Checklist of UK Recorded Encyrtidae (Walker 1846)

Description & Statistics

Encyrtidae is one of the largest of, and structurally most diverse of Chalcidoidea families, especially in structure of the head and antennae. The family currently includes 460 genera and 3735 species placed in 2 subfamilies as follows: Encyrtinae (353 genera and 2920 species), Tetracneminae (107 genera 815 species), and are a cosmopolitan family, next only to Pteromalidae in the number of recently accepted valid genera, of which 153 genera with more than 500 species have been discovered in the Soviet Union. It is also one of the best known chalcidoid families in Africa and with Aphelinidae, it is one of the few families that is in an active stage of being studied. They also share a number of important characters with the aphelinids, and the two groups have been placed in a single family, the Encyrtidae. Encyrtids are more common and varied in dry habitats.

It contains relatively small forms, usually 1-2 mm long, though they may range up to 4 mm. Some of the genera and species resemble the Pteromalidae, Eupelmidae, Tanaostigmatinae, and Aphelinidae.

The majority of encyrtids are primary endoparasitoids of insects, and also Arachnida, with a preference for the larval stages of their hosts, although the eggs, and very rarely the adults, are also attacked. A number of genera are hyperparasitoids, whereas a few are polyembryonic, in which case the hosts are usually Lepidoptera, but occasionally also aculeate Hymenoptera.

Otherwise, encyrtids are so remarkable for the enormous number of structural modifications exhibited in their morphological characters that they are taxonomically a difficult group to study. These morphological variations are indeed a major source of trouble in analyzing the relationships of the genera with certainty.

Classification: The family Encyrtidae, as delimited here, excludes Aphelinidae, Eupelmidae and Signiphoridae. Some authors still include aphelinids and signiphorids as subfamilies of Encyrtidae. The classification of Encyrtidae is still a more or less unsettled problem. In the present review a summary of the existing systems of classification is presented and the classification of the genera adapted in the recent review by Noyes & Hayat (1984) is given.

Identification

Extremely diverse in form, they are easily recognized by the following characters: Body generally compact, though somewhat flattened to strongly flattened forms and antlike forms occur.
Both sexes with mesopleuron integument often without a metallic lustre, but some species also highly refringent, with fine sculpture, very enlarged and inflated, broadly convex, often occupying more than half the thorax in side view, without impressed lines or grooves.
Mesepisternum and mesepimeron not differentiated
Mid coxae usually widely separated from hind pair and inserted level with middle or slightly in front of midline of mesopleuron in side view.
Middle legs usually with greatly enlarged and thickened mid tibial spur (saltatorial), which helps distinguish at least the females of Encyrtidae from those of other families of Chalcidoidea (other than Eupelmidae), and with enlarged basitarsus as in eupelmids, tarsi five-segmented with the exception of two genera in which they are four-segmented.
The protibia is without dorsoapical spicules, with relatively long, curved tibial spur. The mesotibia has a row of pegs along the anteroapical margin, with robust, usually elongated spurs. The tarsi are usually with 5, less often 4, tarsomeres. The mesotarsus is usually with pegs in various patterns along the ventral or anteroventral surfaces of tarsomeres.
The cercal plates are advanced, rather than at the apex of the gaster and frequently in anterior two-thirds
The linea calva present and distinct
The mesoscutum with parapsidal sulci seldom developed, it is transverse and without notauli, or if notauli are present then they are very shallow and curved, never deep and straight
Axillae large, their inner angles meeting mesally; sometimes very short or reduced and are almost always transverse-triangular, usually with contiguous anteromedial angles (sometimes appearing separate because of overhanging posteromedial curvature of the mesoscutum).

Sexual dimorphism is common, with the female antennae usually 11-segmented, the male antennae usually 9-segmented, occasionally branched. Most species are solitary, although gregarious species are known.

The mesothorax in ventral view is with or without a membranous area anterior to each mesocoxa, but the mesocoxa is unable to rotate completely out of the fossa. The mesocoxa is inserted at or anterior to the midline of the acropleuron.
The acropleuron is convexly enlarged to the metapleuron. It is completely enlarged along with the advanced position of the mesocoxae.
The mesosternum is transverse.
Mesonotum almost always uniformly convex, with notauli either shallow or entirely absent.
Prepectus divided into 2 triangular sclerites. The prepectus is flat posterior to the pronotum, but internally there is a prepectal strut between the ventral angle of the prepectus and the mesoscutum.
The pronotum is usually visible in Encyrtidae and usually transverse in dorsal view.
The mesoscutum is usually without notauli but if present then they are linear and somewhat sinuately V-shaped. The mesoscutum is articulated to the scutellar-axillar complex only laterally, with a very slender membranous area and/or depressed anterior margin of the scutellar-axillar complex visible between the sclerites if the mesonotum is arched.
The gaster is sessile, broadly joined to propodeum (very rarely petiolate in non-African species);
The metasoma has a cercus advanced anteriorly, usually distinctly so, and then the apical tergum is large, triangular or U-shaped, and/or with one or more terga M-like between and around the cerci
The ovipositor rarely protruding strongly caudally. Outer plates of ovipositor not demarcated from dorsal part of 9th syntergite.
The wing venation and position of the mid coxae are useful in separating the Encyrtidae from the Eupelmidae. Most species are also characterized by their very short marginal vein and obviously advanced cerci. Members of this family may sometimes be apterous or brachypterous, but in fully winged forms the marginal, stigmal and postmarginal veins relatively short, the marginal vein usually very short to punctiform. The marginal vein is much shorter than the submarginal. Tendency toward reduction in wings evident, associated with the development of jumping ability.

The two recognized subfamilies, Encyrtinae and Tetracneminae, are differentiated by characters that are usually difficult to observe and that are not possessed by all members of each subfamily. These include dentition of mandibles, presence or absence of metasomal paratergites, and the presence or absence of differentiated setae along the linea calva of the forewing (Gibson 1993).

Yoshimoto (1984) provided a key to the subfamilies for most Canadian Encyrtidae: Tetracneminae and Encyrtinae.

Tetracneminae.

Trjapitcyn (1971, 1973a, 1973b) classified Encyrtidae, dividing the family into two subfamilies. The Tetracneminae are recognized by the characters given in the key by Yoshimoto (1984), which was divided into 12 tribes and 11 subtribes (Trjapitcyn & Gordh 1978a, 1978b). Gordh (1979) included 6 tribes and 25 genera under Tetracneminae. Tachikawa (1981) listed the hosts of encyrtid genera of the world. This subfamily is divided into three tribes. In the first, Anagyrini, the body is much flattened, the mouth parts are directed either downward or forward (hypognathous or prognathous), the antennae and legs are elongated, and the basal segment of the female antenna is almost cylindrical. This tribe comprises genera Anagyrus Howard, Anathrix Burks, Leptomastix Förster, and Leptomastidea Mercet. They are principally parasitoids of Pseudococcidae.

