Dasineura mali identification

Dasineura mali specimens from Ontario, British Columbia, and Nova Scotia were identified using morphological features by a taxonomic expert (B. Sinclair, Canadian National Collection of Insects, Arachnids, and Nematodes) and matched the concept of D. mali described by Gagné; no morphological differences between the populations were found. Molecular analysis of D. mali specimens from Ontario (n = 182) and British Columbia (n = 710) showed very little genetic variation across the Cytochrome Oxidase Subunit 1 (COI) barcode region, ranging from 0% to 1.6% nucleotide variation (average p-distance = 0.54%) and were assigned the same barcode index number (AAL1254), indicating that they represent a single operational taxonomic unit or species. Deoxyribonucleic acid barcodes generated from D. mali specimens collected in Nova Scotia (n = 352) also showed low intraspecific variation in the COI barcode region, ranging from 0% to 1.4% (average p-distance = 0.02%). However, specimens from Nova Scotia represent a separate operational taxonomic unit and generated a different barcode index number (ACV3346) than those from Ontario and British Columbia, suggesting evidence of a cryptic species. The two barcode index numbers (AAL1254 and ACV3346) are the most closely related operational taxonomic units in BOLD (nearest neighbours) with a mean distance (K2P) of 3.7%, which is higher than the maximum intraspecific distances within each barcode index number (1.61% in AAL1254 and 0.71% in ACV3346). This may indicate separate species; however, a thorough analysis of D. mali populations from across Europe (the area of origin) may help clarify the level of intraspecific genetic variation within what is currently considered as D. mali and may shed light on whether or not these two barcode index numbers represent cryptic species.

Dasineura mali phenology

In the spring, emerging D. mali adults form mating swarms prior to ovipositing and they die three to five days later (Barnes Reference Barnes1948; Todd Reference Todd1959; Helson Reference Helson1972). In each year, D. mali appeared in pheromone traps beginning in April or May and were present until late September or early October, depending on location and year. For example, in 2016, first catches were recorded in April in Summerland, British Columbia, early May in Beamsville, southwestern Ontario, mid-May in Ottawa, eastern Ontario, and late May near Kentville, Nova Scotia (Fig. 2). The first egg-infested apple terminals were recorded the week of 18 April 2016 (232 degree days), in Summerland, British Columbia, 12 May 2016 (240 degree days) at Mountain View, southwestern Ontario, 25 May 2016 (260 degree days) in Ottawa, eastern Ontario, and 1 June 2016 (308 degree days) near Kentville, Nova Scotia (Fig. 3). From 2014 through 2016, oviposition of the first generation of D. mali eggs peaked in May (288–387 degree days) at Summerland, British Columbia, in late May and early June (321–391 degree days) in eastern Ontario (2015 and 2016 only) and the Niagara region of southwestern Ontario (363–403 degree days), and in mid-June (370–425 degree days near Kentville, Nova Scotia (Table 1, Fig. 3). MacPhee and Finnamore (Reference MacPhee and Finnamore1978) recorded spring emergence of D. mali adults during the second week of June in New Brunswick, consistent with our findings in Nova Scotia.

Fig. 2.

Pheromone trap captures of Dasineura mali males in 2016. A, Nova Scotia orchards; B, Ontario orchards.

Fig. 3.

Percentage of terminal apple-growing shoots infested with Dasineura mali eggs in 2014–2016 over accumulated degree days based on a threshold of 5 °C. A, Nova Scotia sentinel orchards; B, eastern Ontario sentinel orchards; C, southwestern Ontario sentinel orchards; D, British Columbia sentinel orchards.

Table 1.

Calendar dates and degree days (base 5 °C) to Dasineura mali oviposition peaks in single orchards in Nova Scotia, Ontario, and British Columbia 2014–2016.

* None – 50 random shoots were collected but no eggs were recovered.

NA, data not available.

Traps were hung too late in the spring in British Columbia to record the first emergence of adults. Low numbers of adults were caught (< 100) in the first two generations in British Columbia, this was deemed unrealistic based on the high egg and gall counts in the orchard, and therefore trap counts were not used to determine generational peaks in that province. The second-generation D. mali adult flight began in late June in Ontario (eastern and southwestern) and late July in Nova Scotia (Fig. 2). MacPhee and Finnamore (Reference MacPhee and Finnamore1978) recorded second-generation adults emerging in the last half of July in New Brunswick, again, similar to our findings in Nova Scotia. Second D. mali generation oviposition peaked from early to mid-July in Ontario (818 degree days to 934 degree days) and in late June in British Columbia (853 degree days to 931 degree days) (Fig. 3). Again, this second-generation ovipositional peak occurred later in Nova Scotia in late July (1058 degree days) (Table 1, Fig. 3).

