Introduction

In North America, the pepper weevil Anthonomus eugenii (Cano) (Coleoptera: Curculionidae) poses a significant threat to production of cultivated pepper Capsicum annuum Linnaeus (Solanaceae) crops. Originating from Mexico, the current geographic range of A. eugenii includes southern mainland USA and Hawaii (Elmore et al. 1934; Patrock and Schuster 1987), Central America (Bartlett 1978) and the Caribbean (Abreu and Cruz 1985). In addition, this pest has sporadically occurred in Canada (Costello and Gillespie 1993; Labbé et al. 2018) and was briefly present in Italy (Speranza et al. 2014) and the Netherlands (van der Gaag et al. 2020). In the USA, an estimated $23 M US in annual crop losses is attributed to A. eugenii (USDA 1995). In southern Ontario Canada, a single A. eugenii outbreak in 2016 caused an estimated $49 M US in crop losses (Chiu 2020).

Some of the attributes central to the successful establishment and expansion of this pest species into new geographic ranges include its relatively broad host range, its cryptic biology, and the scarcity of effective management tools. Adult A. eugenii can spread quickly by flying between fields, hitchhiking on vehicles or other transported items, reproducing on wild nightshades (Solanaceae) which serve as alternate host plants, or through the transport of infested produce (Patrock and Schuster 1987; Fernández et al. 2021). In addition, while A. eugenii cannot survive exposure to temperatures lower than − 10 °C, it may persist on protected crops during the winter, making it of particular concern to greenhouse pepper producers at northern latitudes (Fernández et al. 2017).

While adult A. eugenii can cause feeding damage to fruit, flowers and leaves, the most damaging life stage are the larvae that feed and develop within fruit, where they contribute to fruit contamination, premature decay and abscission (Campbell 1924; Elmore et al. 1934; Patrock and Schuster 1992; Riley and Sparks 1995). Damage from A. eugenii can be so severe, in both greenhouses and fields, that it may become more economical to remove entire crops than to harvest any remaining uninfested fruit (Riley and Sparks 1995; Fernández et al. 2021).

Currently, intensive crop scouting, physical and cultural management, and the application of chemical insecticides are the main tools and tactics used to manage A. eugenii populations (Costello and Gillespie 1993; McCreary et al. 2017; AAFC 2020). However, the concealed larvae of A. eugenii within fruit are largely unaffected by insecticides and thus only target the adult (Stansly and Konstyk 2015; Labbé et al. 2020). In addition, insecticides can jeopardize existing biological control programs applied to effectively manage other greenhouse pepper pests. Finally, due to the labour-intensive and costly nature of existing physical and cultural management tools geared to this pest, there is a pressing need to develop alternative tactics for A. eugenii control (AAFC 2020).

Among the alternative methods for mitigating A. eugenii infestations is biological control by the parasitoid wasp Jaliscoa (= Catolaccus) hunteri (Crawford) (Hymenoptera: Pteromalidae), a species first identified as an associate to A. eugenii and its congeneric species in 1912 (Dwight-Pierce et al. 1912). In subsequent surveys of Mexico and Canada, J. hunteri represented an important natural enemy of A. eugenii (Rodríguez-Leyva et al. 2007; Labbé et al. 2018). Females attack all instars of A. eugenii larvae as well as the egg stage. However it is most effective at suppressing L3 instar weevils (Rodríguez-Leyva et al. 2000; Gómez-Domínguez et al. 2012). Female J. hunteri can lay up to 466 eggs over their lifetimes and have an intrinsic rate of increase of 0.18 d−1, which is greater than that observed for A. eugenii (0.14 d−1) (Rodríguez-Leyva et al. 2000; Seal et al. 2002; Rodríguez-Leyva 2006). Thus, J. hunteri may represent a good candidate for the biological control of A. eugenii.

