Introduction
The cabbage seedpod weevil, Ceutorhynchus obstrictus (Marsham) (Coleoptera: Curculionidae), is an invasive alien species of European origin and an important pest of canola, Brassica napus Linnaeus and Brassica rapa Linnaeus (Brassicaceae). It has become widespread in canola-growing regions of North America (Dosdall and Mason Reference Dosdall and Mason2010) and costs Canadian canola producers in excess of $300 million (Canadian dollars) in losses annually (Colautti et al. Reference Colautti, Bailey, Overdijk, van Amundsen and MacIsaac2006). A classical biological control project is assessing the need and potential for introduction of exotic parasitoids to suppress populations of C. obstrictus and provide a reduced-risk option to prevent losses. Previous studies documented the Hymenoptera parasitoid communities associated with C. obstrictus across Canada (Dosdall et al. Reference Dosdall, Ulmer, Gibson and Cárcamo2006, Reference Dosdall, Gibson, Olfert and Mason2009; Gillespie et al. Reference Gillespie, Mason, Dosdall, Bouchard and Gibson2006; Mason et al. Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011) and in Georgia, United States of America (Gibson et al. Reference Gibson, Gates and Buntin2006). These show that the communities vary between regions and consist of multiple species, most being native or putatively Holarctic, although overall parasitism was low. Three Palaearctic species of Pteromalidae (Hymenoptera) were also reported: Stenomalina gracilis (Walker), introduced into British Columbia, Canada in the 1940s (Gillespie et al. Reference Gillespie, Mason, Dosdall, Bouchard and Gibson2006) and Mesopolobus gemellus Baur and Muller and Trichomalus perfectus (Walker), both accidental introductions into eastern Canada (Mason et al. Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011).
An important component of any biological control project is to assess the host-range of candidate agents as a measure of their safety (Bigler et al. Reference Bigler, Babendreier and Kuhlmann2006; van Lenteren et al. Reference van Lenteren, Bale, Bigler, Hokkanen and Loomans2006). In an area of introduction, in this case Canada, an important first step is to document the parasitoid assemblage associated with potential nontarget species to provide baseline data that can be compared with post-release monitoring data to assess impacts on target and nontarget species should the candidate agent be released. Thus, the objectives of this study were to determine the larval parasitoid complexes associated with Ceutorhynchus Germar weevils that are potential nontarget hosts of candidate agents for biological control of C. obstrictus in Ontario and Québec, and to assess the host range of these parasitoids.
Materials and methods
Surveys were conducted from 2006 to 2011 in canola-growing areas in southern Ontario and southern Québec, Canada where C. obstrictus was present or could potentially occur (Mason et al. Reference Mason, Baute, Olfert and Roy2004, Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011). Locations sampled for weevils associated with Brassicaceae other than B. napus are shown in Figure 1.
Figure 1 Sample locations in 2006–2011 for Brassicaceae plant hosts of Ceutorhynchus species (A) Erysimum cheiranthoides – Ceutorhynchus omissus: • (present), ○ (absent); (B) Rorippa species – Ceutorhynchus neglectus: • (present), ○ (absent); Erucastrum gallicum – Ceutorhynchus neglectus: ★ (present), ⋆ (absent); (C) Sinapis arvensis – Ceutorhynchus americanus: • (present), ○ (absent); Rorippa species – Ceutorhynchus americanus: ★ (present), ⋆ (absent); (D) Capsella bursa-pastoris – Ceutorhynchus typhae: • (present), ○ (absent); Capsella bursa-pastoris – Ceutorhynchus erysimi: ★ (present), ⋆ (absent).
Sites were visited when the plants (see Table 1) were in the silique stage of development (similar to canola Stage 73–79 of Lancashire et al. Reference Lancashire, Bleiholder, van den Boom, Langelüddeke, Strauss and Weber1991). Plant samples were taken at different times during the season, depending on plant component and plant species. For example, wild mustard, Sinapis arvensis Linnaeus (Brassicaceae) is an annual with a single generation each year and siliques were present in early to mid-July (Mulligan and Bailey Reference Mulligan and Bailey1975). In contrast, shepherd's-purse, Capsella bursa-pastoris (Linnaeus) Medikus (Brassicaceae) flowers from early spring to late fall (Alex Reference Alex1998) and siliques were present from June to September. Presence of adult weevils was determined with a standard sweep net when the plants were flowering (Stage 61–65) by sweeping 25–100 times (depending on stand size) in a 180o arc while walking through a stand of plants. Samples were placed in 70% ethanol and processed in the laboratory.
