Edge effects on parasitoid hymenopteran biodiversity in canola fields

Abstract

Parasitoid hymenopterans are highly valuable pest control agents. Despite their necessity within the agriculture industry, they remain one of the most understudied groups of insects. Non-crop habitats adjacent to farmland provide critical ecological services for these beneficial insects and can act as a corridor from uncultivated land to fields; this interaction is a form of edge effect. Here, we evaluate how 3 types of non-crop field edges affected the diversity, richness, and abundance of parasitoid hymenopterans at varying edge distances in canola (Brassica napus L.; Brassicales: Brassicaceae). We also compiled all sampled genera into a survey of parasitoid Hymenoptera found in southern Alberta canola. We conclude: (1) non-crop boundaries have a weak effect on parasitoid communities within our study system; (2) variability in parasitoid diversity and composition was most pronounced among farms, suggesting a dominant role of farm management practices; (3) parasitoid diversity is highest at edge boundaries and decreases toward the field interior; (4) Diolcogaster spp. (Hymenoptera: Braconidae)—predominantly Diolcogaster claritibia Papp, first reported in Canada in 2014—was the dominant group of parasitoids in these fields, comprising 74% of collected specimens. Overall, the composition of parasitoid communities is likely more sensitive to individual farm management practices or landscape composition and configuration, rather than to the type of bordering non-crop habitat. Further research is needed to determine which of these farm-level drivers are influencing parasitoid communities.

Introduction

Contemporary agricultural practices have a widespread, negative impact on global biodiversity (Campbell et al. 2017). Loss of biodiversity, including insect diversity, is a major concern globally (Wagner 2020), and arthropods in arable land have been especially susceptible to decline (Seibold et al. 2019, Raven and Wagner 2021). A loss of arthropod species can lead to the loss of critical ecosystem and agriculture services such as biocontrol and pollination (Losey and Vaughan 2006, Vanbergen et al. 2013, Wagner 2020). In agroecosystems, 1 major driver of biodiversity decline is the loss of non-crop habitats (Benton et al. 2003). Non-crop habitats, such as field margins, hedgerows, forests, grasslands, etc., provide key resources for beneficial insects including shelter and mating sites, protection for overwintering, and food sources (Thomson and Hoffmann 2010, Zemenick et al. 2019, Niedobová and Kudláček 2026). These uncultivated habitats can also facilitate access to cropland, enabling natural enemies to contribute pest-control services (Thies et al. 2003).

Of the natural enemies of crop pests, parasitoid hymenopterans are one of the most effective groups of pest-suppressors (La Salle and Gauld 1991, Quicke 1997, Narendran 2001). Parasitoids provide host-specific suppression of pests, including those resistant to pesticides, as well as secondary pollination, all while maintaining crop quality and yield (Narendran 2001, Thies et al. 2003, Sarfraz et al. 2005, Zemenick et al. 2019). Despite their agricultural importance, profound speciosity, and cosmopolitan distribution, parasitoid hymenopterans remain one of the most understudied groups of insects (La Salle and Gauld 1991), and are especially vulnerable to decline, particularly due to a lack of conservation strategies (Shaw and Hochberg 2001, Shaw 2006).

In canola agroecosystems, there are a number of major insect pests that are hosts to parasitoid Hymenoptera. For example, wasps of the family Braconidae are major parasitoids of Lygus bugs (Peristenus) and Pieris species (Cotesia). The cosmopolitan diamondback moth, Plutella xylostella is attacked by another braconid, Diolcogaser claritibia. Other Lepidoptera such as cutworms of the family Noctuidae are attacked by a number of Ichneumonidae, for example Apanteles species. In some regions of the Prairies where the exotic cabbage seedpod weevil (Ceutorhynchus obstrictus) has invaded canola, it has become a host for a suite of native Chalcidoidea parasitoids (Dosdall et al. 2006). Besides this handful of known parasitoid–host interactions, there are other key crop pests such as flea beetles (Phyllotreta species), which await detailed study to explore their suite of parasitoids that contribute to pest management.

Habitat complexity is a key factor influencing diversity, as structurally diverse environments are associated with increased biotic and abiotic resources, which in turn promote biodiversity through niche availability (Pianka 1966, Benton et al. 2003). In agricultural landscapes, variation in structure and type of non-crop habitats may influence beneficial insect communities in adjacent fields, but to varying degrees and distances. For example, Morandin et al. (2014) found that parasitoid abundance increased and crop damage decreased in tomato crops bordering hedgerows (compared to weedy non-crop habitats), and that these benefits extended up to 200 m from the edge. Whitehouse et al.’s (2018) edge study in blueberry fields found that natural enemy populations concentrated along edge borders in intensive agricultural systems and did not disperse into the field interior. Furthermore, Kent et al. (2025) reported mixed results on the influence of edges on spider communities in canola; their study found that spider abundance was higher at field borders, but there was no change in species richness compared to the edge and field interior. These discrepancies suggest that edge effects may be context-dependent, influenced by variables such as crop type, type of non-crop habitat, and agricultural management practices.

Canada is the largest producer of canola (Brassica napus L., B. rapa L., B. juncea (L) Czern.; Brassicales: Brassicaceae), a major cash crop with an economic impact of C$43.7 billion on the Canadian economy (GlobalData 2024). The province of Alberta accounts for 29% of Canada’s canola oil production (Statistics Canada 2025a, 2025b). While common canola pests such as Lygus bugs (Hemiptera: Miridae), cabbage seedpod weevil (C. obstrictus Marsham; Coleoptera: Curculionidae), and flea beetles (Phyllotreta Newman spp.; Coleoptera: Chrysomelidae) are well documented, the diversity, abundance, and composition of parasitoid communities found in Alberta fields are less well known (Dosdall and Mason 2010, Cornelsen et al. 2024). To address this, we tested for effects of non-crop field edge-type and distance from edges on the abundance, genus richness, diversity, and community composition of parasitoid hymenopterans in canola fields. We hypothesized that abundance, richness, and diversity would increase with increasing habitat complexity and that these variables would have highest values closest to non-crop boundaries, regardless of edge type, and decrease with distance. Our study objectives were to (1) determine which parasitoid hymenopterans are found in southern Alberta canola, (2) determine if different non-crop edges affect parasitoid community diversity and composition, and (3) determine if parasitoid communities vary with distance from the edge boundary.