Despite these characteristics, past studies of J. hunteri as a biological control agent of A. eugenii have presented unclear results (Bartlett 1978; Schuster 2007; Esteban Rodríguez-Leyva, personal communication). Releases of J. hunteri in Hawaii from 1934 to 1937 resulted in successful wasp establishment but their impact on A. eugenii populations was not verifiable (Bartlett 1978). In 2012, field trials in Florida, USA demonstrated that J. hunteri could significantly reduce the number of weevil-infested field bell pepper fruit (Schuster 2007) but did not demonstrate a significant difference in the number of adult A. eugenii in treated relative to untreated plots. In addition, preliminary greenhouse trials performed by Rodríguez-Leyva (Colegio de Postgraduados, Moncecillo, Mexico, personal communication) found non-significant differences in A. eugenii floral and fruit damage between J. hunteri treated and untreated greenhouses.

These unclear results may be due to our poor understanding of the factors that affect this parasitoid-host relationship in agricultural systems, which could be better elucidated through a series of controlled environment trials. For instance, the premature fruit abscission triggered by A. eugenii can affect wasp host searching efficacy and parasitism success rate and remains a key factor yet to be investigated (Ramalho et al. 2000). As well, due to the limited range of the up to 2.27 mm long J. hunteri ovipositor, larval placement within fruit and fruit size may also be important factors in parasitism success (Riley and Schuster 1992; Gómez-Domínguez et al. 2012). Finally, it is still unclear how various parasitoid exposure periods or weevil larval instar affect parasitoid attack rates and ultimately pest management.

Given these ambiguities, this study aimed to assess the impact of these multiple factors that could affect A. eugenii management potential of J. hunteri. To this end, we performed controlled environment trials first comparing ornamental pepper plant cultivars as weevil hosts, whose production of multiple small fruit are ideal for studying A. eugenii–J. hunteri dynamics. We then evaluated the effects of wasp exposure period, weevil larval instar, fruit size and abscission status to improve our knowledge of the interaction between A. eugenii and its biological control agent, J. hunteri.

Materials and methods

Pepper weevil colony

In 2016, a laboratory colony of A. eugenii was established at Agriculture and Agri-Food Canada’s (AAFC) Harrow Research and Development Centre (HRDC; GPS coordinates 42° 2′ 4.844″ N, 82° 53′ 56.899″ W) from local field and greenhouse collected infested pepper fruit. Weevils were maintained in a controlled environment cabinet at 29 °C, 60% RH, and a L:D 14:10 photoperiod and were maintained as previously described (Fernández et al. 2021; Leo 2022).

Parasitoid colony

A laboratory colony of J. hunteri was established at the HRDC in 2017, from individuals obtained from Koppert Mexico and Canada under the terms of Material Transfer Agreement AGR-14521. Rearing methods were adapted from those established by Vásquez et al. (2005). The wasps, along with a colony of its alternative cowpea weevil host, Callosobruchus maculatus Fabricius (Coleoptera: Bruchidae) (Rojas et al. 1998), were maintained in a controlled environment cabinet set at 27 °C, 60% RH and a L:D 16:8 photoperiod. Approximately 400 g of dry chickpeas and 400 adult C. maculatus were placed into a 1 l Mason jar with a metal mesh screened lid and filter paper. After a one week oviposition period, adult C. maculatus were removed from the jar by sifting chickpeas through a US mesh size 3 sieve. Remaining infested chickpeas were returned to the controlled environment for 14 days until C. maculatus reached L4 (14–21 days old), at which point ~ 90% were divided equally into three J. hunteri oviposition cages. Infested chickpeas were placed into elevated handmade coarse metal (3-10 mm mesh) trays. Remaining chickpeas served to propagate the C. maculatus colony.

Jaliscoa hunteri oviposition cages consisted of 4 l clear plastic jugs (~ 25 × 15 × 13 cm) with a mesh sleeve glued to its 9 cm wide opening and closed at the end with a large freezer clip. Each cage had a cotton-wicked cup containing a 10% honey-water solution, and pure liquid honey was streaked across the top of oviposition cages (1 mm width × 30 cm length streaks) using a syringe. This additional honey greatly improved wasp survival over time (Harvey et al. 2017). The oviposition cages contained adult parasitoids previously aspirated from emergence containers. Fourth instar C. maculatus-infested chickpeas were removed from parasitoid oviposition cages after four days and placed into 4 l clear plastic containers similar to the oviposition cages for parasitoid and C. maculatus emergence. Empty trays in oviposition cages were then refilled with new L4 C. maculatus-infested chickpeas. Twice a week, emergence containers were checked for the presence of parasitoids and C. maculatus, which were then aspirated into separate vials. Containers were checked for parasitoid and C. maculatus three times over ten days, then the contents were frozen and discarded. After three weeks, the contents of an oviposition cage were frozen and discarded.