The 2016 third-generation adults began to emerge in Ontario from mid-August (1126 degree days) through mid-September, whereas they did not begin to fly in Nova Scotia until early September (1546 degree days) (Fig. 2). Third-generation ovipositional peaks occurred in British Columbia in late July–August (1425–1598 degree days) and in Ontario in August (1377–1634 degree days) (Table 1, Fig. 2). Third-generation ovipositional peaks occurred in Nova Scotia from late August–September in the three years (1539–1673 degree days) (Table 1, Fig. 3). MacPhee and Finnamore (Reference MacPhee and Finnamore1978) did not describe a third generation in New Brunswick. They may have discontinued their observations before the egg numbers increased because Eaton and Bent (Reference Eaton and Bent1997) later recorded two to three generations in Nova Scotia. The authors recognise that because the incidence of eggs infesting the growing terminals and pheromone trap captures in this study were recorded only weekly or biweekly, the accumulated degree days can only be used to approximately time the D. mali phenological stages. Overall, there are differences among the provinces in the D. mali adult flight and the degree-day units for each ovipositional peak, and this variability may generally be explained by within-week variability of temperatures (i.e., a few days of high temperatures between sampling times in one region but not the others) at monitored locations.

Larvae collected from galls in British Columbia ranged from 0.3 mm to 3.3 mm in length. Larvae measuring over 2.0 mm that were orange in colour were considered to be third instars (Smith Reference Smith2000). The proportion of this developmental stage consistently increased three times in the weekly samples from 2015 and 2016 in British Columbia and in 2015 in southwestern Ontario, supporting the hypothesis of three generations (Fig. 4). Dasineura mali have been reported to develop through three to six generations per year in northern and southern areas of Europe, respectively (Cross et al. Reference Cross, Solomon, Babendreier, Blommers, Easterbrook and Jay1999), and up to five generations in New Zealand (Smith Reference Smith2000; Shaw et al. Reference Shaw, Wallis, Alspach and Sandanayaka2005).

Fig. 4.

Percentage of third instar Dasineura mali within weekly gall samples. A, southwestern Ontario in 2015; B, Summerland, British Columbia in 2015 and 2016.

The mean number of infested growing tips in the first D. mali generation in each province was smaller than that in the following two generations, except in eastern Ontario (Fig. 5). This is also characteristic of D. mali populations in New Zealand and has been attributed to the relatively smaller number of vulnerable apple-growing tips in the spring versus later during the second D. mali generation which coincides with an abundance of actively growing terminals (Todd Reference Todd1959; Shaw et al. Reference Shaw, Wallis, Alspach and Sandanayaka2005) and the high fecundity of D. mali females (156 eggs per female) (Whitcomb Reference Whitcomb1941). Infestation levels in the third generation were higher than the previous generations in Nova Scotia and British Columbia (except 2016) but lower in both Ontario locations. Samples were less than 100 terminals when D. mali infestations were difficult to find (Nova Scotia) and where there was seasonal limited new terminal growth (Ontario). Smith (Reference Smith2000) found that a sample size of 50 shoots gave low precision at hypothetical infestation rates of 60% or below. Mean infestation levels were below 60% in all three generations each year from 2014 to 2016 in Nova Scotia and below 30% in eastern and southwestern Ontario during the same period.

Fig. 5.

Dasineura mali infestation rate expressed as the mean per cent terminal apple-growing shoots infested with D. mali eggs in each of three generations in orchards in Nova Scotia, eastern Ontario (E Ontario), southwestern Ontario (SW Ontario), and the Okanagan and Similkameen valleys of British Columbia in the growing seasons of 2014 through 2016. 2014: n = 10 (Nova Scotia and British Columbia), 4 (SW Ontario); 2015: n = 4–10 (Nova Scotia), 1 (E Ontario), 9–10 (SW Ontario), 10 (British Columbia); 2016: n = 6–8 (Nova Scotia), 1 (E Ontario), 10 (SW Ontario and British Columbia). ± SE, ± standard error.