Materials and Methods

Our study focused on 3 non-crop habitats bordering southern Alberta canola fields: road verges, shelterbelts, and coulees. These habitats were selected because they are among the most representative surrounding landscapes around farm fields and they differ in structure, levels of disturbance, and resource availability, representing a putative gradient of low to high habitat complexity. Road verges are strips of groundcover, typically grasses and wildflowers, mostly dominated by invasive species, located along field edges. In some cases, these areas are mowed and sprayed with insecticides for grasshopper control. Shelterbelts are rows of shrubs or trees traditionally planted to act as windbreaks and to minimize soil erosion. Coulees are spatially complex basins or ravines that have been eroded by waterways with a variety of plant species within them, including trees, shrubs, wildflowers, and wild grasses. We sampled hymenopterans belonging to the subclade of Parasitoida, which consists of the 3 major lineages of Ichneumonoidea, Ceraphronoidea, and Proctotrupomorpha (including Chalcidoidea, Cynipoidea, Diaprioidea, Mymaro­mmatoidea, Platygastroidea, and Procto­trupoidea) (Polaszek and Vilhemsen 2023). We focused on this paraphyletic subclade as many groups of parasitoid hymenopterans are understudied, and a vast majority of biocontrol agents reside within the superfamilies of Ichneumonoidea and Chalcidoidea (La Salle and Gauld 1991, Goulet and Huber 1993, Gibson et al. 1997, Shaw and Hochberg 2001). Furthermore, to our knowledge, no contemporary surveys of parasitoid hymenopterans in southern Alberta exist, demonstrating a significant knowledge gap regarding Canadian Prairie entomofauna. We expected higher diversity and abundance of Hymenoptera in canola transects near coulees compared to tree shelters, which have lower plant diversity, and relative to road verges that are sometimes mowed or receive insecticide applications.

Study Sites

We selected 15 study sites (Fig. 1) from 7 farms with permission from the landowners. We visited each farm to identify study sites that fit the following criteria: (a) fields were seeded with canola for the 2023 growing season, (b) fields were geographically separated by >2 km, and (c) edge treatments (road verge, shelterbelt, and coulee) were each represented in at least 5 fields. We then selected 1 to 4 study sites from each farm and designated each farm with a single-letter identifier: “A,” “C,” “E,” “G,” “K,” “L,” and “W.” All sites were located within 75 km of Lethbridge, Alberta (49°41′59′′ N, 112°49′06′′ W) (Fig. 1) within the Moist-mixed Grass Ecoregion of the Canadian Prairie Ecozone. This region experiences annual precipitation between 350 and 450 mm and a mean annual temperature of 5 °C, with the summer averaging 15.5 °C (Ecological Stratification Working Group 1995). Surrounding ecoregions include Fescue Grassland to the south and Mixed Grassland to the east. The Fescue Grassland Ecoregion receives annual precipitation of 400 to 500 mm, a mean annual temperature of 3.5 °C, and a mean summer temperature of 14 °C, and Mixed Grasslands exhibit an annual precipitation rate of 250 to 350 mm, an average annual temperature of 5 °C, and a mean summer temperature of 16 °C (Ecological Stratification Working Group 1995). Eleven out of 15 study sites were in Moist-mixed Grasslands, 3 of which bordered Fescue Grasslands and one of which bordered a Mixed Grassland. One site was in Fescue Grassland that bordered Moist-Mixed Grasslands, and the 3 remaining sites were located within the Mixed Grassland Ecoregion.

Fig. 1.

The locations of the 15 canola fields used in this study, including their edge type and ecoregion. Each collection site and associated non-crop edge are indicated by different shapes. Ecoregions are represented in different colors. Farms are denoted by their single letter identifier.

Plant species found in each edge type were documented using a 50 cm by 50 cm sampling quadrat, placed at random along the edge boundary. The process was repeated 4 times at each sample site, and plants were identified via a field guide (Looman 1983). “Common plant species” were those present in at least 4 out of 5 sites with the same edge type, and are documented in Supplementary Information S1.

Sample Collection

Each study site had 2 transects placed 50 m apart, running perpendicular to its designated edge treatment. We placed sampling points along each transect at 10 m, 50 m, 100 m, and 200 m (Fig. 2). Collections occurred between May and August of 2023. We sampled each field 4 times throughout the growing season, corresponding with key canola development stages as outlined by the Biologische Bundesanstalt, Bundessortenamt and CHemische Industrie (BBCH) phenological scale (Lancashire et al. 1991). Crops were sampled at leaf germination and development (BBCH 10 to 29), stem elongation (BBCH 30 to 53), flowering (BBCH 55 to 69), and pod development/ripening (BBCH 71 to 89). Two sampling methods were used: pan traps during the leaf development and stem elongation stages (Fig. 2A), and sweep net sampling during the flowering and pod development stages (Fig. 2B). We opted for 2 separate methods, as seedlings in early development stages are too short and fragile to be swept, and in later development stages, parasitoids would be more present in the crop canopy where they could be collected with sweep nets.

Fig. 2.

An example of the transect design in a sampled shelterbelt field. (A) The pan trap sampling layout, occurring during the seedling and bolting stage of canola. Each dot represents a pan trap placement, with 3 pan traps occurring per sampling point, spaced 25 m apart. (B) The sweep net sampling layout, occurring during canola’s flowering and pod stages. The dotted line represents the path that was followed while sweeping between each sampling point.

Yellow pan traps (22 cm × 15 cm × 6 cm, L × W × H) were filled with 40% propylene glycol and placed at soil level. Three pan traps, 25 m apart, were spread across each of the sampling points between both transects, and insects collected in all 3 of these pan traps were then counted as a single sample for that specific field depth (eg 10 m). As there were 4 field depths (10 m, 50 m, 100 m, and 200 m), there were a total of 12 pan traps, or 4 samples, collected per field. Pan traps remained in the fields for the duration of each early development stage. Sweep net sampling consisted of 200 sweeps at 180° arcs between both transects at each sampling point, leading to a total of 800 sweeps per field; a standard 38 cm diameter sweep net was used. Sweep net samples were collected once at the peak of each late development stage, under favourable weather conditions (temperatures ranging from 15 °C to 30 °C, and wind speeds not exceeding 20 km/h). Samples were then transferred to emergence cardboard boxes, 30 cm by 25 cm by 30 cm (L × W × H) with sealed edges to deprive the inside of light (Dosdall et al. 2006). One face of the box was covered with opaque black fabric to allow the deposit of samples. A mesh screen with 3.5 mm openings was inserted into each box to separate large insects from smaller, flying insects (ie parasitoid hymenopterans). A small opening was cut into the box on the other side of the mesh filter, attracting the flying insects to light, where they were subsequently captured in a 50 ml Falcon tube filled with 95% ethanol. Parasitoids were given 4 days to exit the boxes. After this period, the boxes were manually searched to retrieve any missing specimens.