Table 1 Weevil species collected in Ontario and Québec, Canada in 2006–2011, host plants and feeding niche.
*Accidentally introduced and established in North America.
To determine larval weevil infestation levels and parasitism levels, whole plants at the silique stage (73–79) were randomly selected in each stand, bagged and processed in the laboratory where root crowns, stems, and siliques harvested from racemes were separated from plants. As described in Mason et al. (Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011) samples (root crowns, stems, siliques) were placed in 30×30×30 cm cardboard emergence boxes containing moistened fine vermiculite as a pupation substrate. A 2-cm hole in one side of the box allowed adult weevils and parasitoids to exit into plastic containers. The emergence boxes were placed in a rearing room maintained at 22 ± 1 °C and 18L:6D (light:dark hours). Collection containers were inspected every one to two days until insects ceased to emerge from the boxes for 30 consecutive days. Emerged insects were placed in vials containing 70% ethanol, and weevils and parasitoids were sorted and counted. Percent parasitism was estimated based on numbers of dead weevil larvae and parasitoid pupae found in the vermiculite substrate and emerged weevil and parasitoid adults. Because of the large numbers of pods collected, it was impractical to dissect each pod to determine if weevil remains or parasitoid pupae were present. Thus, estimates of parasitism may be slightly lower than actual. Parasitoids were critical-point dried and identified by G.A.P. Gibson, and weevils identified by P. Bouchard. Voucher specimens of parasitoids and weevils are deposited in the Canadian National Collection of Insects, Arachnids, and Nematodes in Ottawa, Ontario, Canada. Specimen labels include site codes.
Means and standard errors were calculated using PROC Means of the SAS statistical software (SAS 2010).
Results
Seven weevil species, Ceutorhynchus americanus Buchanan, Ceutorhynchus erysimi (Fabricius), Ceutorhynchus neglectus Blatchley, Ceutorhynchus omissus Fall, Ceutorhynchus pallidactylus (Marsham), Ceutorhynchus rapae Gyllenhal, and Ceutorhynchus typhae (Herbst) (Coleoptera: Curculionidae) were reared from siliques, stems, and root crowns of various Brassicaceae plants (Table 1). Among these species, C. americanus, C. neglectus, and C. omissus are native to North America and are associated with the native plants marsh-cress, Rorippa palustris (Linnaeus) Besser, dog mustard, Erucastrum gallicum (Willdenow) Schultz, and wormseed mustard, Erysimum cheiranthoides Linnaeus (Brassicaceae). Ceutorhynchus americanus was also associated with the nonnative wild radish, Raphanus raphanistrum Linnaeus (Brassicaceae) and wild mustard. The nonnative, C. erysimi, C. rapae, and C. typhae, were associated with shepherd's-purse, a nonnative weed associated with these species in Europe. Ceutorhynchus rapae was also reared from the native wormseed mustard and tall wormseed mustard, Erysimum hieraciifolium Linnaeus (Brassicaceae). Ceutorhynchus pallidactylus was reared from wild radish.
Collections of host plants did not yield Ceutorhynchus species at all sites and in all years (Fig. 1, Tables 2 and 3), indicating that distributions are not uniform and/or occurrence varied during the season (i.e., samples collected from the same site at different times did not always yield specimens). Density was variable where weevils were present (Tables 2–3), possibly a reflection of small sample sizes or variable levels of attack. The stem-feeding C. americanus was rare, found in Ontario in marsh-cress in 2008 and 2010, and in the invasive wild radish in 2008, although it was found in wild mustard, in Ontario in 2007, 2008, and 2010, and in Québec in 2007. Ceutorhynchus rapae, also a stem feeder, occurred in shepherd's-purse in 2008, in wormseed mustard in 2010, and tall wormseed mustard in 2011 in Ontario, and in 2007 in Québec.
