Effect of three species of host tree on the cold hardiness of overwintering larvae of Anoplophora glabripennis ( Coleoptera : Cerambycidae )

The Asian long-horned beetle, Anoplophora glabripennis, is a serious destructive pest of forests throughout China as it attacks a wide range of host plants. The effect of host trees on the cold hardiness of A. glabripennis larvae could be the basis for predicting the performance of this forest pest on different common hosts.To evaluate the effect of different species of host plant on the cold hardiness of overwintering larvae of A. glabripennis, we measured the supercooling point (SCP), fresh mass, protein content and concentrations of l ow molecular weight substances in overwintering larvae collected from three different host species (i.e., Populus opera, Populus tomentosa and Salix matsudana). Mean SCPs and protein contents of larvae from these three hosts differed signifi cantly. The SCPs and protein contents of the larvae collected from P. opera and P. tomentosa were signifi cantly higher than those collected from S. matsudana. The concentrations of glycerol, glucose and trehalose in overwintering larvae collected from these host species also differed signifi cantly, but there were no signifi cant differences in the concentrations of sorbitol and inositol. The larvae that were collected from S. matsudana had the highest concentrations of glycerol and trehalose and those from P. opera the lowest contents of glycerol, whereas those from P. tomentosa had the lowest concentrations of trehalose but the highest concentrations of glucose. Because of the signifi cant differences in the quantities of these biochemical substances in their bodies, the cold hardiness of overwintering larvae of A. glabripennis was signifi cantly dependent on the tree they fed on. These effects on the cold hardiness of the overwintering larvae might affect the selection of a host tree and therefore the spread of this beetle. * Corresponding authors; e-mails: zongsx@126.com, ouyangf@ioz.ac.cn INTRODUCTION The cold hardiness of insects is associated with genetically based traits and is also closely correlated with their physiological status, external environment, developmental stage and water content. Other factors also have important effects on cold resistance, including the composition of cold resistant materials, active ice-nucleation, host plant, temperature and humidity, and latitude, among others (Jing & Kang, 2002; Danks, 2006; Li et al., 2008). Before overwintering, phytophagous insects store energy by feeding on a wide variety of host plants (Sakurai et al., 1992; Zhong et al., 2009). After feeding on different host plants, many changes occur in the metabolic processes that affect the concentrations of glycogen, lipids, moisture, glycerol and low molecular weight substances, and these changes in physiological and metabolic processes affect the cold hardiness of insects (Hunter & McNeil, 1997). In different Eur. J. Entomol. 113: 212–216, 2016 doi: 10.14411/eje.2016.026


INTRODUCTION
The cold hardiness of insects is associated with genetically based traits and is also closely correlated with their physiological status, external environment, developmental stage and water content.Other factors also have important effects on cold resistance, including the composition of cold resistant materials, active ice-nucleation, host plant, temperature and humidity, and latitude, among others (Jing & Kang, 2002;Danks, 2006;Li et al., 2008).Before overwintering, phytophagous insects store energy by feeding on a wide variety of host plants (Sakurai et al., 1992;Zhong et al., 2009).After feeding on different host plants, many changes occur in the metabolic processes that affect the concentrations of glycogen, lipids, moisture, glycerol and low molecular weight substances, and these changes in physiological and metabolic processes affect the cold hardiness of insects (Hunter & McNeil, 1997).In different

Determination of SCP and FP
Before dete rmining the SCP the body te mperature of the larvae was rapidly reduced by placing them in a box full of an icewater mixture.Each larva was attached by parafi lm to the tip of a thermocouple, which was linked to an automatic data recorder (SCP measuri ng instrument).With the thermocouple attached and surrounded by absorbent cotton, each larva was placed into a freezing chamber (high-low temperature test chamber); the temperature in this chamber was reduced from 20°C to -30°C (1°C/ min); and the body tem perature of the larva recorded.A physical p henomenon, supercooling occurs when aqueous solutions (or various aqueous systems) remain liquid.The SCP is t h e instantaneous temperature recorded when the body fl uids of an insect begin to freeze.At this poi n t, the body temperat ure suddenly increases because of the emission of latent heat of crystallization; the SCP is the fl ex point and is recorded by the thermocouple and computer.The SCPs of 10-20 larvae collected from each species of host tree were measured.