Comparability of Sampling Methods

Four sites were selected at random to compare collection methods between pan traps and sweep net sampling. Pan traps in these 5 fields were left for an additional 2 weeks to collect specimens during the flowering phase. Upon removal of the comparison pan traps, the fields were then swept as per the outlined methodology. Of the 32 samples used in this comparison, 3 were excluded: 1 pan trap sample and 1 sweep net sample at the same distance from the same site, due to mislabeling, and 1 pan trap sample due to loss of specimens. Pan trap samples used in this comparison were excluded from the final biodiversity analyses; however, they were included in the overall survey of parasitoid genera.

Identification of Samples

Specimens were identified to the level of genus; if genera could not be identified, specimens were identified to the lowest taxonomic level possible and listed as a morphotype (eg “Pteromalinae, Morphotype A”). While identification to species is desirable, parasitoid hymenopterans are notoriously difficult to identify due to their small sizes (generally <3 mm), the large number of species, complicated anatomy, and few available keys. Identification was done primarily through taxonomic keys (eg Goulet and Huber 1993, Gibson et al. 1997), as well as through DNA barcoding via the Canadian Centre for DNA Barcoding (CCDB, University of Guelph) using the COI-5P barcoding marker (complete methodology is available at Ratnasingham et al. 2024). Ninety-five specimens were submitted for DNA barcoding, of which 83 were successfully sequenced; DNA sequences and corresponding specimen identifications are publicly available at https://doi.org/10.5883/DS-ABBE23.

A level of uncertainty regarding specimens within the Proctotrupomorpha clade and lower-level classification (subfamilies and genera) of Pteromalidae (Chalcidoidea) should be noted. No lower-level taxonomic keys exist for Nearctic Cynipoidea, for example, and thus sample identification relied heavily on DNA barcoding and could not be confirmed morphologically. Few keys exist for other superfamilies in Proctotrupomorpha. Likewise, pteromalids are highly varied and notoriously difficult to identify via taxonomic key (Gibson et al. 1997), and not all specimens could be identified via DNA barcoding. After identifying all specimens, we calculated the genera richness, total abundance of individuals, and Shannon diversity index (SDI) of each sample. Richness and Shannon diversity indices were then converted to Hill numbers (q = 0 and q = 1, respectively) in order to standardize biodiversity measurements (Hill 1973). For simplicity, we continue to refer to these measurements as genera richness and SDI throughout this paper.

Comparison of Edge Types and Distances from the Edge

Each field was sampled 4 times throughout the growing season. During each sampling period, specimens were collected at 4 distances from the edge; thus, a total of 240 samples were collected (excluding the additional 20 samples that were used to compare collection methods between pan traps and sweep net sampling; see Comparison of Sampling Methods section). Of the 240 samples collected, 7 were excluded due to mislabeling (n = 3) or loss of specimens (n = 4).

Analyses

We first used generalized linear mixed models (GLMMs) to test for differences in richness, abundance, or SDI between pan trap and sweep net sampling methods and to test for fixed effects (distance from the edge, edge type, crop stage, and farm), interactions between distance from the edge and edge type, and random effects (repeated sampling of both study site and distance from the edge) on (a) parasitoid richness, (b) Shannon diversity, and (c) total abundance. We used GLMMs with a negative binomial distribution to model genera richness and abundance, as the data were over-dispersed and right-skewed, and a GLMM with the Gaussian distribution to model SDI as the data were continuous and normally distributed. After fitting each model, we used likelihood ratio tests (LRTs) to test the significance of each fixed effect. When there was a significant effect of edge type or distance from the edge, we used post-hoc pairwise tests of least squares means, with the Tukey method to adjust for multiple comparisons. Because individuals of the genus Diolcogaster were highly abundant in most samples, we used a Spearman correlation analysis to measure the influence of the presence of Diolcogaster on other genera and community evenness. We also repeated our analysis without the dominant Diolcogaster, but the results were qualitatively the same, thus we do not include them further.

We used 2 methods to assess the influence of edge type and distance from the edge on parasitoid community composition. First, we used non-metric multidimensional scaling (NMDS) ordination to visualize differences among samples in community composition, calculating community dissimilarities using the Bray-Curtis dissimilarity metric on the untransformed sample-by-species-abundance matrix. We then conducted a permutational multivariate analysis of variance (PERMANOVA) using the Bray-Curtis dissimilarity metric with 9,999 permutations to test for differences in community composition at different distances and edge-types. Next, we used an indicator species analysis (in this case, indicator genus) to identify genera that were significantly associated with different edge types. This analysis uses permutation to identify taxa that are significantly more frequent and/or abundant near 1 edge type than others (Dufrêne and Legendre 1997). Results for this analysis are given in Supplementary Information S2.

We used R version 4.4.1 for all analyses (R Core Team 2025), packages “glmmTMB” to build GLMMs (Brooks et al. 2017, McGillycuddy et al. 2025), “emmeans” for pairwise post-hoc tests, (Lenth and Piaskowski 2025), “car” for model testing, diagnosis, and visualization (Fox and Weisberg 2019), “labdsv” to conduct indicator species analysis (Roberts 2013), “vegan” for diversity analysis and ordination (Oksanen et al. 2015), “visreg” to create partial regression plots showing model predictions (Breheny and Burchett 2017), “ggplot2” for plot creation (Wickham 2016), “ggbreak” to create a y-axis break in Fig. 3 (Xu et al. 2021), “hillR” to convert diversity measurements to Hill numbers (Li 2018), “dplyr” for data management (Wickham et al. 2026a), and “tidyr” for data organization (Wickham et al. 2026b).

Fig. 3.

The distribution of the total number of individual specimens collected, categorized by family within each superfamily. Note the y-axis break between 1,500 and 12,500.

Results

Survey of Parasitoida

We collected 15,057 specimens, of which 12,781 belonged to the superfamily Ichneumonoidea (85% of all wasps), 682 belonged to the superfamily Chalcidoidea, and the remaining 1,594 were spread between the superfamilies in Proctotrupomorpha: Ceraphronoidea (59), Cynipoidea (1,148), Diaprioidea (20), Platygastroidea (362), and Proctotrupoidea (5) (Fig. 3; Table 1). The dominant genera (total catch, rank) with over 100 specimens listed in order of dominant family were the following: Diolcogaster (11,149, 1st), Cotesia (269, 6th), Peristenus (193, 8th), Microctonus (122, 10th) (Braconidae); Campoplex (486, 2nd), Erigorgus (137, 9th) (Ichneumonidae); Hexacola (417, 3rd), Trybliographa (364, 4th), Ganaspis (286, 5th); Trimorus (196, 7th) (Scelionidae); Brachymeria (106, 11th).