Table 2 Mean density (±SE) and mean parasitism (±SE) of Ceutorhynchus species weevils collected from Capsella bursa-pastoris (Linnaeus) Medikus, Erysimum cheiranthoides Linnaeus, Erucastrum gallicum (Willdenow) Schulz, Rorippa palustris (Linnaeus) Besser, and Sinapis arvensis Linnaeus in Ontario, Canada 2006–2011Footnote * (n indicates the number of samples from which the value was calculated; and “.” signifies data not collected).
* Not shown are results for: C. americanus from Raphanus raphanistrum Linnaeus in 2008 [3/stem (n = 1)]; C. rapae from Capsella bursa-pastoris in 2008 [0.3 ± 0.1/stem (n = 2)]; C. rapae from Erysimum cheiranthoides in 2010 [0.5 ± 01/stem (n = 2)]; C. rapae from Erysimum hieraciifolium Linnaeus in 2011 [0.1/stem (n = 1)].
† Number per root crown, per stem or per 10 siliques.
Table 3 Mean density (±SE) and mean parasitism (±SE) of Ceutorhynchus species weevils collected from Capsella bursa-pastoris (Linnaeus) Medikus, Erysimum cheiranthoides Linnaeus, Rorippa palustris (Linnaeus) Besser, and Sinapis arvensis Linnaeus in Québec, Canada 2007–2010Footnote * (n indicates the number of samples from which the value was calculated; and “.” signifies data not collected).
* Not shown are results for Ceutorhynchus rapae from Erysimum hieraciifolium in 2007 [0.2/stem (n = 1)].
Parasitoids were recovered from all host plant species associated with all Ceutorhynchus species (Tables 2–3), except C. americanus and C. pallidactylus from wild radish. Parasitism levels were variable and the small sample sizes indicate that these do not necessarily reflect actual levels. At least 18 larval parasitoid species were collected from six putativeFootnote 1 host weevil species (Table 4). Among these, the native C. neglectus and C. omissus, and the adventive C. typhae were associated with the highest numbers of parasitoids. Parasitoid species known to be associated with weevils were also reared from unidentified hosts in collections of siliques of the Brassicaceae species field pepper-grass, Lepidium campestre (Linnaeus) Aiton (Necremnus tidius Footnote 2 (Walker) (Hymenoptera: Eulophidae) and Lyrcus Walker species (Hymenoptera: Pteromalidae)), tumble mustard, Sisymbrium altissimum Linnaeus (M. gemellus Baur and Muller (Hymenoptera: Pteromalidae)) and siliques and stems (one species each) of yellow rocket, Barbarea vulgaris Aiton (Eupelmus vesicularis (Retzius) (Hymenoptera: Eupelmidae) and Chlorocytus Graham species (Hymenoptera: Pteromalidae)).
Table 4 Parasitoid species (number of specimens) reared from Ceutorhynchus species hosts collected from Brassicaceae plant hosts 2006–2011.
*Putatively Holarctic.
†Accidentally introduced and established in North America.
Mesopolobus gemellus was associated with the greatest number of Ceutorhynchus host species (six), followed by, an unnamed species of Eulophidae and E. vesicularis (five each), N. duplicatus, Chlorocytus species, and Mesopolobus moryoides Gibson (four each), Trimeromicrus maculatus GahanFootnote 3 and Pteromalus Swederus species 1 (Pteromalidae) (three each), N. tidius, Eurytoma tylodermatis Ashmead, and Eurytoma Illiger species (Eurytomidae), and Eurydinoteliodes perdubia (Girault)Footnote 4 and Trichomalus lucidus (Walker) (Pteromalidae) (two each), an unnamed Encyrtidae species, Euderus glaucus Yoshimoto (Eulophidae), Lyrcus nigroaeneus Ashmead, Lyrcus species, and T. perfectus (Walker) (Pteromalidae) (one each). Not all species associated with weevil hosts were reared from that host in each year.
Parasitoids associated with native weevils
Ceutorhynchus americanus was host to six species of parasitoid. Eupelmus vesicularis and T. lucidus were reared from stems of marsh-cress collected in Ontario, whereas Eulophidae species, Chlorocytus species, M. gemellus, and M. moryoides were reared from stems of wild mustard (Table 4). Among these, M. gemellus was the most abundant, accounting for 50% of the 18 individuals reared from collections of wild mustard in Ontario. The only species reared from Québec collections of wild mustard was an unidentified Eulophidae.
