Introduction
Egg-laying decision-making by females is critical to the survival of offspring for insects whose larvae have little or no possibility to move from their initial food resources, e.g. seeds, fruits, and trees (Anbutsu and Togashi, Reference Anbutsu and Togashi1997a; Deas and Hunter, Reference Deas and Hunter2012; Mitchell, Reference Mitchell1975). In many insect species, females utilise and benefit from conspecific cues for oviposition choice (Prokopy and Roitberg, Reference Prokopy and Roitberg2001). By joining conspecifics and engaging in group oviposition, some insects enjoy fitness benefits such as protection from natural enemies and overcoming a plant’s defensive response (Desurmont and Weston, Reference Desurmont and Weston2011; Fletcher and Miller, Reference Fletcher and Miller2008; Raitanen et al., Reference Raitanen, Forsman, Kivelä, Mäenpää and Välimäki2014). Conversely, for some insects, avoiding conspecifics and engaging in spaced oviposition also realises fitness benefits such as reducing resource competition (Anbutsu and Togashi, Reference Anbutsu and Togashi1996; Kiflawi et al., Reference Kiflawi, Blaustein and Mangel2003; Zhang et al., Reference Zhang, Li, Gao, Liu, Dong and Xiao2019). Studying how information from the behaviour of conspecifics affects egg-laying decision-making is crucial for understanding the biology of insects, which can lead to novel strategies for pest management. However, at present, such interindividual interactions are poorly studied aspects of the research on pest science.
Invasive longhorn beetles are significant threats, causing economic and ecological damage through tree mortality in forests, agricultural lands, and urban areas. The Asian long-horned beetle Anoplophora glabripennis (Coleoptera: Cerambycidae) (fig. 1a), which is native to China and the Korean Peninsula, has become a high profile invasive pest species in North America, Europe, and Japan (Haack et al., Reference Haack, Hérard, Sun and Turgeon2010; Hu et al., Reference Hu, Angeli, Schuetz, Luo and Hajek2009; Wang et al., Reference Wang, Li, Luo, Wang, Dou, Ui Haq, Shang and Cui2023). This species uses various broadleaf trees such as genera Acer, Populus, Salix, and Ulmus as its hosts and kills them through mass infestation (Sjöman et al., Reference Sjöman, Östberg and Nilsson2014; Straw et al., Reference Straw, Fielding, Tilbury, Williams and Inward2015; van der Gaag and Loomans, Reference van der Gaag and Loomans2014; Yasui et al., Reference Yasui, Fujiwara-Tsujii, Kugimiya, Shibuya, Mishiro and Uechi2024). Eggs are laid within ‘oviposition scars’ made by females through the bark of host branches (fig. 1b,c), and larvae tunnel into the woody tissues, causing physical damage through their feeding activities (Faccoli et al., Reference Faccoli, Favaro, Smith and Wu2015; Keena and Sánchez, Reference Keena and Sánchez2018; Li et al., Reference Li, Pei, Wang, Tian, Ren and Luo2024). In the United States, for example, potential damage to urban sugar maples was valued at up to approximately $669 billion (Nowak et al., Reference Nowak, Pasek, Sequeira, Crane and Mastro2001). To develop management strategies for A. glabripennis, it is necessary to understand their biological characteristics, including reproductive behaviour and egg-laying patterns.
The Asian long-horned beetle Anoplophora glabripennis and its reproductive behaviours. (a) A pair in copula, where the male mounts on the female back. (b) A female chewing an oviposition scar through the bark. (c) An oviposition scar (enclosed by a white dashed line) on the Chinese elm Ulmus parvifolia. A single egg is laid just under the bark in the cambium. (Online version in colour.)
