Effects of three host pine species on diapause induction and life-history traits of Monochamus alternatus alternatus (Coleoptera: Cerambycidae)

ABSTRACT

Monochamus alternatus alternatus Hope is distributed in China and Taiwan and transmits the pinewood nematode, the causative agent of pine wilt disease. Unlike univoltine and semivoltine Monochamus alternatus endai Makihara native to Japan, M. a. alternatus has two generations a year (bivoltinism). Some polyphagous insect herbivores exhibit that the incidence of diapause is affected by the host plant species on which they feed, resulting in a host plant dependent variation in the number of generations a year (voltinism). For evaluating the effects of alien pine species on the colonization of M. a. alternatus, its larvae were reared on bolts of Pinus densiflora and P. thunbergii native to Japan and P. echinata native to North America in the laboratory. This study indicated that the three pine species did not increase the incidence of larval diapause to 1.0, suggesting that any pine species did not change the voltinsim. It also determined that the pine species did not influence the ovariole number of females that forwent diapause. The impact of feeding on P. densiflora and P. thunbergii on adult body mass was variable and there was no difference in adult body mass between P. densiflora and P. echinata. Thus, this study suggested that the voltinism of M. a. alternatus was not altered by some pine species native to unoccupied regions such as Japan and North America, i.e. its reproductive rate per year did not decrease owing to reduced number of generations per year, when it arrived there and then colonized.

KEYWORDS:

• Biological invasion
• host-plant effects
• Pinus densiflora
• Pinus echinata
• Pinus thunbergii

Introduction

Diapause is physiologically arrested development and controlled neuro-hormonally in insects (Tauber et al. 1986; Danks 1987). It is separated into facultative and obligate diapause. Facultative diapause is induced by environmental conditions, primarily by photoperiodic and temperature conditions, while obligate diapause is necessarily induced regardless of environmental conditions (Tauber et al. 1986; Danks 1987). When polyphagous insect herbivores exhibit facultative diapause, the induction, duration, or termination of diapause can be affected by the host plant species on which they feed. This can cause a host plant dependent variation in the number of generations a year (voltinism) (e.g. Hunter and McNeil 1997; Wedell et al. 1997; Tanzubil et al. 2000; Liu et al. 2010). In that case, the size of overwintering population may be closely related with the survival and fecundity of nondiapause individuals, which is sometimes affected by host plant species (e.g. Haack et al. 2010; Togashi 2021).

Monochamus alternatus alternatus Hope (Coleoptera: Cerambycidae) is distributed in China and Taiwan and the primary vector of the pinewood nematode, Bursaphelenchus xylophilus (Steiner et Buhrer) Nickle, which causes pine wilt disease and has been devastating pine forests in East Asia and westernmost Europe (Makihara 2004; Akbulut et al. 2017). Unlike univoltine and semivoltine Monochamus alternatus endai Makihara native to Japan, this insect has two generations a year (bivoltinism) because of facultative diapause (Togashi 1989, 2019a, 2019b; Enda and Makihara 2006; Togashi and Toki 2018); All larvae enter diapause under a photoperiodic regime of 12 h light and 12 h dark (LD 12:12 h) at 23°C, whereas some larvae forgo diapause under photoperiods between LD 0:24 h and LD 20:4 h at 25–29°C (Enda and Kitajima 1990; Togashi 2014, 2017). Interestingly, the probability of diapause decreases as the per capita amount of food available decreases or as the initial number of larvae per unit of bark surface area increases at LD 16:8 h and 25°C (Togashi 2014, 2021; Togashi and Toki 2018).

In the same Monochamus species, adult body size sometimes differs among host-plant species that they feed on in the larval stage. Naves et al. (2006) reported that the elytron length of M. galloprovincialis (Olivier) was significantly longer when the larvae were reared on P. pinaster Aiton than on three other species, P. halepensis Miller, P. pinea L. and P. radiata D. Don in the laboratory. However, Akbulut (2009) described no difference in adult body mass of M. galloprovincialis between two host-plant species, P. sylvestris L. and P. nigra Arnold, in the laboratory.