Measurement of fresh mass
To measure the fresh weight of larvae from each species of host tree, 14-34 larvae were selected.The fresh mass of each larva was determined three times using an electronic balance (± 0.001 g).

Measurement of low molecular weight substances
The determination of low molecular weight substances was based on the method reported by Feng et al. (2014).From each of fi ve individuals from the same species of tree, 100 μL of hemolymph was extracted and mixed with 0.4 ml of 79% ethanol (including 10 g of erythritol as an internal standard) in 2 ml Eppendorf tubes.After centrifugation twice at 10,000 g for 5 min, the extracted supernatant was stored at −20°C until analyzed.In the pre-analytical phase, the sample liquids were evaporated to dryness using nitrogen at 40°C.The dried samples were dissolved in 1.0 ml of pyridine.A mixture of hexamethyldisilazane and chlorotrimethylsilane was added to an ice-water bath, which stood for 30 min at 20°C.After high-speed centrifugation, 1.0 μl of the supernatant was directly injected into a Meteorological Chromatography Detector (Agilent 7890 GC, America) to analyze the components.The components and the concentrations of the low molecular weight substances were determined based on comparisons with the peak times and areas of fi ve different concentrations of a standard that ranged from 0 to 1.0 μg/μl in incremental steps of 0.2 μg/μl.

Measurement of protein content
From the larvae, 10 μl of hemolymph fl uid was extracted and mixed with 10 μl of PBS in Eppendorf tubes.The supernatant was collected after centrifugation at 3,000 g for 10 min.To measure protein content, 2 ul of protein extract was analyzed in an ND2000 (NanoDrop 2000, USA) at 595 nm.For each group this was replicated three times.

Statistical analyses
The statistical analyses of all data were performed using SPSS statistical software package version 18.0 for Windows.One-way analysis of variance (ANOVA) and least signifi cant difference tests (LSD, α = 0.05) were used to test the differences in SCP, fresh mass, protein content and concentrations of low molecular weight substances in the larvae collected from the three species of host tree.Linear regression was used to test the relationships between SCP and the physiological and biochemical substances in the overwintering larvae.tant pests of forests throughout China (Qin & Jin, 1959).The larvae bore into and feed on the trunk of the tree causing serious damage to the conducting tissues of the host, which results in the death of large branches.The beetles can kill trees, which eventually leads to a reduction in the productivity of forested lands and enormous economic loss (Hu et al., 2009).In China, the life cycle of A. glabripennis is one to two years long.The larvae overwinter inside the trunks of trees in most areas, although eggs or pupae overwinter in a few areas (Luo & Li, 1999).Currently, this beetle is recorded feeding on 18 genera of plants, including Populus, Salix, Ulmus and Acer (Luo & Li, 1999).The 3 species of trees, Populus opera, Populus tomentosa and Salix matsudana are the most common hosts and widely distributed.In addition, the damage caused by A. glabripennis to these three host trees has changed in recent years.The ability of the larvae to overwinter in different hosts is an important factor that affects A. glabripennis populations the following year.The effects of different species of host tree on the cold hardiness of overwintering larvae of A. g labripennis, however, remain to be investigated.
In this paper, the focus of the research is on larvae of A. glabripennis that fed and overwintered on P. opera, P. tomentosa and S. matsudana, respectively.The SCP, fresh mass, protein content and concentrations of low molecular weight substances in overwintering larvae were measured and used to compare their cold hardiness.Understanding the effect that different host plants have on the overwintering of A. glabripennis larvae will help determine the cause of outbreaks and how they can be controlled, and provide a solid foundation for the effective control of this beetle.

