A number of root maggot fly species from the large genus Delia Robineau-Desvoidy (Diptera: Anthomyiidae) are important pests of cultivated crucifers and many other field and vegetable crops. The present work provides identification keys in English and French to the adults, third instar larvae, puparia, and eggs of all pests of cultivated crucifers and other vegetable crops in Canada, namely Delia antiqua (Meigen), D. floralis (Fallén), D. florilega (Zetterstedt), D. planipalpis (Stein), D. platura (Meigen) and D. radicum (Linnaeus). DNA barcodes are provided for all species except D. planipalpis and new data on larval host associations in southern Québec are presented.

Delia platura
Photo by Michael Richardson

Feeding damage by root maggot flies (Delia Robineau-Desvoidy spp. (Diptera: Anthomyiidae)) can result in important crop losses in Canada, being especially detrimental to vegetable crops (Griffiths 1986, 1991, 1993; Howard et al. 1994) and, more recently, to the production of a variety of canola cultivars (Brassica rapa Linnaeus and Brassica napus Linnaeus (Brassicaceae)) (Broatch 1993; Soroka et al. 2004; Soroka and Dosdall 2011). As their common names indicate, certain phytophagous Delia species such as the onion maggot, Delia antiqua (Meigen) and the cabbage maggot Delia radicum (Linnaeus), have a relatively narrow range of hosts, in this case feeding exclusively on Allium Linnaeus (Amaryllidaceae) or Brassicaceous species, respectively. However, other primarily saprophagous taxa such as the seedcorn maggot, Delia platura (Meigen), and the bean seed maggot Delia florilega (Zetterstedt) have a much wider range of larval plant food, including decomposing (and sometimes fresh) cultivated crucifers and Allium species as well as many other types such as legumes, cucurbits, a wide variety of garden crops and even certain cereals (Griffiths 1993; Howard et al. 1994).

Because most crops can be infested by more than one species of root-maggot flies, the correct identification of species involved in plant damage is important for the adequate development and assessment of pest-control strategies, as each may respond differently to treatments. This is especially relevant to targeted control methods such as the mass release of sterile insects, a strategy currently used for the control of the onion maggot in southern Québec (Fournier 2014). Much can be learned about the ecology and population dynamics of Delia flies by studying adults  (Broatch 1993; Broatch et al. 2006; Ellis and Scatcherd 2007; Soroka and Dosdall 2011); however, using adults as a proxy for the abundance and species composition of the taxa actually causing the damage can be misleading. Different trap designs will capture different proportions of each species (Finch 1989; Broatch and Vernon 1997); some species can be very abundant as adults but rarely involved in damage, and the relative contribution of different species to plant damage can be over or underestimated (see DNA barcoding and larval host associations section).

Specimen identification of even the most common Delia pest species can be difficult for anyone without extensive training and this challenge is more pronounced when dealing with females or immature specimens which may lack obvious diagnostic features. While excellent, the keys to adult Delia flies published by Griffiths (1993) in his monumental revision of Nearctic Anthomyiidae rely heavily on genitalic characters and include over 150 couplets for each sex, being therefore virtually incomprehensible to non-specialists. The key to third instar larvae by Dahlem and Thompson (1991) remains a good resource for Delia species of economic importance but it contains some inaccuracies (see key to larvae for additional details). The work of Brooks (1951) offers identification keys to the different life stages of Delia pests of cruciferous crops in Canada and remains the most comprehensive taxonomic tool currently available for the group. However, Brooks (1951) includes pests of crucifers only, is poorly illustrated, does not discuss intraspecific variability for certain key structures, and uses obsolete name combinations for most taxa.

The main objective of this work is, therefore, to provide a simple and comprehensive taxonomic resource (identification keys to the different life stages and reference DNA barcodes) for economically important species of Delia flies in cultivated crucifers and other vegetable crops in Canada. New host association data for various crops grown in southern Québec are also presented.

