Hyphantria cunea(Drury)is a serious invasive defoliator pest in China. It originally distributed in North America, ranging from Canada to Mexico, and was one of the few insect pests introduced from North America into other continents. Introduced to what was formerly Yugoslavia in the 1940s(firstly recorded in 1949)(Wagner, 2005; Douce, 2003), it now has occupied probably its entire range in Europe from France to the Caspian Sea in the east, as well as penetrated into Central Asia: Turkmenistan(from 1990—1993), Uzbekistan(Fergana Valley from 1996—1997), Kyrghyzstan, and southeastern Kazakhstan. It was also introduced into Japan in 1945, and has adjusted its number of generations per year since its arrival(Gomi et al., 1996).

It was firstly discovered in 1979 at D and ong City, Liaoning Province(FAO, 2007; Zhang, 2008). Since then, H. cunea began to spread into Northern China. At present, it had spread to Liaoning, Sh and ong, Henan, Shaanxi, Hebei, Tianjin, Beijing and other provinces(Liu et al., 2008; Liu et al., 2005; Takeda, 2005; Yang et al., 2008). Frequent outbreaks occurred in these areas and great economic losses have been caused subsequently. As an important invasive pest, it usually lives and attacks cultivated plants, particularly the ornamentals, man-made forests and fruit trees around residential areas, towns and cities(Yang et al., 2007; Sullivan et al., 2011). The pest has become a big problem to the planting projects in northern China.

To date, insecticide use was still the most widely practiced management method and played an important role in the control of H. cunea. Although some effects had been achieved, insecticides also caused the pollution of the environment, killed a large number of natural enemies.

Under the increasing concern over pesticide efficacy and safety to humans and the environment, alternative control options need to be fully evaluated. Under natural conditions, fungi are frequent and important mortality factors in the insect populations. Entomopathogenic fungi have shown their great potential to control insect pests both in the field and under greenhouse conditions(Inglis et al., 2001; Franci et al., 2012; Butt et al., 2001). Beauveria bassiana is one of the most common pathogens infecting insects. It has long been recognized as a potential biocontrol agent and was actively being developed to control various pest insects(Zibaee et al., 2013; Edison et al., 2006; Özdikmen et al., 2004; Aurelien et al., 2004; Eken et al., 2006; Liu et al., 2007; Lu et al., 2008; Wang et al., 2007; Qin et al., 2012; Liu et al., 2013; Bextine et al., 2002).

There was significant difference in the mortality between the different strains of B. bassiana. And an obvious host specificity exited in different strains. So It was feasible to enhance the specialization and the pathogenicity of B. bassiana to some kind of pest through artificial orientation training(Guo et al., 2010). In this experiment, the pathogenicity of some isolates from different hosts or habitats to larvae of H. cunea was studied, and a high virulent strain was screened through laboratory and field experiments.

The eggs of H. cunea were collected in June of 2012 from the field in Baoding, Hebei. Larvae hatched from the eggs in the laboratory were bred in the artificial climate chamber at a fluctuating temperature of 25 to 28 ℃, humidity 80%-85% and a photoperiod of L: D (12: 12)to 5-instar larvae with leaves of Populus tomentosa which had been surface sterilized.

Fungi used in the experiments(Tab. 1)were obtained from the Forest Pathology Laboratory of Hebei Agricultural University. BH01 was isolated from larva’s cadaver of H. cunea obtained from Qinhuangdao of Hebei province. Cadavers of H. cunea seemed to be fungal diseases were washed in a solution of 1% sodium hypochlorite for one minute and twice in sterile distilled water, for one minute each time. Then they were dried on sterile filter paper. After drying, they were transferred to PDA plates and incubated at 25 ℃. The fungi were identified by microscopically inspecting of the sporulating structures and conidial morphology. BI05 was isolated from a larva’s cadaver of Apriona germari from Dingzhou County, Baoding City, Hebei, China(Li et al., 2007). BS04, BS05 and BS08 were bated from soil used the old larvae of Tenebrio molitor as bate insects(Li et al., 2006).

