Red pines (Pinus resinosa) and scots pines (Pinus sylvestris) with resin beads characteristic of S. noctilio attack (Dodds et al. 2010) were located from October to December in New York and Pennsylvania. These trees were often suppressed and patchily distributed. In spring, infested trees were felled, areas of trunks with resin beads were cut into bolts, and ends of bolts were coated in sealant (Waxlor, Willamette Valley Company, Eugene, Oregon, USA). For the majority of studies, 70 cm long bolts were placed in fiber barrels (91 cm high × 61 cm diam) covered securely with window screening, in an unheated barn. Once adult wasp emergence began (in early to mid-July), barrels were checked daily for two months and then subsequently 3–4 times per week. Sirex that emerged were placed in 29 ml clear plastic cups at 4°C. All Sirex emerging were identified using characters described in Schiff et al. (2012). To cover the emergence periods for both Sirex species, checking for emergence finished in early November, and, at that time, barrels were completely checked for any additional specimens.

For studies comparing the abundance of Sirex species by tree species, Sirex were reared from 50 cm long bolts placed in horizontal cardboard rearing tubes. Bolts were spaced from tube sides using nails and were housed at ambient environmental conditions in a screened outdoor insectary. Glass emergence jars were fitted to the tube ends, oriented facing screened windows for a natural photoperiod, and emerging Sirex were collected daily.

To quantify the sizes of emerging adults, we followed Madden (1974) and Nahrung (2016) in using the apical width of the prothorax. Measurements were made using digital calipers (Traceable Products, Webster, Texas, USA). Infection with the parasitic nematode Deladenus siricidicola negatively impacts adult size (Kroll et al. 2011; Haavik et al. 2016a), so we diagnosed whether each of the adults being measured was parasitized by these nematodes. Males and females were dissected alive under a dissecting microscope at 10× magnification. The abdomen was removed and cut lengthwise to open it, without damaging the venom glands in females, and presence of nematodes within the abdominal cavity and reproductive tissues was recorded. Eggs from females were counted at 63× magnification and venom glands were removed and immediately weighed on a microscope cover slip.

To compare sizes of S. noctilio adults with (n = 92) and without (n = 412) nematodes, we used randomly selected individuals emerging from barrels in 2012–2014. There were far fewer S. nigricornis emerging during this period and nematode infections were uncommon in S. nigricornis, so analyses included only S. nigricornis of both sexes without nematodes (n = 36).

To assess weights of venom glands, a total of 51 S. noctilio were randomly selected in 2014–2016 for dissection and 21 of these contained nematodes. Since few female S. nigricornis emerged from rearings from the northeast during 2014–2016, 30 S. nigricornis females from Arkansas and Louisiana that had been caught in panel traps were used for venom gland weights. We were concerned that these S. nigricornis from the southeast both had been flying before being trapped and were from a different geographic region. To test the accuracy of using southeastern S. nigricornis as replacements for northeastern S. nigricornis, we compared the relationship between numbers of eggs and body size for S. nigricornis females collected from traps in the southeastern vs. emerging from wood from the northeastern USA. Southeastern S. nigricornis carried the same number of eggs in relation to prothorax width as S. nigricornis emerging from wood in the northeastern US (t = -1.10; P = 0.2713), so we used venom gland weights from southeastern S. nigricornis for comparisons.

Use of P. resinosa versus P. sylvestris by S. nigricornis and S. noctilio was quantified using infested trees in northern New York State, with site information in Foelker et al. (2016; Table 1) (2010: 10 P. resinosa, 12 P. sylvestris; 2011: 22 P. resinosa, 10 P. sylvestris). Trees were harvested from 21 April – 24 May in 2010 and 31 March – 7 May in 2011 and Sirex were reared from trees as described above.

On 4 June 2013, 12 Sirex-infested P. resinosa were collected from a plantation on River Road, Warren Co., New York (43°31'59.7"N 73°49'30.8"W). Trees were cut into 219 70 cm long bolts and woodwasps were reared as described above, with barrels checked every 1–2 days to collect individuals very soon after emergence; females were never found to be ovipositing when collected this soon after emergence. In November 2013, barrels were thoroughly checked for dead Sirex so that any first year emergers were not mistaken for second-year emergers. Barrels were stored in an unheated barn over the winter and were checked for emergence throughout the 2014 flight season. This procedure was repeated for a third year, through the 2014–2015 winter and 2015 flight season. As a continuation to this study, in Tioga Co., Pennsylvania, in spring 2014 we harvested 23 infested P. resinosa and, in spring 2015, 29 P. resinosa. Wood from trees cut in 2014 and 2015 was maintained and emergence was checked for two years.