Parasitism

Parasitoids collected included Platygaster demades, P. tuberosula Kieffer, Synopeas myles (Walker) (Hymenoptera: Platygastridae), L. nigroaeneus, Aphanogmus vicinus Förster (Hymenoptera: Ceraphronidae), Eurytoma Illiger (Hymenoptera: Eurytomidae) species, and an unidentified Braconidae (Hymenoptera) (Table 2). The parasitoid species complex differed among regions, with P. demades being most abundant in Nova Scotia, L. nigroaeneus in Ontario, and S. myles dominating in British Columbia. Among these, the Platygastridae are endoparasitoids that oviposit in host eggs and/or larvae (He and Wang Reference He and Wang2015) then develop and emerge when the host is pupating (He et al. Reference He, Wang, Walker, Rogers and Lo2010). Aphanogmus Thomson species are either endoparasitoids of larvae (Parnell Reference Parnell1963) or ectoparasitoids of pupae (Luhman et al. Reference Luhman, Holzenthal and Kjaerandsen1999; Evans et al. Reference Evans, Dessart and Glenn2005; Matsuo et al. Reference Matsuo, Ganaha-Kikumura, Ohno and Yukawa2016), and Pteromalidae associated with galls develop as ectoparasitoids that feed externally on larvae, pupae, or even adults of the gall former or as inquilines that feed on the gall tissue (Askew Reference Askew1961).

Table 2.

Parasitoid species associated with Dasineura mali-infested orchards in Nova Scotia, Ontario, and British Columbia 2014–2016.

* Identified as most likely due to poor condition of specimens.

Platygaster demades is a solitary koinobiont endoparasitoid of D. mali that oviposits in eggs and young larvae. It was reared from all three D. mali generations in Nova Scotia and found in most of the Nova Scotia orchards surveyed, indicating that P. demades established and dispersed at least 7 km after its release in 1993 (Table 2). Dasineura mali larvae were difficult to find in Nova Scotia and were only sampled in a small number of the orchards (3–6) that received no or minimal pesticide treatments and therefore would be expected to support parasitoid populations.

Per cent parasitism of third-generation D. mali in Nova Scotia averaged 29% and 34% in 2014 and 2016, respectively (Table 3), and may be responsible for the low incidence of the midge in this province. Platygaster demades plays a significant role in suppressing D. mali populations in both New Zealand and Europe (Cross et al. Reference Cross, Solomon, Babendreier, Blommers, Easterbrook and Jay1999; Tomkins et al. Reference Tomkins, Wilson, Thomson, Bradley, Cole and Shaw2000). The ecology of P. demades has been extensively studied in New Zealand where the parasitoid was successfully introduced in 1925 and established to control the pear leaf-curling midge Dasineura pyri (Bouché) (Diptera: Cecidomyiidae). In the 1950s, P. demades was found to also parasitise D. mali in New Zealand (Todd Reference Todd1959). Upon emergence, P. demades females have a mean preovipositional load of more than 1300 eggs and live for up to 35 days at temperatures of 13–20 °C (Sandanayaka and Ramankutty Reference Sandanayaka and Ramankutty2007). Platygaster demades eggs hatch, and the larvae feed and complete development after the mature D. mali larval hosts begin to form cocoons. Parasitism can be determined by dissecting the host larvae and counting the number of P. demades eggs (He et al. Reference He, Wang, Walker, Rogers and Lo2010). In New Zealand, D. mali parasitism levels based on larval dissections vary between 83% in the first D. mali generation to between 50% and 60% in the third and fourth generations (Shaw et al. Reference Shaw, Wallis, Alspach and Sandanayaka2005). Seasonal highs of over 80% P. demades parasitism are common in New Zealand (Todd Reference Todd1959; He and Wang Reference He and Wang2011). Per cent parasitism determined by dissection can be significantly higher than that recorded from emergence, which relies heavily on successful rearing, and it is possible that the P. demades parasitism rates recorded in Nova Scotia are lower than the actual levels of parasitism as the hosts were not dissected.

Table 3.

Mean per cent total Dasineura mali parasitism recorded from infested gall samples in Nova Scotia, Ontario, and British Columbia from 2014–2016.