More than 5000 adults and 3500 immatures from dozens of Nearctic and Palaearctic localities were examined to document intraspecific variability and host associations; however, all specimens photographed in the keys except for those belonging to Delia planipalpis (Stein) were reared from laboratory cultures to ensure the correct association of the various life stages. Live puparia of Delia floralis (Fallén) (cultured from wild Norwegian stocks) were provided by R. Meadows (Norwegian University of Life Sciences, Ås, Norway) and K. Hillier (Acadia University, Wolfville, Nova Scotia, Canada), while laboratory cultures of D. platura, D. florilega, D. antiqua and D. radicum descending from wild Québec populations were maintained by F. Fournier (Collège Montmorency, Laval, Québec, Canada). All wild larvae collected in the province of Québec are housed at the Bishop’s University Insect Collection, Sherbrooke, Québec, Canada (BUIC). Additional adult and/or immature specimens from BUIC, the Canadian National Collection of Insects, Ottawa, Ontario, Canada (CNC), the National Museum of Natural History, Smithsonian Institution, Washington DC, United States (NMNH) and the E.H. Strickland Museum, Edmonton, Alberta, Canada (UASM) were also examined. Provincial acronyms used in the distribution section of each species are based on Canadian provincial and territorial postal abbreviations.

Adults and their external structures were photographed with a Canon DSLR camera equipped with a Canon MP-E65 lens and a Canon MT24EX flash. Male genitalia and immature specimens and structures were photographed using a Luminera 1 digital camera mounted on a Leica M125 dissection microscope. Multiple images taken at different focal points were stacked using Helicon Focus 6.3.0 (Helicon Soft Ltd., Kharkov, Ukraine) and final image editing was completed in GIMP (v.2.8; http://www.gimp.org). All larvae were immersed in near-boiling water for a minute before they were measured and photographed.

DNA barcodes (Folmer region of COI gene) were sequenced for multiple adult specimens of all species except D. planipalpis (lack of recently collected material) as well as for the eggs and larvae (1st and 3rd instars) of most other species included in this work. A neighbor joining tree including a subset of high quality sequences from various life stages is presented in Fig. 1. Images, collection details and individual COI sequences for a subset of barcoded specimens can be retrieved from the Barcode of Life Data System (BOLD) (www.barcodinglife.org) in the public dataset: Delia pests of Canada (dx.doi.org/10.5883/DS-DPOC). All vouchers were deposited in the BUIC.

The DNA barcodes for D. radicum, D. floralis, D. antiqua and D. florilega all clustered in single, well-defined Barcode Index Numbers (BIN, Ratnasingham and Hebert 2013) (Fig. 1). Delia platura sequences, however, formed two distinct BINs (AAA3453 and AAG2511) separated by a minimum distance of 4.45% (Fig. 1). More than 100 adults (both sexes) and 50 third instar larvae from each of these two D. platura BINs were examined but no morphological differences was found. There is, however, a distinctive pattern in the geographical distributions of these BINs based on thousands of records found in the Barcode of Life Data System (BOLD, Ratnasingham and Hebert 2007): AAA3453 is almost entirely restricted to the New World (transcontinental distribution in Canada and the United States, Costa Rica) with a single record from South Africa whereas AAG2511 is found throughout Europe, Asia, Greenland, eastern Canada (NB, NS, NL, ON, PEI, QC), with a single record from Alaska. Delia platura is found on all continents except Antarctica (Griffiths 1993) and little is known of the origins of the North American fauna. The global distribution pattern reported here for the two BINs suggests, however, that the New World fauna includes at least two distinct populations, one possibly native (AAA3453) and one introduced (AAG2511). The phylogeography of these two groups of D. platura will be further investigated in a separate project as it will require additional data beyond the scope of the present work.

Fig. 1. Neighbor-joining tree based on Kimura 2-parameter distances (K2P, Kimura 1980) for representative sequences from each species. Each line includes species name, BOLD sample ID, read length, number of ambiguities and Barcode Index Number.