Pepper plants

Capsicum plants cv. Blaze (815H) were grown in rockwool substrate (Grodan Delta; Roermond, The Netherlands), and maintained under greenhouse conditions. Prior to the start of trials, plants were pruned of their first crown fruit to encourage greater fruit production on side branches, which resulted in plants with a higher total number of fruit of various sizes during the trials. Once these plants produced a full cohort of unripened fruit that were light yellow-green in colour with some approaching full size (seven weeks old and ~ 25–30 cm3 volume), these plants were then placed into 13.7 cm wide saucers (#059–4434-6 Planters’ Pride, Canadian Tire, Toronto, ON, Canada) and transferred into a controlled environment cabinet set to 27 °C, 60% RH and a L:D 16:8 photoperiod where they were watered three times a week and maintained there for the duration of trials.

Effect of plant cultivar on parasitoid potential for Anthonomus eugenii suppression

Trials were conducted at the HRDC to assess the impact of the parasitoid, J. hunteri, on A. eugenii established on ornamental pepper plants. For these trials, six plants of each of three ornamental C. annuum pepper cultivars (Blaze, Medusa, Wicked, Stokes Seeds, Canada) were individually enclosed within a 28 × 46 cm food-safe high-clarity microperforated polypropylene bag (1 mm diameter perforations with 27 perforations cm−2; Chantler Packaging Inc, Mississauga, ON), sealed at the main stalk base using a twist tie around a dry cotton wrap. Each individually enclosed plant represented an experimental unit. Six A. eugenii (three males, three females) of between seven and 12 days following their adult emergence were released into the cage, where they remained for the duration of the trial. This age of weevils was chosen as females are known to have a pre-oviposition period of 2.9 days at 26 °C (Seal and Martin 2017). These A. eugenii were sexed by microscopic observation of enlarged mucrones (spurs) present on the hind legs of male A. eugenii as well as the longer rostrum of females (Eller 1995; McCreary et al. 2017).

Seven days after the initial A. eugenii release, five adult male and five female J. hunteri (3–4 days old) were removed from a mixed (and assumed mated) colony and released into three treated plant cages for each pepper cultivar. A 10% honey-water solution (honey from Sun Parlor Honey, Kingsville, ON, Canada) at a volume of ~ 0.45 ml total within three sprays, was then sprayed onto the leaves of these J. hunteri treated plants to provide wasps with a source of carbohydrates to prolong their survival and improve their reproduction (Morales-Ramos et al. 1996). This sprayed solution dried relatively quickly and so did not represent a substrate for sooty mold. One and two weeks after the parasitoid and A. eugenii introductions, respectively, experimental plants were destructively sampled. All insects, fruit, and buds were removed from each plant. Buds and fruit were counted and placed into 10 cm wide Petri dishes with ventilated lids (with ~ 3 cm wide 0.01 mm gauge mesh screened opening) and were kept in a controlled environment cabinet (Conviron GEN1000 Reach-In, Controlled Environments Inc., Pembina, ND, USA) set at 27 °C, 60% RH and a L:D 16:8 photoperiod. Pepper weevil and parasitoid emergence were monitored 2–3 times a week for two weeks (i.e., four weeks since the start of the trial) which was sufficient time to account for all weevils and wasps that could emerge from each experimental plant.

