Bark beetles in the genus Tomicus (Coleoptera: Curculionidae, Scolytinae), with the exception of Tomicus puellus (Reitter) that infests primarily Picea trees, are well-known pests of pine (Pinus) trees in most Eurasian countries where pine is native (Bakke 1968, Ye 1991, Faccoli 2007, Kirkendall et al. 2008, Li et al. 2010, Lieutier et al. 2015). These univoltine bark beetles are commonly called shoot beetles because the newly emerged adults feed inside living shoots of their host trees during the summer months as they become sexually mature, which often results in death of the infested shoots (Långström 1983, Ye 1994a, Kohlmayr et al. 2002). The Palaearctic genus Tomicus contains eight recognized species worldwide, of which five occur only in Asia, one only in Europe, and two in Eurasia (Kirkendall et al. 2008, Li et al. 2010, Lieutier et al. 2015). The Eurasian species, Tomicus piniperda (L.) is the only Tomicus species to be introduced beyond its native range, being first found in the Great Lakes region of North America in 1992 (Czokajlo et al. 1997, Haack and Poland 2001, Lieutier et al. 2015).

Tomicus brevipilosus (Eggers) is native to Asia where it has been reported to occur in China, India, Japan, Korea, and Philippines (Kirkendall et al. 2008, Lu et al. 2014). This bark beetle has been recorded to infest several species of Pinus, including Pinus yunnanensis (Franchet), Pinus koraiensis Siebold & Zucc., and Pinus kesiya Royle ex Gordon, but severe damage has seldom been reported (Murayama 1959, Kirkendall et al. 2008, Lu et al. 2014). In China, T. brevipilosus has been generally regarded as a secondary pine pest, mainly infesting pine trees weakened by drought, defoliation, and other environmental stressors (Kirkendall et al. 2008, Lu et al. 2014). However, since the late 1990s, T. brevipilosus along with Tomicus yunnanensis Kirkendall and Faccoli and Tomicus minor (Hartig) have reached outbreak levels, resulting in widespread tree mortality with more than 200,000 ha of infested P. yunnanensis forests in southwestern China (Ye 1991, Duan et al. 2004, Liu et al. 2010, Lu et al. 2014).

In Yunnan, T. brevipilosus completes its life cycle in Yunnan pine (P. yunnanensis) (Lu et al. 2014). After exiting the host material in which the brood developed, newly emerged adults fly to the crowns of nearby pine trees, usually beginning in June, where they feed in shoots for the next 9–10 mo and become sexually mature (Lu et al. 2014). Each adult usually feeds inside and kills 3–5 shoots (Lu 2011). Extensive shoot feeding by Tomicus adults can cause growth loss and lower a tree's natural resistance to the point where individual trees are predisposed to infestation (i.e., breeding attacks) during the beetles' reproductive phase (Lieutier et al. 2003, Lu et al. 2014).

Sexually mature adult beetles eventually depart the shoots to reproduce in the trunks of P. yunnanensis trees, primarily in April and May (Lu et al. 2014). It is not known what initiates the reproductive phase in T. brevipilosus. Adult females initiate attack by constructing individual longitudinal galleries in the phloem tissue and depositing eggs in niches along the gallery walls. After hatching, larvae feed transversely in individual galleries in the phloem and outer sapwood and then pupate at the ends of the galleries. Larval feeding disrupts nutrient flow within the phloem tissue (Långström 1983, Fernández et al. 1999, Ye and Ding 1999, Långström et al. 2002).

Gallery construction and reproduction by T. brevipilosus and at times other Tomicus species are regarded as the direct causes for P. yunnanensis tree mortality, particularly when mass attack occurs on individual trees (Ye and Ding 1999, Lu et al. 2014). As one aspect of developing a T. brevipilosis management program, the objective of our study was to clarify the timing and duration of the beetle's reproductive period and within-tree attack pattern, as well as other aspects of the beetle's reproductive biology. A further aim of this study was to improve our understanding of the reproductive ecology of T. brevipilosus with respect to other Tomicus species that infest P. yunnanensis in southwestern China.