Table 1.

Complete survey of parasitoid Hymenoptera genera collected in 2023, including total number of genera and individual specimens collected per superfamily

Ichneumonoidea

Total genera: 67

Total specimens: 12,781

BRACONIDAE: Agathinae

Agathus sp. [2]

BRACONIDAE: Alysiinae

Asobara sp. [3]

Aspilota sp. [13]

Chorebus sp. [11]

Dinotrema sp. [60]

Orthostigma sp. [1]

Phaenocarpa sp. [8]

BRACONIDAE: Aphidiinae

Aphidius sp. [3]

Diaeretiella sp. [6]

D. rapae

Trioxys sp. [1]

BRACONIDAE: Braconinae

Bracon sp. [44]

B. cephi, B. lissogaster

Habrobracon sp. [1]

Morphotype A [56]

BRACONIDAE: Cheloninae

Chelonus sp. [24]

BRACONIDAE: Euphorinae

Meteorus sp. [1]

Microctonus sp. [123]

Peristenus sp. [193]

P. dayi, P. howardi, P. mellipes

Perilitus sp. [1]

BRACONIDAE: Microgastrinae

Apanteles sp. [2]

Cotesia sp. [269]

Diolcogaster sp. [11,149]

D. claritibia

Microgaster sp. [1]

Microplitis sp. [2]

BRACONIDAE: Opiinae

Opius sp. [66]

Morphotype A [4]

BRACONIDAE: Rhyssalinae

Acrisis sp. [4]

Oncophanes sp. [3]

Morphotype A [4]

Morphotype B [3]

BRACONIDAE: UNKNOWN SUBFAMILY

Morphotype A [1]

Morphotype B [6]

Morphotype C [3]

ICHNEUMONIDAE: Anemabninae

Erigorgus sp. [137]

ICHNEUMONIDAE: Banchinae

Banchus sp. [2]

ICHNEUMONIDAE: Campopleginae

Bathyplectes sp. [8]

Campoletis sp. [11]

Campoplex sp. [486]

Diadegma sp. [61]

Eriborus sp. [2]

Sinophorus sp. [2]

Morphotype A [4]

Morphotype B [1]

ICHNEUMONIDAE: Cryptinae

Aritranis sp. [3]

ICHNEUMONIDAE: Diplazontinae

Diplazon sp. [78]

D. laetatorius

Homotropus sp. [2]

ICHNEUMONIDAE: Ichneumoninae

Aoplus sp. [1]

Cratichneumon sp. [2]

Diadromus sp. [32]

Diphyus sp. [3]

Ichneumon sp. [6]

Spilichneumon sp. [2]

Morphotype A [1]

Morphotype B [1]

ICHNEUMONIDAE: Mesochorinae

Astiphromma sp. [1]

Mesochorus sp. [14]

ICHNEUMONIDAE: Ophioninae

Ophion sp. [3]

ICHNEUMONIDAE: Oxytorinae

Morphotype A [1]

ICHNEUMONIDAE: Phygadeuontinae

Gelis sp. [2]

Phygadeuon sp. [16]

Morphotype A [2]

Morphotype B [1]

ICHNEUMONIDAE: Pimplinae

Morphotype A [3]

Morphotype B [1]

ICHNEUMONIDAE: Tryphoninae

Erromenus sp. [6]

ICHNEUMONIDAE: UNKNOWN SUBFAMILY

Morphotype A [2]

Morphotype B [2]

Morphotype C [1]

Chalcidoidea

Total genera: 50

Total specimens: 682

CHALCIDIDAE: Chalcidinae

Brachymeria sp. [106]

Conura sp. [9]

CHALCIDIDAE: Haltichellinae

Psilochalcis sp. [1]

ENCYRTIDAE: Encrytinae

Copidosoma sp. [83]

Morphotype A [3]

Morphotype B [1]

EUCHARITIDAE: Oraseminae

Orasema sp. [15]

EULOPHIDAE: Eulophinae

Diglyphus sp. [28]

Euplectrus sp. [1]

Necremnus sp. [8]

Morphotype A [21]

Morphotype B [13]

Morphotype C [32]

EULOPHIDAE: Tetrastichinae

Aprostocetus sp. [54]

Baryscapus sp. [1]

Oomyzus sp. [1]

Tetrastichus sp. [26]

Morphotype A [3]

EULOPHIDAE: UNKNOWN SUBFAMILY

Morphotype A [2]

Morphotype B [1]

EUPELMIDAE: Eupelminae

Morphotype A [2]

EURYTOMIDAE: Eurytominae

Eurytoma sp. [22]

Morphotype A [2]

Morphotype B [12]

Morphotype C [21]

MYRMARIDAE: Mymarinae

Gonatocerus sp. [3]

Stephanodes sp. [1]

Lymaenon sp. [1]

PERILAMPIDAE: Perilampinae

Euperilampus sp. [2]

Perilampus sp. [153]

PIRENIDAE: Pireninae

Macroglenes sp. [54]

M. penetrans

Morphotype A [2]

PTEROMALIDAE: Miscogastrinae

Halticoptera sp. [13]

Merismus sp. [11]

Morphotype A [1]

PTEROMALIDAE: Pteromalinae

Catolaccus sp. [1]

Homoporus sp. [1]

Pteromalus sp. [21]

Trichomalus sp. [16]

Morphotype A [1]

Morphotype B [4]

Morphotype C [2]

Morphotype D [2]

Morphotype E [76]

Morphotype F [23]

SYSTASIDAE: Systasinae

Systasis sp. [1]

TORYMIDAE: TORYMINAE

Torymus sp. [3]

Morphotype A [1]

Morphotype B [1]

PROCTOTRUPOMORPHA

Ceraphronoidea

Total genera: 5

Total specimens: 59

CERAPHRONIDAE: Ceraphroninae

Aphanogmus sp. [6]

Ceraphron sp. [41]

MEGASPILIDAE: Megaspilinae

Conostigmus sp. [1]

Dendrocerus sp. [7]

Megaspilus sp. [4]

Cynipoidea

Total genera: 14

Total specimens: 1,148

CYNIPIDAE: UNKNOWN SUBFAMILY

Morphotype A [1]

FIGITIDAE: Eucoilinae

Ganaspis sp. [286]

Hexacola sp. [417]

Striatovertex sp. [51]

Trybliographa sp. [364]

Morphotype A [3]

Morphotype B [2]

Morphotype C [6]

Morphotype D [2]

Morphotype E [1]

Morphotype F [1]

FIGITIDAE: Figitinae

Figites sp. [1]

