The invasive cactus moth, Cactoblastis cactorum (Berg) (Lepidoptera: Pyralidae) is native to Argentina, Uruguay, Paraguay, and southern Brazil. Its larvae are oligophagous on most species of Opuntia Mill. (Caryophyllales: Cactaceae) (Mann 1969). Although originally introduced to the Caribbean as a biological control agent for indigenous Opuntia spp., C. cactorum became an invasive species in the region (Zimmermann et al. 2001). An established population was found in the Florida Keys in 1989, and by 2009 it had spread as far north as North Carolina and as far west as Louisiana (Dickel 1991; Hight & Carpenter 2009). Cactoblastis cactorum poses a threat to ecologically important and endangered cactus species in Florida, and has the potential to become a serious economic pest in Texas and Mexico, where Opuntia species are used for forage, food, fuel, and dye production (Zimmermann et al. 2004).

Chemical control of C. cactorum is not recommended due to its presence in ecologically sensitive natural areas (Pemberton & Cordo 2001). Eradication via sanitation, pheromone trapping, and sterile insect technique is effective over small areas, but the current range of C. cactorum is too large for this to be economically feasible (Varone et al. 2015). Classical biological control can provide a sustainable method of suppressing widespread populations of invasive species. However, imported biological control agents must be assessed for host specificity to prevent undesirable non-target effects (Strand & Obrycki 1996). Apanteles opuntiarum Martinez & Berta (Hymenoptera: Braconidae) is a gregarious larval endoparasitoid of C. cactorum and Cactoblastis doddi Heinrich (Lepidoptera: Pyralidae) in Argentina (Martinez et al. 2012; Varone et al. 2015). This narrow host range makes it a promising candidate for biological control of C. cactorum in North America.

A successful parasitoid rearing program serves to ensure the identity of the beneficial organisms; to determine their reproductive ability in the laboratory; and to allow the conduct of research on their biology, including host specificity, to determine suitability for release (Ertle 1993). The development of a laboratory rearing protocol for A. opuntiarum will allow future workers to obtain consistent reproduction, optimal parasitism rate, and sufficient numbers to continue with the biological control evaluation process. Information related to optimal sex ratio is important for the culture of successive generations in the laboratory. Hymenoptera have haplodiploid sex determination, meaning that unfertilized eggs become males (haploid) and fertilized eggs become females (diploid). Female parasitoids may adjust offspring sex ratio in response to environmental conditions, host quality, and inbreeding status (Godfray 1994).

The effects of inbreeding are particularly important to quarantine colonies, as obtaining new genetic material from outside the country may be expensive or otherwise difficult. The natural mating structure of a species, especially the frequency of local mating, can influence its response to inbreeding in the laboratory (Hardy 1994). The theory of local mate competition predicts a positive correlation between inbreeding and proportion of female offspring (Hamilton 1967). Local mate competition tends to occur in gregarious species whose hosts are patchy in distribution. In these species, male siblings often compete to mate with their sisters immediately after emerging. Therefore, production of male offspring should be lower under high-inbreeding conditions (Werren 1987; Hardy 1994). Little is known of the natural mating structure of A. opuntiarum, but it is a gregarious parasitoid, males emerge before females, and hosts may be patchily distributed (Varone et al. 2015). If local mate competition applies to A. opuntiarum, inbreeding in captivity theoretically should promote a high proportion of females in the colony.

The primary goal of this study was to describe a laboratory rearing method for A. opuntiarum, with special attention to sex ratio and the effects of inbreeding. We hypothesized that sibling mating of A. opuntiarum produces higher proportions of female offspring than nonsibling, outcrossed mating.

Materials and Methods

APANTELES OPUNTIARUM REARING

The A. opuntiarum rearing method described below was developed in 2015 as a component of this project. Host-parasitoid ratios and numbers follow Mengoni Gonalons et al. (2014). Parasitized C. cactorum larvae were collected in Dec 2015 and Mar 2016 from field sites in Cordoba, Catamarca, and Santiago del Estero, Argentina. Laboratory populations of A. opuntiarum originated from this material.