Tab.1 Isolates of B. bassiana for the experiments

Tab.1 Isolates of B. bassiana for the experiments Isolates Host species or habitats Locality Collect date Viability(%)(mean±SE) BS04 Soil Yixian, Baoding 2005-06 90.39 ±3.21 BS05 Soil Baoding 2005-06 96.10 ±1.89 BS08 Soil Yixian, Baoding 2005-06 95.29 ±1.78 BH01 H. cunea Qinhuangdao 2010-09 95.23 ±2.91 BI05 A. germari Dingzhou, Baoding 2003-07 91.09 ±2.99

The fungi were grown on PDA plates in Petri dishes and maintained at 25 ℃. Conidia were harvested from 10 days old surface cultures directly by scraping. Spore suspensions were made by suspending conidia in 20 mL sterile distilled water in little beaker containing 0.05% Tween-80. Beakers were agitated on a vibrant shaker for 10 min to produce a homogenous conidial suspension. The spore concentrations were then adjusted to defined concentrations using a Neubauer hemocytometer. Viability of the conidia was checked by a germination test prior to the experiment and assured to be >90% for all isolates.

The mortalities of larvae inoculated with the B. bassiana isolates showed in Tab. 1 were compared at 1×10⁸ conidia·mL^-1. Larvae of H. cunea were sprayed with conidial suspension using POTTER spray tower, drained of excess suspension by filter paper. The sterile distilled water containing 0.05% Tween-80 was used as the control.

After treatment, insects were transferred individually into a transparent plastic vial and fed on leaves of Pinus tomentosa which had been surface sterilized. The leaves were changed daily. Mortality was recorded daily for eight days. Dead insects were surface sterilized and transferred into a Petri dish lined with moistened filter paper. The mortality due to fungi was confirmed by microscopic examination of hyphae and spores on the surface of the dead insect. All test insects were maintained in the artificial climate chamber as described above. For one treatment, 50 insects and 3 mL conidial suspension were used.There were four replicates for one isolates.

Data from the experiment were subjected to variance analysis(ANOVA), and differences are presented by the results of LSD multiple range test. The LT₅₀ value was calculated using the probit procedure of the SPSS statistical package.

For the highly virulent isolates identified in the above bioassay, a further investigation was designed. To determine the lethal concentration(LC50)of the selected highly virulent isolates, larvae were sprayed with five concentrations of conidia(1×10⁴, 1×10⁵, 1×10⁶, 1×10⁷ and 1×10⁸ conidia·mL^-1).The control was the same solution without the fungus. For each conidial concentration, 50 larvae were used. The insects were inoculated and reared as described in 2.4, and dead insects were treated as described above to confirm the cause of death. The LC₅₀ was calculated using of the probit procedure of the SPSS statistical package.

The field experiments were carried out in October of 2012 and 2013, respectively in Baoding. At this time, H. cunea was at the mature larval stage and would crawl down from the tree to pupate. Three treatments, including 8×10⁸ conidia·mL^-1 of B. bassiana BH01, 4.5% β-cypermethrin 2 000 times liquid and the water control were arranged in the experiment. There were 4 replicates for each treatment, and 3 trees for one replicate. In this test, Taishan ft-796 sprayer was used to routinely spray. A thin film was laid under the tree to catch the drops of dead H. cunea. The dead and live insects on the tree and film were counted at 3, 5 and 7 d post experiment respectively. The control effect was calculated using the following formula:

Mortalities and LT50 of H. cunea larvae inoculated with the tested isolates were shown in Tab. 2.

Tab.2 Mortalities and LT₅₀ of H. cunea larvae treated with various isolates of B. bassiana in the laboratory^①

Tab.2 Mortalities and LT50 of H. cunea larvae treated with various isolates of B. bassiana in the laboratory① Isolates tested Percent mortality(corrected)/% LT 50(95% fiducial limits)/d 1 2 3 4 5 6 7 8 BS04 0 0 0 0 10.42 14.89 25.53 32.61c 8.524(7.830-9.654)a BS05 0 2 0 8.33 14.58 21.28 23.40 30.43c 8.80(7.926-10.279)a BS08 0 4 8.16 22.92 31.25 55.32 65.96 71.74b 5.939(5.582-6.342)b BH01 0 8 12.24 33.33 56.25 76.60 80.85 84.78a 4.94(4.627-5.259)c BI05 0 4 12.24 43.75 50.00 57.45 70.21 71.75b 5.44(4.67-6.36)bc CK 0 0 2 4 4 6 6 8 ①Values followed by the same letter in the same column are not significantly different by LSD(P=0.05).The same below.

All tested isolates were significantly different in their virulence against H. cunea. The isolate BH01 was the most virulent towards H. cunea, causing approximately 84.78% mortality eight days post-infection. This isolate was followed by isolates BS08 and BI05 that resulted in 71.74% and 71.75 mortality. And BS04 and BS05 caused only 32.61%, 30.43% mortality, respectively. For the control, cumulative larval mortality was 8% after eight days, and no fungi were observed on the dead larvae.