Variation in parasitism between the generations appears to depend upon the synchrony of the parasitoid with the host. Peak P. demades adult emergence coincided with D. mali oviposition peaks in generations one, three, and four in New Zealand (Shaw et al. Reference Shaw, Wallis, Alspach and Sandanayaka2005), although P. demades parasitism of the second D. mali generation is generally low (Todd Reference Todd1959; Shaw et al. Reference Shaw, Wallis, Alspach and Sandanayaka2005; Sandanayaka and Ramankutty Reference Sandanayaka and Ramankutty2007). He et al. (Reference He, Wang, Walker, Rogers and Lo2010) and He and Wang (Reference He and Wang2011, Reference He and Wang2014) found that P. demades parasitism levels in the field demonstrated a strong density-dependent type III functional response. Platygaster demades can also produce both aestivating and nonaestivating embryonic phenotypes over the season and in this way is able to manipulate the number of generations per season to maximise survival (He and Wang Reference He and Wang2012). Aestivation of P. demades eggs increased from late spring to mid-summer and decreased in late summer and early fall in New Zealand (He and Wang Reference He and Wang2011). Should a similar phenomenon occur in the Nova Scotia population, it could explain the observed higher percentage of parasitism recorded in the third generation in Nova Scotia. He and Wang (Reference He and Wang2012) recorded P. demades aestivation induction to begin when the photoperiod reached 13 hours of light to 11 hours of dark and 22.8 °C in New Zealand. Examination of data from Nova Scotia in 2016 found these conditions to occur shortly after the first peak of D. mali-infested tips. Aestivation termination in New Zealand occurred at 12:12 hour (light:dark) and 15 °C (He and Wang Reference He and Wang2012) conditions, which typically occur in September in Nova Scotia orchards.

No P. demades were recovered from Ontario D. mali collections. In British Columbia, five specimens of P. demades were found for the first time in 2016 from second-generation and third-generation D. mali larvae from two different apple orchards in the Fraser Valley of British Columbia (one in Abbotsford and the other in Delta; Table 2). There is no record of P. demades introductions into British Columbia or the western United States of America. Although parasitism surveys in the previous two years in the same orchards did not yield P. demades, the host collections from these orchards were not always conducted at optimal developmental periods and sample sizes were often low which may have compromised the accuracy of the assessments.

Lyrcus nigroaeneus is a solitary koinobiont ectoparasitoid that consumes several, at least two to three already partly mature D. mali larvae, during its development (A.M.B., personal observation). It is native to the Nearctic region and was the most abundant parasitoid emerging from D. mali collected in Ontario orchards, making up 67–100% and 83–100% of the parasitoids collected in eastern and southwestern Ontario, in 2015 and 2016, respectively (Table 2). Overall parasitism levels in Ontario were low (Table 3), with only 0–6.8% in each generation in eastern Ontario, and 1.5–22.2%, 14.8–29.9%, and 0.2–41.9% in first, second, and third generations, respectively, in southwestern Ontario. Parasitism in southwestern Ontario dropped in the first and third D. mali generations in 2016, despite a general increase in the number of larvae reared (Table 3). Rearing methods were improved from 2015 to 2016; however, the extremely hot dry summer of 2016 may have been a factor in the lower D. mali parasitism.

In 2015, 99.1% (n = 545) of the L. nigroaeneus in Ontario emerged from the galls prior to D. mali larvae dropping into the petri dishes, indicating that the parasitoid pupates in the gall after consuming larvae (Table 4). In Ontario, the number of apple-growing tips infested with D. mali was low in all years, compared to infestations in Nova Scotia and British Columbia suggesting that L. nigroaeneus may be an important mortality factor (Fig. 5). Lyrcus nigroaeneus was occasionally found in Nova Scotia outside of the 2014–2016 survey (identified by G.A.P. Gibson, 2008, 2015, 2017).

Table 4.

Parasitoid emergence from Dasineura mali larvae and galls in Ontario orchards in 2015.

Synopeas myles (Walker) (Hymenoptera: Platygastridae), a solitary koinbiont endoparasitoid, was the most abundant species reared from D. mali in the Fraser Valley (Abbotsford), British Columbia in each of 2014–2016 (Table 2). In the interior of British Columbia (Okanagan and Similkameen valleys), no parasitoids were found in 2014 (n = 17 340 D. mali sampled); however, a total of 172 and 2819 S. myles emerged from D. mali hosts from Okanagan and Similkameen orchards in 2015 and 2016, respectively (Table 2).