Chrysoplatycerini

The second tribe has a compact body which is not flattened and the antenna is almost always flattened. Eg. Chrysoplatycerus splendens (Howard), a parasitoid of pseudococcoids.

Ericydnini

has a somewhat elongated body, and the expansion of the parastigma is not triangular. Eg. Clausenia purpurea Ishii, parasitoid of Pseudococcus comstocki (Kuwana).

Tetracnemini

has a body that is not very compact, and the male antennae have long funicle branches. Eg. Paraleurocerus bicoloripes Girault, parasitoid of Cameraria caryaefoliella (Clemson) and Lithocolletis sp..

Encyrtinae.

This subfamily includes a diversified group which is divided into 36 world tribes and 30 subtribes (Trjapitcyn & Gordh 1978a).

Subfamily Encyrtinae Walker is separated into 3 generalized groups based on whether the body is flattened, compact or elongated. Flat-bodied forms with the mouth parts turned downward and backward (opisthognathous) are represented by Ixodiphagus Howard and Hunterellus Howard (Ixodiphagini), which are parasitoids of hardbacked ticks (Ixodoidea), and Habrolepis Förster, Anabrolepis Timberlake, and Adelencyrtus Ashmead (Habrolepidini), which are principally parasitoids of diaspine scales (Yoshimoto 1984).

Yoshimoto (1984) stated that "The elongate body forms are confined to two tribes. The first tribe (Cheiloneurini) is distinguished by the stigmal vein of the fore wing, which is short and straight, by the submarginal vein, which in most instances has a triangular dilation or bend in the apical third, and by the apex of the scutellum, which often has a cluster of hairs. Eg. Cheiloneurus Westwood, a hyperparasite on Coccoidea (Homoptera) and also Chrysidoidea, on Chalcidoidea, and Apoidea (Hymenoptera)."

"In the second tribe (Encyrtini), the stigmal vein of the fore wing is curved, the apex of the submarginal vein is not modified, and the apex of the scutellum has a cluster of hairs. The genus Encyrtus Latreille is the only representative and is a parasite of coccid scales (coccidae, Homoptera)."

"Most encyrtids belong to the genera with compact, or more or less compact, body forms and these are separated into eight groups. The first group (tribe Trechnitini) is distinguished by the strong metallic lustre and hyaline wings. Eg. Prionomitus Mayr, Trechnites Thomson, and Psyllaephagus Ashmead; all are parasites of psyllides (Psyllidae, Homoptera)."

"The second group (tribe Bothriothoracini) is distinguished by the fifth sternite reaching the apex of the gaster, and contains two subgroups. The group is distinguished by the body frequently being coarsely sculptured and sometimes more or less flattened. Eg. Bothriothorax Ratzeburg, a parasite of syrphid flies (Syrphidae, Diptera)."

"The third group (tribe Homolotylini) is distinguished by the obliquely truncate female antennal club and the mesoscutum with notauli. Eg. Homalotylus Mayr, parasite of coccinellid larvae (coccinellidae, Coleoptera), and Isodromus Howard, parasite of hemerobiid larvae (Hemerobiidae, Neuroptera)."

"The fourth group (tribe Copidosomatini) is distinguished by the occipital margin being sharp and by the antenna being usually inserted at the mouth margin. Eg. Copidosoma Ratzeburg, Coelopencytus Timberlake, Paralitomastix Mercet, Ageniaspis Dahlbom, and Pentacnemus Howard. They are polyembryonic parasites of various families of Lepidoptera larvae and some bee and wasp larvae (Hymenoptera)."

"The fifth group (tribe Pseudorphopini) is distinguished by the submarginal vein of the fore wing being short, the postmarginal and stigmal veins rudimentary, the first segment of the mid tarsi short, and the body lacking metallic luster. Eg. Pseudorphopus Timberlake is a parasite of coccid scales (Coccidae, Homoptera)."

"The sixth group (tribe Aphycini) is distinguished by the body lacking metallic luster (except in Blastothrix Mayer), the antenna short, and the mesoscutum sometimes having notauli. Eg. Blastothrix Mayer, Aphycus Mayr, Pseudophycus Clausen, Stemmatosteres Timberlake, Tetracyclos Kryger, and Metaphycus Mercet. The members of this group have been relatively well studied because they are useful parasites in suppressing harmful scale insects (Compere & Annecke 1969). Gibson & Yoshimoto (1981) redescribed Tetracyclos boreios Kryger and discussed its anatomy, its placement, and its association with Pseudococcidae (Homoptera)."

"The seventh group (tribe Microteryini) is distinguished by the antenna not being broadened, the flagellum uniformly segmented, the periphery of the antennal scrobe rounded, and the mesoscutum without notauli. Eg. Tachinaephagus Ashmead, parasite of Stomoxyidae and Muscidae pupae (Diptera), Microterys Thomson, parasite of Coccidae, Dactylopiidae and Ortheziidae (Homoptera); Aphidencyrtus Ashmead, parasite of Aphididae (Homoptera), and hyperparasite on Braconidae and Aphelinidae (Hymenoptera); Ooencyrtus Ashmead, egg parasite of Lepidoptera, Neuroptera, Orthoptera, Coleoptera and Hemiptera [also found on Diptera: Chloropidae; Pseudencyrtus Ashmead, parasite of Cecidomyiidae (Diptera), and Cerchysius Westwood, parasite of Coleoptera and Diptera."

"The eighth group (tribe Arrhenophagini) is distinguished from all other groups by the four-segmented tarsi except in Tetracylos boreios Kryger, which almost certainly is related to Stemmatosteres Timberlake (Asphycini). Eg. Arrhenophagus Aurivillius, parasite on Diaspididae (Homoptera)."

