Materials and Methods

A colony of E. scutellatus was established from parasitized C. ethlius caterpillars collected in residential gardens in southern Florida, where Canna is common as a perennial ornamental plant. Three main sites in the Fort Lauderdale and Pompano Beach area with persistent E. scutellatus wasp populations were identified between May 8 and June 30, 2000.

Elachertus scutellatus was reared in an incubator at 25°C (77°F), the average annual temperature for Fort Lauderdale, Florida as recorded by the National Oceanic and Atmospheric Administration (1998), and high RH under a L:D 14:10 regime.

Host C. ethlius caterpillars were reared in a greenhouse at 22°C, high RH and under long-day conditions. This caterpillar is a near-monophagous leaf-rolling species and was fed ad libitum on leaves of the Canna lily. Caterpillars were exposed to adult wasps early in the third larval stadium (MacDonald & Caveney 2004).

Pre-adult E. scutellatus Rearing

After exposure to wasps, the caterpillars were placed under a dissecting microscope to examine whether eggs had been laid on their integument. Parasitized caterpillars were placed individually in 100-mm diameter Petri dishes together with a damp piece of filter paper and a fresh piece of Canna leaf as food. Fresh leaf material was given daily as long as the hosts continued to feed. As a check for parasitism, apparently non-parasitized caterpillars were placed in 100-mm diameter Petri dishes with a damp filter paper and leaf material. These caterpillars were fed and observed until they completed the molt to the next instar. As molting confirmed the caterpillars were not parasitized, they were discarded.

Each Petri dish was sterilized daily with 70% ethanol, rinsed with distilled water and a new piece of damp filter paper was placed in the bottom. The host leafroll was kept folded to conserve moisture for parasitoid larvae and only opened briefly for observation each day. Upon pupation of the parasitoid, the leafroll was opened and stapled to filter paper to prevent curling of the leaf as it dried, and to allow easy observation. Observations of the immature wasp stages were made each day and the development of the parasitoid on the host was described.

Adult E. scutellatus Rearing

Adult wasp broods were held in Nalgene® transparent styrene-acrylonitrile utility boxes (7 × 6 × 12 cm). The wasps were provided with pure Billybee® clover honey brushed onto small yellow “Post-it”® notes (3.7 × 5 cm) stuck to the inside of the box (Hagley & Barber 1992; Pitcairn & Gutierrez 1992; Zaviezo & Mills 1999). Humidity was kept high in the container by placing moist filter paper in the bottom of the dishes. Moisture was not allowed to collect into droplets on the container walls, as the adults drowned easily within them. The wasps were removed daily with a mouth aspirator and the container was sterilized with 70% ethanol and rinsed with distilled water. The adults were sexed, mortality recorded and the healthy wasps then returned to the container with fresh honey.

One early third instar host caterpillar was provided per one or two adult female wasps. Host caterpillars were removed from the colony and exposed to adult wasps in the breeding population several times a week for an average of 24 h, although oviposition occasionally was completed within a few h.

External Morphology and Development of E. scutellatus

Egg: The egg of E. scutellatus is a large, yolk-rich anhydropic structure with a thick, tough chorion (Fig. 1a). A simple anchor on the egg, possibly an extension of the chorion (Clausen 1940; Kasparyan 1981), penetrates the host cuticle and secures the egg to the host integument. The egg has the shortest developmental duration of the immature stages. Embryogenesis is completed in about two days (2.1 ± 0.2, n = 39).

Larva: E. scutellatus larvae are hymenopteriform, 13-segmented (Gauld & Bolton 1988) and lack any tubercles or spines (Fig. 1b). The number of instars was not determined. Following parasitoid hatching, the host caterpillar ceases to feed but continues to expel fecal pellets. The parasitoid larvae feed externally on the caterpillar for six days (6.0 ± 0.6, n = 13). During this time, the host turns from green to a dull yellow color, presumably due to the absence of gut contents and hemolymph normally visible through its translucent cuticle. Initially, the gut contents of the wasp larvae appear white, but as feeding ensues the gut turns yellow or brown. The host caterpillar dies several days after the parasitoids begin feeding and is often entirely consumed before the wasps pupate. Frequently the hosts died prematurely, usually within two days of wasp feeding, from what appeared to be a pathogenic infection. This caused the death of the associated parasitoid larvae.