Although C. cactorum can be reared efficiently on artificial diet, hosts fed with Opuntia spp. cladodes were significantly more attractive to female parasitoids (Varone et al. 2016,2017). To serve as optimal hosts for A. opuntiarum, 35 C. cactorum larvae (second to third instar) were transferred from artificial diet blocks to a slice of Opuntia cladode (20 × 7 cm) and allowed to feed for 24 h before exposure to the parasitoid.

Six adult A. opuntiarum (2 females and 4 males) were placed in a 1.9 L glass canning jar with a modified 80-mesh lid (Fig. 1). A small slice of Opuntia (2 × 2 cm) and a damp strip of paper towel (1 × 3 cm, Kimberly-Clark Wypall L30, Kimberly-Clark Corp., Roswell, Georgia, USA) were added to provide moisture. Honey was applied to the mesh lid with a paintbrush to provide an energy source. A dollop of frass from C. cactorum was applied to the bottom of the jar. To avoid oviposition before mating, A. opuntiarum males and females remained in the jar without C. cactorum for 24 h. After this 24 h mating period, the slice of Opuntia and 35 C. cactorum larvae were added to the jar (Fig. 2). Apanteles opuntiarum and C. cactorum remained together in the jar for 6 d.

Fig. 1.

Initial mating jar arrangement for Apanteles opuntiarum. A 1.9 L glass canning jar contained 2 adult female and 4 adult male A. opuntiarum. The wasps were left to mate for 24 h before the introduction of host material. (a) Mesh lid with honey drop. (b) Paper towel strip with water. (c) Frass of cactus-fed Cactoblastis cactorum. (d) Small slice of Opuntia tissue.

After the 6 d exposure period, C. cactorum larvae and Opuntia were transferred to a plastic tub (19 × 29 × 9 cm) with a ventilated lid (2 circular holes, covered with 80 mesh, 6.5 cm diam). Each tub contained a wire grid (10 × 27 cm), where the long ends were bent at right angles to form a support structure for the Opuntia cladodes (Fig. 3). Approximately 150 mL of unscented cat litter (Special Kitty Unscented Natural Clay Cat Litter, WalMart Inc., Bentonville, Arkansas, USA) was added to each container to absorb frass and cactus fluids.

Fig. 2.

Mating jar arrangement, 24 h after initial set-up. Host material was introduced in the form of an Opuntia cladode containing 35 Cactoblastis cactorum larvae (instars 2 and 3). The position and size of cladode are shown relative to the 1.9 L glass canning jar.

Fig. 3.

Rearing container for Cactoblastis cactorum that were exposed to Apanteles opuntiarum. After the 6 d exposure period, Opuntia cladode and C. cactorum larvae were moved to this plastic container. (a) Plastic lid with mesh cut-outs. (b) Opuntia cladode containing 35 C. cactorum larvae. (c) Wire rack to elevate host material above substrate. (d) Plastic tub containing unscented cat litter.

Every 2 to 3 d, C. cactorum frass was removed, Opuntia tissue was replaced, clean litter was added, and pupae of C. cactorum and A. opuntiarum were counted and removed. Pupae of C. cactorum were discarded. Clutches of A. opuntiarum pupae that had erupted from their C. cactorum host larva were removed intact, and placed into plastic Petri dishes. Each Petri dish contained a drop of honey on the lid and a strip of damp paper towel. A small slice of Opuntia (2 × 2 cm) was added to each Petri dish to provide moisture over weekends and holidays.

Upon emergence, adult A. opuntiarum were counted, sexed, and individually separated into 15 mL plastic test tubes. A minute drop of honey was placed on each lid. A strip of damp paper towel was placed over the edge of each tube. When the paper became dry, water was added to the external edge of the paper, allowing the provision of moisture without opening the tube.

Results

SIBLING MATING TESTS

Kruskal-Wallis analysis of clutches by parental relatedness (Table 2) demonstrates a significant relationship between the proportion of female offspring and inbreeding status (H = 13.925; df = 2; P < 0.001). The mean female offspring per jar did not vary between treatments (P > 0.4), although the crossbreeding treatments produced noticeably higher (but statistically insignificant) numbers of male offspring (P > 0.2). Clutch size and clutch number did not differ significantly between treatments (P > 0.05).

Table 1.