Monochamus alternatus larvae primarily exploit trees of the genus Pinus, although they have a wide host range including species in the genera Abies, Cedrus, Picea, and Pinus of the family Pinaceae (Nakamura 2019). Thus, host-plant species may affect the incidence of diapause as well as life-history traits. In this study, influences of host-plant species on such traits of M. a. alternatus were investigated using three Pinus species, P. densiflora Sieb. et Zucc., P. echinata Mill. and P. thunbergii Parl. Pinus densiflora and P. thunbergii are common species in Japan and are included in the subsection Pinus (Sylvestris) while P. echinata is widely distributed in the eastern and southern United States and is included in the subsection Austrates (Mirov 1967; Takeuchi 2008). The genetic distance between the two subsections is second farthest among eight subsections in the subgenus Pinus on the molecular phylogenetic tree (Liston et al. 2003). In addition, the degree of susceptibility to pine wilt disease differs among them; P. densiflora and P. thunbergii are very susceptible whereas P. echinata is resistant (Woo et al. 2008). Pine species differing in some characteristics may exert different impacts on the diapause induction and life-history traits of insects.

Monochamus a. alternatus will be able to colonize easily in unoccupied regions such as Japan and North America if pine species native to those regions do not inhibit the first generation larvae from forgoing diapause, i.e. if they do not prevent the occurrence of second generation of this insect. In that case, high survival rate and fecundity will enhance the likelihood of colonization. Thus this study investigated some life-history traits of larvae that were fed with bolts of the three pine species and compared them between P. densiflora and P. thunbergii taking account of the risk of invasion of Japan by M. a. alternatus.

The aim of this study was to determine the effects of the three pine species on the diapause induction and performance of nondiapause insects in M. a. alternatus and to evaluate whether they make it difficult for the insect to colonize unoccupied regions.

Materials and methods

Insects

Six larvae and 6 adults of M. a. alternatus were collected by Dr. W. Toki on 16 June 2010 in Yangmingshan, Taipei City, Taiwan. A laboratory population was then established from five of the 6 field-collected larvae and 9 offspring of the field-collected adults (generation 0). This was done with the permission of the Minister of Agriculture, Forestry and Fisheries, Japan (21Y1218). Adults of generation 0 were reared on excised twigs of P. densiflora and P. thunbergii and after sexual maturation they were paired and provided with branch sections of the two pine species as oviposition substrate. Eggs were taken out of the branch sections and placed on wet tissue paper in Petri dishes. Newly hatched larvae within one day of hatching were inoculated singly into P. densiflora bolts because the larvae feed on the inner bark (phloem). The population produced 70 adults that had forgone diapause and 36 that had undergone diapause (generation 1). After that, the laboratory population was maintained using adults that had forgone diapause. Consequently, the incidence of diapause decreased as the generation advanced, indicating that nondiapause was a heritable trait.

Adults of generation 3, 8 and 15 of the laboratory population were reared to obtain larvae to be used in experiment 1, 2 and 3, respectively. Eggs excised from branch sections and larvae within one day of hatching were used in the experiments. All generations were reared on P. densiflora or P. thunbergii under constant conditions of 25 °C and LD 16:8 h.

Pine bolts

Two healthy P. densiflora and four healthy P. thunbergii trees were felled and the stems were cut into 120-cm long logs on 24 February 2012 in Nishi-Tokyo City, Tokyo (mean ± SE of diameter at breast height = 9.28 ± 0.43 cm, 8 years old for P. densiflora; 8.76 ± 0.41 cm, 8 years old for P. thunbergii). The logs were placed in plastic bags and held at 7 − 13°C. The stems were cut into 15-cm long bolts and cut ends of bolts were coated with paraffin wax (melting point of 56 − 58°C) on 8 or 11 March 2012. Bolt length and diameter at the midpoint were determined to calculate the bark surface area of each bolt which was used to estimate the amount of food available for larvae. Bolts were placed in plastic bags and held at 20 − 25°C before use in experiment 1 (Table 1). For experiment 2, one healthy P. densiflora and one healthy P. thunbergii trees were felled on 9 May 2014 in Nishi-Tokyo City (7.0 cm diameter at breast height, 12 years old for P. densiflora; 9.2 cm diameter, 12 years old for P. thunbergii). On 11 May 2014, bolts were prepared from the stems in the same manner as those in experiment 1 (Table 1). The bolts were held at 10°C before use.

