Citizen science reveals the establishment of Chamaesphecia empiformis (Esper) (Lepidoptera: Sesiidae), a long-lost biological control agent for Euphorbia cyparissias (Euphorbiaceae), in Ontario, Canada

Abstract

The European root-boring moth, Chamaesphecia empiformis (Lepidoptera: Sesiidae), was released in Ontario, Canada, in 1971 and 1989 as a biological control agent for the perennial invasive weed cypress spurge, Euphorbia cyparissias Linnaeus (Euphorbiaceae). Until recently, the moth was believed to have failed to establish. Beginning in 2015, images posted on citizen science platforms, including BugGuide, iNaturalist, and Facebook groups, indicated that C. empiformis was in fact established. We have confirmed its presence in eastern Ontario by morphological and molecular identification of field-collected adults and eggs. This may be the longest-known lapse between the release of a weed biological control agent and confirmation of its establishment. Citizen science may provide valuable records documenting the establishment and dispersal of biological control agents.

Citizen science and social media platforms are providing new opportunities for nonprofessional naturalists to make valuable contributions to scientific knowledge (e.g., Osawa et al. 2017; Hiller and Haelewaters 2019; Martel et al. 2022; Angelidou et al. 2023). For entomological observations, popular platforms include BugGuide and iNaturalist. In July 2023, BugGuide (www.bugguide.net) held more than 1.9 million images representing more than 44 000 species of arthropods from the United States of America and Canada, while iNaturalist (www.inaturalist.org) held more than 132 million observations of all taxa around the world, including 18.6 million observations of 42 856 arthropod species from the United States of America and more than 3.1 million observations of 17 453 species from Canada. Numerous groups on Facebook (www.facebook.com) also report on natural history observations for particular geographic areas or taxonomic groups. Here, we report that observations by users of these citizen science platforms have revealed the establishment of a European moth, released in Canada as a biological control agent, that had gone undetected for possibly more than 40 years since its release.

Cypress spurge, Euphorbia cyparissias Linnaeus (Euphorbiaceae), is a perennial European herb that is widely distributed in Ontario and elsewhere in North America (Fig. 1) and is considered weedy or invasive (Stahevitch et al. 1988). It occurs in cultivated fields, in waste areas, and along roadsides and is toxic to livestock, including horses, goats, cattle, and sheep; milk from cattle grazing on E. cyparissias is tainted with a reddish colour and bitter taste (Ontario Ministry of Agriculture, Food and Rural Affairs 2021). The plant is listed as a noxious weed in Ontario (Government of Ontario 2014), Manitoba (Province of Manitoba 2017), and Saskatchewan (Saskatchewan Minister of Agriculture 2010), in Canada, and as a prohibited invasive species in New York (State of New York 2020) and Connecticut (Connecticut General Assembly 2019), in the United States of America. Euphorbia cyparissias, along with leafy spurge, Euphorbia virgata Waldstein and Kitaibel, was identified as a target for biological control in Canada in 1961. Leafy spurge was traditionally referred to in most North American literature as Euphorbia esula Linnaeus, but the correct name for the North American invasive taxon is now considered to be E. virgata (Riina et al. 2016).

Figure 1. Release sites, iNaturalist observations, and collection sites of Chamaesphecia empiformis in Ontario, Canada, with distribution of Euphorbia cyparissias from Global Biodiversity Information Facility (2021).

Several root-feeding moths in the genus Chamaesphecia, including Chamaesphecia empiformis (Esper) (Lepidoptera: Sesiidae), have been identified and screened as potential biocontrol agents (Schröder 1970). Ten species of Chamaesphecia in Europe feed on Euphorbia spp., most being specific to one or a few Euphorbia species (Toševski et al. 1996). Chamaesphecia empiformis is highly host-specific to E. cyparissias. It was for a time synonymised with C. tenthrediniformis (Denis and Schiffermüller), which feeds on Euphorbia esula Linnaeus and E. salicifolia Host in Europe, but the two species were confirmed as distinct by Naumann and Schroeder (1980). Chamaesphecia empiformis is native across Europe and southern Russia to Kazakhstan and is common in dry sandy habitats. It is univoltine, adults being on the wing from mid-May to early August. Females lay eggs singly on the lower leaf surface of E. cyparissias. After hatching, larvae bore into the stems, first mining upwards and then turning down until they reach the root system, where they feed and overwinter. In early spring, the larvae make an exit tunnel reinforced with silk and root fibres in which they pupate (Toševski et al. 1996).