Materials and Methods

Study Area. The field study was carried out (primarily by P.C., J.L., and H.Y.) in a P. yunnanensis stand in Anning County (24.97° N, 102.33° E, 1,800 m a.s.l), approximately 80 km west of Kunming, the capital of Yunnan, in southwestern China. The stand was along the lower slope of a hillside and covered about 300 ha. The P. yunnanensis trees in this stand were originally planted by aerial seeding in the mid-1970s. Most of the pine trees were 24–28 cm in diameter and 7.5–8.5m tall at the time of this field study (2007–2009). Low numbers of T. yunnanensis and T. minor were first recorded infesting trees in this stand in the early 1990s, followed by T. brevipilosus within a few years (Duan et al. 2004). All these three Tomicus species have been active in this stand since the late 1990s (Duan et al. 2004, Lu et al. 2014). It is important to note that in Yunnan, T. yunnanensis was considered as T. piniperda due to their morphological similarities before 2008 (e.g., Ye 1991, 1994a,b, 1995; Ye and Lieutier 1997; Ye and Ding 1999; Långström et al. 2002; Lieutier et al. 2003; Duan et al. 2004), when T. yunnanensis was first described as a new Tomicus species (Kirkendall et al. 2008).

Experimental Procedures. During January to July in each year from 2007 to 2009, 3–4 P. yunnanensis trees with evidence of recent bark beetle breeding attacks were sampled every 10 d, for a total of 10 trees per month. Most of the sampled trees were cut from the interior portion of the stand and were among the larger trees present. We selected trees that appeared to have been recently infested based on the texture and color of the frass and resin that was present near the entry hole where the Tomicus beetles entered the bark along the trunk surface (Lu 2011). The sampled trees were felled by cutting the trunk near the ground with a chainsaw, and then we cut the trunk into 50-cm-long logs, starting at the base and stopping near mid-crown. The logs were marked in the field to identify from which tree and which part of the trunk they were cut and then transported to the laboratory. In the laboratory, usually within 1–2 d after cutting, we measured the surface area of each log and then carefully removed the outer bark to look for bark beetles and their galleries.We recorded several parameters related to beetle reproduction and development, including length and width of the maternal gallery using a Vernier calliper (HMCT 6202-01, Harbin, China), number of maternal galleries for each Tomicus species, and the number of eggs, larvae and at times pupae for each maternal gallery. All parent bark beetles were collected from the maternal galleries, placed individually into labeled plastic bags, and identified to species under a stereomicroscope (Nikon-smz 500) based on morphological characteristics of the various Tomicus species (Kirkendall et al. 2008, Li et al. 2010). We recorded our observations for each Tomicus species on a per log basis, using data from all the bark beetles galleries that were present, no matter their age.

In addition, detailed observations were made on 106 T. brevipilosus maternal galleries (in which the parent female was still present) at 10-d intervals from March to July 2009, from which the timing of first occurrence of eggs, larvae, and pupae were recorded. These parameters were used to estimate the duration of egg, larval, and pupal development in the field. At the same time, several other maternal gallery parameters were measured and recorded including gallery length, gallery width, and numbers of eggs or larvae present (for details see Lu 2011).

Data Analysis. Correlation analysis was conducted between the length of the maternal egg gallery and the number of eggs present in the galleries constructed by T. brevipilosus. Statistical analyses were performed using SPSS (version 13 for Windows) with an alpha level of 0.05. Temperature data were obtained from Yunnan Meteorological Bureau, which had a weather recording station about 8 km from our field site.

Results

Breeding Attacks. T. brevipilosus breeding attacks, including sister broods, on P. yunnanensis trees were initiated in early March and ended in early June, spanning a period of about 3 mo (Fig. 1). The timing and duration of breeding attacks were similar over the 3-yr study period from 2007 to 2009 (Fig. 1).

The attacking population of T. brevipilosus parent females varied from month to month, based on the total number of maternal galleries found each year for each Tomicus species while debarking the logs but formed an approximately normal distribution curve each year during the period from March to June (Fig. 1). Typically, breeding attacks in March accounted for about 10–25% of the annual attacks (10.25% in 2007, 24.44% in 2008, and 18.39% in 2009). From March to May, the number of attacks increased rapidly, peaking in April in 2007 and 2009, or in May in 2008 (Fig. 1). The peak month of breeding attacks in each year of study accounted for a relatively high percentage of the total annual attacks: 65.38% in 2007, 37.78% in 2008, and 58.62% in 2009. In June, when new Tomicus infestations ended, the June breeding attacks accounted for only 2.56% of all annual attacks in 2007, 8.89% in 2008, and 11.49% in 2009 (Fig. 1).