Table 1. Number, size and storage period of Pinus densiflora, P. thunbergii and P. echinata bolts used for rearing Monochamus alternatus alternatus larvae in each of three experiments

Exper-iment	Bolts used for rearing
	Pine species	Number of used	Length (cm)*	Diameter (cm)*	Period of bolt storage*^†
1	P. densiflora	50	15.0 ± 0.0	7.2 ± 0.3	52.1 ± 2.3
	P. thunbergii	48	15.0 ± 0.1	6.7 ± 0.3	50.1 ± 2.2
2	P. densiflora	31	15.1 ± 0.1	6.0 ± 0.2	30.0 ± 0.6
	P. thunbergii	45	15.0 ± 0.1	6.8 ± 0.3	24.1 ± 0.6
3	P. densiflora (J)^§	27	15.2 ± 0.1	3.9 ± 0.2	36.4 ± 1.0
	P. densiflora (K)^§	34	15.0 ± 0.1	8.3 ± 0.2	15.8 ± 0.7
	P. echinata (L)^§	42	14.9 ± 0.1	6.3 ± 0.4	59.0 ± 1.8

* mean ± SE.

^†Number of days from tree felling to larval inoculation

^§Letter in parentheses is used for identification of tree.

In experiment 3, one healthy P. densiflora (tree J) and one healthy P. echinata (tree L) were felled on 27 February 2017 in Nishi-Tokyo City (4.9 cm diameter at breast height, 11 years old for P. densiflora; 8.7 cm diameter, 10 years old for P. echinata). On 4 March 2017, bolts were made of the stems in the same manner as those in experiment 1. The bolts were placed in plastic bags and held at 10°C before use. Excessive exudation of oleoresin from wounds made it difficult to inoculate larvae on P. echinata bolts even ca one month after felling and the inoculation on all bolts from tree L was completed 82 days after felling. One additional P. densiflora (tree K) was felled and the stem were placed in a plastic bag on 6 April 2017 in the Koishikawa Arboretum, Bunkyo-ku, Tokyo (10.5 cm diameter at breast height, 15 years old). After the stem was held at 6 − 26°C indoors, bolts were made of the stem on 16 April 2017 and held at 10°C before use (Table 1).

Procedures

A bark disc of ca 8 mm diameter was removed from each bolt and a small depression was excavated on phloem or xylem with a carving knife. The first instar larvae were placed individually into a depression. The bark disc was replaced in its original position on the bolt and fastened with adhesive cloth tape. Bolts inoculated with larvae were placed in transparent containers of 39 × 28 × 24 cm. They were held at constant conditions of 25°C, 100% RH and LD 16:8 h. Bolts were sometimes misted to keep the xylem and phloem moist. Adult emergence from bolts was monitored daily to record the date of emergence and sex. The body mass was measured to the nearest 0.1 mg with an electronic balance. Adult females were dissected in 0.8% saline solution under a microscope to record the number of ovarioles in each of two ovaries. Development time was defined as a period of time from larval inoculation to adult emergence. To determine the amount of phloem consumed by each larva in experiment 3, bark was removed carefully from some randomly selected bolts after adult emergence. Profiles of solid, moist excrement (frass) on xylem were traced on a sheet of transparent OHP film (OHP-S100, A4, 3 M Japan, Yokohama) with a thin black felt pen and photocopied on white kent paper for drafting (Tochiman Silver Hill Kent, No. S160, A3, 238 g/m², Sakae Technical Paper, Tokyo). Paper was cut along the profiles with scissors and the pieces of paper showing the area of excrement were weighed for each larva. Five square sheets (10 cm × 10 cm) of paper were weighed separately, averaging 2391.7 mg/100 cm² (SD = 16.0 mg/100 cm²). The area of excrement (area of phloem consumed by each larva) was estimated by dividing the weight of paper pieces showing the area of excrement by the mean area density of paper.