Prinsloo & Annecke (1979) provide a key to 106 African genera, but it is estimated that the actual number of genera occurring in the Ethiopian region may perhaps be twice as high. There is little agreement on the subfamily classification of the encyrtids, but it is probably justified to divide the known African fauna into three subfamilies. The Arrhenophaginae include minute species with reduced number of antennal segments and reduced wing venation. Of the three genera in this subfamily, Arrhenophagus is one of two encyrtid genera in which the tarsi are four-segmented. The Antheminae includes only Anthemus, the only other genus with four-segmented tarsi; it is separated from Arrhenophagus, and all other encyrtid genera, by the rudimentary axillae. The remainder of the encyrtids, which form the large majority of the fauna, may be placed in the subfamily Encyrtinae, which again may be divided into distinct groups or tribes. The tribe Tetracnemini forms a large group of genera often referred to as the "mealybug genera," as the members of this group are exclusively parasitic in Pseudococcidae. Species of this tribe are characterized by the mandibles which have two teeth and by the cercal plates on the abdomen which are usually placed near the base of the gaster. Anagyrus is cosmopolitan and includes many species important to biological control; it is perhaps the best known genus in this group, and its species are generally characterized in the female by a greatly expanded antennal scape which is marked with black and white. Another common genus of this group is Leptomastix, the species of which are usually yellowish with long slender antennae, and which are often collected when sweeping grass, where they parasitize grass-inhabiting mealybugs. The largest tribe of the Encyrtinae is the Bothriothoracini. It includes the bulk of the African fauna of this family, more than 70 African genera displaying an extremely wide range of forms, differing greatly in structure... A few of the more common genera are: Habrolepis, Metaphycus, Microterys, Anicetus, Cheiloneurus, Comperiella, Psyllaephagus, Aphidencyrtus and Homalotylus. The third tribe of the Encyrtinae is the Encyrtini which includes three genera, in all of which the ovipositor sheaths (gonostyli) are absent. The best known genus in this group is the cosmopolitan Encyrtus, genus of large encyrtids characterized by a semi-erect tuft of bristles on the scutellum."

top of page

Biology & Behavior

Encyrtidae usually develop as internal parasitoids, the only exceptions being the occasional species of Microterys which are predaceous on the eggs of Coccidae. This habit was observed by Silvestri (1919b) in the case of M. sylvius Dalm. in relation to Eulecanium coryli L. in Italy. DeBach (1939) found M. titiani Gir. to have the same habit.

The hyperparasitic species are usually direct in their relationship, though Syrphophagus aphidivorus Mayr is an indirect secondary parasitoid of Macrosiphum cornelli Patch through Aphelinus juncundus Gahan (Griswold 1929).

Species attacking Homoptera obtain their food mostly from the honeydew secretions of their hosts, though quite a few feed directly on host body fluids. Several species of Ooencyrtus feed at punctures made in the host egg in which their progeny will develop.

In Tachinaephagus zealandicus Johnston & Tiegs (1921) suggested that mating occurred both inside and outside the host puparium. They isolated emerging females with hosts, and the resulting progeny contained females, thus demonstrating pre-emergence mating. However, when Olton & Legner (1974) selected females at emergence from Musca domestica puparia, only male progeny were produced indicating that they were unmated. However, mating was repeatedly observed immediately after emergence from the puparium. Males and females usually issued through the same aperture cut in the host puparium and tended to remain close to the host for several seconds while grooming. Once the females moved away from the host, males pursued rapidly from the posterior and moved sideways with the female. Females were quite active and males were often buffed in mounting. Mating proceeded rapidly, lasting only circa 15-20 sec. after mounting. Mating was observed in light and darkness; however, activity of both sexes was reduced in darkness.

Adult females generally contain fully developed eggs at the time of emergence, and they are consequently able to oviposit immediately. Among certain of the multibrooded species parasitic in single brooded Coccidae, there is a period of diapause at the beginning of adult life of the females of the summer brood that may be of advantage to the parasitoid during the period when the host is not in a suitable stage of development for parasitization. Silvestri (1919a) found that the second brood of adults of Blastothrix sericea refused to oviposit in young host scales, and an examination of their ovaries showed them to be much reduced during the summer and early autumn. Reproductive activity is thought to revive during the autumn, for 1st instar larvae were found in 2nd instar hosts in November.

Female parasitoid host examination, feeding and oviposition are well exemplified in Microterys clauseni Comp. parasitic in Ceroplastes floridensis Comst. in Japan. The preliminary activities are sharply distinguished from insertion of the ovipositor for oviposition. The female first examines the host scale carefully with the antennae, after which she mounts upon its dorsum. After a further examination in that position, she inserts the ovipositor perpendicularly into the body. This insertion and the probing that follows are not for oviposition but may serve two purposes: to determine whether the host is in a suitable physical condition and whether it is already parasitized. Sometimes feeding takes place on fluids that exude from the wound. When examination is complete and the host is judged satisfactory. the female dismounts from the scale dorsum, approaches the caudal end of the body, and inserts the ovipositor by a backward thrust between the anal plates. The egg is deposited in the intestines and lies with the anterior portion of the stalk extruded between the plates (Clausen 1940/1962). Egg placement is constant and has been recorded also for Eusemion corniger Wlk. (Martelli 1910) and Diversinervus elegans Silv. (Compere 1931).

The unusual habit of inserting the ovipositor beneath the margin of the body of the half-grown Saissetia female and piercing the body wall from beneath is shown by Metaphycus lounsburyi How. Normally the species attacking Coccidae insert the ovipositor through the dorsum. The oviposition habit of hyperparasitic Syrphophagus aphidivorus is distinctive in that the female stands upon the dorsum of the living aphid, inserts the ovipositor perpendicularly, and probes around until the young Aphelinus larva is located. Then the body is pierced and the egg deposited. Surprisingly the aphid host appears not to be inconvenienced or to suffer any discomfort during this act, although it may be already sluggish as a result of the activities of the Aphelinus larva within its body.

Adult females feeding on host body fluids usually does not result in serious host injury. In several other families it is known that certain species which oviposit in the host egg or whose larvae are egg predators completely consume the contents of the eggs on which they feed, and this habit is probably of greater importance in reducing the host population than is the parasitic or predaceous habit of the larvae. Observations on Ooencyrtus johnsoni How. indicate that its stinging of the eggs of Murgantia histrionica Hahn incident to feeding may similarly result in heavy mortality (Maple 1937). It was found that a very large number of eggs that had been punctured with the ovipositor failed to hatch, even though oviposition did not take place. The death of the embryo is due not to the abstraction of appreciable quantities of fluids from the egg but most likely either to mechanical injury or to the injection of some toxic agent at the time of stinging (Clausen 1940/1962).

Little is known about the reproductive potentials of the monoembryonic Encyrtidae. The maximum recorded is circa 250 for Microterys speciosus Ishii, and the majority of species do not seem to deposit greatly in excess of 100-150 eggs. Ishii found an average of 172 mature eggs in the ovaries of gravid females of Aphycus timberlakei Ishii. Crossman (1925) found that virgin females of Ooencyrtus kuvanae How. deposit a smaller number of eggs than do those which are mated. Maple found that mating had no influence on the oviposition activities of the females of O. johnsoni. In fact, one unmated individual produced 224 male progeny, a number in excess of that secured from the mated females. The oviposition period of this species covers circa three weeks (Clausen 1940/1962).