Ceutorhynchus neglectus yielded parasitoids in all years, although the assemblage and proportions of species varied (Table 4). Overall, 12 parasitoid species were reared from this silique specialist. All species were present in Ontario and the greatest diversity was observed in 2010 when seven species were reared. Over all years, T. lucidus was most prevalent, representing 22.1% of the 12 parasitoid species (n = 77 individuals) reared from marsh-cress in Ontario. In Québec, 71 individuals of six species, N. duplicatus, Eulophidae species, E. vesicularis, M. gemellus, M. moryoides, and T. lucidus were also reared on marsh-cress. Among these, M. moryoides constituted 45.1% of the parasitoids reared, followed by N. duplicatus at 33.8%. The dominant species differed among years, for example T. lucidus was most abundant in 2007, N. duplicatus in 2008, and M. moryoides in 2009. Parasitoids were only reared (n = 76 individuals) from dog mustard collected in Ontario and of the five species M. moryoides represented the greatest proportion at 75%.
Ceutorhynchus omissus was host to the greatest diversity of parasitoids, accumulating 13 species in Ontario and five in Québec populations (Table 4). Mesopolobus gemellus was the dominant species, comprising 41.6–75% of parasitoids reared from siliques collected in Ontario in 2006 and 2008–2011. In 2010, 142 of 300 individuals reared were M. gemellus, while M. moryoides (77 individuals) and N. duplicatus (56 individuals) were also abundant. Necremnus duplicatus was reared in 2006 and 2008–2011 in Ontario. A single individual of T. perfectus was reared from siliques of wormseed mustard in Ontario in 2011. In Québec, samples in 2007 yielded five parasitoid species (n = 26 individuals), M. gemellus (42%), N. duplicatus (15%), an unidentified Eulophidae and T. lucidus (11.5% each), and E. tylodermatis (4%). Only M. gemellus was collected from Québec samples in other years (2008 and 2009) and overall, of the 46 individuals reared, it was the most prevalent parasitoid at 63%. Trichomalus lucidus was collected in Québec in 2007 (three individuals) and 2010 (two individuals).
Parasitoids associated with nonnative weevils
Ceutorhynchus erysimi yielded three parasitoid species (Table 4), all from Ontario collections in 2010 and 2011, and all introduced species. Among the 26 individuals reared, N. tidius was most prevalent at 92.3%. The native N. duplicatus was not reared from C. erysimi. Collections were not made from 2006–2009 as the feeding niche was not known.
Ceutorhynchus rapae yielded single specimens of E. perdubia and M. gemellus from stems of shepherd's-purse and tall wormseed mustard, respectively, from two sites in Ontario and one in Québec. In addition, two Eulophidae species specimens were reared from stems of shepherd's-purse in Ontario.
Ceutorhynchus typhae was host to 10 parasitoid species (Table 4); the entire assemblage was reared from samples collected in Ontario in 2008. Necremnus duplicatus and N. tidius as well as M. gemellus were present in all years. In 2006 M. gemellus dominated, comprising 43.3% of the 127 parasitoid individuals reared, followed by N. duplicatus plus N. tidius at 33.9%, T. maculatus at 18.1%, E. vesicularis (2.4%), Chlorocytus species (1.6%), and M. moryoides (0.8%). Similarly, in 2008 N. duplicatus plus N. tidius dominated, comprising 36.8% of the 133 parasitoid individuals reared, followed by M. gemellus at 18%, T. maculatus at 16.5%, M. moryoides at 13.5%, E. perdubia (4.5%), unidentified Eulophidae (3.0%), Chlorocytus species (2.3%), Lyrcus species (2.3%), Pteromalus species (2.3%), and E. vesicularis (0.8%). In Québec, only M. gemellus (24 individuals in 2008) and an unidentified Eulophidae (one individual in 2007) were reared from shepherd's-purse siliques.