Impact of parasitoid exposure period, Anthonomus eugenii life stage, host fruit size or abscission status on A. eugenii offspring emergence

Nine ornamental pepper plants (cv. Blaze) of an equal age cohort (8–10 weeks old based on cohort specific-season-dependent development) were randomly selected for trials, each producing fruit of equivalent developmental stages (size up to 2 cm diameter) and in nearly equivalent numbers. All fruit were counted on each plant prior to the start of trials. The foliar and fruiting parts of each plant were then enclosed within a rectangular tulle mesh sleeve (~ 25 × 50 cm with 0.01 mm gauge) to create an experimental plant cage. For each of these cages, six adult male and six female A. eugenii were added. All A. eugenii adults released in these trials were between two and 14 days old from the day of their adult emergence and were sourced from a mixed-sex colony container. After two days, A. eugenii adults were removed from all plants by placing them into a 4 °C refrigerated chamber (KeepRite Refrigeration, Brantford, ON, Canada) for 20–30 min, which caused weevils to fall off without affecting their survival as determined through prior assessments. Plants were then returned to ambient laboratory conditions (~ 23 °C) and were briefly shaken upside down to dislodge any remaining weevils which were removed by aspirator. Adult weevils were counted, and every individual initially placed onto plants was accounted for.

For parasitoid treatment cages, newly emerged adult J. hunteri – ten females and ten males, were released into each treatment plant cage. Parasitoids were sourced from the colony, either between 1–24 h or 1–48 h after adult emergence for females and males, respectively. This short time frame served to reduce parasitoid-host contact time to minimize preference for parasitizing cowpea weevil hosts. For these trials, J. hunteri were released onto A. eugenii infested plants when A. eugenii were at the L1 (PW1 cohort) or L3 (PW3 cohort) life stages. For PW1 cohorts, parasitoids were released two days following the initial A. eugenii exposure, which corresponded to the presence of L1 A. eugenii based on their known temperature-dependent development time (Toapanta et al. 2005). Parasitoids were removed from plants either five or nine days after A. eugenii release, resulting in a three or seven days exposure period, respectively. Accordingly, for PW3 treatments, parasitoids were released for trial days 7 to 14 for a 7-days exposure and on days 9 to 12 for a three days exposure. Along with parasitoid releases, plants were sprayed with a 10% honey-water solution (Sun Parlor Honey, Kingsville, ON, Canada) at a volume of ~ 0.45 ml total within three sprays to improve parasitoid survival (Morales-Ramos et al. 1996). Each treatment combination had three experimental replicates and three assessments were conducted in time (technical replicates) for a total of nine replicates per treatment and condition. Control plants which had A. eugenii without J. hunteri exposure were assessed in three experimental replicates repeated over time through six trials for a total of 18 technical replicates.

At the end of each parasitoid exposure period, all wasps were removed by aspirator from J. hunteri treatment plants. Following these exposure periods, 14 days after the initial A. eugenii introduction, fruit were removed from each plant and sorted based on whether they were abscised or remained attached to plants at the time of fruit collection, as well as by their size (three fruit diameter categories: small: < 1 cm, medium: 1 cm-1.5 cm, large: > 1.5 cm). The fruit from each of these groups were placed separately into ventilated plastic emergence containers (4.5 × 12 cm diameter Ziploc, Dow Chemical, USA) and maintained within the previously described controlled environment cabinet. Each emerged A. eugenii and parasitoid from these fruit were counted 2-3 times per week for two weeks.

Statistical analysis

All analyses were performed in SAS v 3.8.1 (SAS 2020). For initial trials comparing the effects of J. hunteri treatment on A. eugenii emergence in three ornamental pepper plant cultivars, weevil offspring count data were first examined for normality using both a Shapiro-Wilks test and residuals plots and data were found to be normally distributed. Adult A. eugenii emergence data were then compared for all cultivars using a PROC Mixed ANOVA with the parasitoid treatment modeled as a fixed effect and replicate and pepper variety as random effects, followed by a Tukey’s Honest Significant Difference (HSD) test at a confidence level of 5% to identify significant differences among treatment means.

For subsequent assays examining the impacts of immature pepper weevil life stage and parasitoid exposure period (three days; n = 71 and seven days; n = 77) on A. eugenii offspring emergence for a single pepper cultivar, data were first assessed for normality through both Shapiro-Wilks test and residuals plots. Finding these data to be normally distributed, the mean number of adult A. eugenii emerging from parasitoid treatment and control plants were compared using a PROC Mixed ANOVA with wasp exposure length, weevil instar and their interaction modeled as fixed effects, and replicate and trial block as random effects, followed by a Tukey’s Honest Significant Difference (HSD) test at a confidence level of 5% to identify significant differences among treatment means.