Bolts that did not produce beetle adults were dissected to record the developmental stages of living and dead insects in them ca 150 days after larval inoculation. Living and dead adults, pupae and prepupae were included in individuals that forwent diapause (Togashi 2017). When living larvae were whitish yellow to yellow in color and had no fecal materials in the intestine, they were judged to be in diapause (Togashi 1991, 1995). When they had fecal materials in the intestine, they were judged to be in pre-diapause state. Larval instar was determined by head capsule width: 1st instars = 0.585–1.170 mm wide, 2nd = 1.260–1.935 mm, 3rd = 1.980–3.150 mm, and 4th = 3.150–4.365 mm (Kozima and Katagiri 1964; Ochi 1975; Togashi 1991).

Statistical method

For combined data from experiments 1 and 2, the analysis of covariance (ANCOVA) was applied for comparing adult body mass and development time between two food resources, P. densiflora and P. thunbergii. Explanatory nominal variables in the model were experiment (E), pine species (P) and beetle sex (S) while two covariates (continuous variables) were the bark surface area of bolt (B) and the period of bolt storage (R). Three interactions between P and E, between P and B, and between P and R were included in the full model to evaluate the interactive effects containing pine species on the variance of response variables. In this analysis, experiment 2, P. densiflora and male were treated as references. When full ANCOVA model showed that some interactions containing a covariate had no effects on the response variable at the 5% significance level, reduced ANCOVA model excluding the interactions was used to determine the effects of explanatory nominal variables. Pairwise comparisons of means were performed at the 5% significance level adjusted by Bonferroni’s method. The same method was applied for the number of ovarioles using four explanatory variables and three interactions: E, P, B, R and interactions between P and E, between P and B, and between P and R.

In the case of experiment 3, the adult body mass, development time and area of phloem consumed were compared between P. densiflora and P. echinata trees using ANCOVA with the tree (T) and beetle sex (S) as nominal variables, the bark surface area of bolt (B) as a covariate, and the T × B interaction. Male and tree L were used as references in the ANCOVA model. In the ANCOVA on the ovariole number, beetle sex was excluded from the explanatory variables. The ANCOVAs in experiment 3 were implemented in the same manner as those in experiments 1 and 2.

Because only two larvae in diapause were found in experiment 2 or 3, two tailed Fisher’s exact test was used to determine if each host-plant species or each tree made all larvae induce diapause. It was applied for comparing the incidence of larval diapause between pine species or between experiments for combined data obtained from experiments 1 and 2. In experiment 3, it was also applied for comparing the incidence of larval diapause among three trees. The 5% significance level adjusted by Bonferroni’s method was used for multiple comparisons of diapause incidences.

Computation was performed using Systat 13 software (Systat Software).

Results

Comparison of growth and development between insects fed with P. densiflora and P. thunbergii (Experiments 1 and 2)

Twenty-three and 19 adults emerged from P. densiflora and P. thunbergii bolts, respectively during a mean incubation period of 146.4 days at 25°C in experiment 1: 11 females and 12 males for P. densiflora bolts and 11 females and 8 males for P. thunbergii bolts. For experiment 2, 29 and 36 adults emerged from P. densiflora and P. thunbergii bolts, respectively during a mean of 158.3 days at 25°C: 23 females and 6 males for P. densiflora bolts and 18 females and 18 males for P. thunbergii bolts.