In this study, we investigated the effects of information from the behaviour of conspecifics (specifically, whether conspecifics have already oviposited in the site) on egg-laying decision-making in A. glabripennis. Considering the observations of intraspecific predation in longhorn beetle larvae (Dodds et al., Reference Dodds, Graber and Stephen2001; Ware and Stephen, Reference Ware and Stephen2006), we hypothesised that females of this species avoid sites already containing conspecific cues. First, we observed the oviposition pattern in the field, by measuring the distance between neighbouring oviposition scars. Because the field survey suggested that selection of oviposition sites by females is not random, we then performed laboratory bioassays to examine the response of females to conspecific (and their own) oviposition scars.
Materials and methods
Sample collection
Trees infested by A. glabripennis were obtained at three locations in the Kanto region, Japan, during 2023. Logs of the goat willow Salix caprea (Malpighiales: Salicaceae) were collected in Sakuragawa-shi, Ibaraki (29 logs from one tree). Logs of the katsura tree Cercidiphyllum japonicum (Saxifragales: Cercidiphyllaceae) were collected in Kasama-shi, Ibaraki (9 logs from one tree) and Hama-cho, Chuo-ku, Tokyo (5 logs from one tree). All logs were placed in outdoor cages at the Forestry and Forest Products Research Institute in Tsukuba-shi, Ibaraki. To collect emerged adults, we monitored the cages between late May and early August in 2024. The beetles were individually maintained in plastic cups (∼130 mm in diameter and 100 mm in height) under 14L:10D at 25°C in the laboratory and fed fresh twigs with leaves of the Chinese elm Ulmus parvifolia (Urticales: Ulmaceae). The sex of the adult was distinguished by the morphology (males have longer antennae relative to their body and are generally smaller than females) following a previous study (Meng et al., Reference Meng, Hoover and Keena2015). To avoid the decline in fecundity with age, all individuals were used for the experiments within 30 days of collection, referring to past research that performed similar oviposition assays (Keena and Sánchez, Reference Keena and Sánchez2018; Smith et al., Reference Smith, Bancroft and Tropp2002). All animal care procedures were authorised by the Ministry of the Environment of Japan (registered under number 24000004), and the experiments described here complied with relevant institutional, national, and international guidelines and legislation.
Oviposition pattern in the field
To investigate whether the presence of conspecific cues affects the egg-laying decision-making by females, we observed oviposition patterns in the field by focusing on the distance between neighbouring oviposition scars (in this species, females make scars through the bark of host branches for laying eggs). We measured the distance from a given oviposition scar to the nearest scar using a tape measure. This was performed on all oviposition scars found on the surface of the obtained logs after the emergence of adults: 29 logs of S. caprea measuring 49.186 ± 1.360 cm (mean ± SE) in length and 14 logs of C. japonicum measuring 63.193 ± 6.909 cm (mean ± SE) in length. Oviposition scars were easily distinguished from mere chew marks not containing room for eggs or bark cracks due to the deposition of larval frass. If the distance between neighbouring oviposition scars does not follow a normal distribution, even with a sufficient sample size, we can expect that the selection of oviposition sites by females is not random. Here, we did not use typical analysis for determining spatial patterns (e.g. uniform distribution, clumped distribution or Poisson distribution) in ecological studies because the obtained logs were only a part of each infested tree.