Melanips sp. [9]

Morphotype A [4]

Diaprioidea

Total genera: 4

Total specimens: 20

DIAPRIIDAE: Belytinae

Morphotype A [10]

Morphotype B [2]

DIAPRIIDAE: Diapriinae

Morphotype A [3]

DIAPRIIDAE: UNKNOWN SUBFAMILY

Morphotype A [5]

Platygastroidea

Total genera: 12

Total specimens: 362

PLATYGASTRIDAE: Platygastrinae

Inossotemma sp. [2]

Platygaster sp. [7]

Synopeas sp. [1]

Morphotype A [86]

PLATYGASTRIDAE: UNKNOWN FAMILY

Morphotype A [1]

SCELIONIDAE: Scelioninae

Calliscelio sp. [47]

Gryon sp. [59]

SCELIONIDAE: Teleasinae

Teleas sp. [1]

Trimorus sp. [196]

Morphotype A [3]

SCELIONIDAE: Telenominae

Telenomus sp. [27]

Trissolcus sp. [10]

PLATYGASTROIDEA: UNKNOWN FAMILY

Morphotype A [1]

Morphotype B [2]

Proctotrupoidea

Total genera: 2

Total specimens: 5

HELORIDAE: Helorinae

Helorus sp. [4]

PROCTOTRUPIDAE: Proctotrupidae

Proctotrupes sp. [1]

Unknown specimens are identified by morphotype letter code. Certain specimens identified to species are included.

Two hundred and six individuals were discarded, either due to decomposition (occurring in some pan trap samples), important features such as heads or wings missing, failure of DNA sequencing (due to specimen contamination or failure to process), or mislabeling. The majority of specimens were identified to genus, with the remaining specimens listed as morphotypes in their respective subfamilies or families, accounting for 142 specimens (see Table 1).

The majority of the samples were identified as Diolcogaster spp., consisting of 74% of all specimens collected. Of the 100 Diolcogaster specimens that were identified to species, all were classified as Diolcogaster claritibia. While Diolcogaster was extremely abundant, it had a weak negative correlation with the abundance of other genera, with correlation coefficients ranging between −0.0833 and −0.0009, excluding Torymus Dalman (Cynipoidea: Torymidae), which had a negative correlation of r_s = −0.1361. As expected, Diolcogaster abundance had a strong negative correlation with community evenness (r_s = −0.8597).

The indicator species analysis suggested that a number of parasitoid genera were indicators of certain field border types (Table S2). The most abundant among the indicator genera were Peristenus wasps (188 individuals) and Campoplex (172), from the Braconidae and Ichneumonidae families, respectively. Both of these genera were indicators of road verge habitats. Peristenus also had the highest prevalence among the native genera, as they occurred in 68% of the samples.

Comparison of Sampling Methods

Pan trap and sweep net sampling did not differ significantly across both diversity measurements (GLMM_richness, LRT: χ² = 1.7831, df = 1, P = 0.1818; GLMM_SDI, LRT: χ² = 0.5551, df = 1, P = 0.4562). However, individual abundance was significantly different between both sampling methods (GLMM, LRT: χ² = 4.3842, df = 1, P = 0.0363), where sweep net sampling resulted in a greater collection of individuals than pan traps (β  =  5.47, SE = 2.51, t = 2.175, P = 0.0406).

Effects of Crop Development Stages

Genus richness (GLMM, LRT: χ² = 28.7948, df = 3, P < 0.0001), SDI (GLMM, LRT: χ² = 25.0236, df = 3, P < 0.0001) and abundance (GLMM, LRT: χ² = 24.610, df = 3, P < 0.0001) significantly differed across all crop development stages. However, because abundance was significantly different between sampling methods measured within the same crop stage, results pertaining to this dependent variable should be interpreted with caution. Post-hoc analyses are reported in Table 2. While half of the crop stages analyzed were significantly different, consecutive crop stages did not differ across diversity measurements (eg flowering and pod development; LS-means_richness: β  =  0.2230, SE = 0.099, P = 0.1072; LS-means_SDI: β = −0.3910, SE = 0.249, P = 0.3976). Conversely, temporally distant crop stages such as seedling germination and pod development had significantly different amounts of diversity (LS-means_richness: β  =  0.4990, SE = 0.094, P < 0.0001; LS-means_SDI: β  =  0.6490, SE = 0.248, P = 0.0443), where early crop development stages showed greater parasitoid diversity and genera richness compared to late stages.

Table 2.

Pairwise comparisons of richness, Shannon diversity indices, and abundance across all crop stages, where germination = BBCH 10 to 29, shoot = BBCH 30 to 53, flower = BBCH 55 to 69, and pod = BBCH 71 to 89 (Lancashire et al. 1991)

	Richness			Shannon diversity index			Abundance
Crop stages	Estimate	SE	P-value	Estimate	SE	P-value	Estimate	SE	P-value
Germination—Shoot	0.1230	0.083	0.4540	0.1750	0.247	0.8948	0.0263	0.148	0.9980
Germination—Flower	0.2760	0.087	0.0083*	1.0400	0.247	0.0002*	−0.3345	0.141	0.0836
Germination—Pod	0.4990	0.094	<0.0001*	0.6490	0.248	0.0443*	0.4163	0.155	0.0370*
Shoot—Flower	0.1540	0.090	0.3152	0.8660	0.248	0.0028*	−0.3608	0.141	0.0520
Shoot—Pod	0.3760	0.096	0.0005*	0.4750	0.249	0.2267	0.3900	0.154	0.0541
Flower—Pod	0.2230	0.099	0.1072	−0.3910	0.249	0.3976	0.7508	0.151	<0.0001*

Significantly different pairwise comparisons are designated with an asterisk.

Effects of Edge Type and Distance

There were no significant interactions between edge type and distance from the edge across all diversity (GLMM_richness, LRT: χ² = 3.951, df = 6, P = 0.6834; GLMM_SDI, LRT: χ² = 7.6322, df = 6, P = 0.2663) and abundance (GLMM, LRT: χ² = 1.890, df = 6, P = 0.9296) measurements. We then adjusted the models to remove the interaction between these 2 dependent variables, analyzing them as separate fixed effects.

There was no significant difference in the richness (GLMM, LRT: χ² = 0.8878, df = 2, P = 0.6415; Fig. 4A), SDI (GLMM, LRT: χ² = 3.5198, df = 2, P = 0.1721; Fig. 4D), and abundance (GLMM, LRT: χ² = 4.4680, df = 2, P = 0.1071; Fig. 4H) of parasitoid hymenopterans across any of the edge treatments.

Fig. 4.