Chamaesphecia empiformis was approved in 1970 for release in Canada as a biological control agent for E. cyparissias and was released in Ontario in 1970, 1971, and 1989 (Harris 1984; Bourchier et al. 2002). According to Alex (1992), these releases were made using eggs imported from the then Commonwealth Institute of Biological Control (now CABI; www.cabi.org) in Switzerland. These eggs were hatched in quarantine, and larvae were transferred to potted plants, which were then transplanted to the field. The original shipment records on file at Agriculture and Agri-Food Canada show that 23 boxes of infested E. cyparissias roots were buried for overwintering at the Belleville, Ontario laboratory in December 1970. In 1971, three releases were made, presumably of the next generation reared from these imported larvae. At Picton, Ontario, southeast of Belleville, 800 eggs and 100 larvae were released on 29 June 1971. (This release was listed as C. tenthrediniformis by Harris (1984), but the original release forms show C. empiformis.) The shipment record shows the target at this site as Euphorbia esula. In May 2022, a visit to this site by ASM confirmed that the spurge species growing there is leafy spurge (E. virgata) and not E. cyparissias. On 12 July 1971, 648 eggs and 17 larvae were released on E. cyparissias at the Field Station in Sidney Township, Ontario, and on 19 July 1971, 1184 eggs were released at Ewing Farm; the target weed was not identified on the release form but likely was E. cyparissias. We have not been able to identify the exact locations of the “Field Station” or “Ewing Farm.” A further shipment of C. empiformis collected in Lower Austria in July 1983 was sent to J. Alex at Guelph University (Guelph, Ontario) in September 1983 but do not appear to have been used for release. A second series of releases took place in 1989, when 596 larvae reared from material collected in Austria and Yugoslavia were transplanted in roots of potted E. cyparissias plants to a release site at Melancthon, Dufferin County, Ontario in July and August by Alex.

There was no confirmation that the Ontario releases had resulted in the establishment of C. empiformis when biological control projects in Canada were reviewed up until 2000 (Alex 1992; Bourchier et al. 2002), and no further monitoring efforts have been reported since. Thus, until recently, it was assumed that C. empiformis had failed to establish in Canada.

In 2015, a photograph taken near Tweed, Ontario, submitted to BugGuide by user Jason King (https://bugguide.net/node/view/1164121), appeared likely to be C. empiformis, suggesting that the species might in fact be established in Ontario. A further observation of a mating pair of Chamaesphecia at Tweed on 25 June 2019 was posted on iNaturalist by user Joe Bartok (https://www.inaturalist.org/observations/35198558). In July 2020, one of us (CEM) photographed a sesiid near Burnstown, Ontario, and posted the image to the Insects and Arachnids of Ontario Facebook Group (https://www.facebook.com/groups/121814901322513). ASM recognised that this photograph was also possibly of a Chamaesphecia and reposted it to BugGuide (see https://bugguide.net/node/view/1855092).

Following up on these observations, ASM and CEM collected one adult Chamaesphecia sp. on E. cyparissias near Burnstown, Ontario, on 20 July 2020 and three further specimens on 1 August 2020. Several eggs were collected on leaves of E. cyparissias on each of these dates. Eggs were found laid singly on the lower leaf surfaces, as described by Toševski et al. (1996). Two further adults were collected by ASM on 11 June 2022 on E. cyparissias at Burnt Lands Provincial Park, Ontario. The pinned adult specimens were provided to JFL for confirmation of identification, and the eggs were sent to IT for molecular identification. Several additional observations of C. empiformis have since appeared on iNaturalist, as well as one from Haliburton County that was identified as C. tenthrediniformis. Given the uncertainty over whether C. tenthrediniformis was ever actually released in Ontario (see above), we tentatively include this record as C. empiformis. See Figure 1 and Supplementary material, Table S1 for locations of release and recovery sites mentioned in the text.

Details of collected Ontario specimens preserved in the Canadian National Collection are as follows:

• 1M: Burnstown, 45.375° N, 76.601° W, 20.vii.2020, on Euphorbia cyparissias, Alec S. McClay (specimen CNC759002).

• 3F: Burnstown, 45.375° N, 76.601° W, 1.viii.2020, on Euphorbia cyparissias, C.E. Morris and Alec S. McClay (specimens CNC759003–CNC759005).