The daily minimum and maximum air temperatures for the recording station about 8 km from our field site are shown in Fig. 2 for 1 January through 30 June during 2007–2009. Considering areas of the world where Tomicus species experience cold winters, spring flight is usually initiated when air temperatures exceed 10–12°C (Bakke 1968; Salonen 1973; Långström 1983; Haack et al. 2000, 2001; Haack and Poland 2001; Ye et al. 2002; Lieutier et al. 2015). Considering a flight threshold of 12°C at our study site, it is clear from the data presented in Fig. 2 that this temperature was exceeded on almost every day from January through June during 2007–2009. Even air temperatures of 15°C were exceeded nearly every month (Fig. 2).

T. brevipilosus Breeding Attacks in Relation to Other Tomicus Species. The log dissections during this 3-yr study indicated that breeding attacks by T. brevipilosus occurred in P. yunnanensis trees both with and without coinfestation by the other two Tomicus species that were present in the area (Figs. 3 and 4). For example, in March, when T. brevipilosus commenced breeding attacks, there already existed several Yunnan pine trees that were infested by T. yunnanensis and T. minor, which had initiated breeding earlier in the year than T. brevipilosus (Lu et al. 2014). In this situation, T. brevipilosus only attacked trees that were already infested by T. yunnanensis and T. minor (Fig. 3). In April, as the number of trees newly infested by T. yunnanensis and T. minor decreased, T. brevipilosus continued to infest pine trees in which T. yunnanensis and T. minor already occurred as well as initiated attack by itself on other pine trees (Fig 3). During May and June, when there were no other pine trees with new T. yunnanensis or T. minor breeding attacks, T. brevipilosus attacked only by itself on previously uninfested pine trees (Fig. 3).

Fig. 1.

Percent of annual T. brevipilosus breeding attacks that were initiated monthly on P. yunnanensis trees (n=10 trees per month) that were sampled at the study site in Yunnan from January through July in 2007–2009.

Within-Tree Distribution on Breeding Attacks. T. brevipilosus was able to colonize the entire trunk of P. yunnanensis trees, usually starting about 50 cm above the ground level and reaching the mid-crown level (Fig.3). However, the colonization pattern varied dramatically, depending if the other two Tomicus species were already present on the tree trunks or not. In those cases when both T. yunnanensis and T. minor were already present along the trunk at the time of T. brevipilosus infestation, which usually happened in March to April (Fig. 3), T. brevipilosus tended to colonize the open spaces of inner bark (phloem) where the other two Tomicus species and their brood had not already occupied. In most cases, T. yunnanensis appeared to be the first species to colonize the trees, occupying primarily the upper portions of the trunk, whereas T. minor tended to colonize trees already infested by T. yunnanensis, infesting primarily the lower portions of the trunk (Ye and Ding 1999) (Fig. 4A). There was some overlap of both T. yunnanensis and T. minor in the mid-trunk region of most trees (Fig. 4A). Given this situation, the early-season T. brevipilosus attacks were scattered along the entire trunk between the gallery systems of the two other Tomicus species but being concentrated mostly in the mid- and lowertrunk samples (Figs. 1 and 4A). Therefore, during the early portion of the T. brevipilosus flight season, T. brevipilosus attacks tended to overlap mostly with T. minor galleries in the lower trunk and to a much lesser degree with T. yunnanensis galleries in upper trunk and with both of these Tomicus species in the mid-trunk sections (Fig. 4A).

Fig. 2.

Maximum, minimum, and mean daily air temperatures from 1 January through 30 June in 2007–2009 in Anning County (24.97°N, 102.33°E, 1800 m a.s.l) that were recorded at an official weather station about 8 km from our study site.

Fig. 3.

Number of P. yunnanensis trees with recent attacks by various combinations of three Tomicus species (T. brevipilosus, Tb; T. minor, Tm; T. yunnanensis, Ty) based on sampling 10 trees per month in Yunnan that were cut from January to July, 2007–2009.

In those cases where T. brevipilosus was the only Tomicus species present to initiate attack, which usually happened from mid-April to early June (Fig. 3), T. brevipilosus tended to initiate attack along the lower trunk first and then move progressively upward along the trunk (Fig. 4B). When the tree trunks were fully colonized by T. brevipilosus, the attack density was broadly similar over much of the trunk surface area (Fig. 4B).