Pines with resin beads indicative of S. noctilio attack were harvested in central and northern New York State and north-central Pennsylvania and Sirex were reared from them. Even within the same region, it was rare that the exact same site was sampled more than one year. Pinus resinosa were harvested from plantations in Tioga Co., Pennsylvania yearly, from 2011 and 2013–2015, with 20–30 trees harvested each year. Sirex were also reared from mixtures of P. resinosa and P. sylvestris from natural forests in northern New York State and from mature plantations in central New York State. In 2007 only P. sylvestris was sampled as described in Long et al. (2009). Trees included in analyses were those from which either one or the other or both Sirex species emerged.

A general linear mixed model with year as a random effect was used to compare sizes of Sirex. To compare body size versus venom gland mass, the significance of difference between slopes was calculated (Cohen et al. 2003). Comparisons of body size across species, sex and nematode parasitism were conducted using mixed models with year as a random effect and least squares means were used for post hoc comparisons (SAS 2014). Because very few S. nigricornis were reared that were parasitized by nematodes, comparing body sizes for individuals that were parasitized versus not parasitized was only possible for S. noctilio. Tree species use analysis was conducted using a mixed model with (ln +1)-transformed densities by volume of wood for S. nigricornis versus S. noctilio (SAS 2014). Sirex species, tree height, tree diameter and tree species were explanatory variables, with tree nested within site and site as random effects. The interaction of Sirex species with tree species was initially included but was removed as it was not significant. Wilcoxon signed ranks tests were used to compare numbers of S. noctilio versus S. nigricornis adults emerging from trees in 2007–2014.

Weights of venom glands increased with increasing body size (measured as prothorax width) (Fig. 1) but this relationship differed by species. The slope of the regressions of body size against venom gland weight for S. nigricornis without nematodes was less steep compared with S. noctilio without nematodes (t = 4.527; df = 56; P < 0.0001); this slope was close to twice as steep for S. noctilio compared with S. nigricornis. The weight of the venom glands for S. nigricornis females without nematodes also was significantly less than for S. noctilio with nematodes (t = 3.3824; df = 47; P = 0.0015). Sirex noctilio parasitism by nematodes did not significantly affect the relationship between the weight of the venom gland and body size, when compared with non-parasitized S. noctilio individuals (t = 1.1141; df = 47; P = 0.2709).

Figure 1.

Relationships between venom gland mass and body size, measured as pronotum width, for S. noctilio and S. nigricornis. Sirex noctilio data were analyzed by presence or absence of parasitism by nematodes (Deladenus siricidicola) while numbers of S. nigricornis parasitized by nematodes were too low for analysis. A Sirex noctilio females without parasitism by nematodes B S. noctilio females parasitized by nematodes, and C S. nigricornis without nematode parasitism.

For collections across 2012–2014, on average non-parasitized S. noctilio females (prothorax width: 3.90 + 0.83 mm) were larger than S. nigricornis females (2.46 + 0.35 mm) (t = 6.27; P < 0.0001) and non-parasitized S. noctilio males (2.92 + 0.83 mm) were larger than S. nigricornis males (2.38 + 0.56 mm) (t = 3.00; P = 0.0028). For male S. noctilio, nematode parasitism unexpectedly resulted in larger body sizes (t = -2.62; P = 0.0091) while this relationship was reversed for females (t = 2.34; P = 0.0195) (Fig. 2).

Figure 2.

Mean body size (+ SE), measured as pronotum width, for S. noctilio males or females parasitized by the nematode D. siricidicola or not. Different letters (comparing either capital letters or comparing lower case letters) demonstrate significant differences within sexes.

Sirex densities by tree species did not differ between the two Sirex species (F_1,52.6 = 0.01; P = 0.9395). The only main effect that was significant in the model was the comparison of densities of the two Sirex species (F_{1, 56} = 32.11; P < 0.0001; S. nigricornis density = 8.9 ± 2.6/m³, S. noctilio density = 44.3 ± 11.5/m³).