The potential for parasitoids to suppress adult A. eugenii offspring emergence were next determined, first on a single treatment versus control plant basis, as well as for plants distinctly within each of the two parasitoid exposure period treatments. These suppression data were calculated by taking the mean number of A. eugenii emerged from control plants and subtracting from these the mean number of A. eugenii emerged from treatment plants, then dividing this number by the mean A. eugenii emergence for control plants. In the rare instance where individual suppression datapoints fell below zero, due to variance in emergence between treatment and control plants, these negative data points were set to a minimum of zero, as it is logically not relevant to have negative pest suppression values. These weevil offspring suppression data were next analyzed first on a per plant basis, as well as for each of the 7- and 3-days exposure periods. Prior to analyses, these datasets were examined for normality using Shapiro-Wilks tests and residuals plots and were all normally distributed except for suppression per fruit data from the 7-days exposure treatments which required an arcsine-square root transformation to achieve normality.

These three suppression datasets were then analyzed by PROC Mixed ANOVA and Tukey’s HSD post-hoc tests at a confidence level of 5%. For suppression per plant analyses, exposure length, weevil instar and their interaction were set as fixed effects, while replicate and block were analyzed as random effects. For suppression analyses for either 7- or 3-days parasitoid exposure datasets, fruit size (small, medium large), fruit status (on plant or abscised) and their interaction were modeled as fixed effects as these were designated as the main factors under investigation in these analyses. Whereas weevil instar, replicate and block were analyzed as random effects.

Results

Effect of pepper cultivar on parasitoid Anthonomus eugenii suppression potential

Releases of J. hunteri successfully reduced the overall number of A. eugenii adult offspring emerging from infested ornamental pepper plant buds by approximately 64.3 ± 3.2% compared to control/untreated plants (p = 0.004; Fig. 1; Table 1). However, no significant difference was observed in mean number of A. eugenii offspring emergence between cultivars (p = 0.305, Fig. 1; Table 1).

Fig. 1
figure 1

Mean number of adult pepper weevil, Anthonomus eugenii offspring (± SE) that emerged from buds of ornamental pepper Capsicum annuum plants of three varieties (Blaze, Medusa and Wicked) previously unexposed (control, black columns) or exposed to parasitoid Jaliscoa hunteri (white columns). Significant differences in A. eugenii emergence between treated and control plants for each variety was designated by a line and asterisks, Tukey’s HSD, α = 0.05

Full size image
Table 1 Summary statistics for mixed model analyses of variance for effects of Jaliscoa hunteri exposure, and other plant or host parameters, on the emergence of adult pepper weevil, Anthonomus eugenii, from ornamental pepper Capsicum annuum
Full size table

Impact of parasitoid exposure period, Anthonomus eugenii life stage, host fruit size, abscission status on A. eugenii offspring emergence and suppression

In subsequent assessments of how J. hunteri exposure period and A. eugenii larval life stage might affect adult weevil emergence at the whole plant level for a single pepper cultivar, it was found that exposure length, but not instar nor its interaction with exposure period, significantly affected weevil offspring emergence (Table 1). Specifically, exposure of L3 instar weevils, to J. hunteri over seven days resulted in 61.7 ± 11.9% (± SE) lower adult A. eugenii offspring emergence (8.1 ± 2.5 weevils per plant) relative to untreated control plants (21.2 ± 3.1 weevils per plant; p = 0.032; Fig. 2). However, despite the appearance of a trend to reduced weevil suppression for either larger L3 weevils or those exposed to wasps for the longer 7-days period, none of the other treatment conditions incurred significantly fewer offspring relative to the control (Table 1; Fig. 2). For instance, despite an apparent 51.2 ± 7.8% reduction in emergence (10.4 ± 1.7 weevils per plant) observed for L3 weevils exposed to wasps for three days relative to the control, this treatment was not significantly different from controls (p = 0. 1135), as were 3- or 7-days exposed L1 weevil treatments averaging even higher offspring weevils per plant (15.2 ± 3.5; p = 0.691; and 12.4 ± 2.1; p = 0.288), respectively (Table 1; Fig. 2).