Reproduction. T. brevipilosus adult females initiated oviposition within a few days of starting gallery construction on the trunks of P. yunnanensis trees given that the first eggs were observed in early March. Overall, considering all T. brevipilosus breeding attacks observed in 2009, oviposition occurred from early March to early June (Table 1), about 3–3.5 mo, peaking in April to May. The period of active oviposition closely matched the period of tree colonization (i.e., breeding attacks; Fig. 1).

In 2009, the first T. brevipilosus eggs were found in early March and the first larvae in mid-March, indicating that the egg incubation period was less than 2wk in the field during March. The first pupae were recorded in early May, with most pupating by mid-May, when average daily air temperature was about 18°C (Fig. 2). Therefore, T. brevipilosus larvae appeared to require about 40–50 d to complete larval development. The pupation period lasted about 10–20 d at the study site. Callow adults of T. brevipilosus were first observed under the bark of the sampled trees in late May, with emergence of the new brood adults starting in early June and peaking during mid- to late June (Table 1).

The typical maternal gallery of T. brevipilosus is a single longitudinal tunnel with an entrance hole at the bark surface that leads to a slightly enlarged mating chamber and then a single gallery in which eggs are deposited along the gallery walls. The average length of apparently fully constructed T. brevipilosus maternal galleries was 7.2±2.3 (mean±SE, n=106) cm long (14.0 cm maximum) and varied in width from 4.7 to 5.4 mm. The average number of eggs laid per gallery was 50.9±18.7 (mean±SE, n=106) and ranged from 22 to 128 eggs. The number of the eggs per gallery was significantly and positively correlated with maternal gallery length (R2=0.486, n=106, P<0.0001) (Fig. 5).

Fig. 4.

Average attack density of T. brevipilosus, T. minor, and T. yunnanensis at various heights along the main trunk of P. yunnanensis trees in Yunnan when sampled monthly from January through June 2007–2009. (A) The vertical distribution for the three Tomicus species individually when all three species were present on the same trees (March and April). (B) The vertical distribution for T. brevipilosus when it was the only beetle present on the sampled trees (April–June). The data were pooled for all months and years.

Table 1.

Typical life cycle of T. brevipilosus in Anning County (24.97° N, 102.33° E, 1,800 m a.s.l), Yunnan, China, based primarily on field work conducted in 2009

Fig. 5.

Linear relationship between number of eggs within a single T. brevipilosus maternal gallery and the length of the corresponding maternal gallery.

Discussion

The presence of two Tomicus species co-occurring in the same pine forest has been commonly reported in many Eurasian countries (Bakke 1968, Långström 1983, Ye and Ding 1999), but the co-occurrence of three Tomicus species infesting the same pine trees has seldom been reported in the world (Lu et al. 2014). In Yunnan, in southwestern China, where three Tomicus species can coexist in the same P. yunnanensis forest, it was not known how these three Tomicus species would interact. That is, would they compete with each other for the same resource or would they reduce competition by partitioning their breeding resource either spatially or seasonally (Paine et al. 1981, Flamm et al. 1987, Wilkinson and Haack 1987, Haack et al. 1989, Amezaga and Rodrı´guez 1998, Ayres et al. 2001)? Obviously, answering this question will help in understanding the damage potential caused by the various Tomicus species in southwestern China.

This study, along with the results presented by Lu et al. (2014), indicated that the three Tomicus species that coexist in P. yunnanensis forests are in large part separated seasonally in their reproductive cycles, and when they do overlap seasonally, they partition themselves spatially along the main trunk. For example, Lu et al. (2014) noted that T. yunnanensis reproduces from early November to March of the next year, whereas T. minor started to reproduce about 1–2wk later, from late November to late March or early April, and usually only infested trees already under attack by T. yunnanensis (Ye and Ding 1999, Chen et al. 2010). In contrast, breeding duration of T. brevipilosus was primarily from early March to early June, thus overlapping with that of the other two Tomicus species by only about 1 mo (Fig. 3). These differences in the peak breeding season for the above three Tomicus species in Yunnan likely reduce interspecific competition among these Tomicus species on the one hand but also lengthen the total period of time that pine trees can be under attack by Tomicus beetles each year and thereby can result in more serious damage to Yunnan pine stands (Lu et al. 2014).