Studying oviposition and fecundity, Olton & Legner (1974) found that females walked or flew when approaching prospective M. domestica host larvae. The host was usually mounted postero-dorsally and inspected with the antennae and tip of the ovipositor. If the host was accepted by the parasitoid, the ovipositor was inserted inter- or intrasegmentally just beneath the integument. The eggs were deposited in 20-45 sec. Chemical stimuli present in the breeding medium apparently were important in releasing the ovipositional response, as females were induced to oviposit in 10 host integuments that were washed in distilled water and re-contaminated with medium, while no oviposition was obtained with 10 clean integuments. Multiple ovipositions were common under laboratory conditions, where an individual female that had just deposited a cluster of 3-6 eggs would return to deposit another cluster. Under unusually high parasitoid/host densities (>50 parasitoids/host) different females usually attacked an individual host simultaneously.

Newman & Andrewartha (1930) stated that T. zealandicus preferred to oviposit in fully developed blow fly maggots, but if this stage were unavailable they readily parasitized earlier instars or pupae. This also was true with M. domestica except that only white to light tan puparia were accepted for oviposition. T. zealandicus could not pierce a hardened puparium. Olton & Legner (1974) compared the number of viable eggs deposited and total number of larvae parasitized per female at three host densities. Females generally deposited most of their eggs during the first 12 h., and there was considerable variation between consecutive periods of oviposition within and among treatments. However, egg number ranged from 11 to 148 (Olton & Legner 1974).

Female T. zealandicus longevity was significantly different between high and low host densities; they lived an average of 50 h at a host density of 5 and 67.2 h at a host density of 30 (Olton & Legner 1974). To determine whether this variation in longevity occurred in the adult life span, the reproductive behavior was divided into prereproductive, reproductive and nonreproductive periods. Six-hour intervals in these categories were totaled, multiplied by 6, and divided by the number of females observed to give the average time per category. There was an apparent trend towards a longer prereproductive period at the low host density. The prereproductive period at the low host density comprised a considerable part of the average longevity. Reproductive periods were not significantly different among all three host densities, the greatest variation occurring among nonreproductive periods which were considerably longer than at the highest host density.

Olton & Legner (1974) studying the extent of parasitization and distribution of viable eggs at three host densities found no significant difference in the average number of hosts parasitized nor the average number of viable eggs laid per female among treatments. But there was a significant difference in the average number of viable eggs laid per host between the high and low density. A tendency towards multiple oviposition or more eggs per insertion of the ovipositor was shown at the low density. Averages of 4.58 eggs/host at a density of 30 hosts and 6.14 eggs/host at a density of five hosts were well below levels resulting in superparasitization. It was concluded that T. zealandicus is a short-lived species capable of depositing a relatively fixed number of eggs during a short reproductive period, regardless of host density. The probability of an increased number of eggs per host was a function of host density.

Under constant conditions cultures of parasitic Hymenoptera may display a periodicity and time-spread in adult emergence (Beck 1968, Legner 1969). In T. zealandicus peak adult eclosion usually occurred during the early morning hours. Emergence seemed independent of a transition from darkness to light as shown by 0400 and 0800 h peaks in all light regimes, suggesting that adult eclosion showed a circadian rhythm (circa a 24-h period) with a free-running period that was similar in all treatments. Analyzing the 12:12 LD treatment to determine if the numbers and sex of parasitioids that emerged per host changed during a 3+ day period, Legner & Olton (1974) regressed the average number of parasitoids emerging per host (Y) on time of emergence (X) and found a highly significant coefficient of -0.859. This indicated that the higher the average density of parasitoids per host (within limits) the shorter the developmental period. Host contents were consumed sooner at the higher parasitoid/host densities. There was no obvious change in the proportion of emerged females and males during periods of peak eclosion (early morning), and females predominated 2:1. Parasitoids issued from a single exit host within circa 3 min. after breakthrough. Both males and females initiated emergence; therefore, there was no differential rate of development between sexes at various densities as found by Legner (1969) for pupal parasitoids.

A high rate of multiple oviposition was thought to occur with high parasitoid:host ratios in T. zealandicus (Olton & Legner 1974). Results of reproductive potential tests indicated that females deposited 4-6 eggs per host during their life span. There was enough food present in a single standardized host to sustain development of progeny resulting from at least 3-4 average egg depositions; however, the progeny usually were small and stunted at the upper limit.

Solitary encyrtid parasitoids of mealybugs produce a pronounced inflation of the host body, causing it to become circular in cross section, cylindrical, and smoothly rounded at both ends. The interior of the shell is smooth and often highly polished, as if by a secretion provided by the parasitoid itself. These parasitized mealybugs are usually referred to as "mummies" and bear a superficial resemblance to certain dipterous puparia. The gregarious species, such as Acerophagus notativentris Gir., produce a similar inflated condition, and each of the surface cells is distinctly outlined externally. In some hyperparasitic species, pupation takes place within the larval skin of the primary host. This is shown by Quaylea whittieri Gir., a solitary internal parasitoid of the mature larvae of Scutellista, Metaphycus and other parasitoids and egg predators of Saissetia and related Coccidae. The larval skin of the host parasitoids becomes very distended and darkened, and also resemble dipterous puparia (Clausen 1940/1962).

The effect of the stage of host development at the time of attack on the cycle of the parasitoid is shown in the case of Hunterellus hookeri How. (Ixodiphagus caucurtei Buy.), which develops internally in many species of ticks in various parts of the world. The eggs are deposited in the nymphal instars of the host, and development of the larvae is delayed until the host becomes engorged with blood. This was called "latent parasitism," and under some conditions a period of six months may elapse from the time of oviposition until the host shows evidence of parasitization. The obligatory diapause in the early larval stage is imposed by the host and is apparently due to the nutritional requirements of the parasitoid larva not being met until the host is fully fed (Clausen 1940/1962; Cooley 1928, Cooley & Kohls 1933, Brumpt 1930).

Regarding sex ratio and parthenogenesis, the majority of Encyrtidae reproduce bisexually, and there is usually a slight preponderance of females among the progeny, the extreme in this respect being a ratio of 5.3:1 in Zarhopalus sheldoni Gir. (Clausen 1924). An exception to this rule may be in Tetracnemus pretiosus in which the males are in excess in the ratio of 1.4:1 (Clancy 1934). This record is based on laboratory rearings and may differ from the field ratio, which in Ooencyrtus johnsoni is circa 4:1.