Discussion
The larval host plant associations and feeding niches determined for C. neglectus and C. omissus are reported for the first time. The association of C. americanus with the native marsh-cress and the alien wild mustard are also new. Previously this species was reported to feed on roots and stems of turnip, B. rapa Linnaeus, cultivated radish, Raphanus sativus Linnaeus (Brassicaceae) (Beirne Reference Beirne1971), whitetop, Lepidium draba Linnaeus (Cripps et al. Reference Cripps, Schwarzländer, McKenny, Hinz and Price2006), all alien species, and the native poor-man's-pepper-grass, L. virginicum Linnaeus (Buchanan Reference Buchanan1937). Further work is needed to clearly define the life cycles of C. americanus, C. neglectus, and C. omissus in order to develop rearing methods for host-range studies of potential biological control agents and to track the impact of accidentally or intentionally introduced natural enemies of the cabbage seedpod weevil, C. obstrictus. The documentation here of C. americanus emerging from wild mustard combined with earlier reports of this species feeding on turnip, cultivated radish and whitetop suggests that some native weevils are capable of adapting to new host plants. Although as Cripps et al. (Reference Cripps, Schwarzländer, McKenny, Hinz and Price2006) report, C. americanus was the least abundant of the 11 species encountered during surveys for natural enemies of whitetop. Of the 73 known Ceutorhynchus species in North America details of the biology of only a handful of species are known, as Scheibner (Reference Scheibner1963; 5) stated “… primarily because many of the species are apparently uncommon…. Knowledge is more nearly complete for the species which occur more commonly, among which are those of economic importance”. However, even for these latter species, such as the invasive C. erysimi, C. pallidactylus, C. rapae, and C. typhae, comprehensive information is not available and documentation that the species was reared from the host plant(s) noted is lacking. In addition to C. obstrictus, details of the life cycle of the stem-feeding C. subpubescens LeConte, a western species, and the silique-feeding C. neglectus, both on the alien flixweed, Descurainia sophia (Linnaeus) Webb (Brassicaceae), were provided by Dosdall et al. (Reference Dosdall, McFarlane and Palaniswamy1999, Reference Dosdall, Ulmer and Bouchard2007). The C. neglectus reported in this study is probably a different species than that studied by Dosdall et al. (Reference Dosdall, McFarlane and Palaniswamy1999) and studies using molecular methods are under way to verify this hypothesis (P.B., personal observation; see also Laffin et al. Reference Laffin, Dosdall and Sperling2005).
The parasitoid assemblages of the Nearctic weevil species, C. americanus, C. neglectus, and C. omissus in Ontario and Québec comprised complexes of species that vary from year to year. Of the 19 species reared, 16 are also associated with the cabbage seedpod weevil, C. obstrictus (Mason et al. Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011), which is an invasive alien species established in these provinces in the 1990s (Brodeur et al. Reference Brodeur, Leclerc, Fournier and Roy2001; Mason et al. Reference Mason, Baute, Olfert and Roy2004). Among the 16 species shared with C. obstrictus, 12 are putatively Nearctic and may form an assemblage that has adapted to this nonnative invader. These include N. duplicatus, now determined to be the only species of Necremnus reared from C. obstrictus in North America (G.A.P.G., personal observation). Recognition that N. duplicatus and not N. tidius parasitises C. obstrictus in North America helps explain the absence of Necremnus species from rearings of C. obstrictus in Europe. Among the remaining four parasitoid species, two, E. vesicularis and T. lucidus, are putatively Holarctic or Palaearctic species that invaded North America some time ago. We hypothesise that M. gemellus and T. perfectus are recent introductions, the latter thought to have been introduced into Québec with C. obstrictus (Mason et al. Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011).