The abundance of different Delia species in canola and/or vegetable crops is often monitored using passive traps (yellow and/or blue sticky cards, cone traps, water traps, etc.) to capture adults (Finch 1989). While D. radicum and D. antiqua are often assumed to be the main contributors to crop damage in crucifers and Allium, respectively, adults of D. platura and/or D. florilega can be much more abundant in passive traps than primary pests (Broatch and Vernon 1997; Villegas et al. 2009), as they are attracted to organic matter and fresh crop residues (Hammond 1990) and can breed in previously damaged and/or healthy cultivated plants. Little is known, however, of their actual involvement in plant damage. To determine the species composition of Delia species feeding on vegetable crops in southern Québec, more than 3500 third instar larvae were collected from 85 vegetable fields distributed in 21 farms in the Montérégie region (Fig. 2) between 15 May 2015 and 23 September 2015. Each field was sampled at least once and sampling protocol varied according to crop type. In crucifers (Brassicaceae), 150 seedlings each of broccoli (Brassica oleracea Linnaeus var. italica Plenk), cauliflower (B. oleracea var. botrytis Linnaeus) and Napa cabbage (Brassica napus Linnaeus subsp. pekinensis (Lour.) Hanelt) were examined, and 300 seedlings of radishes (Raphanus sativus Linnaeus) were collected. Regarding onions, all yellow onion (Allium cepa Linnaeus) seedlings along 30 lines of 1 meter each were examined, and all green onion (Allium fistulosum Linnaeus) seedlings along 15 lines of 1 meter each were examined. Broccoli and cauliflower were grown in mineral soils while Napa cabbages, radishes and all onions were grown in muck soils. The larvae were identified morphologically and those of D. platura and D. florilega were grouped together since they can’t be distinguished morphologically (see identification key to larvae for details). The abundance of species in samples where at least one larva was collected was compared in each crop using a two-tailed Wilcoxon-signed rank test on paired data. Green and yellow onion samples were divided into two groups according to the presence or absence of mass release (since 2012) of sterile D. antiqua on the farms where they were collected (Fournier 2014). The statistical analysis was computed using the R package “stats” 3.2.4 (R Development Core Team, 2016) and statistical significance was assessed using an alpha-level of P < 0.05 and 0.95 confidence intervals.

As expected, the diversity of collected larvae was low: Delia platura/D. florilega was collected in all crops, D. radicum was found only in crucifers and D. antiqua only in onions (Table 1). In crucifers, most larvae were collected from broccoli (n=2089) and Napa cabbage (n=858). The larval abundance of each species was quite variable but Delia radicum was significantly more abundant than Delia platura/D. florilega only in Napa cabbage (V = 108.0, P = 0.007) and radishes (V = 242.5, P = 0.001) (Table 1), even if D. platura and D. florilega adults were always much more abundant than D. radicum on both sticky cards and emergence traps placed in the fields (A.-M. Fortier and J. Savage, unpublished). Liu and Sparks (1999) reported substantial damage to Napa cabbage by D. florilega in Texas based on rearing data but our results suggest that the pattern may not apply to more northern localities, and that the attribution of feeding damage to Napa cabbage by D. florilega and D. platura in southern Québec based on the high abundance of adults observed laying eggs in the field by Lafontaine et al. (2008) may have been incorrect.

In spite of intensive sampling efforts in onions, infestation rates were low in 2015, with damages averaging 0.7% and 1.6% in yellow and green onions respectively, independently of the control method used (F. Fournier, unpublished) and relatively few Delia larvae were collected, both in fields where sterile D. antiqua had been released and those where they had not (Table 1). Overall, we collected more larvae of Delia platura/D. florilega (n=360) than of D. antiqua (n=213), but found no significant differences in species abundance (Table 1). In onions, D. platura is often reported in joint infestations with D. antiqua (Merrill 1951; Merrill and Hutson 1953; Finlayson 1956). However, in the present work, D. platura/D. florilega larvae were collected alone in 38 of 72 yellow onion samples (53%) and in 17 of 37 green onion samples (46%), suggesting primary infestations.

Since the larvae of D. florilega are identical to those of D. platura and because of the presence of two distinct BINs for D. platura (see DNA Barcoding section), we used DNA barcodes to determine larval host associations of 288 randomly selected D. platura/D. florilega larvae collected in various vegetable crops in southern Québec (Table 2). Both BINs of D. platura were recorded from all hosts – broccoli, cauliflower, Napa cabbage, radish, green and yellow onions. In crucifers, the larvae from BIN AAG2511 (n=139) were more than 2.5x as abundant as those from BIN AAA3453 (n=52) but the trend was reversed in onions, where 49 out of 64 sequenced individuals belonged to BIN AAA3453 (Table 2). Delia florilega was always less abundant than D. platura (BINs combined) but recorded from all sampled hosts with the exception of cauliflower (Table 2).

The host association data presented here is based on a large number of sampled plants and localities throughout the season, but it was all collected over a single year (2015) in southern Québec. It should therefore be used as baseline data to further document host preference (and its variability) in Delia flies found in vegetable crops in Canada.