The mean body mass of adults that had forgone diapause was smallest when they had been reared on P. thunbergii bolts in Experiment 1 (reduced ANCOVA) (Tables 2 and 3). The adult body mass was not affected by the bark surface area of bolt, the period of bolt storage or sex (Table 3). The development time of nondiapause individuals was longest when larvae were inoculated into P. thunbergii bolts in Experiment 2 (reduced ANCOVA) (Tables 2 and 3). The development time increased with increasing period of bolt storage (coefficient = 0.408, p = 0.026) and with decreasing bark surface area of bolt (coefficient = −0.040, p = 0.024) while it was not affected by sex (Table 3). The ovariole number was not affected by experiment (E), pine species (P), bark surface area of bolt (B), period of bolt storage (R) or the P× E, P× B or P× R interactions, indicating an invariable ovariole number (Tables 2 and 3).

Table 2. Effects of Pinus densiflora, P. thunbergii and P. echinata bolts and insect sex on adult body mass, development time, ovariole number and food consumption for nondiapause individuals and incidence of diapause in Monochamus alternatus alternatus

Table 3. Analysis of variance tables for effects of experiment, pine species (Pinus densiflora and P. thunbergii), insect sex, bark surface area of bolt and period of bolt storage on adult body mass, development time and ovariole number in Monochamus alternatus alternatus (experiments 1 and 2)

Response variable	ANCOVA model*	Explanatory variables^†	F value	df	pvalue
Body mass	Full	P × B, P × R	< 0.687	1, 98	> 0.409
	Reduced	Experiment (E)	9.09	1, 100	0.003
		Pine species (P)	3.36	1, 100	0.070
		Insect sex (S)	0.78	1, 100	0.380
		Bark surface area (B)	2.26	1, 100	0.136
		Period of bolt storage (R)	2.95	1, 100	0.089
		P × E	7.20	1, 100	0.009
Development time	Full	P × B, P × R	< 0.395	1, 98	> 0.531
	Reduced	Experiment (E)	2.90	1, 100	0.092
		Pine species (P)	18.9	1, 100	< 0.001
		Insect sex (S)	2.07	1, 100	0.153
		Bark surface area (B)	5.24	1, 100	0.024
		Period of bolt storage (R)	5.12	1, 100	0.026
		P × E	8.57	1, 100	0.004
Ovariole number	Full	P × B, P × R	< 1.24	1, 51	> 0.270
	Reduced	Experiment (E)	0.47	1, 53	0.494
		Pine species (P)	0.18	1, 53	0.673
		Bark surface area (B)	0.19	1, 53	0.664
		Period of bolt storage (R)	0.29	1, 53	0.594
		P × E	1.74	1, 53	0.192

* Full ANCOVA (analysis of covariance) model included three interactions between experiment, pine species, bark surface area and period of bolt storage. When some interactions containing bark surface area or period of bolt storage showed no significant effects at the 5% significance level, reduced ANCOVA model excluding the interactions was used to determine the effects of explanatory variables. So, only insignificant interactions are described for explanatory variables in the full model in Table 3.

^†Cross indicates the interaction between explanatory variables.

A mean of 146.4 days (SE = 0.7 days) after larval inoculation, dissection of P. densiflora bolts detected 1 female and 1 male dead adult, 1 dead pupa, 14 living larvae in diapause, 3 pre-diapause living larvae and 12 dead larvae, whereas dissection of P. thunbergii bolts found 1 dead male adult, 2 living pupae, 2 living prepupae, 18 living larvae in diapause and 1 living pre-diapause larva in experiment 1. One of the four pre-diapause living larvae was 3rd instar and the remaining three were 4th instar. In the case of 32 larvae in diapause, two were 3rd instar and 30 were 4th instar. In experiment 2, there were two 4th-instar living larvae in diapause in xylem after the incubation of a mean of 158.3 days (SE = 1.3 days). One living larva was in a P. densiflora bolt and the other in a P. thunbergii bolt.