The sex ratios of the overwintering and spring generations of Microterys clauseni in Ceroplastes are markedly different. The first generation is solitary in young scales, and the adults that emerge in the early spring are predominantly female, in the ratio of 3:1. In the second generation, which is gregarious in the mature scales, an average of 3.15 individuals develop in each scale and the ratio is increased to 9:1. Parthenogenetic reproduction occurs in male progeny only. The peculiar aspect about the reproductive habits of this species is that the "brood" in each scale consists of only one sex (Clausen 1940/1962). In a series of 73 Ceroplastes females isolated individually for parasitoid emergence, not a single exception to this rule was found. The explanation of this is not clear, because the eggs are deposited singly, and the females show no hesitation in ovipositing in hosts that already contain one or more eggs. Therefore, the parasitoid content of a scale is often the result of successive ovipositions by several females over a period of days.

Unisexual reproduction occurs in a number of species. Embleton secured only a single male among 1,000 adults of Encyrtus infelix, and only a single one has been secured among the extensive rearings of the same species in Hawaii. Timberlake (1919) recorded the same reproduction habit in Adelencyrtus odonaspidis Full., Blepyrus mexicanus How., Pauridia peregrina Timb., and Saronotum americanum Perk. To this list may be added Anagyrus subalbicornis Gir., Habrolepis dalmanni and Comperiella unifasciata. Occasional males have been reared in most of these species, but they seem to play no part in normal reproduction. However, Ishii believed that virgin females of Microterys speciosus produce only female progeny, whereas those which are mated produce both sexes. Parker & Thompson (1928) mentioned that their rearings of polyembryonic Copidosoma thompsonii Mercet have not yielded a single male, though they do not state that reproduction is unisexual.

Polyembryony.

The Encyrtidae is the only family demonstrating polyembryonic reproduction among Chalcidoidea. It is mostly confined to a group of closely related genera comprising Ageniaspis, Copidosoma, Paralitomastix and Copidosoma. Most if not all species of these genera probably reproduce by polyembryony. Hosts are exclusively Lepidoptera. Some of these species attack small hosts (eg Holcothorax testaceipes attacking Phyllonorycter spp.) and produce about half a dozen individuals, but others (eg some Copidosoma spp.) that attack large hosts (such as Apamea monoglypha) produce up to two thousand individuals from one parasitoid egg. In many species there are two morphs of larvae, a smaller one which develops normally, and a larger "guard" morph which emerges from the embryonic envelope first but fails to ecdyse and eventually disintegrates (Cruz, 1981). Polyembryonic species often cause the host larvae to become grotesquely deformed and twisted when they are killed as prepupae

Marchal (1898, 1904) first studied polyembryony on A. fuscicollis Dalm. and A. testaceipes Ratz. Other early investigations were Silvestri (1906, 1908, 1914a) on L. truncatellus Dalm., A. fuscicollis praysincola Silv., C. buyssoni Mayr., C. tortricis Waterst., and C. nanellae Silv.; Martin (1914) on A. fuscicollis; Patterson (1915, 1917, 1918, 1921a) on C. gelechiae How. and L. floridanus Ashm.; Leiby (1922) on C. gelechiae; and Ferriere (1926b) on L. kriechbaumeri Mayr. Bugnion (1891) first observed the presence of embryo "chains" of A. fuscicollis Dalm. in the larvae of Hyponomeuta, but he did not attribute them to polyembryonic development. It was not until the classic series of papers by Marchal, the first of which appeared in 1898 and which dealt not only with Ageniaspis but also with several platygasterid genera of similar habit, that the true explanation of their origin was revealed.

There is much uniformity in habit among the members of this group. All lay the egg within the embryo of the host egg, and the host attains the mature larval stage before it dies. In some species a few host individuals may reach the pupal stage. At the time the parasitoid larvae attain maturity, the body of the host becomes much distended, being often twice or more the size of a healthy larva. It is frequently considerably distorted, with a mummified appearance and the uneven surface shows each of the outer parasitoid pupation cells distinctly. In other species the bodies are distended but not deformed. The number of parasitoids that are able to complete development in a single host is dependent on the size of the latter. The species that are known to be of polyembryonic habit, with their hosts and the number of individuals developing in each one, were listed by Clausen (1940/1962).

The genus Encyrtus, as now restricted, is limited in its host preferences to nondiaspine Coccidae Clausen (1940/1962). Ishii (1932a) recorded what he believed was an embryo chain of Syrphophagus sp. in the larva of a syrphid fly. The parasitoid egg is deposited within the embryo in the host egg and during the remainder of incubation of the latter, and in the body of the developing caterpillar, it multiplies into a varying number of cells, forming an elongate, asymmetrical body, enveloped in a membrane of parasitoid origin, the whole mass being generally referred to as an "embryo chain" (Clausen 1940/1962). This chain, which is free-floating in the body of the caterpillar, finally breaks up into its component parts, each of which becomes attached to a host organ and develops into an embryo and finally into a larva. This latter stage is attained only after the host larva has become mature. In hosts that produce a very large number of parasitoid individuals, several of these embryo chains may be found, each one of which has developed from a single egg.

The sexual, or normal, larva is hymenopteriform and presents no distinctive features. Among the larvae arising from an embryo chain, there are a varying number that are characterized by the lack of the reproductive, the respiratory, and possibly the circulatory system. These, which have been designated as asexual larvae (Clausen 1940/1962) were first observed by Silvestri in Copidosoma truncatellus and have since been found among the broods of many other species. They develop in advance of the sexual larvae and are of greater size. They are unable to feed directly and begin to disintegrate shortly after emergence from the embryonic envelope. Silvestri expressed the opinion that these larvae serve a definite purpose in lacerating or breaking up the host tissues for the feeding of the sexual larvae, but Parker & Thompson (1928) considered them to be mere monstrosities and of no special significance. Silvestri also presented the hypothesis that the asexual larva is an ancestral form, harking back to the time when the normal larva was free-living and somewhat vermiform.

The length of the life cycle of the polyembryonic Encyrtidae is dependent on that of the host, insofar as larval feeding is not completed until the host larva is in its final instar. Most of the species listed have only one generation per year, though several have two or three. In the latter case, the cycle of the summer broods, from egg to adult, is circa 30 days.