Pupa: When feeding is complete, approximately one day prior to pupation, E. scutellatus larvae leave the host carcass but remain within the leafroll (Fig. 1c). They then expel their gut contents and pupate. E. scutellatus does not spin a cocoon for pupation (Fig. 1d) but instead, as seen in most chalcidoid wasps, the pupa is obtect (appendages are held against the body by a secretion produced at the last larval molt and the exposed surfaces of the appendages are heavily cuticularized) (Richards & Davies 1964) and adecticous (has reduced mandibles not used in adult eclosion) (Richards & Davies 1964). The pupae actively pump a clear fluid from their anus (Fig. 2) prior to tanning and hardening of their exterior. When dry, this fluid becomes fibrous and attaches the pupae to the inner surface of their host’s leafroll.

One to two days prior to eclosion in the female wasp, a dark ovipositor becomes visible through the pupal casing as it stands out against her yellow abdomen. Adult wasps emerge within one week of pupal ecdysis (6.5 ± 0.5, n = 10).

Adult: The adult wasp is black, approximately 1.5 mm long, with reddish eyes and ocelli (Howard 1897). The anterior portion of the ventral side of the adult male abdomen is translucent (Fig. 1e). The anterior portion of the ventral side of the female abdomen is honey-yellow (Howard 1897). The clearly visible ovipositor is positioned on the ventral midline of the female abdomen, but does not extend past its tip (Fig. 1f). Females were typically larger than males within a brood, although between broods the size range of the sexes overlaps considerably.

Discussion

As seen in many ectoparasitoids, E. scutellatus wasps lay anhydropic eggs that have a large enough quantity of yolk that the embryo can complete its development to the larval stage without the need to sequester protein and nutrients from the host (Quicke 1997). The yolk of anhydropic eggs is rich in protein, of which vitellin is the major component, and has numerous lipid droplets (Quicke 1997). A large amount of egg protein is associated with the consumption of host fluids by female wasps (Flanders 1950; Le Ralec 1995). The eggs of most parasitoids are roughly oval and many ectoparasitoid eggs have stalks or pedicels to anchor the egg to the host (Quicke 1997). The morphology of E. scutellatus eggs is much like the eggs of the ichneumonid subfamily Tryphoninae, another group of ectoparasitic koinobiont wasps (Mason 1967; Kasparyan 1981).

The egg of E. scutellatus is unusual in that it is very large with a tough chorion. This presumably protects against both desiccation (Mason 1967) and physical damage that may occur when the host is allowed to remain active and mobile after attack by a parasitoid wasp (Mason 1967). To protect against the dehydration of eggs, female wasps often will attack only hosts living in humid or partially concealed situations (Mason 1967). To protect the eggs from the active host, female E. scutellatus wasps most often place the eggs between the host prolegs where they are difficult for the host to detect and destroy. These constraints are claimed to account for the rarity of ectoparasitic koinobionts (Gauld 1988; Godfray 1994; Mayhew & Blackburn 1999).

The larval morphology of E. scutellatus resembles that of other eulophids (Gauld & Bolton 1988). The white appearance of first instars of the wasp probably is due to the prenatal yolk resorbed during embryogenesis (Anderson 1973) and as feeding ensues, the larvae take on the color of the consumed host fluids, which can be seen through the translucent cuticle. Although the number of instars for E. scutellatus was not determined, Hanson & Gauld (1995) determined that in other eulophids there are three to five instars.

Parasitoid larvae are capable of altering the physiology of their active hosts, increasing or arresting developmental rates (Harvey et al. 1998; Doury et al. 1995; Hemerik & Harvey 1999) and changing host feeding habits (Shaw 1981; Powell 1989). Oral secretions from larval E. scutellatus may be the cause of the observed cessation in caterpillar feeding (Godfray 1994; Morales-Ramos et al. 1995; Richards & Edwards 1999) as well as the lack of a host phenoloxidase reaction at larval feeding sites (Richards & Edwards 1999). Interestingly, the larvae of pimpline ichneumonids that parasitize lepidopteran pupae are capable of producing toxic anal secretions from modified Malpighian tubules that inhibit host phenoloxidase activity (Godfray 1994). Although the presence and targets of larval secretions have not been closely studied, it seems parasitoid larvae possess specific adaptations that may alter host physiology and result in an increase in their own chance of survival.