Fig. 2
figure 2

Mean number of adult pepper weevil, Anthonomus eugenii offspring (± SE) emerged from all infested pepper fruit of an experimental Capsicum annuum cv. Blaze plant in response to exposure of first (white columns) or third (black columns) instar A. eugenii to Jaliscoa hunteri for either three or seven days relative to no parasitoid control (hatched column). Columns with different letters indicate a significant difference in A. eugenii emergence, Tukey’s HSD, α = 0.05

Full size image

However, when weevil instar and wasp exposure period were analyzed as main effects and as a proportion of control untreated plants by analyzing weevil suppression per plant, it was found that weevil instar (p = 0.026), but not exposure period (p = 0.067) significantly affected A. eugenii offspring emergence (Table 1). In this analysis, as with the previous one, only longer 7-days period wasp exposures of the largest L3 weevils resulted in significantly greater weevil emergence suppression relative to 3-days exposures of L1 weevils (p = 0.027), all remaining treatments being not significantly different from one another.

Examining weevil suppression potential from a different vantage point, focusing on fruit size and abscission status as main effects, it was clear that fruit size consistently dictated wasp suppression potential, for either 7-days (p = 0.035) or 3-days (p < 0.001) wasp exposure analyses (Table 1; Fig. 3). Following a 7-days wasp exposure, significantly greater weevil suppression was achieved in small relative to large (p = 0.032), but not medium sized fruit (p = 0.273). Whereas following the shorter 3-days wasp exposure, smaller fruit generated greater weevil suppression levels than either the medium (p = 0.004) or larger (p < 0.001) sized fruit, but the difference between medium and large fruit however was not significant (p = 0.182; Fig. 3).

Fig. 3
figure 3

Mean percentage (± SE) suppression of pepper weevil, Anthonomus eugenii emergence from large (> 1.5 cm), medium (1–1.5 cm), or small (< 1 cm) infested fruit of ornamental Capsicum annuum pepper plants cv. Blaze, following either 3-days (black bars) or 7-days (white bars) exposures to Jaliscoa hunteri wasps. Columns with different letters indicate a significant difference in the mean percent weevil emergence suppression for either 3-days (lower case), or 7-days wasp exposures (upper case), Tukey’s HSD, α = 0.05

Full size image

Whether fruit were attached or separated from the plant following abscission also significantly affected suppression, but only following 7-days wasp exposures (p = 0.025; Table 1). In this longer treatment group, attached fruit rendered on average slightly greater levels of weevil suppression (69.7% ± 0.04) relative to abscised fruit (54.6% ± 0.07).

Throughout the single cultivar trials in this study, a low cumulative total of 48 individual J. hunteri offspring wasps were observed emerging from J. hunteri treated and A. eugenii infested plants. As such, parasitoid emergence was pooled by treatment group and differences in emergence among three fruit sizes, or based on fruit abscission status, parasitoid exposure period and weevil instar were not statistically analyzed due to this low total number of parasitoids. However, it is worth noting that no wasp emerged from fruit in which L1 A. eugenii were exposed to J. hunteri for three days, nor, as expected, from fruit on control plants. For all other treatment groups, wasps emerged out of both abscised fruit and those remaining attached to plants, as well as from fruit of all size classes. An average of 3.0 ± 1.0 wasps emerged per plant when L3 A. eugenii were exposed to J. hunteri for three days compared to only 1.2 ± 0.5 and 1.1 ± 0.0 wasps per plant for both L1 and L3 A. eugenii seven days wasp exposure treatments, respectively.

Discussion

In this study, both ornamental pepper cultivar comparisons as well as subsequent evaluation of several insect and plant host characteristics supported the potential value of J. hunteri in reducing A. eugenii infestations of pepper plants. We showed that releases of this wasp onto A. eugenii infested plants could significantly reduce the number of A. eugenii offspring that emerged, depending on the length of the parasitoid exposure period and the life stage of A. eugenii being attacked. Overall, we saw the lowest offspring weevil emergence when plants infested by L3 instar A. eugenii were exposed to wasps for the longest exposure period tested of seven days. These reductions also corroborate with results of other commercial greenhouse trials we have previously conducted, during which J. hunteri releases reduced A. eugenii offspring emergence and overall pest pressure relative to neighbouring untreated (no J. hunteri) but A. eugenii infested greenhouses (unpublished results).