Hibernation is also host influenced. C. gelechiae, parasitic in Gnorismoschema salinaris, passes the winter in the adult stage and oviposits in the spring as soon as host eggs become available, while when parasitic in G. gallaesolidaginis, this period is passed as an egg in the host embryo. Other species are found during the winter in their early stages of development within the partly grown host larvae, and still others are in the mature larval stage within the host carcass. In every species, there is a close synchronization with the cycle of the host. The parasitoid brood that emerges from a single host may all be of the same sex, or they may be mixed. No males have been found in C. thompsoni and in only one case in L. kriechbaumeri. These two species may reproduce unisexually. In most of remaining species the broods are of one sex only, with, in some species, occasional broods containing a few individuals of the other sex also. Silvestri noted that the broods of L. truncatellus in Phytomctra gramma are usually of one sex, but Leiby (1926) found that the great majority of those from P. brassicae are mixed. L. floridanus and Paralitomastix variicornis usually produce mixed broods. The brood may arise from a single parasitoid egg, or it may be the result of several ovipositions. It was noted in several species that several eggs are deposited at one insertion of the ovipositor. Silvestri thought that about 100 normal larvae are produced from each egg of L. truncatellus. The several thousand individuals in each brood must thus be the result of a considerable number of ovipositions. Marchal and others considered that the mixed broods are the result of oviposition by both mated and virgin females in the same host individual, the former producing female progeny and the latter male progeny only. But Patterson and others do not accept this explanation.

Parasitism Effects on Host Reproduction.--Many encyrtid parasitoids of Coccidae attack the adult females, but, if oviposition is in the nymphs, host death does not occur until after maturity is reached. The host thus may be able to realize a portion of its reproductive potential, and the value of the parasitoid is correspondingly reduced. Ishii thought that Microterys speciosus exercised little repressive effect on the increase of Ceroplastes rubens Mask., for death of the parasitized female seldom occurred until the full quota of eggs had been laid. In contrast, M. clauseni, in its spring generation upon adult C. floridensis Comst., very largely inhibits oviposition after the parasitoid eggs are laid, and the portion of the oviposition potential that is realized in the field is small (Clausen 1940/1962).

Imms (1918a) estimated that 71.9% of the females of Eulecanium coryli L. parasitized by Blastothrix sericea deposit about their normal quota of eggs. Females of Lecanium kunoensis parasitized by Encyrtus infidus in Korea were estimated to deposit circa 50% of the normal number, and it was noted that oviposition frequently takes place even after the parasitoids within the body have attained the pupal stage (Clausen 1940/1962). The effect of parasitism on the diaspine scales is more quickly evident. Taylor (1935) stated that the disintegration of the body of Aspidiotus destructor Sign. begins immediately after hatching of the eggs of Comperiella unifasciata Ishii and Spaniopterus crucifer Gahan; therefore, oviposition ceases within five days after attack by these parasitoids.

Host species

Encyrtidae is very important in biological control practice, having been used effectively against a number of pests, particularly scale insects, aphids and whiteflies. Most Encyrtidae are primary, nymphal endoparasitoids of Coccoidea (Homoptera), but also of the eggs or larvae of Coleoptera, Diptera, Lepidoptera, Neuroptera, Orthoptera, Hemiptera and other genera parasitize spiders Arachnida. A few species are hyperparasitic on Hymenoptera (as primary parasitoids of hyperparasitoids). At least two genera, Hunterellus and Ixodophagus are parasitic on the nymphs of ticks. One tribe, Copidosomatini, whose members are primary parasitoids of Lepidoptera, has an unusual process of multiplication of specimens from one egg within the host larva. The egg in this polyembryonic type of development divides into an irregularly branched chain of cells, each of which turns into a separate embryo. The resulting endoparasitic larvae consume the host larva and pupate within its swollen and distorted skin, with the emerging adults being all of one sex, unless more than one egg was deposited initially (Gibson 1993). The encyrtiform type egg with its aeroscopic plate, is unique to Encyrtidae. A number of polyembryonic species are egg-larval or egg-pupal parasitoids of Coleoptera and Lepidoptera.

The Encyrtidae is one of the most important chalcidoid families for the biological control of insect pests (see summary in Noyes & Hayat, 1994). Many species have been used successfully against a variety of economically important pests, especially species of Coccoidea infesting long-lived woody plants. Species which have been particularly successful include: Habrolepis dalmanni introduced from North America into New Zealand for the control of Asterolecanium variolosum, a serious pest of oak; Anagyrus dactylopii introduced into Hawaii from Hong Kong to control Nipaecoccus vastator, a pest of citrus; Tetracnemoidea brevicornis introduced into North America and New Zealand from Australia to control Pseudococcus fragilis, a pest of citrus.

Almost all species belonging to the Tetracneminae are parasitoids of Pseudococcidae, whilst species of Encyrtinae are known to be parasitoids of a wider variety of coccoids (occasionally also of Pseudococcidae) and other insects, mites, ticks and spiders (Tachikawa, 1981).

Members of this family are most often found to attack the homopterous family Coccidae, especially the Lecaniinae and Dactylopiinae, although other families such as Aphididae and Cercopidae may also serve as hosts. In the hemipterous family Pentatomidae and in closely related forms, only the egg stage is attacked. Many Lepidoptera are parasitized, some by species that develop in the eggs and others in the larvae. Several genera are in the latter group that are capable of polyembryonic reproduction, it being possible that several thousand individuals emerge from a single host. Among Coleoptera, the larval and pupal stages of Coccinellidae and Chrysomelidae are frequent hosts. Dipterous pupae, in particular those of the Syrphidae and Cecidomyiidae, are often parasitized. Several species are recorded from neuropterous cocoons, principally of the genus Chrysopa. On occasion instances are known of attack on other orders and families such as Ooencyrtus submetallicus Howard attacking the chloropid dipteran, Hippelates pusio Loew. Some encyrtids are internal parasitoids of the nymphs of ticks (Ixodidae), represented by the genera Huntrellus and Ixodiphagus.

Regarding secondary parasitism, some representatives of the family develop in Coccidae and Aphididae and in lepidopterous eggs, the primary hosts being the immature stages of other Chalcidoidea, etc. Some genera are known to parasitize the immature stages of the Dryinidae in their cocoons.

A large number of species of Encyrtidae have been utilized in the biological control of crop pests, in particular of scale insects. Four species which have adequately controlled their hosts in one or more geographic areas are Anagyrus dactylopii How. (on Pseudococcus filamentosus Ckll.); Blastothrix sericea Dalm. (on Eulecanium coryli L.); Habrolepis dalmanni Westw. (on Asterolecanium variolosum Ratz.); and Pseudaphycus utilis Timb. (on Pseudococcus nipae Mas.) (Clausen 1940/1962). There are many other cases where an appreciable reduction in the host population occurred, although not sufficient to eliminate entirely the need for other controls.