The initiation of Tomicus breeding attacks in northern latitudes that experience cold winters is closely related to maximum daily air temperatures. For example, in northern Europe and North America, T. piniperda typically initiates flight and breeding attacks when daily air temperatures reach 10–12°C, and T. minor when temperatures reach 12–14°C(Bakke 1968; Salonen 1973; Långström 1983; Schlyter and Löfqvist 1990; Haack et al. 2000, 2001; Haack and Poland 2001; Ye et al. 2002). In northeastern Italy, Tomicus destruens (Wollaston) initiates spring flight when daily mean temperatures exceed 12°C (Faccoli et al. 2005), but in central Italy, T. destruens flies to breeding sites in October and November when the maximum daily temperatures are 18–22°C and daily minimum temperatures are 7–14°C (Sabbatini Peverieri et al. 2008).

In contrast, in Yunnan, little is known about the circumstances that trigger spring flight in the four local Tomicus species. In the case of T. yunnanensis, adults initiate breeding attacks in November apparently by simply departing the shoots and walking along the branches to the main trunk of the same tree on which they shoot fed (Liu et al. 2010). Later, usually beginning in January, T. yunnanensis adults will fly to new hosts after first experiencing periods of cold and then having mean maximum air temperatures once again reaching or exceeding 12°C (Liu et al. 2010). The threshold temperature for spring flight of T. minor has not been studied in Yunnan, but it is likely slightly higher than that of T. yunnanensis given that T. minor initiates spring flight soon after T. yunnanensis (Lu et al. 2014).

The threshold temperature for spring flight of T. brevipilosus is not known, but it is likely considerably higher than that of T. yunnanensis and T. minor given that T. brevipilosus spring flight starts about 2 mo later, in March (Lu et al. 2014; Fig. 3). Perhaps initiation of T. brevipilosus spring flight is more closely linked to daily mean temperature rather than the daily maximum as reported for T. destruens in Italy (Faccoli et al. 2005). An examination of the temperature data in Fig. 2 shows that in March of each year, the daily mean air temperature ranged over 11.8–19.6°C in 2007, 8.7–18.5°C in 2008, and 11.9–18.2°C in 2009. If we consider a mean air temperature of 15°C as an arbitrary threshold value for T. brevipilosus flight, then there were 15 d that could have supported flight in March and 16 d in April 2007, 10 and 25 d in 2008, and 18 and 26 d in 2009, respectively. Of course, other factors besides simply air temperature could influence initiation of spring flight in T. brevipilosus, such as day length, ovarian development, or some physiological change within the host tree that is detected by adults as they overwinter in the shoots.

As shown in Fig. 1, T. brevipilosus peak flight was apparently shifted 1 mo later in 2008 when compared with 2007 and 2009. It is not clear if this shift is simply an artifact of the trees that were sampled in the different years or if this apparent shift reflects some major temperature differences among the three sampling years. If the difference in the flight season is related to temperature, it is difficult to identify a major difference in the 2008 temperature data shown in Fig. 2 when compared with the data from 2007 and 2009. One difference, asmentioned above, is that there were fewer days in March 2008 when the mean daily temperature exceeded 15°C when compared with March in 2007 or 2009. However, mean temperatures in April 2008 were broadly similar to April 2009 but warmer than April 2007. Such year to year variation in the timing of the initial spring flight of overwintering Tomicus adults is common and has been shown to be related to annual differences in spring temperatures (Bakke 1968; Salonen 1973; Långström 1983, 1986; Haack and Lawrence 1995; Haack et al. 2000; Faccoli et al. 2005).