LT₅₀ values differed significantly among the isolates when applied at a concentration of 1×10⁸ conidia·mL^-1(Tab.2). The LT₅₀ values of B. bassiana BH01, BS08 and BI05 were 4.94 d, 5.939 d and 5.44 d, respectively, and shorter significantly than that of the other two isolates.

Based on the above bioassays, the isolates BH01, BS08 and BI05 were selected to determine the LC₅₀value. By dose-mortality analyses, the mortality of H. cunea larvae differed significantly when inoculated with different concentration of B. bassiana. The LC₅₀ values of three isolates were shown in Tab. 3.

Tab.3 LC₅₀ values of isolates tested against larvae H. cunea eight days after treatment in laboratory experiment

Tab.3 LC50 values of isolates tested against larvae H. cunea eight days after treatment in laboratory experiment Isolates tested LC 50(95% fiducial limits)/(conidia·mL -1) Dose regression equation(correlation coefficient) BS08 1.34 × 10 7(5.26 × 10 6-4.86 × 10 7)b Y=-3.203 + 0.450X(0.942 4) BH01 1.39 × 10 6(7.58 × 10 5-2.55 × 10 6)c Y=-4.377 + 7.13X(0.927 8) BI05 2.11 × 10 7(7.23 × 10 6-1.06 × 10 8)a Y=-2.940 + 0.401X(0.952 1)

The LC50 value for isolate BH01 was 1.39 × 10⁶ conidia·mL^-1, proved to be the most virulent against 5-instar H. cunea. The LC₅₀ values of the other two isolates BS08 and BI05 were higher than that of isolate BH01 significantly.

The results in the field test(Tab. 4)showed that 4.5% β-cypermethrin could kill the larvae of H. cunea quickly. The control effect of 4.5% β-cypermethrin to H. cunea reached to 89.43% 3 d after treatment, which has no significant difference with that at 5 d and 7 d, 94.14% and 93.29% respectively. To BH01, the rate of control H. cunea was noticeably slower than that of 4.5% β-cypermethrin, and the control effect increased gradually with time. The control effect to H. cunea were 38.21% and 56.98% respectively at 3 d and 5 d post control, which were significantly lower than that of 4.5% β-cypermethrin. But till seven days after control, the control effect of BH01 increased to 88.84%, which was no significant different with 4.5% β-cypermethrin.

Tab.4 The control effect in field experiment of B. bassiana BH01^①

Tab.4 The control effect in field experiment of B. bassiana BH01① Control effect(%) 3d 5d 7d 4.5% β-cypermethrin 89.43a 94.14a 93.29a BH01 38.21a 56.98b 88.84c ①Values followed by the same letter in the same line are not significantly different by LSD(P=0.05).

In conclusion, our laboratory and field experiments suggest that B. bassiana BH01 can provide an effective and acceptable level of control against larvae of H. cunea at concentration of 10⁸ conidia·mL^-1. But the control effect to egg, pupa and adult are needed because of the generation overlap in the life cycle of H. cunea. And before the use in the field, other fungal characteristics such as spore production, germination, hyphal growth rates and effects varying environmental conditions must also be evaluated. Further research is currently being conducted along these lines.

In the bioassay, the common method was immersion and spraying. But H. cunea was difficult to soak by water because of the overabundance of chaeta. And the amount of spore suspension remained among the chaeta after immersion was uncontrollable, which made inter-individual difference of inoculum. So in the test, a spray tower was used to lower this difference.

In northern China, the weather was very dry. H. cunea, a sun-loving insect, occurred mainly in the space where there was good light. The space was unfavourable to the survival and germination of spores of B. bassiana because of the dry, hot and intense ultraviolet. So H. cunea in the leaf feeding period can not be infected easily by B. bassiana. Furthermore, H. cunea in the high crown was often difficult to contact with B. bassiana limited by spraying equipment. The mature larvae of H. cunea would crawl down from the tree to search pupation site which mainly occurred under such surface coverage as rubble masonry, plant litter and surface deposit, etc. This process provides a big advantage for contact of B. bassiana with H. cunea. So control H. cunea using B. bassiana in mature larvae stage was more feasibility than in the other stages.

The mature larvae of H. cunea were selected as the test insects in this study. The results showed that controlling mature larvae of H. cunea using B. bassiana was a feasible control approach. But to avoid the effect of hot, dry and ultraviolet on the B. bassiana, it was better to spray at the peak of pupation.