In Africa a single genus, Hunterellus, is known to attack ticks, and the two species both parasitize the nymphs of species belonging to the Ixodidae. Proleurocerus and Amira, each represented in Africa by a single species, develop from the egg sacs of spiders. The insect host spectrum of the Encyrtidae is extremely wide, with a preference for hosts belonging to the Hemiptera, especially those of the suborder Homoptera, but there are not many of the major insect orders that are free from parasitism by the encyrtids. Isodromus is the only genus in the region which develops in the nymphs of Neuroptera; Comperia species parasitize the oothecae of cockroaches. Some five families of the Coleoptera have been recorded as hosts, and the most common encyrtid parasitoid of beetles is Homalotylus, which attacks a number of coccinellid species. Species attacking Diptera are in the minority. Of these, Tachinaephagus zealandicus Ashmead, a species originally described from New Zealand, is a parasitoid of the puparia of various species of important Muscidae, Calliphoridae and Sarcophagidae, including Sarcophaga haemorroidalis (Fallén) and species of Chrysomyia. A large number of encyrtid genera parasitize the various families of Coccoidea or scale insects, and their hosts are often restricted to a specific family. Anagyrus, Leptomastix, Leptomastidea and Clausenia, all common and important in biological control, are exclusively parasitic in Pseudococcidae (mealybugs); Habrolepis, Adelencyrtus and Zaomma are perhaps the most common parasitoids of armoured scale insects (Diaspididae), whereas the species of Metaphycus, Encyrtus, Anicetus, Paraceraptrocerus and Microterys are often reared from soft scale insects (Coccidae). Lac insects (Lacciferidae) are represented in Africa by a single genus namely Tachardina, which is usually heavily parasitized in nature by Tachardiaephagus and Adencyrtus. Other well known homopterous families, such as the Psyllidae, Aphididae, Cercopidae and Cicadellidae also serve as hosts of a number of encyrtids. Only three genera have been recorded as parasitoids of Hymenoptera, of which Coelopencyrtus is the most common. Four genera are known from this region as being polyembryonic: Paralitomastix, Litomastix, Copidosoma and Coelopencyrtus, the first three genera being parasitic exclusively in Lepidoptera. Of these, Litomastix is the most common, attacking important species such as the army worm and species of bollworm. Copidosoma is known from a single species introduced into South Africa from South America as a natural enemy of the potato tuber moth, Phthorimaea operculella (Zeller). Coelopencyrtus is the only known genus with this mode of reproduction that parasitizes Hymenoptera, and it is represented in Africa by a number of species, all of which have been reared from species of the bee family Anthophoridae. The most common hyperparasitoids belong to the genera Aphidencyrtus and Cheiloneurus: the former is usually a secondary parasitoid of aphids, the latter usually of mealybugs. The most common encyrtid egg parasitoids belong to the genus Ooencrytus; they usually develop from eggs of Lepidoptera, Heteroptera and sometimes Coleoptera."

top of page

Life Cycle.

Most species of Encyrtidae produce several generations each year. This is especially true of those whose hosts are in a proper stage for attack throughout the season. Under optimum temperature, the cycle from egg to adult is complete in 2-7 weeks. Ooencyrtus malayensis Ferr., parasitic in the eggs of Pentatomidae in Java, completes its cycle in 12-13 days. In many cases the duration of the life cycle is strictly dependent on that of the host. This is especially true of the polyembryonic forms that oviposit in the host egg and emerge from the mature larva or the prepupa. A given species may have a single annual generation on one host and several on another, depending upon the habits of the hosts. In each case, the cycles are closely synchronized (Clausen 1940/1962).

Among those species that are parasitic in scale insects, several generations per year is the rule though notable exceptions occur. Microterys clauseni, which passes through its spring generation in Ceroplastes in circa one month, nevertheless has only two generations annually. These adult must persist in the field for several months until the young scales are sufficiently developed for attack in autumn. Hibernation may be in any larval or even in the pupal stage within the host. Many of the species that parasitize insect eggs have a life cycle entirely independent of that of the host. Thus, Ooencyrtus kuvanae has six and a partial seventh in the eggs of the gypsy moth, though the latter has an annual cycle, the greater portion of the year being passed in the egg stage. The parasitoid itself hibernates as an adult (Clausen 1940/1962).

top of page

Immature Stages of Encyrtidae

The Encyrtidae reveal an exceptional diversity in form of the immature stages, and many of the modifications are strictly adaptive. These are made necessary not only by the wide range of hosts attacked, but by the varied conditions under which development takes place.

The Egg.

The egg is characteristically dumb-bell shaped and is laid inside the host. In many cases the stalk of the egg may remain protruding through the body wall of the host thus enabling the larva, when it hatches, to utilise atmospheric oxygen.

Two general types of egg are produced by the Encyrtidae, there being the stalked and the encyrtiform, the latter representing an adaptive modification of the first. In both forms, the ovarian egg is two bodied, and the contents of the anterior body, or bulb, are forced into the egg proper at the time of oviposition, leaving the stalk as a slender tube at the anterior end. In the stalked form, this stalk is func­tionless after deposition but may, in some instances, serve to attach the egg to the integument or to other of the internal organs of the host. Representative genera having this type of egg are Aphidencyrtus, Cerapterocerus, Eusemion and Anarhopus. In Tetracnemus pretiosus, according to Clancy, the stalk is reduced to a broad, blunt petiole one third to one fourth the length of the egg body.

The encyrtiform egg is distinguished from the stalked form by a heavy surface rib, termed the aeroscopic plate by Silvestri (1919), which extends the length of the stalk and of the greater portion of the egg itself. Well known genera having encyrtiform eggs are Encyrtus, Microterys (Fig. 68A, B), Aphycus, Metaphycus, Blastothrix and Ooencyrtus. The plate of O. johnsoni is described by Maple as granulate in appearance and is composed of a mosaic of cells upon the outer surface of the stalk and egg body except for a thickened area near the base of what remains of the bulb of the ovarian egg. In deposited eggs, the plate is much darker than the remainder of the chorion.

In two species of Isodromus parasitic in Chrysopa larvae, it has been found by Clancy that the egg of one, I. niger Ashm., is typically encyrtiform, whereas the other, I. iceryae How. (Fig. 69), lacks the aeroscopic plate and bears merely a melanized ring and a delicate membranous collar on the stalk.

Regarding development of eggs and larvae, in many cases it has been observed that a marked increase in size takes place among the stalked eggs during the course of incubation. The dimensions given by Clancy for the egg of Cheilophagus compressicornis Ashm. are 0.16 X 0.06 mm, for the egg body at the time of deposition and 0.73 X 0.27 mm on the fourth day immediately before hatching. This is exceeded in Tetracnemus pretiosus Timb., the egg of which increases in length from 0.03 or 0.04 mm to 0.25 mm. In Chrysopophagus, the trophic membrane is seen, completely enveloping the embryo, one to two days after deposition. Through this membrane food materials are derived from the body fluids of the host. It envelops the body of the 1st instar larva throughout the stage, allowing the head and tail to protrude.