• 2F: 5 km NE Almonte, 45.4333° N, 76.25° W, 11.vi.2022, swept on Euphorbia cyparissias, Alec S. McClay (specimens CNCLEP00309616–CNCLEP00309617)

In external morphology, the Ontario specimens agree with diagnostic characterisation of C. empiformis as described by Toševski et al. (1996), Špatenka et al. (1999), and Laštůvka and Laštůvka (2001). One of the main diagnostic characters of Chamaesphecia species is a forewing posterior transparent area that is short or abbreviated, extending only to the middle of the anterior transparent area, without reaching the discal spot. Given that no native species of Chamaesphecia are known from North America, this character uniquely distinguishes C. empiformis.

In other genera of Synanthedonini, such as Carmenta and Synanthedon, both of which are species-rich in North America, the posterior transparent area is much longer and reaches, or is extended under, the discal spot. No other North American species with variously extensive black scaling in the forewing has a similar posterior transparent area that is short or abbreviated before the discal spot; therefore, based on this character alone, C. empiformis should be easily recognised even if other colour features are worn. See Supplementary material, Document S1 for further notes on comparative morphology.

The barcoding region of the mitochondrial cytochrome c oxidase subunit 1 gene (mtCO1) was used to confirm the identity of the sampled eggs from Burnstown, and adult specimens collected in Burnstown in 2020 and in Burnt Lands in 2022 were also barcoded (see Supplementary material, Document S1 for methods). The sequence data for eggs were deposited in the National Center for Biotechnology Information (NCBI) GenBank database (www.ncbi.nlm.nih.gov) under accession number ON908965. Specimen data and sequences for adults are available in the Barcoding of Life Datasystems (BOLD) in the public dataset DS-CHAMANA “Chamaesphecia in North America” (dx.doi.org/10.5883/DS-CHAMANA). Additional records were also compared by searching for the species name in the BOLD public data portal and analysed with the tools implemented in BOLD.

Sequence data from the eggs show 99% identity with C. empiformis (NCBI accession number KP253719) originating from Thueringen, in Vorarlberg state, Austria, (Huemer and Hebert 2015). Barcode sequences of Ontario adult specimens show at least 98.92% identity with several C. empiformis from Austria (TLMF Lep 09883), Germany (BC ZSM Lep 28573), and Bulgaria (CCDB-04608 F08). All Canadian and European specimens cluster into barcode index numbers (BIN) BOLD:AAC0677, within which is a fair amount of haplotype variation (0–3.26%, mean 1.70%, in pairwise comparisons). The Ontario specimens fall into two haplotype groups with a maximum divergence of 1.71%, one from each of the two different collecting sites. Haplotype variation among conspecific European specimens from Austria, Bulgaria, Germany, and Italy is also relatively high, ranging from 0–3.26% (mean 1.41%).

We have not had an opportunity to search for C. empiformis in Dufferin County, where the 1989 releases were made, and to date, no citizen science observations of the species have been posted from this area. However, given that the site where the first observations of C. empiformis were made in 2015 in Tweed, Ontario, is approximately 35 km northeast of the area where releases were made in 1971, it seems plausible that these 1971 releases were the source of the established population. If this is the case, the established population went undetected for 44 years. As far as we are aware, this is the longest such interval in the history of biological control of weeds. The closure of the Belleville Research Centre in 1972 (Castonguay 1998; MacQuarrie et al. 2016), with the subsequent transfer of Agriculture and Agri-Food Canada’s biological control of weeds research programme to the Regina Research Station in Saskatchewan, likely contributed to this long detection interval.

Our observations do not allow us to make any assessment of the impact of C. empiformis on the density or vigour of E. cyparissias populations other than to observe anecdotally that the collections were made in fairly dense stands of the plant that did not show obvious signs of damage. Further studies would be useful to evaluate the impact of C. empiformis and the benefits of rearing and distributing it as a biological control agent for E. cyparissias. Nor are we able to say if our findings of C. empiformis in eastern Ontario represent a recent upsurge in numbers and dispersal or if the species has long been widespread but unnoticed.

Given the rapid growth in the popularity of citizen science platforms for natural history observations, we expect that this will not be the last case where they provide valuable data on the establishment and spread of biological control agents. Lack of postrelease monitoring has often been identified as a failure in biological control programmes for invasive plants (e.g., Carson et al. 2008). We encourage biological control researchers and practitioners to be aware of the potential of citizen science to help fill this gap.