The incidence of larval diapause was significantly smaller than 1.0 for each of four populations in experiments 1 and 2 (Fisher’s exact test, p < 0.001 for all populations). There was no difference in the incidence of diapause between two larval groups provided with different pine species in experiment 1 or between those in experiment 2 (Fisher’s exact test, p = 0.652 for experiment 1; p = 1.000 for experiment 2) (Table 2). By contrast, there was significant difference in the diapause incidence of larvae fed with P. densiflora between experiments 1 and 2 (Fisher’s exact test, p = 0.002). That was the case for larvae fed with P. thunbergii (p < 0.001).

Comparison of growth and development between insects fed with P. densiflora and P. echinata (Experiment 3)

Fifty and 35 adults emerged from P. densiflora and P. echinata bolts, respectively during a mean incubation period of 152.6 days: 17 females and 8 males for P. densiflora tree J, 11 females and 14 males for P. densiflora tree K, and 18 females and 17 males for P. echinata.

Unlike observed mean body mass, adjusted mean body mass of nondiapause individuals was significantly greater when larvae fed on bolts from tree J than when those on bolts from trees K and L because of the effects of bark surface area (coefficient = 0.575 for tree J; coefficient = −0.339 for tree K; coefficient = 0.649 for reference) (Tables 2 and 4). The difference in the adjusted mean development time of nondiapause insects among three trees showed the same trend as that in the adjusted mean body mass (coefficient of bark surface area = 0.059 for tree J; coefficient = 0.0009 for tree K; coefficient = 0.092 for reference) (Tables 2 and 4). By contrast, there was no difference in the number of ovarioles among three trees (Tables 2 and 3). The bark surface area of bolt did not affect the number of ovarioles. The adjusted mean food consumption by larvae differed among three trees (coefficient of bark surface area = 0.250 for tree J; coefficient = −0.050 for tree K; coefficient = 0.521 for reference) (Tables 2 and 4). Taken together, there was no difference between P. densiflora and P. echinata in the four beetle traits (Table 2).

Table 4. Analysis of variance tables for effects of tree (Pinus densiflora and P. echinata), insect sex and bark surface area of bolt on adult body mass, development time, ovariole number and food consumption in Monochamus alternatus alternatus (experiment 3)

Response variable	ANCOVA model*	Explanatory variables^†	F value	df	pvalue
Body mass	Full	Tree (T)	1.78	2, 78	0.175
		Insect sex (S)	0.02	1, 78	0.880
		Bark surface area (B)	35.3	1, 78	< 0.001
		T × B	5.41	2, 78	0.006
Development time	Full	Tree (T)	1.16	2, 78	0.318
		Insect sex (S)	0.20	1, 78	0.659
		Bark surface area (B)	19.5	1, 78	< 0.001
		T × B	3.46	2, 78	0.036
Ovariole number	Full	T × B	1.75	2, 39	0.188
	Reduced	Tree (T)	13.0	2, 41	0.846
		Bark surface area (B)	< 0.001	1, 41	0.990
Food consumption	Full	Tree (T)	3.76	2, 33	0.034
		Insect sex (S)	2.22	1, 33	0.146
		Bark surface area (B)	70.8	1, 33	< 0.001
		T × B	4.82	2, 33	0.015

* Full ANCOVA (analysis of covariance) model included the interaction between tree and bark surface area. When the interaction showed no significant effect at the 5% significance level, reduced ANCOVA model excluding the interaction was used to determine the effects of explanatory variables. So, only the insignificant interaction is described for explanatory variables in the full model in Table 4.

^†Cross indicates the interaction between explanatory variables.

Dissection of bolts detected three dead male adults and two 4th-instar living larvae in diapause in P. densiflora bolts and a dead adult female and a living pupa in P. echinata bolts after a mean incubation period of 152.6 days (SE = 0.5 days). The incidence of diapause was significantly smaller than 1.0 for each tree (Fisher’s exact test, p < 0.001 for all trees). There was no difference in the incidence of larval diapause among three trees (Fisher’s exact test, p = 0.514) (Table 2).

Discussion

This study showed that three pine species, P. densiflora, P. thunbergii and P. echinata, did not influence the ovariole number of females that forwent diapause. Thus, the lifetime fecundity is thought to be influenced by traits such as adult body size and longevity (Togashi 1997).