The parasitoid assemblage of the European invader C. typhae consists of 10 species. It is the only Ceutorhynchus species from which both the Palaearctic N. tidius and Nearctic N. duplicatus were reared. The only other Ceutorhynchus species from which N. tidius was reared is C. erysimi, suggesting that N. tidius was introduced into North America along with one of these two Palaearctic weevil species and that it has not adapted to attack native weevils. The two Necremnus species along with M. gemellus (a probable introduction) are the dominant parasitoids of C. typhae. The Nearctic species T. maculatus and M. moryoides are also abundant and appear to have adapted to C. typhae. Muller et al. (Reference Muller, Dosdall, Mason and Kuhlmann2011) studied the parasitoids associated with C. typhae in Germany, Switzerland, and France, and found that only M. gemellus and S. gracilis were associated with this seed-feeding weevil. The former species comprised 98% of individuals, whereas only a single specimen of S. gracilis was reared. Our results show that the Palaearctic N. tidius and E. vesicularis, both widely distributed and highly polyphagous, also parasitise C. typhae and suggest that these parasitoids may not occur in the region where Muller et al. (Reference Muller, Dosdall, Mason and Kuhlmann2011) conducted their European study. Furthermore, our study supports the Muller et al. (Reference Muller, Dosdall, Mason and Kuhlmann2011) findings that T. lucidus is not associated with C. typhae.
The parasitoid assemblages of the stem-feeding C. americanus (six species) and C. rapae (three species),Footnote 5 and the root crown/leaf mid-rib feeder C. erysimi (three species) are small compared with the diversity associated with weevils that feed in the siliques of their host plants (10–14 species). This suggests that the latter weevil species are more vulnerable to attack by parasitoids, although the small sample sizes for these species may have underestimated the diversity of associated parasitoids.
The finding that M. gemellus is associated with all Ceutorhynchus species reared during our study is remarkable. This European species, when originally described was thought to be associated with C. typhae, except for a single female from Ceutorhynchus turbatus Schultze (Coleoptera: Curculionidae) (Baur et al. Reference Baur, Muller, Gibson, Mason and Kuhlmann2007). It was recently associated with C. obstrictus in eastern Canada (Mason et al. Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011) and our results suggest M. gemellus is a generalist parasitoid of silique-infesting and stem-infesting and possibly root crown-infesting Ceutorhynchus species weevils. Mesopolobus gemellus may have been introduced with C. typhae, although molecular studies are needed to substantiate this hypothesis. Given this broad host range, that M. gemellus has not been reported elsewhere in North America (including from C. obstrictus [Mason et al. Reference Mason, Miall, Bouchard, Gillespie, Broadbent and Gibson2011]) suggests it may be a recent invader and tracking the spread and impact of this species on native Ceutorhynchus species will be important. Furthermore, if its primary host is indeed C. typhae and attack of other Ceutorhynchus species is a spillover effect due to high populations of this host, M. gemellus may not establish in western North America because C. typhae does not occur there (Bousquet et al. Reference Bousquet, Bouchard, Davies and Sikes2013), despite the ubiquitous distribution and abundance of its host plant, C. bursa-pastoris (United States Department of Agriculture 2013).
Mesopolobus gemellus, thought to be a recent adventive introduction, has a broad host range in North America, attacking all host species collected, including the native C. americanus, C. neglectus, and C. omissus. In contrast, only a single specimen of T. perfectus, a candidate biological control agent, was collected from the native C. omissus. Trichomalus perfectus may be a more recent adventive introduction than M. gemellus and thus yet to manifest its potential impacts fully. Tracking of M. gemellus and T. perfectus is recommended to monitor their spread and impacts on nontarget species.
The findings reported here provide new information on the host plants and feeding niches of the native C. americanus, C. neglectus, and C. omissus and the first documentation of parasitoids associated with each species. The results also show that the parasitoid assemblage of silique feeders is more diverse than those of stem or root crown feeders (although additional data is needed to verify this hypothesis), that native parasitoids adapt to introduced weevil hosts, and some, not all, introduced parasitoids adapt to native weevil hosts. These have important implications for programmes involving the introduction of weevils for biological control of weeds as well as for biological control of pest weevils.
Acknowledgements
Sara Abdallah, Margaret Aslanian, Ahmed Badiss, Lisa Bartels, Ana Maria Farmakis, Gabriel Gregorich, Tom Hay, Shashi Juneja, Melanie Lacroix, Kathryn Makela, and Michael Wogin provided technical assistance. Suzanne Warwick and Gerry Mulligan are gratefully acknowledged for expert verification of plant identifications. This project was funded in part by the Agriculture and Agri-Food Canada, Pest Management Centre Improving Farming Systems and Practices Initiative Project PRR03-370.