The Eulophidae are exceptional within the Chalcidoidea in that silk cocoon formation is rare (Quicke 1997). Elachertus scutellatus do not produce a cocoon but instead harden and tan their cuticle (Richards & Davies 1964) and fasten the pupa to the substrate with a clear fluid that becomes fibrous when dry. Fidgen and Eveleigh (1998) state that Elachertus cacoeciae pupae attach themselves by their meconium [“waste material excreted by an insect upon eclosion as an adult (a few parasitic wasps) or, in the case of the majority of parasitic wasps, just prior to pupation” (Quicke 1997)]. It seems unlikely that the fluid is simply a waste product in E. scutellatus, as the prepupae expel their gut contents far in advance of the active production of the clear fluid they use for attachment. It has been reported that many chalcidoids produce a cocoon-like structure from a liquid secreted by the anal end (and occasionally also the oral end) of the final instar (Flanders 1938; Quicke 1997). In several endoparasitoids, the fluid flows over the larvae under the pressure of the host hemolymph, then dries and hardens into an amorphous sheath often associated with host tracheae (Flanders 1938; Colazza & Bin 1992; Ceresa-Gastaldo & Chiappini 1994). In the Chalcididae, the origin of the liquid is the iliac glands, structures homologous to the specially modified Malpighian tubules found in many ichneumonoid larvae (Quicke 1997). Freshly-ecdysed E. scutellatus pupae may not need to produce a protective cocoon with this liquid (as they are ectoparasitic) and instead make use of the host shelter. In E. scutellatus the excreted fluid is restricted to the anal area and is used to fasten the pupa to its substrate. This behavior seems to be intermediate between the formation of endoparasitic or silk cocoon produced anally and a lack of anal silk excretion in most parasitoids. The hindgut of chalcid larvae is very muscular and adapted to active expulsion of cocoon-forming secretions (Flanders 1938).

Generally, adult male wasps emerge before females (Harvey et al. 1998; Hanson & Gauld 1995; Jyothi et al. 1999; Zaviezo & Mills 1999). This was also seen in E. scutellatus, and is likely an advantage to the male in securing a female for mating immediately upon her emergence. The observed size differences between broods is likely due to differences in available resources and competition for food as the size of parasitoids is often related positively to host size (Wen et al. 1995; Fidgen et al. 2000;) and negatively to brood size (Hardy et al. 1992; Harvey et al. 1998).

Population Mortality

The rearing of E. scutellatus in the laboratory ended eight months after the colony was first established. Despite close care, the size of the population declined slowly without an obvious cause until the entire population died out. The highest mortality of E. scutellatus occurred in the larval stage. In a closely related species, Euplectrus plathypenae (Eulophidae), a similar pattern of mortality was observed (Parkman et al. 1981).

The demise of the colony may have been a result of adult female behavior or physiology. According to Schneider (1941), ovarial regression (ovarialkrise) “is a tendency toward ecological castration and may so decrease the readiness to oviposit that under laboratory conditions egg deposition does not continue” (in Flanders 1950). Ovarialkrise is considered characteristic of tropical wasps (in Flanders 1950). Similarly, phasic castration (atrophy of ovaries resulting in sterility) was frequently observed in laboratory mating studies of Chrysocharis laricinellae (Eulophidae) (Quednau 1967).

Most laboratory-reared wasp populations are sustainable when the females are provided with a source of honey and hosts for host-feeding, and the males are provided with honey water (Hagley & Barber 1992; Pitcairn & Gutierrez 1992; Zaviezo & Mills 1999). The brand of honey routinely used to rear E. scutellatus was replaced for a short duration of time with another brand, whereupon oviposition decreased. When the original brand of honey was restored, oviposition increased but failed to recover fully. It is possible the colony did not recuperate from the decline in female egg laying, but it seems unlikely that this was the cause of the death of the population. Synovigenic wasps such as E. scutellatus are capable of resorbing mature eggs (oösorption) should a food source become limited (Jervis & Kidd 1986).

Prior to the complete extirpation of the wasp colony, host caterpillars often turned black or bright red with yellow patches and sometimes developed a sweet odor, like that of citronella. Death of the host consequently resulted in the early death of the parasitoids. Several studies have concluded that the act of parasitoid oviposition can increase the host’s susceptibility to viral infections resulting in the death of the host and associated parasitoids (Irabagon & Brooks 1974; Levin et al. 1981; Eller et al. 1988). Ectoparasitic larvae also have been shown to compromise host immune systems (Richards & Edwards 1999). The larvae of an ectoparasitic koinobiont eulophid wasp, Eulophus pennicornis, are known to produce factors that decrease the number of host hemocytes in order to reduce the host immune response (Richards & Edwards 1999). A reduction in host defences could result in pathogenic infection and death of both the host and the larval parasitoids. Although no direct evidence of infection was obtained, this is a possible explanation for the death of the parasitoid colony.