Model-adjusted fitted boxplots of the study’s dependent (richness of genera, A-C, Shannon diversity indices, D-F, abundance of individuals, H-J) and independent (edge type, distance from the edge, and sampled farm) variables. Blue lines represent group means, and gray shading represents the 95% confidence interval range. Code letters denote pairwise comparisons of independent variables and apply within panels only.

Parasitoid richness was significantly affected by distance from the field edge (GLMM, LRT: χ² = 15.1176, df = 3, P = 0.0017). After adjusting for multiple comparisons, the post-hoc analysis found that richness was highest closest to the edge at 10 m compared to 50 m (LS-means: β  =  0.2752, SE = 0.0932, z = 2.954, P = 0.0031), 100 m (β  =  0.3410, SE = 0.0946, z = 3.601, P = 0.0003), and 200 m (β  =  0.3148, SE = 0.0940, z = 3.350, P = 0.0008) across all treatments (Fig. 4B). Diversity tended to be highest at the edge and decreased further into the field (Fig. 4E), but we found no significant difference in SDI at any distance (GLMM, LRT: χ² = 5.9382, df = 3, P = 0.1147). Abundance followed a similar pattern but was not statistically different: highest at the edge and lower further into the field (Fig. 4I, GLMM LRT: χ² = 6.036, df = 3, P = 0.1099).

The ordination suggests little variation in community composition across edge types and distance from the edge (Fig. 5A and B). Consistent with this visual overlap, the PERMANOVA test showed no significant effect of treatment (F_{2, 233} = 0.732, R² = 0.0109, P = 0.1798) and distance (F_{3, 232} = 0.852, R² = 0.0127, P = 0.4328) on community composition.

Fig. 5.

NMDS ordination plots for edge treatment and distance into the field based on the abundance of parasitoid hymenopterans. The stress level in 3 dimensions is 0.19 (only the first 2 dimensions are plotted). (A) The ordination plot for edge boundaries with samples and corresponding 95% CI ellipses colored by treatment. (B) The ordination plot for distances into the field with samples and their corresponding 95% CI ellipses colored by distance. (C) The ordination plot for farms with samples and their corresponding 95% CI ellipses.

While edge treatment and distance from the edge had little effect on both diversity measurements, GLMM models showed significant differences in all biodiversity measurements among farms (GLMM LRT, SDI: χ² = 29.5251, df = 6, P < 0.0001; richness: χ² = 30.0742, df = 6, P < 0.0001; abundance: χ² = 39.656, df = 6, P < 0.0001) (Table 3, Fig. 4C, F, and J). In addition, the PERMANOVA test showed significant differences in parasitoid community composition between farm sites (F_{6, 229} = 3.5849, R² = 0.0859, P < 0.0001; Fig. 5C). Pairwise comparisons of all farms across all treatments can be seen in Table 3.

Table 3.

Pairwise comparisons of richness, Shannon diversity indices, and abundance across all sampled farms

	Richness			Shannon diversity index			Abundance
Farms	Estimate	SE	P-value	Estimate	SE	P-value	Estimate	SE	P-value
A–C	−0.1756	0.154	0.9155	−0.0155	0.440	1.0000	0.1076	0.245	0.9995
A–E	0.4349	0.173	0.1562	1.3763	0.439	0.0287*	0.3115	0.286	0.9319
A–G	0.2737	0.117	0.2285	1.6189	0.307	<0.0001*	−0.2645	0.210	0.8705
A–K	0.2256	0.142	0.6914	0.2881	0.373	0.9875	0.4938	0.211	0.2240
A–L	−0.1116	0.141	0.9860	1.2144	0.387	0.0285*	−0.5888	0.251	0.2237
A–W	0.5126	0.156	0.0181*	1.3299	0.384	0.0095*	0.4248	0.268	0.6908
C–E	0.6105	0.209	0.0542	1.3918	0.554	0.1541	0.2039	0.340	0.9969
C–G	0.4493	0.159	0.0714	1.6344	0.440	0.0039*	−0.3721	0.277	0.8312
C–K	0.4012	0.148	0.0947	0.3036	0.415	0.9907	0.3862	0.253	0.7285
C–L	0.0640	0.186	0.9999	1.2299	0.520	0.2130	−0.6964	0.314	0.2845
C–W	0.6882	0.198	0.0090	1.3454	0.517	0.1256	0.3172	0.328	0.9608
E–G	−0.1612	0.145	0.9240	0.2426	0.355	0.9935	−0.5760	0.211	0.0908
E–K	−0.2093	0.200	0.9437	−1.0882	0.502	0.3122	0.1823	0.315	0.9974
E–L	−0.5465	0.140	0.0018*	−0.1618	0.365	0.9994	−0.9002	0.205	0.0002*
E–W	0.0777	0.156	0.9989	−0.0464	0.364	1.0000	0.1134	0.228	0.9989
G–K	−0.0481	0.148	0.9999	−1.3308	0.373	0.0066*	0.7583	0.245	0.0327*
G–L	−0.3853	0.113	0.0119*	−0.4045	0.311	0.8512	−0.3243	0.166	0.4425
G–W	0.2389	0.132	0.5399	−0.289	0.307	0.9657	0.6893	0.191	0.0056*
K–L	−0.3372	0.176	0.4687	0.9263	0.464	0.4168	−1.0825	0.285	0.0028*
K–W	0.2870	0.188	0.7310	1.0418	0.461	0.2642	−0.0689	0.299	1.0000
L–W	0.6242	0.135	<0.0001*	0.1154	0.340	0.9999	1.0136	0.200	<0.0001*

Significantly different pairwise comparisons are highlighted. df_richness: Inf; df_SDI: 219; df_abundance: Inf.

Discussion

Parasitoid Hymenoptera are important components of the arthropod community in agricultural landscapes, yet we know little about what influences their abundance, diversity, and composition. Our study shows that the abundance and richness of this group in canola fields tend to be highest near non-crop borders and decline moving into the field but that the type of non-crop habitat has little influence. Overall, the identity of the farm had the largest influence on parasitoid communities in canola fields, suggesting that local-scale factors, such as farm management practices, may be major determinants of parasitoid diversity and composition in southern Alberta canola. Broader scale factors (climate, topography, ecoregion) or intermediate-scale factors (landscape structure) were not considered in this study and their influences on Hymenoptera communities remain to be disentangled from local farm (field-level) scale factors.