Oviposition-choice bioassays
Two behavioural assays were conducted to examine how the presence of exisiting oviposition scars affects the egg-laying decision-making by females. In the first experiment, a female was given a choice between a control branch and a branch containing oviposition scars from a different female. As egg-laying substrates, fresh branches of U. parvifolia (150 mm in length) were used. Because this species shows a preference for branch diameter when laying eggs (Huang et al., Reference Huang, Wang, Hai, Wang and Lyu2024), we confirmed there were no significant differences in the branch diameters between the treatments (see the Results section). The branches used in each replication were contiguous from the same tree to provide similar bark thickness, phloem quality, and moisture. To obtain the treated branches (i.e. those with oviposition scars), a pair of adults were introduced into a plastic container made by connecting two plastic cups (∼130 mm in diameter and 100 mm in height) with a U. parvifolia branch (for oviposition) and excised twigs (as food). After 24 h, we removed the males and allowed the females to lay eggs for 1 week. The sequence of oviposition events by A. glabripennis was carefully described in previous studies (Faccoli et al., Reference Faccoli, Favaro, Smith and Wu2015; Keena and Sánchez, Reference Keena and Sánchez2018; Li et al., Reference Li, Pei, Wang, Tian, Ren and Luo2024); females make scars through the bark of host branches and lay single eggs at the cambial interface in some, but not all of the scars. Before the choice tests, we removed the female and examined the treatment branch to determine the number of oviposition scars. We then added a control branch into each plastic container. Next, we introduced a different female from the same population of origin into the plastic container and allowed them to lay eggs for 1 week, after which we counted the oviposition scars on both branches. The second experiment consisted of a choice between a control branch and one that contained her own oviposition scars under the same conditions as the above method. We performed 20 replicates for each experiment. During the oviposition-choice bioassays, 16 of 20 females in the choice between a control branch and a treated branch with conspecific scars and 11 of 20 females in the choice between a control branch and a treated branch with own scars made oviposition scars. Individuals who did not make any oviposition scars during the experiments were removed from the dataset.
Data analysis
All statistical analyses were performed using R software (v4.4.2) (R Core Team, 2024). In the field survey, the distribution normality of the distance between neighbouring oviposition scars was tested using the Shapiro–Wilk test. We also examined the effect of the perimeter length of logs on the distance between neighbouring oviposition scars using Spearman’s rank correlation coefficient. In the laboratory oviposition-choice bioassays, we analysed the numbers of oviposition scars using generalised linear mixed models (GLMMs) with a Poisson distribution, and the branch diameters using GLMMs with a Gaussian distribution. In the GLMM, treatment (i.e. scar-present or scar-absent) was a fixed factor and female ID was a random factor. We also examined the effect of the numbers of initial oviposition scars on the numbers of oviposition scars added using generalised linear models (GLMs) with a Poisson distribution. Likelihood ratio tests were conducted to determine the significance of each fixed effect. A significance value of p < 0.05 was considered to indicate statistical significance. Experimental data analysed during this study are included in the Supplementary Information file (Dataset S1).
Results
Oviposition pattern in the field
The distance between neighbouring oviposition scars did not follow a normal distribution in both tree species. In S. caprea, the distance from a given oviposition scar to the nearest scar was 4.383 ± 0.091 cm (mean ± SE, range 1.0–14.7 cm, n = 594 oviposition scars from one tree, fig. 2a). The data were significantly deviated from a normal distribution (Shapiro–Wilk test, p < 0.001), showing a skewness of 1.730 (positive skew). No significant differences were observed in the distance between neighbouring oviposition scars depending on the perimeter length of the logs (Spearman’s rank correlation, ρ = 0.069, p = 0.094). In C. japonicum, the distance from a given oviposition scar to the nearest scar was 7.516 ± 0.416 cm (mean ± SE, range 2.0–26.2 cm, n = 116 oviposition scars from two trees, fig. 2b). The data was significantly deviated from a normal distribution (Shapiro–Wilk test, p < 0.001), showing a skewness of 1.380 (positive skew). No significant differences were observed in the distance between neighbouring oviposition scars depending on the perimeter length of the logs (Spearman’s rank correlation, ρ = −0.073, p = 0.435).
Oviposition pattern of Anoplophora glabripennis in the field. Each histogram shows the frequency distribution of the distance from a given oviposition scar to the nearest scar by tree species. (a) Distance between neighbouring oviposition scars found on the branches of the goat willow Salix caprea. (b) Distance between neighbouring oviposition scars found on the branches of the katsura tree Cercidiphyllum japonicum. (Online version in colour.).