Among the eight species of Tomicus worldwide, T. piniperda usually breeds in wind-blown trees, fire-damaged trees, stumps, and other severely stressed pine trees (Bakke 1968; Ye 1991; Haack et al. 2000, 2001; Kirkendall et al. 2008). Tomicus armandii Li and Zhang may infest weakened Pinus armandii Franchet trees in Yunnan (Li et al. 2010). T. yunnanensis is regarded as one of the most aggressive species of Tomicus, being able to attack and kill live, apparently healthy P. yunnanensis trees (Ye and Lieutier 1997, Lu et al. 2014). T. minor is often considered to be more of an opportunist, usually infesting host trees that are already infested by other Tomicus species such as T. piniperda or T. yunnanensis, which could facilitate tree death (Bakke 1968, Långström 1983, Eidmann 1992, Ye and Ding 1999, Lieutier et al. 2003, Chen et al. 2009). In this study, we discovered that T. brevipilosus preferred to breed in trunks of P. yunnanensis trees that were already infested by both T. yunnanensis and T. minor. However, later in the season, T. brevipilosus was able to attack apparently healthy P. yunnanensis trees on its own, indicating that T. brevipilosus is more similar to T. yunnanensis in its aggressiveness when compared with T. minor (Chen et al. 2009, 2010).

The spatial distribution of T. brevipilosus breeding attacks in the trunks of P. yunnanensis trees was greatly influenced by the presence or absence of T. yunnanensis and T. minor breeding in the same tree (Fig. 4). In trees that were already infested by the other two Tomicus species, T. brevipilosus colonized areas of the trunk that were not already occupied, resulting in a scattered distribution of T. brevipilosus attacks along the trunks. The ability of T. brevipilosus to adjust its infestation pattern in response to other Tomicus species likely decreases interspecific competition among Tomicus species, as well as better utilizes the limited inner bark resources. In addition, it is possible that the early-season attack pattern of T. brevipilosus could enhance the reproductive success of T. yunnanensis and T. minor by attacking unoccupied areas of the tree trunk and thereby further reducing any residual host tree resistance. However, later in the flight season, when T. yunnanensis and T. minor brood are already well advanced, the arrival of T. brevipilosus would likely not help much in further reducing host vigor.

In those cases when T. brevipilosus initiates attack by itself, the first attacks are generally located along the lower trunk with subsequent attacks occurring throughout the entire trunk. The within-tree attack pattern varies among other Chinese Tomicus species as well (Ye and Ding 1999). In Yunnan, the first attacks by T. yunnanensis tend to be along the upper trunk, whereas T. minor attacks are more concentrated along the lower trunk (Ye 1995, Ye and Ding 1999, Chen et al. 2010). In contrast, in northern Europe, Långström (1983) reported that T. piniperda tended to initiate attack along the lower trunk of local pines (mostly Pinus sylvestris L.), where the bark was thicker, whereas T. minor, which flies later, tended to colonize the upper portions of the trunk and branches where the bark is thinner. In the above two situations, it appears that T. minor is simply colonizing trees already infested by other Tomicus species but concentrating its attack on those portions of the trunk that are less occupied and thereby reducing interspecific competition. It is reasonable to assume that the attack pattern in any one world area is a reflection of the Tomicus and host-tree species present, as well as differences in beetle aggressiveness, threshold temperatures for flight, and adult body size (Kirkendall et al. 2008, Li et al. 2010, Lieutier et al. 2015). For the three Tomicus species infesting P. yunnanensis in Yunnan, the adult beetles are broadly similar in body size, with T. yunnanensis being slightly larger, and T. brevipilosus and T. minor being very similar (Kirkendall et al. 2008, Lieutier et al. 2015).

It is unclear why the attack pattern differs between T. yunnanensis (starting in the upper trunk) and T. brevipilosus (starting in the lower trunk) when each is the first bark beetle to infest the same species of pine, although such variation in attack pattern is well-recognized among bark beetles (Rudinsky 1962). Nevertheless, both attack patterns can result in tree death. This topic still requires further exploration.

The average length and number of eggs in T. brevipilosus maternal galleries reported in this study are broadly similar to findings reported for other Tomicus species that make longitudinal galleries (all species except for T. minor). For example, in comparison to the 7.2-cm-long average T. brevipilosus maternal gallery in this study, others have reported mean values of 4–8 cm for T. destruens (Faccoli 2007), 7.1–11.0 cm for T. piniperda (Långström and Hellqvist 1985), and 9.6–10.6 cm for T. yunnanensis (Ye and Ding 1999). However, an inverse relationship has been reported between attack density and average length of maternal galleries in some Tomicus species (Salonen 1973, Långström 1984, Faccoli 2009), but this relationship has not yet been investigated in T. brevipilosus. Clearly much more research is needed to fully elucidate the life history of T. brevipilosus and its interactions with other Tomicus species in Yunnan.