Our finding of significantly higher genera richness but not other diversity measurements at edge borders may be explained by a variety of factors. For example, Beketov et al. (2010) found that environmental variables such as aquatic vegetation, water quality, and terrestrial habitat composition may influence mosquito communities by increasing richness through resource availability but decreasing abundance through activities such as predation. Similar effects could be occurring in this study, where sampled non-crop habitats provide enough resources or niches for multiple unique parasitoid genera, but factors such as pesticide use, predation from birds, or disturbances from farm machinery may be lowering the overall abundance of individuals within sampled fields. This may also explain why abundance, while not significantly different across sampled distances, was still observed to be higher closest to the edge compared to further distances (Fig. 4I). If some variable decreased parasitoid abundance within a field, then non-crop habitats could provide a level of protection for the insects, resulting in spillover effects from non-crop habitat into arable land and therefore more insects along the edge boundaries. These trends of higher diversity and richness closer to the edge parallel other studies (eg Rand et al. 2006, Nguyen and Nansen 2018, Whitehouse et al. 2018) that suggest components of insect biodiversity peak at the edge of non-crop habitats and decline with increasing distance into cropland.

There are at least 3 possible explanations as to why edge type did not differ in diversity measurements: (1) the 3 edge types examined may not differ sufficiently in functional complexity to produce distinct parasitoid responses; (2) large-scale landscape factors such as topography, water bodies, and microclimates have greater influence on parasitoid communities compared to habitats immediately adjacent to sampled fields; (3) external factors, such as surrounding crops, crop rotation, or farming practices, play a stronger role in parasitoid biodiversity than edge types. The underlying cause may be a combination of all 3; however, the fact that farms had significantly differing levels of SDIs and genera richness implies that the individual farming practices of producers likely affect community diversity. Agricultural intensification—such as heavy tillage, excessive pesticide use, and shortened crop cycles—is well-known to decrease biodiversity (Tscharntke et al. 2005, Campbell et al. 2017). Each producer that took part in the study practices their own farming philosophy and engages in varying levels of agricultural intensification. For example, farms “A” and “L” do not till their fields, which may have contributed to their higher biodiversity markers, as certain species of parasitoid hymenopterans (and their hosts) overwinter in the field soil (Nilsson 2010). While pesticide use could not be accurately documented or applied to the results of this study due to timing differences and lack of communication from some producers, spraying frequency likely also played a role in diversity measurements. Farms “E,” “G,” “W,” and “K” were all documented to be sprayed at least twice throughout the growing season to combat various pest insects (ie Lygus bugs, flea beetles, and cabbage seedpod weevils), which would have likely had non-target effects on beneficial parasitoids. These 4 farms, particularly “G” and “W,” had lower overall SDIs, abundance of individuals, and richness of genera compared to the other farms sampled. Again, while farming practices and land management could be contributing to the observed low levels of parasitoid biodiversity, other factors such as surrounding crops and coarse-scale landscape effects should be investigated.

Finally, parasitoid diversity and abundance significantly differed across the majority of crop development stages. This overall trend is to be expected and is most likely due to parasitoid phenology, as their emergence and activity levels are temporally linked to those of their hosts. As resource type and availability changes throughout the crop development cycle, so too will the host communities present and, subsequently, their parasitoids. However, there is some overlap in the feeding patterns of pest insects and consecutive crop development stages; this may explain why we observed non-significant differences between certain consecutive crop stages, such as the seedling germination to shoot development stages, and the flowering to pod ripening stages. Flea beetles, for example, prefer to feed on seedlings in early spring and may continue feeding until early summer (Burgess 1977), a timeline that aligns with the transition from seedling germination to shoot development in canola. Similarly, Lygus bugs prefer to feed on the flowers and early pods of host plants, up until pods have ripened and are too difficult to pierce (Kelton 1975). Thus, while overall we expect insect diversity to shift significantly over the crop development cycle, there is still a level of overlap of host feeding preferences between consecutive development stages.

Ichneumonoidea was the most abundant superfamily of parasitoids in this study. Ichneumonoidea is well studied for its biocontrol agents (Quicke 1997), and indeed, many of the collected genera included natural enemies such as Peristenus Förster (Hymenoptera: Braconidae), parasitoids of Lygus Hahn (Hemiptera: Miridae); Bracon Gahan (Hymenoptera: Braconidae), parasitoids of the wheat stem sawfly, Cephus cinctus Norton (Hymenoptera: Cephidae); and Campoplex Gravenhorst (Hymenoptera: Ichneumonidae), parasitoids of various lepidopterans.

The indicator species analysis, applied to genera in this study, suggested that members of Ichneumonoidea can have indicator value in canola. Peristenus (Braconidae) had the highest prevalence and abundance near road verges. These wasps are known to parasitize Lygus bugs in various crops, but were considered rare in canola habitats (Fernandez et al. 2018). However, a recent study showed that they have now become widespread in many crops, including canola, where the most common species is likely to be P. howardi attacking L. keltoni (Floate et al. 2024, Pawluk 2024). Their higher abundance near the edges may suggest that they are migrating to the fields from surrounding fields that had canola, or other hosts of Lygus bugs, the previous year where they overwintered in the soil. Also, it is possible that road verges harbor feral alfalfa, which is a host of Lygus bugs and helps to maintain a population of Peristenus wasps along the road edges. The fact that Peristenus were highlighted in this analysis can be used to encourage growers to avoid spraying close to road verges where these parasitoids can be relatively more abundant than at further distances into the field.

Most wasps in this study belonged to the genus Diolcogaster, another wasp belonging to the family Braconidae. Over 100 species of Diolcogaster are currently known, and it is considered one of the most common genera in the subfamily Microgastrinae (Moghaddam et al. 2019), although New World Diolcogaster are poorly understood, with a complicated history of taxonomic classification. These wasps are endoparasitoids of various Lepidoptera larvae and have a global distribution (Fernández-Triana 2023). In particular, D. claritibia Papp, a parasitoid of the diamondback moth (P. xylostella), is likely the dominant species that was collected in this survey; however, only 100 of >11,000 specimens were identified to species and confirmed to be (in the case of DNA barcoding) or likely to be (in the case of morphological identification) D. claritibia. Thus, there is a probability that other species of Diolcogaster (eg D. facetosa Weed; Fernández-Triana 2010) exist in the samples. D. claritibia was first officially reported in North America by Fernández-Triana et al. in 2014 and is believed to have originated from Europe. It may have existed in Alberta as far back as 1994 (eg Finnamore 1994); however, no concrete evidence exists to confirm this. Moeinadini et al. (2025) suggest that D. claritibia’s range will continue to expand due to its capacity to adapt to various habitats, emphasizing the need for ongoing surveys to monitor the population.