Insect
In July 2012, P. opera (mean height and diameter approximately 9 m and 10 cm, respectively) and S. matsudana (approximately 5 m and 8 cm, respectively) were used to raise A. glabripennis in the fi eld.These trees were in a farmland shelterbelt in Lingwu City (38°19´N, 106°26´E), Ningxia Hui Autonomous Region.The trunks of the host trees were ringed with a wire cage (approximately 1.0-1.5 m) approximately 1.5-3 m above the ground.The adults were fi rst collected in the fi eld in Yanchi City in the middle of July and placed inside the wire cage on each trunk in the proportion of 1 : 1 male to female.Approximately 20-30 pairs of A. glabripennis were caged on each tree to lay eggs and produce offspring.The larvae that had fed on P. tomentosa were collected from a natural P. tomentosa forest in Yanchi City (37°78´N, 107°22´E), Ningxia Hui Autonomous Region.The different species of brood trees were transported to a laboratory at Beijing Forestry University during January 2014.The overwintering larvae were removed from the brood trees and maintained in 10 ml transparent test tubes at outdoor temperatures.Holes were made in the walls of the tubes using an electric iron to ensure an adequate air supply.The larvae collected from each host species were the ones used in this study.
The head widths of larvae collected from P. opera ranged between 3.80 and 5.08 mm, from P. tomentosa between 3.88 and 4.88 and from S. matsudana between 3.37 and 4.88 mm.As previously it was the head widths of fourth and fi fth-instar overwintering larvae that were measur (He & Huang, 1992).

SCP
The SCPs of the overwintering larvae that were collected from the three species of host tree differed signifi cantly (F = 8.92; df = 2, 39; P = 0.001; Table 1).The larvae collected from S. matsudana had the lowest SCP (−16.02 ± 1.08°C) and those collected from P. opera the highest SCP, both of which were signifi cantly higher than that of the larvae from the other host tree (P 1 = 0.029; P 2 < 0.001).The SCPs of larvae col lected from the three species of host trees differed from one another, and the difference between the maximum and the minimum value was 8°C.In general, the supercooling capacity of overwintering larvae from the three trees was ranked as follows: P. opera < P. tomentosa < S. matsudana.

Fresh mass
In terms of fresh mass the overwintering larvae from the three species of trees were signifi cantly different (F = 4.15; df = 2, 61; P = 0.021; Table 2).The fresh mass of overwintering larvae collected from P. opera was signifi cantly lower than that of those from the other two host species (P 1 = 0.012; P 2 = 0.008).However, there were no signifi cant differences in the fresh mass of the larvae from P. tomentosa and S. matsudana, which had a fresh mass of approximately 1.55 g.

Concentration of low molecular weight substances
Glycerol, glucose, sorbitol, trehalose and inositol were identifi ed in the hemolymph of overwintering larvae.The concentrations of glycerol (F = 6.51; df = 2, 8; P = 0.030), glucose (F = 5.93; df = 2, 8; P = 0.004) and trehalose (F = 5.21; df = 2, 8; P = 0.049) in the hemolymph of overwintering larvae from the three species of trees differed signifi cantly (Table 3).The concentrations of glycerol and trehalose in the overwintering larvae from S. matsudana were the highest, whereas the glycerol content of those from P. opera (26.58 ± 1.18 μg/μl) was signifi cantly lower than that of those from P. tomentosa and S. matsudana (P 1 = 0.036; P 2 = 0.014).The concentration of glucose in the overwintering larvae from P. tomentosa was signifi cantly higher than that of those from the other two species of trees (P < 0.05).Nevertheless, the trehalose content of the overwintering larvae from P. tomentosa was signifi cantly lower than that of those from P. opera and S. matsudana (P 1 = 0.032; P 2 = 0.031).There were no signifi cant differences in the concentrations of sorbitol and inositol in the larvae collected from the three species of trees.

Protein content
The protein content of overwintering larvae collected from the three trees was signifi cantly different (F = 6.25; df = 2, 8; P = 0.034; Table 4).The concentration of protein in the overwintering larvae from S. matsudana (96.87 ± 6.80 mg/ml) was signifi cantly lower than that of those from P. opera and P. tomentosa (P 1 = 0.015; P 2 = 0.043).The content of protein in the larvae from P. opera was higher than that of those from P. tomentosa, but not signifi cantly so.