Oviposition-choice bioassays
Females avoided both conspecific and their own oviposition scars during the choice tests. In the first experiment, females made less oviposition scars on branches containing scars made by other females than those without scars (likelihood ratio test, df = 1, χ 2 = 63.80, p < 0.001, fig. 3a). Females showed stronger avoidance of the treated branches when they had more conspecific oviposition scars (likelihood ratio test, df = 1, χ 2 = 83.14, p < 0.001, fig. 3b). We also confirmed that there were no significant differences in the branch diameters (likelihood ratio test, df = 1, χ 2 = 1.07, p = 0.294, fig. S1a). In the second experiment, females also avoided making oviposition scars on branches containing already-made own scars (likelihood ratio test, df = 1, χ 2 = 55.24, p < 0.001, fig. 3c). Females showed stronger avoidance of the treated branches containing more of their own oviposition scars (likelihood ratio test, df = 1, χ 2 = 52.65, p < 0.001, fig. 3d). We also confirmed that there were no significant differences in the branch diameters (likelihood ratio test, df = 1, χ 2 = 1.15, p = 0.273, fig. S1b).
Effects of already-made oviposition scars on the egg-laying decision-making in Anoplophora glabripennis. (a) Comparison of the number of oviposition scars added by a female between a branch without scars and a branch containing scars from a different female. (b) Correlation between the number of initial oviposition scars made by a female and the number of scars added by a different female. (c) Comparison of the number of oviposition scars added by a female between a branch without scars and a branch containing scars from the same female. (d) Correlation between the number of initial oviposition scars by a female and the number of scars added by the same female. In bar plots (a, c), each straight line represents that two connected data points belong to the same replication. Error bars represent the standard error of the mean (SE). Asterisks indicate significant differences (likelihood ratio test, *p < 0.05, **p < 0.01, ***p < 0.001). In scatter plots (b, d), GLM-fitted curves are shown together. (Online version in colour.).
Discussion
We demonstrated that females avoid conspecific cues during egg-laying decision-making in A. glabripennis. This is the first study to report interindividual interactions in this invasive longhorn beetle when females chose an oviposition site. Oviposition deterrence caused by conspecific cues can be interpreted as a part of social information use (Prokopy and Roitberg, Reference Prokopy and Roitberg2001). Examples from various taxa illustrate that joining (Desurmont and Weston, Reference Desurmont and Weston2011; Fletcher and Miller, Reference Fletcher and Miller2008; Raitanen et al., Reference Raitanen, Forsman, Kivelä, Mäenpää and Välimäki2014) or avoidance (Anbutsu and Togashi, Reference Anbutsu and Togashi1996; Kiflawi et al., Reference Kiflawi, Blaustein and Mangel2003; Zhang et al., Reference Zhang, Li, Gao, Liu, Dong and Xiao2019) behaviour differs among species. Moreover, within a given species, the decision to join or avoid conspecifics can be affected by the physiological and informational state of the individual and by contextual response thresholds to resource availability (Otake and Dobata, Reference Otake and Dobata2018). The fitness benefit of oviposition deterrence associated with conspecific cues has often been discussed in terms of avoiding food shortage and cannibalism. In longhorn beetles, which are internally feeding insects, newly hatched larvae have little or no possibility to move from their initial food resources. In such cases, if an adult lays eggs on a site already occupied by conspecific larvae, their progeny could experience increased mortality (Anbutsu and Togashi, Reference Anbutsu and Togashi2000). Previous studies reported that females avoid hosts already occupied by conspecific cues (presence of oviposition scars, eggs, larvae, and their frass) in some longhorn beetles such as Monochamus alternatus (Anbutsu and Togashi, Reference Anbutsu and Togashi1996, Reference Anbutsu and Togashi2000, Reference Anbutsu and Togashi2002), M. saltuarius (Anbutsu and Togashi, Reference Anbutsu and Togashi1997b), M. scutellatus (Peddle et al., Reference Peddle, De Groot and Smith2002), and Paraglenea fortunei (Wang et al., Reference Wang, Zeng and Li1990). In this study, females avoided both conspecific and their own oviposition scars, in contrast to Peddle et al. (Reference Peddle, De Groot and Smith2002), who concluded that females did not avoid their own eggs in M. scutellatus. The adaptive advantage for females to avoid their own oviposition scars may be different depending on the species.