Fig. 2. Location of farms from Montérégie, southern Québec, Canada, where Delia larvae were collected in 2015.

Table 1. Mean, standard deviation (SD) and range of Delia species abundance per sample (n) of larvae collected in crucifers and onions in 2015 in Montérégie, southern Québec, Canada, with a comparison of mean species abundance for each crop type. S = samples collected in fields where sterile males of D. antiqua were released, NS = samples collected in fields where no sterile males of D. antiqua were released.

* The larvae of these two species cannot be distinguished morphologically.
^αV value from two tailed Wilcoxon-signed rank test on paired data with confidence interval of 0.95.

Table 2. Number of Delia platura and D. florilega larvae per species (based on DNA barcodes) and Barcode Index Number (BIN) for a selected subset of the larvae collected in 2015 in Montérégie, southern Québec, Canada (n = number of farms from which at least one larva was sequenced).

The species included in this work and listed below were selected based on published records of crop damage in Canada (Brooks 1951; Griffiths 1991, 1993; Broatch 1993; Howard et al. 1994; Soroka and Dosdall 2011) and the identification of over 3500 larvae collected in damaged crucifers and onions in southern Québec. While Delia echinata (Séguy) has been reported as a pest of spinach in the Palaearctic Region (Miles 1953; Suwa 1974) it was excluded from the present work as damage caused by this species in the Nearctic Region is restricted to carnations (Griffiths 1991). Due to numerous nomenclatural changes and the large number of publications focused on some of the most important pest species, a list of most commonly used synonyms and combinations found in the literature as well as the French and English common names recognized by the Entomological Society of America (http://www.entsoc.org/pubs/common_name) and/or the Entomological Society of Canada (http://www.esc-sec.ca/d13/CNDBHome.php) are provided below.

Delia antiqua (Meigen, 1826) (En Français)
– English common name: onion maggot; French common name: Mouche de l’oignon.
– Commonly used synonyms or combinations: Hylemyia antiqua, Hylemya antiqua.

Delia floralis (Fallén, 1924) (En Français)
– English common name: turnip maggot; French common name: Mouche du navet.
– Commonly used synonyms or combinations: Hylemyia crucifera Huckett, Hylemya crucifera, Hylemya floralis.

Delia florilega (Zetterstedt, 1845) (En Français)
– English common name: bean seed maggot; French common name: none.
– Commonly used synonyms or combinations: Hylemya trichodactyla (Rondani), Hylemyia trichodactyla, Delia liturata (Meigen), Hylemya liturata.

Delia planipalpis (Stein, 1898) (En Français)
– Common name: none.
– Commonly used synonyms or combinations: Hylemya planipalpis, Hylemyia planipalpis.

Delia platura (Meigen, 1826) (En Français)
– English common name: seedcorn maggot; French common name: Mouche des semis.
– Commonly used synonyms or combinations: Hylemya platura, Chortophila cilicrura Rondani, Hylemya cilicrura, Hylemyia cilicrura.

Delia radicum (Linnaeus, 1758) (En Français)
– English common name: cabbage maggot; French common name: Mouche du chou.
– Commonly used synonyms or combinations: Hylemya brassicae (Bouché), Hylemyia brassicae, Erioischa brassicae.

Keys to Delia pests of cultivated crucifers and other vegetable crops in Canada

Botanophila fugax (Meigen) is an anthomyiid species associated with decaying plant matter (Ferrar 1987, p. 72 as Pegohylemyia fugax) whose immature stages are quite distinctive from those of Delia species included in the present work (see Brooks 1951 and Miles 1952 for descriptions of immature stages (as Pegohylemyia fugax)). However, the adults (and especially the females) are remarkably similar to those of certain Delia species. According to Brooks (1951), B. fugax is only occasionally encountered on cruciferous crops in Canada. The examination of  3220 adults caught by conical traps and sticky traps and more than 3500 larvae collected in various Allium (green and yellow onions) and cruciferous crops (radish, Napa cabbage, broccoli, cauliflower) in southern Québec, Canada, in the summer of 2015 (Fig. 2) yielded only 15 adults of B. fugax (0.5%) and no larvae. Botanophila fugax is therefore distinguished from Delia spp. only in the key to adults in the present work.

Key to adults
– (En Français)

Key to larvae
– (En Français)

Key to puparia
– (En Français)

Key to eggs
– (En Français)