Unlike ovariole number, combined data from experiments 1 and 2 in this study showed that the adult body mass of nondiapause individuals varied depending on the interaction between pine species (P. densiflora and P. thunbergii) and experiment. However, body mass tended to be heavier in experiment 2 than in experiment 1. Extended periods of bolt storage before larval inoculation may have reduced the adult body mass in experiment 1, because the nutritional quality of phloem decreases after the death of trees (Haack and Slansky 1987; Hanks 1999). Thus, the period of bolt storage in this study was regarded as the indicator of food quality for M. a. alternatus larvae, although bolts in experiment 1 may have decreased in the nutritional quality more quickly than those in experiment 2 because of the storage at warm temperatures. When larvae were reared on P. densiflora or P. echinata bolts (experiment 3), the observed mean adult body mass decreased as the mean period of bolt storage was extended, suggesting that the body mass decreased as the food quality decreased.

Development time from larval inoculation to adult emergence from bolts tended to be longer when larvae were fed with P. thunbergii bolts than when fed with P. densiflora bolts in experiments 1 and 2, suggesting nutritionally lower quality of P. thunbergii phloem. Furthermore, the development time increased as the period of bolt storage was extended in experiments 1 and 2 combined. In experiment 3, there was no difference in development time between P. densiflora and P. echinata. However, the adjusted mean development time was longer for P. densiflora tree J than for P. densiflora tree K. As a result, the adjusted mean amount of phloem consumed by larvae was greater for tree J than for tree K. Since the mean period of bolt storage was longer for tree J than for tree K, the nutritional quality of phloem was suggested to affect the development time.

The above mentioned discussion suggests a negative correlation between the adult body mass and development time through different nutritional quality of phloem. In Monochamus carolinensis (Olivier), when the larvae were reared on Pinus banksiana Lamb. logs harvested in different seasons, a negative correlation was found between the adult body mass and development time due to different nutritional quality of phloem (Akbulut and Linit 1999).

In this study, the diapause incidence was significantly higher for experiment 1 than for experiments 2 and 3. It was considered that the genetic change in the used beetle population caused different incidences of diapause among three experiments. The beetle population used for experiments was maintained by selecting for nondiapause individuals. As a result, the diapause incidence decreased from 0.54 to 0.03 over 6 generations (Togashi unpublished data). Reversely, the selection for diapause individuals over 3 generations caused the diapause incidence to increase from 0.54 to 0.73 (Togashi unpublished data).

This study showed that the feeding on three pine species did not cause the diapause incidence to increase to 1.0, suggesting that those pine species did not alter the voltinism or life cycle of M. a. alternatus populations with different genetic structures. In addition, Zhang and Linit (1998) maintained a Taiwanese strain of M. a. alternatus on P. banksiana native to North America over multiple generations at a constant temperature of 27 − 28°C. Thus, the previous and present studies indicated that the voltinism of M. a. alternatus was not altered by some pine species native to unoccupied regions such as Japan and North America, i.e. its reproductive rate per year did not decrease owing to reduced number of generations per year, when it arrived there and then colonized.

This study also showed that there was no difference in the diapause incidence between P. densiflora and P. thunbergii or between P. densiflora and P. echinata when M. a. alternatus populations had the same genetic structure. Thus, it is considered that there is no difference in the impact on colonization of M. a. alternatus between P. densiflora and P. thunbergii, although the nutritional quality of phloem might affect the adult body mass and development time.

Acknowledgments

The author is grateful to Dr. W. Toki of Nagoya University for providing M. a. alternatus, Associate Professor N. Matsushita of the University of Tokyo for identifying P. echinata, the staff of Tanashi University Forest, University of Tokyo, for helping transfer pine logs and Professor M.J. Linit of the University of Missouri for invaluable comments and improving English. Introduction of prohibited species was permitted under special conditions by Minister of Agriculture, Forestry and Fisheries, Japan (21Y1218).

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the Japan Society for the Promotion of Science [22380081 and 26292080].