In this study, three visually and morphologically different ornamental pepper plant cultivars were compared for their effects on A. eugenii emergence and exposure to J. hunteri. Blaze cv. produced red–orange-yellow fruit with a sizeable diameter (max ~ 2 cm at base) relative to buds (~ 2–3 mm), compared to peppers of Medusa cv. which were narrow, slender (~ 1 cm diameter at base) yellow-orange-red fruit that were densely clustered. These two cultivars also had a higher A. eugenii emergence than a third cultivar, Wicked cv., which had a similar shape and size to Blaze cv. but instead produced fruit whose predominant colours, were red and purple when fully ripened.

Given the results of initial cultivar trials, Blaze cv. plants were subsequently used for all further trials. Relative to more prominently cultivated A. eugenii host plants such as the bell pepper fruit, each of these ornamental plant cultivars had many more small fruit per plant, which facilitated examination of how J. hunteri might reduce emergence of A. eugenii. Through our research, we showed that even pepper buds as small as 3 mm in diameter, can support A. eugenii development to adulthood. It was also determined that releases of J. hunteri onto A. eugenii-infested plants could significantly reduce A. eugenii emergence under certain conditions.

Previous work with J. hunteri established that this wasp parasitizes the larvae of A. eugenii (Rodríguez-Leyva et al. 2000), but none has explicitly compared early and late developing A. eugenii larvae for their susceptibility to the parasitoid. We have determined that the parasitoid most successfully suppresses A. eugenii emergence when the developing pest is attacked as a third instar larvae. However further work investigating longer oviposition periods with more mature wasps may also show a reduction for earlier weevil instars.

While emergence of parasitoid offspring from A. eugenii-infested fruit was overall quite low in this study, there was nonetheless a small amount from treatments in which parasitoids were provided either L1 and L3 weevils for seven days, confirming that the wasp can successfully complete its development on larvae at either instar, and by extension likely also on L2 or pupal A. eugenii (Toapanta et al. 2005; Gómez-Domínguez et al. 2012). In contrast, no parasitoid emerged from shorter, 3-days L1 exposures. Given this low rate of parasitoid emergence, the potential for establishment of a persistent parasitoid population following initial wasp release within an A. eugenii-infested greenhouse is unlikely. As such, repeated releases of the parasitoid over time would be required for continued suppression of this pest. Alternatively, future work could explore the use of small-fruit banker plants or substrate systems that represent suitable alternative insect hosts from which this wasp can more successfully attack, develop and emerge, thereby establishing a consistent crop-pest suppressing population (Bartlett 1978).

On the other hand, the low rate of parasitoid emergence documented in this study may also result from the artificially short wasp exposure periods assessed here (3- or 7-days exposures). Since a greenhouse-released parasitoid will undoubtedly be exposed to host weevils at multiple life stages for much longer periods of time, the overall parasitoid impact is likely to be greater than what was observed here.

To date, it is assumed that parasitism represents the principal mechanism through which J. hunteri affects A. eugenii survival. However, host mortality may also result from non-reproductive effects including from wasp host probing or unsuccessful parasitism, through which offspring parasitoids may die prematurely (Abram et al. 2019). Furthermore, weevils can also be killed through host feeding, whereby the adult parasitoid feeds on the haemolymph of host larvae, which is a source of nutrients and protein required for wasp oogenesis (Morales-Ramos et al. 1996; Gonzaga-Segura et al. 2022). This can occur especially in cases where the host is not ideal for parasitism (Kidd and Jervis 1991). While host feeding by J. hunteri on immature A. eugenii has yet to be thoroughly described, observations of this phenomenon in this species have previously been reported (Rodríguez-Leyva et al. 2000; Murillo-Hernández et al. 2019; Gómez-Domínguez et al. 2021). Supporting this are several instances of numerical reductions of A. eugenii emergence previously documented following J. hunteri treatments, which may have resulted from parasitoid host feeding on A. eugenii larvae, and possibly at multiple host developmental stages (Rodríguez-Leyva et al. 2000; Murillo-Hernández et al. 2019; Gómez-Domínguez et al. 2021).