Production of Apanteles opuntiarum offspring from rearing jars treated with 3 different parental mating scenarios.

Discussion

Our experiments demonstrated a significant relationship between sibling mating and offspring sex allocation in A. opuntiarum. Sibling-mated A. opuntiarum produced a higher proportion of female offspring per clutch (n = 602) than did those mated with non-siblings. When using the jar as the statistical unit, the results were not significant, possibly due to the low sample number (n = 44). However, the trend was consistent with the hypothesis of local mate competition (Hamilton 1967). We conclude that some degree of local mate competition occurs in natural populations of A. opuntiarum.

Different responses to sibling mating may reflect different natural mating structures, or may occur due to experimental conditions. Experimental inbreeding of other haplodiploid species shows a variety of effects (Table 3). In most of these experiments, male-biased sex ratios resulted from inbreeding treatments, or no effect was observed. Female offspring and inbreeding were correlated only in the eusocial ant Cardiocondyla obscurior Wheeler (Hymenoptera: Formicidae) (Schrempf et al. 2006). It is unclear why C. obscurior should be subject to local mate competition, but it may be linked to the capacity to adopt unrelated queens, or to male dispersal dimorphism (de Menten et al. 2005).

Among parasitoid species, experimental comparisons of inbred and outbred offspring sex ratios are not the only approach to evaluating the influence of local mate competition on sex allocation. Many studies have used a foundress number as a proxy for inbreeding status (Herre 1987; King 1987; Somjee et al. 2011). In the case of Nasonia vitripennis (Walker) (Hymenoptera: Pteromalidae), laboratory inbreeding does not affect offspring sex ratio, but the manipulation of foundress numbers supports the theory of local mate competition (Werren 1983). Therefore, lack of response to laboratory inbreeding does not necessarily mean that local mate competition has no impact. In the A. opuntiarum experiments, foundress numbers and relatedness were controlled (2 female siblings in each mating jar).

There are 2 possible inferences about the natural mating structure of A. opuntiarum based on laboratory sex ratios. First, sibling mating predominates in the field, and the high proportion of male offspring from crossbred pairs is a symptom of outbreeding depression. Second, natural populations are subject to a mixture of inbreeding and outbreeding. In this scenario, an adjustable sex ratio allows female wasps to maximize offspring relatedness according to the availability of sibling vs. non-sibling mates. In the second case, but not in the first, long-term inbreeding may exhibit deleterious effects on laboratory populations. Further observation of the captive colony at the Division of Plant Industry may shed light on this distinction.

Table 2.

Mean (± SD) clutch size (number of parasitoids per clutch), number of clutches per jar, female sex ratio, and number of female or male Apanteles opuntiarum offspring produced from mating jars treated with 3 different parental mating scenarios.

The effects of Wolbachia on sex ratios of A. opuntiarum remain unclear. Mengoni Gonalons et al. (2014) diagnosed Wolbachia infection in both captive and wild individuals. The strain belongs to supergroup B, and has a partial coincidence with a strain known from parthenogenetic South American weevils (Coleoptera: Curculionidae). The reproductive effects of Wolbachia include cytoplasmic incompatibility, which could inhibit fertilization of females by unrelated males (Werren et al. 2008). Experimental manipulation of Wolbachia infection via antibiotics was beyond the scope of this study, but could provide an interesting avenue for future research.

The rearing protocol was successful in producing multiple generations of parasitoids, with sufficient numbers of female offspring to maintain the colony and conduct host range testing. However, successful parasitism occurred in fewer than 62% of jars. Within these, the average parasitism rate was less than 11%. Overall, parasitism should be improved to increase rearing efficiency and reduce costs. Future work may concentrate upon manipulation of environmental variables to increase parasitism rate in captivity.

The A. opuntiarum rearing program represents the first successful application of the local mate competition theory to increase the proportion of female offspring via inbreeding of an ichneumonoid biological control agent. There remains much opportunity for the development and improvement of similar rearing protocols for other parasitoid species of economic importance. Future attempts to manipulate parasitoid sex ratio via sibling mating should focus upon gregarious hymenopteran species with early male emergence.

Table 3.

Review of various studies measuring the effect of experimental inbreeding on sex ratio in haplodiploid Hymenoptera.