The overwhelming dominance of D. claritibia in canola is intriguing and difficult to explain. This species was first reported as an adventive migrant to Canada from samples collected in southern Alberta, Canada in 2003 (Fernández-Triana et al. 2014), but its common occurrence in the region in this crop was not noticed until more than a decade later (Cárcamo, unpublished data 2018). The only host reported for this parasitoid was the diamondback moth (P. xylostella) until Abram et al. (2022) recorded it from its congener, P. armoraciae in southern British Columbia. Diamondback moth outbreaks occur in western Canada when the moths are carried by the southern jet streams from the United States or Mexico into the Canadian Prairies, but these events occur only intermittently (Hopkinson and Soroka 2010). The last 3 major outbreaks in the region occurred in 1995, 2001, and 2017 (Prairie Pest Monitoring Network, https://prairiepest.ca/). It is possible that D. claritibia populations increase substantially the year after an outbreak, which explains the large populations observed in canola fields in 2018. How they manage to persist at high numbers now that their preferred host has been rare for many years is a mystery. One possibility is that they have switched to other Lepidoptera hosts, such as cutworms (Noctuidae) or Pieris species (Pieridae), which are common in canola fields. Alternatively, there could be cryptic related parasitoids that we failed to recognize because we determined to species only a small proportion of all the wasps collected. D. claritibia has been reported elsewhere in the Parkland Ecoregion of north central Alberta in canola-pea intercrops where it appears to be one of the most common Hymenoptera, despite low occurrence of its main host (Jose Correa Ramos, Rowan Rowland, Maya Evenden, Dominic Donkor, Jagroop Kahlon, and Felix Sperling, unpublished data). Interestingly, a recent survey of the genetic variability of D. claritibia in this canola-pea intercrop study suggested that the population was founded by very few individuals because there was almost no variability among COI haplotypes (Rowan Rowland, Maya Evenden, Jose Correa Ramos, and Felix Sperling, unpublished data). The range expansion of this species and its potential impact on native Lepidoptera and indirect effects through competitive displacement of native parasitoids is worth continued study.

It was expected that Ichneumonoidea and Chalcidoidea would be the most common superfamilies, as they are the most populous and speciose (Goulet and Huber 1993, Huber 2017); however, the poorly understood superfamily, Cynipoidea, was the second most abundant group of parasitoids after Ichneumonoidea. Cynipoids include Cynipidae, a family of gall-forming wasps that include many known natural enemies, including some of the few examples of phytophagous hymenopterans (Goulet and Huber 1993). The commonly collected genera in this family were Hexacola Förster, Ganaspis Förster, and Trybliographa Förster, all of which belong to the family Figitidae and include species of known biocontrol agents of dipterans (Wishart and Monteith 1954, Kasuya et al. 2013, Diamond et al. 2001). Trybliographa rapae Westwood, for example, attacks the cabbage root fly, Hylemya brassicae L. (syn. Delia radicum; Diptera: Anthomyiidae), whose larvae attack the roots of Brassicaceae plants such as canola (Wishart and Monteith 1954).

Following Ichneumonoidea and Cynipoidea, Chalcidoidea was the third most collected superfamily. Chalcids in this survey included the most morphotypes collected, primarily due to their small size causing difficulty when identifying features morphologically (eg Gibson et al. 1997). Chalcidoidea was much less frequently collected in this survey compared to previous Albertan surveys (eg Finnamore 1994), perhaps again due to their small size making them difficult to detect when hand-sorting through pan trap and sweep-net samples. Indeed, fewer chalcidoids were caught during the early growing season, implying that some samples may have been missed during pan trap collection and sorting.

Finally, the other superfamilies of Parasitoida were collected sparsely in this survey. Only 20 specimens were collected from Diaprioidea, 5 specimens from Proctotrupoidea, and 362 specimens from Platygastroidea. No specimens were collected from Mymarommatoidea; however, this was expected, as only 2 species in the superfamily are currently recognized in Canada, neither of which are found in Alberta (Huber et al. 2021).

In conclusion, this study demonstrated that edge type has minimal effects on overall parasitoid populations in our study system and that distance from the edge significantly influences parasitoids only at the distance closest to the edge. While many studies report strong edge effects in agricultural systems, the relatively weak responses observed here suggest that such effects may be context-dependent and that local edge effects may be outweighed by coarser-scale landscape factors such as topography, surrounding waterbodies, or farm-level drivers such as frequency and type of pesticide use, tillage, and crop-rotation cycles. Future research should investigate the relationship between parasitoid diversity and abundance to specific farm management practices to determine what practices lead to the greatest abundance of beneficial parasitoids. It should be noted that our study was limited to a sampling timeline of only 1 growing season. As insect populations fluctuate with shifts in climate, host availability, crop rotations, and more, the conclusions presented in this paper should be interpreted within the context of a single study year. Future, multi-year research should be conducted to monitor interannual variability of parasitoid populations in southern Alberta canola, as well as determine if the diversity trends observed in this study are consistent across a larger temporal scale.

Finally, this study provided the first contemporary survey of parasitoid hymenopterans in southern Alberta canola, demonstrating Diolcogaster as the dominant genus. Continuing surveys of parasitoid hymenopterans should be conducted in order to monitor the spread of Diolcogaster spp. and determine (1) if the exotic D. claritibia is the dominant species in southern Alberta canola, and (2) if Diolcogaster spp. are outcompeting native parasitoids.

Author Contributions

Abbe Pawluk (Conceptualization [equal], Data curation [lead], Formal analysis [equal], Funding acquisition [equal], Investigation [lead], Methodology [equal], Resources [equal], Validation [equal], Visualization [equal], Writing—original draft [lead], Writing—review & editing [equal]), Jenny L. McCune (Data curation [equal], Formal analysis [equal], Supervision [supporting], Visualization [equal], Writing—review & editing [equal]), Robert A. Laird (Conceptualization [supporting], Formal analysis [equal], Funding acquisition [equal], Project administration [equal], Resources [equal], Supervision [equal], Validation [equal], Writing—review & editing [equal]), and Héctor A. Cárcamo (Conceptualization [equal], Formal analysis [supporting], Funding acquisition [equal], Investigation [equal], Methodology [equal], Project administration [equal], Resources [equal], Supervision [equal], Writing—review & editing [equal])

Supplementary Material

Supplementary material is available at Environmental Entomology online.

Funding

This research was funded by the Alberta Conservation Association (Grant in Biodiversity to AP; grant no. 1265-1/1208), the Western Grains Research Foundation and the Alberta Canola Producers Commission (granted to HC; grant no. ACPC-133-AR20; AFC2020F093R), and the Natural Sciences and Engineering Research Council of Canada (Discovery Grant to RL; grant no. RGPIN-2021–02967).

Conflicts of Interest

None declared.