Nearly half of adult S. nigricornis emerged from wood during the first season in our rearings, over half emerged during year 2 (Table 1), and no S. nigricornis emerged during year 3. In contrast, emergence of S. noctilio adults primarily occurred in year 1 (Table 1). Numbers of S. noctilio emerging from wood in 2013–2014 were low but among these, 26.7% emerged during the second year. For wood harvested in 2014 or 2015 and reared for two years, from 1.5% to 10.9% of the S. noctilio emerged during year 2.

Between 2010 and 2015, S. nigricornis densities were lower than densities of S. noctilio in central New York State and north central Pennsylvania, where most pines that were sampled had been purposefully planted (Table 2). However, when sites in central and northern New York State were sampled within 2–3 years of the first reports of S. noctilio in those areas (see Table 2 footnotes), numbers of S. nigricornis emerging did not differ significantly from S. noctilio, although they were usually numerically lower. When S. nigricornis emerged, they more commonly co-occurred in the same trees as S. noctilio and rarely were the only Sirex species emerging from a tree. In 2015, no S. nigricornis emerged from infested material from north central Pennsylvania and during this study, densities of S. nigricornis from this same area were always low.

The overall body sizes of S. noctilio as well as the sizes of their venom glands were significantly larger than the bodies and venom glands of the native S. nigricornis. The venom glands of S. noctilio are also larger than those of seven other European siricids (Spradbery 1977) and experimentation has shown that venom from S. noctilio had greater phytotoxic activity than venom of these European siricids (Spradbery 1973). While a key component of the venom in the S. noctilio gland has been characterized as noctilisin (Bordeaux et al. 2014), the identities of compounds in the venom glands of S. nigricornis, or glands of any other siricids, are not known (Wang et al. 2016). Regardless, the fact that the S. noctilio venom glands are larger (relative to adult body size) than venom glands known from other siricids is consistent with the fact that this is the most aggressive siricid, reported as able to kill relatively healthy pines where it is adventive (see Suppl. material 1: A), while other woodwasps are considered relatively benign secondary pests in forests (Schiff et al. 2012).

Haavik et al. (2016a) found that S. nigricornis carries more eggs per body size than S. noctilio and our studies have documented the same trend (AEH and JCH, unpublished data). However, the present study demonstrated that on average S. noctilio females with or without nematode infections were larger than S. nigricornis females and the same relationship holds for males. So, although S. nigricornis has the potential to carry more eggs than S. noctilio, when we calculated the numbers of eggs per female based on sizes of females that emerged from wood in our studies, using the regression equations in Haavik et al. (2016a), the number of eggs for average-sized S. noctilio female was 166 but 138 eggs would be produced by S. nigricornis. Thus, the larger body size of S. noctilio emerging from trees compensated so that on average females of the invasive carried more eggs than S. nigricornis. It was unexpected that while nematode-infected S. noctilio females were smaller than healthy females, nematode-infected males were larger than healthy males, an association also found by Haavik et al. (2016a).

The densities of the two Sirex species emerging from P. resinosa and P. sylvestris did not differ by tree species. Studies have found a trend of P. sylvestris being colonized more frequently by S. noctilio compared with P. resinosa (Dodds et al. 2010; Foelker 2016) and S. nigricornis was not included in those studies. Largely based on the southern hemisphere where many North American pines have been introduced, susceptibility of pine species to S. noctilio varies by tree species (Ryan and Hurley 2012; Nahrung et al. 2015). Our result was somewhat unexpected because P. sylvestris is native to Europe, where it is assumed to have coevolved with S. noctilio (Ayres et al. 2014). However, we do not know exactly where the genotypes of S. noctilio introduced to New York and Pennsylvania originated and, given the broad native geographic range of S. noctilio, it is possible that these genotypes of this woodwasp did not co-evolve with P. sylvestris (Boissin et al. 2012; Bittner et al. 2017). On the other hand, P. resinosa and S. nigricornis co-evolved in North America and in this case there was no preference for the native pine over the introduced pine.