Mean parasitism (primarily S. myles) in interior British Columbia orchards was nil in 2014 (0%), low in 2015 (1.9–7.9%), but increased dramatically in each generation in 2016, reaching 29.6% by the third generation (Table 3). This increase may be responsible for the first observed decrease in the per cent of apple-growing tips infested by D. mali during the third generation of that year (Fig. 4). In 2016, a single female S. myles specimen emerged from D. mali in Nova Scotia and a specimen of poor quality identified as a questionable S. myles male emerged in Ontario in 2015 (Table 2). Synopeas myles is known to parasitise seven gall midge species in the genera Contarinia Róndani (Diptera: Cecidomyiidae) and Dasineura Róndani in Europe, including the swede midge, Contarinia nasturtii Kieffer (Diptera: Cecidomyiidae) (Abram et al. Reference Abram, Haye, Mason, Cappuccino, Boivin and Kuhlmann2012). Contarinia nasturtii is a recent invader in the eastern Canadian provinces, including Nova Scotia and Ontario; however, it is not yet known to occur in British Columbia (Hallett and Heal Reference Hallett and Heal2001; Anonymous 2017). The findings of this study are the first known global records of S. myles parasitising D. mali.

Platygaster tuberosula was recorded for the first time in 2014 from D. mali hosts in Nova Scotia and British Columbia and in 2015 from Ontario hosts (Table 2), although it appears to be of minor importance in Nova Scotia, southwestern Ontario, and British Columbia. It was more prevalent in eastern Ontario, although L. nigroaeneus was much more abundant. Seventy-five per cent of P. tuberosula from Ontario were found to emerge after development in younger than third instar D. mali (Table 4). Due to the D. mali collection protocol used in this study, fewer immature larvae were evaluated for parasitism, however, from 14 112 immature D. mali larvae that dropped from galls collected across three generations in eight Ontario orchards in 2015, 11.7% were found to successfully pupate and emerge as adults, which indicates that associated parasitoids may also survive. Overall adult emergence in this laboratory assessment was generally low, with only 37.6% of the mature third instar D. mali larvae successfully emerging as adults (n = 6689).

Platygaster tuberosula was introduced into Saskatchewan in 1993 and 1994 for classical biological control of wheat midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae) (Olfert et al. Reference Olfert, Doane and Braun2003). The wheat midge is the primary host of P. tuberosula and is present across Canada, including British Columbia, Ontario, and Nova Scotia. Platygaster tuberosula has also been previously recorded to parasitise D. mali (Elliott et al. Reference Elliott, Olfert and Hartley2011; Chavalle et al. Reference Chavalle, Buhl, Censier and De Proft2015). Several Aphanogmus vicinus Förster (Hymenoptera: Ceraphronidae) were only recovered from D. mali collected in British Columbia in the first and second generations of 2015 and 2016.

In summary, the recent invasion of D. mali into the interior of British Columbia stimulated a survey of parasitism of this pest in apple-growing regions of British Columbia, Ontario, and Nova Scotia. Molecular analyses indicated that the Nova Scotia D. mali population is different than populations in British Columbia and Ontario suggesting that at least two invasions of this species occurred, or the occurrence of cryptic species within D. mali. Additional sampling and a thorough population genetic analysis of D. mali across its geographic range in Europe may help clarify the origin of the invasive populations in Canada. Three generations of D. mali occur annually in each province, and population levels vary among years and regions. The parasitoid communities associated with D. mali differ among provinces. Platygaster demades, introduced into Nova Scotia in the early 1990s as a biological control agent, appears to be an important mortality factor in Nova Scotia and was discovered during this study to be present in British Columbia but not Ontario. Additional population genetic studies on P. demades in Canada and Europe may also help clarify the origin of the established populations, particularly in British Columbia where no known introductions have been made. The native L. nigroaeneus is the most abundant parasitoid of D. mali in Ontario. Two other exotic species, P. tuberosula introduced for biological control of wheat midge, and S. myles, a known parasitoid of the invasive swede midge, were associated with apple leaf midge, the latter species for the first time. Synopeas myles is the most abundant parasitoid of D. mali populations in the interior of British Columbia. The parasitoid communities associated with D. mali appear to be dynamic in each province, with new species exploiting this host as time moves forward. Further research is needed to understand the impact of each parasitoid community on D. mali populations, as well as their potential impact on nontarget gall midges.