The deep internal localization of A. eugenii larvae within fruit, and the short length of the parasitoid ovipositor are believed to be key factors affecting parasitoid pest suppression ability, especially in large pepper fruit. In this study, Blaze cv. ornamental peppers were characterized as having a diameter < 3 cm and a thin pericarp, thus it is not surprising that greater parasitoid effects were observed among smaller fruit, in addition to those that remained attached on plants where they may be readily accessed by parasitoids from all angles. While A. eugenii eggs are laid peripherally within the pepper pericarp. As development proceeds, A. eugenii larvae hatch and tunnel deeper into fruit where they may become inaccessible to parasitoids (Elmore et al. 1934). The length of the parasitoid ovipositor (~ 1.4–2.3 mm) may thus limit its ability to target older A. eugenii larvae that have tunnelled deeper into larger sized fruit (Gómez-Domínguez et al. 2012). This may explain why the effect of wasps was greatest in smaller fruit, in which A. eugenii larvae would be readily accessible.

Corroborating with this finding is a previous study on field bell pepper fruit, which only found J. hunteri parasitoids emerging from fruit with a diameter < 2.5 cm (Riley et al. 1992). In that study, it was suggested that the thicker pericarp of larger fruit posed an obstacle for parasitoid access to A. eugenii if the ovipositor could not pierce entirely through the fleshy fruit interior (Riley et al. 1992). Despite this limitation, another published work has modeled typical growth rates for greenhouse-grown fruit of several pepper cultivars, and showed that bell peppers have a diameter ≤ 2.5 cm for between 6.7 and 14 days post-flowering, while several other small-fruit cultivars remain ≤ 2.5 cm wide, for ≥ 60 days or for their entire developmental periods (Teixeira et al. 2023). This highlights that weevils within even the largest of pepper fruit might still be successfully parasitized, as long as the host is targeted within this critical phase of fruit development.

In this study, parasitoids released onto infested ornamental pepper plants were initially sourced from a cowpea weevil rearing system, but were removed immediately after their emergence (0–24 h) so as to avoid them developing a preference for this host species. However by doing so, the mated parasitoids studied here were unlikely to be at their peak fecundity as females ranged in age from 3–4 days old at the beginning of exposure periods and between 6–11 days old at the end. Given that the peak fecundity period for J. hunteri has been documented to start at four days post-adult emergence and continues until 29 days when fecundity begins to drop, our study likely considerably underrepresents the total control potential of wasps (Rodríguez-Leyva et al. 2000). This emphasizes the need for future studies to evaluate J. hunteri females over their known peak fecundity period to more fully appreciate their pest population suppression potential.

This study is unique as it is the first documented to use ornamental pepper plants for carefully assessing the interactions between A. eugenii and J. hunteri in such a whole-plant system. However, it is important to note that none of the cultivars we studied currently consist of major commercial edible greenhouse pepper varieties (AAFC 2020). Instead, such ornamental pepper plants are characterized by production of an abundance of buds and small pepper fruit with a thin pericarp, which makes them ideal systems for studying A. eugenii and parasitoid dynamics without the need for maintenance of very large pepper plant cultivars such as bell peppers. They have consequently facilitated our quantitative evaluation of parasitoid effects in this study and may also facilitate future studies on this topic (Patrock and Schuster 1987, 1992).

Taken together, this work supports the potential of J. hunteri as a novel and valuable tool for targeting A. eugenii, an economically important pest in North America. It has provided key information that can be applied in development of best practices for the biological control of this pest. However, future work is needed to establish the ideal frequency and rate of J. hunteri release in commercial settings, and to clarify the role of wasp host feeding on A. eugenii suppression. Collectively, this additional information could greatly improve the sustainable management of A. eugenii, a challenging and invasive pest in North America and beyond.