Sirex noctilio and S. nigricornis also differed significantly based on the percentages of the populations emerging from wood after 1 vs 2 years. In our studies, S. noctilio mainly emerged in year 1; from 1.5–27.7% of S. noctilio emerged the second year (Table 2) while in Ontario, Ryan et al. (2012b) found 2.2% emergence during year 2. Second-year emergence of S. noctilio was 15–25% in New Zealand and Australia, including Tasmania (Morgan 1968; Taylor, 1978), 1.6% in Eurasia (Spradbery and Kirk 1978), and 24% in Galicia, Spain (Lombardero et al. 2016). In all cases, percent emergence for S. noctilio in the second year of rearing was less than that recorded for S. nigricornis in our study, for which close to 50% emergence occurred in both years 1 and 2. Differences in voltinism between S. noctilio and S. nigricornis could have large impacts on their population dynamics. Based on our results, future studies of S. nigricornis, at least in northeastern North America, should allow two years for emergence from infested wood.

In northeastern North America S. noctilio is now the most abundant woodwasp attacking pines (Long et al. 2009; Ryan et al. 2012a; Foelker et al. 2016). Unfortunately, there are no records of the population densities or tree use by S. nigricornis before S. noctilio arrived but now, use of suppressed pines by S. nigricornis is much less than use of this resource by S. noctilio (e.g., Table 2).

Sirex noctilio has a temporal advantage over S. nigricornis as many emerge 1–2 months before S. nigricornis, although there is overlap in emergence between these species in northeastern North America (Ryan et al. 2012a; Haavik et al. 2013; Hartshorn et al. 2016; Suppl. material 1: B). Sirex noctilio is found colonizing trees at higher densities than S. nigricornis. Therefore, each year S. noctilio will occupy what is considered as being their preferred resources (i.e., suppressed trees that are not yet dead; see Suppl. material 1: A) before S. nigricornis adults have emerged. Among all pines from which Sirex were reared, only S. noctilio emerged from 56% (across the 9 years of this study). However, after the invasive had been present for a few years in an area, when the two species both emerged from trees, numbers of S. nigricornis were lower than S. noctilio and we very rarely found only S. nigricornis emerging from a tree. Based on the low numbers of S. nigricornis emerging from infested trees, we hypothesize that population densities of S. nigricornis emerging could be low because when adults of the native species emerge, the transient resource of suppressed trees had already mostly been exploited. However, S. nigricornis could have an advantage in situations where pines unacceptable to S. noctilio for some reason could be available for use by S. nigricornis or when pines could become weakened (e.g., by lightning strikes) after the main flight of S. noctilio adults and before or during the S. nigricornis flight time. In addition, the fact that the transient resource of suppressed trees are often patchily distributed (e.g., Ayres et al. 2014) could lead to coexistence of these species via differential spatial distribution of use of suppressed pines.

Monceau et al. (2015) have shown that some degree of niche differentiation between native and invasive hornets can minimize competition. We did not sample recently dead trees in this study but experiments have shown that S. nigricornis will oviposit into wood from trees that have recently been cut, although oviposition was minimal as soon as 30 days after cutting the wood (Hartshorn 2012). We do not know to what extent S. noctilio will choose to oviposit in wood from recently cut trees but when siricids were reared from diverse types of wood sampled over eight years from Europe, Turkey and North Africa, S. noctilio only emerged from samples from standing trees and timber and never from fallen trees or wood on the ground (Spradbery and Kirk 1978). While niches of these two Sirex species probably differ to some extent relative to the health or condition of pines that are preferred or acceptable for oviposition and development, there is also niche overlap (see Suppl. material 1: A). Based on relative densities and traits, we hypothesize that after the invasion of S. noctilio, S. nigricornis could develop in suppressed trees less often than prior to the invasion and the native might now more frequently use recently fallen, dead trees, as S. noctilio would have already exploited the majority of standing suppressed pines. Alternatively, suppressed pines further weakened by S. noctilio attack could at times provide an increased resource for S. nigricornis populations. Knowledge about normal densities of S. nigricornis before invasion by S. noctilio is necessary to further understand the impact of the native on the invasive woodwasp. In addition, further studies investigating the health of pines associated with oviposition and development of these two Sirex species are needed to clarify the overlap in niches of these two species with regard to tree health.