DISCUSSION
The supercooling point is an important indicator of cold hardiness in insects and is also considered to be the lowest temperature at which an insect can survive.However, large numbers of some insects die at temperatures above their average supercooling point (Masahiko & Kazuhiro, 1998;Bale & Hayward, 2010), whereas other insects can survive at temperatures below their average supercooling points (Baust & Rojas, 1985).The SCPs of the larvae from the different species of hosts differed signifi cantly, which is consistent with previous results, e.g., for Carposina niponensis, Bactrocera dorsalis and Hemiberlesia pitysophila (Ren et al., 2006;Zhong et al., 2009;Wang et al., 2011).Therefore, the species of host tree had a signifi cant effect on the SCP of overwintering larvae of A. glabripennis.The importance of the host in determining the cold resistance of insects is also revealed by host transfer studies; for example, the SCP of Aphis fabae decreases when it switches from feeding on beans in summer to feeding on euonymus in autumn and spring (Gash & Bale, 1985).The cold hardiness of A. glabripennis larvae collected in different geographical regions differs (Feng et al., 2014), which might be due to the host plants and temperatures differing in the different regions.Thus, we hypothesized that a primary reason for the differences in population dynamics of A. glabripennis recorded in different areas might be due to their selecting different species of host tree.Based on fi eld studies, the degree of damage to S. matsudana caused by A. glabripennis is more severe than that to P. opera.Moreover, following fi eld inoculation, the survival of larvae feeding on S. matsudana was higher than that of those feeding on P. tomentosa.The results of this study indicate that the supercooling capability of the overwintering larvae of A. glabripennis that fed on S. matsudana was higher than that of those that fed on P. opera and P. tomentosa.Thus, we speculated that this increase in the supercooling capacity might affect host selection by A. glabripennis.Host plant is a major factor determining the accumulation of glycerol and small molecule cryoprotectants, especially in polyphagous insects; which affects the metabolism associated with cold hardiness (Hunter & McNeil, 1997;Zvereva, 2002).Based on our research, the larvae of A. glabripennis that fed on different species of host trees differed signifi cantly in their protein contents, concentrations of low molecular weight substances and supercooling capacity.The larvae that came from S. matsudana had a high supercooling capacity, high total concentrations of low molecular weight substances and low contents of protein.By contrast, the larvae from P. opera had a low supercooling capacity, low total concentrations of low molecular weight substances and high contents of protein.These differences may be related to the nutritive quality and levels of secondary substances in the different species of host plant.
In addition, low molecular weight sugars and polyols, including glycerol and trehalose, act as cryo-protective agents in many species of insects during overwintering (Storey & Storey, 1991;Koštál et al., 2007;Liu et al., 2007Liu et al., , 2009)).In this study, larvae from S. matsudana had the highest concentrations of glycerol and trehalose, and lowest SCP.A previous study reports that supercooling is associated with host plants and the concentration of glycerol in H. armigera (Liu et al., 2007).Thus, it is likely that the concentration of glycerol and trehalose are associated with cold resistance.This needs to be confi rmed in future studies.Furthermore, the growth and development of Heliothis armigera that feed on different species of host trees differ signifi cantly, which refl ects the adaptive strategy of the insect in which the host plants that provide adequate nutrition for insect development also enhance their resistance to adverse environments (Tan & Zhao, 1990).The results of our study are consistent with previous research in indicating that larvae that came from P. opera were lower in body weight, and therefore, this host is not a good host and is likely to result in the development of small adults with a low fecundity.
Based on this study, the species of host tree signifi cantly affected the overwintering larvae of A. glabripennis.We only recorded the supercooling points and some of the substances in the overwintering larvae.The associations between survival and the low temperature limits of larvae from different host trees and their cold hardiness strategies are still unclear, mainly because many factors affect the cold hardiness of overwintering larvae of A. glabripennis, such as the nutritional status of the host, developmental stage of the insect and its content of certain substances, among others.Further research is required to clarify how these factors affect the cold hardiness of overwintering larvae and determine the connection between the host plants and the mechanisms of cold resistance.By understanding how P. opera, P. tomentosa, S. matsudana and A. glabripennis interact, we can better understand the cause of outbreaks and forecast the population dynamics of this pest insect in different landscapes in the future.

Table 1 .
Supercooling point (°C) of overwintering larvae collected from three species of trees.

Table 2 .
Fresh mass (g) of overwintering larvae collected from three species of trees.

Table 3 .
Concentration (μg/μl) of low molecular weight substances in the hemolymph of overwintering larvae collected from three species of trees.
Values are means ± SE.Values followed by different letters within a column are signifi cantly different (P < 0.05).

Table 4 .
Protein content (mg/ml) of overwintering larvae collected from three species of trees.