How do females recognise and avoid conspecific oviposition scars in A. glabripennis? It is possible that females use semiochemical signals to mark their own oviposition sites and detect hosts occupied by conspecifics. Oviposition deterrence is induced by a marking pheromone released by females in some insects (Nufio and Papaj, Reference Nufio and Papaj2001; Roitberg and Prokopy, Reference Roitberg and Prokopy1987). In A. glabripennis, the female trail secretions (two major components: 2-methyldocosane and (Z)-9-tricosene, two minor components: (Z)-9-pentacosene and (Z)-7-pentacosene) were identified (Graves et al., Reference Graves, Baker, Zhang, Keena and Hoover2016; Hoover et al., Reference Hoover, Keena, Nehme, Wang, Meng and Zhang2014). These authors proposed that trail secretions may act as a spacing pheromone for females. Intriguingly, in this study, females showed stronger avoidance of the treated branches when they had more conspecific oviposition scars. This suggests that females can recognise oviposition scars themselves or oviposition-related substances. In other longhorn beetles, M. alternatus females were deterred from oviposition by the jelly-like secretion that the conspecific females deposited in the oviposition scars immediately after oviposition, which shows that females’ recognition of the egg-containing scars and departure from such scars were elicited by semiochemicals contained in the jelly-like secretion they deposited (Anbutsu and Togashi, Reference Anbutsu and Togashi2000, Reference Anbutsu and Togashi2001). It is also possible that A. glabripennis females perceive the shape of oviposition scars tactually or visually. Future studies should investigate the detailed mechanisms of oviposition deterrence caused by conspecific cues, e.g. chemical analysis of oviposition scars and further bioassays using artificial scars. Identifying the key to the avoidance can lead to ‘oviposition deterrents’, which discourage the pests from laying eggs on treated trees.
This study provides insights into the reproductive behaviour of this invasive longhorn beetle, which is useful for developing environmentally friendly control methods such as oviposition deterrents. Oviposition deterrents could be a viable option for reducing established and incipient populations in urban and natural forests, where intensive management methods such as spraying pesticides and destructing infested trees may be environmentally undesirable (Prokopy, Reference Prokopy1972, Reference Prokopy and Mitchell1981). Here, we point out that taking account of interindividual interactions is important for future technical innovations in applied entomology.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S000748532500032X.
Acknowledgements
We thank Tadahisa Urano, Midori Iida, Reiko Nishihara, and Ayaka Hashiguchi for their assistance in collecting and rearing insects; Haruo Kinuura, Noritoshi Maehara, Takuma Takanashi, Takeshi Matsumoto, and Eiriki Sunamura for helpful discussion. We also thank all other members of the Department of Forest Entomology, Forestry and Forest Products Research Institute (FFPRI) for inspiring scientific discussions. This study was supported by a grant from the Research Program on Development of Innovative Technology, Bio-oriented Technology Research Advancement Institution (JPJ007097; project ID 04015C1).
Author contributions
TK: conceptualization, data curation, formal analysis, investigation, methodology, resources, validation, visualization, writing-original draft, writing–review and editing; KU: conceptualization, investigation, methodology, resources, validation, writing–review and editing; ST: conceptualization, methodology, resources, writing–review and editing; HT: conceptualization, funding acquisition, methodology, project administration, resources, writing–review and editing; ES: conceptualization, methodology, resources, supervision, writing–review and editing. All authors gave final approval for publication and agreed to be held accountable for the work performed therein.
Competing interests
We declare we have no competing interests.
