

Conflict of interest
The authors declare no conflict of interest in relation with this paper.