The distinctive longitudinal rib or plate on the encyrtiform egg of Microterys and many other genera has received considerable attention. It was generally assumed that the egg stalk, which protrudes through the integument of the host, serves as a tube through which the larva draws its air supply from the outside. Timberlake (1913) stated that the larva "maintains, without the least doubt an intimate and vital connection with the egg stalk, and the latter might properly be called a living part of the organism." However, this explanation ignores the relationships of the rib to this function (Clausen 1940/1962).

Silvestri (1919) made the first observations relating to the manner of respiration of the larva through an egg of this type, in which the distinctive rib was described in detail and considered from the point of view of its function. This was done on a series of species parasitic in lecaniine Coccidae in Italy. It was noted that the interstices of the cells of the rib were filled with air and hence designated the structure as the "aeroscopic plate" but concluded that actual respiration of the larva was through the lumen of the tube itself. The plug at the outer end of the stalk was thought permeable to gases, and exchange was thought to be by diffusion. Maple (1937) studied the structure and manner of functioning of this plate in O. johnsoni. He observed that the single pair of caudal spiracles of the larva are always in direct contact with the plate of the egg body and that the location of this point of attachment has no relation to the egg stalk. No aperture could be found in the external plug of the egg by means of which air can pass through the lumen of the stalk. Stained oil tests showed ready penetration of the interstices of the place and thence into the larva's tracheal system. No penetration into the stalk lumen was found. Maple's work showed that the aeroscopic plate, rather than the stalk lumen, provided the channel through which the outside air reached spiracles of the parasitoid larva within the host's body.

Unclear is how the air supply contained in the rib interstices reaches the larval spiracles. The rib is external whereas the posterior end of the larva, bearing the spiracles, remains within the cup-like egg shell. Either the chorion beneath the plate is permeable, or perforations are made somehow by the larva so that an air supply in the plate becomes accessible (Clausen 1940/1962).

It was generally assumed that all eggs of this type which project through the host integument and to which the larvae are affixed possessed the aeroscopic plate until Clancy found that the egg of Isodromus iceryae How. lacks the plate and that the shell and stalk have no function other than to hold the young larva in position. Silvestri previously mentioned that the 1st instar larva of Aphycus melanostomatus Timb. (A. punctipes Dalm.) lacked the caudal spiracles even though arising from an encyrtiform egg but he still thought that the air supply was derived through the egg stalk.

All accounts of development of encyrtiform eggs and larvae pointed out that the posterior end of the larva is encased in the eggshell, but nothing is mentioned regarding the manner in which that position was attained. The stalk is at the anterior end of the egg, and the head of the embryo would thus be formed near the base of the stalk and the posterior end of the body, bearing the spiracles, at the opposite end. In Maple's illustration of the mature embryo of O. johnsoni within the egg, the opposite orientation was shown, with the posterior end of the body at the anterior end of the egg and the spiracles in contact with the aeroscopic plat at the base of the stalk. His illustration of the newly hatched larva of Anagyrus yuccae Coq. indicates a normal orientation prior to hatching (Clausen 1940/1962).

It seems that the position of the young larva with respect to the eggshell is brought about either by a rotation of the developing embryo within the egg or by a reversal of position of the larva immediately after hatching. The former is obvious in O. johnsoni, and observations on Microterys clauseni indicate a similar movement prior to hatching (Clausen 1940/1962). Encyrtiform larvae usually maintain their connection with the egg stalk until the final larval stage, utilizing it for respiration during this entire period. The successive exuviae are forced back over the body and become a part of the sheath enveloping the posterior end of the body. The number of exuviae is directly related to the larval instar.

First instar Larvae.

Among the monoembryonic species of the family, four forms of first instar larvae may be distinguished, based upon morphological modifications having a functional nature. The first instar larva is caudate, vesiculate or hymenopteriform (or "encyrtiform") varying from spherical to elongate. The tail may be bifurcate. They vary from 10- to 14-segmented and may or may not have functioning spiracles.

The hymenopteriform larva has a body of 12-13 visible segments, is widest in the thoracic or anterior abdominal region, and has no sculpturing or segmental processes. These larvae lie free in the body cavity of the host and lack the open tracheal system. A typical representative of this group is Comperiella bifasciata How. (Compere and Smith, 1927).

The second is the encyrtiform, so called because it hatches from the encyrtiform egg previously described. The number of body segments is reduced, there being only 10-11 visible, and the last segment, which bears the single pair of spiracles, apparently represents several that have fused. The last four or five segments are usually closely enveloped by the eggshell, and this connection persists through the greater portion of the larval stage. The larva of Isodromus iceryae, which hatches from the modified encyrtiform egg already described, is hymenopteriform, for it lacks the posterior spiracles and consequently does not derive its air supply through the stalk. Both the egg and first instar larva of this species appear to represent transi­tional stages between the hymenopteriform and the encyrtiform type. The genus Microterys contains many well known species having encyrtiform larvae; yet it has been shown by De Bach that in M. titiani it is hymenopteriform, with a full com­plement of spiracles.

The caudate larva (Fig. 70) is frequently found among the species attacking Coccidae and Aphididae and is characterized by the development of the last abdominal segment into a tail like organ that may exceed the body proper in length and may bear setae on the distal por­tion. It is associated with the stalked type of egg. These larvae do not possess open spiracles. Many genera have larvae of this type, the best known being Aphidencyrtus, Cerapterocerus, Cheiloneurus, Eusemion and others.

The vesiculate form is similar to the hymenopteriform, except that the proctodaeum is evaginated to form a caudal vesicle, and may also have a ring of fleshy protuberances around each of the first 12 segments. This modification is rare among the Encyrtidae and is at present known only in the genera Anarhopus and Clausenia (Fig. 71C), both of which para­sitize mealybugs. Tetracnemus pretiosus (Fig. 71A, B) may be of the same type, though it is uncertain whether or not the small expanded organ on the caudal segment corresponds to the vesicle in the above named species. Both A. sydneyensis and T. pretiosus are distinguished from other known Encyrtidae by the presence of a ring of fleshy processes or protuberances on the first 12 body segments; and the former has also a single, curved medium process dorsally on the last segment, immediately above the vesicle.

Encyrtiform larvae of M. clauseni, which occurs in the hind intestines of Ceroplastes, seems to limit its feeding to the contents of the digestive tube, and only after the 2nd molt is the intestinal wall broken and direct feeding on the viscera and body fluids occurs. The host usually dies 10-12 days after parasitoid oviposition.

In many species that have stalked eggs, a varying proportion of individuals may also retain the egg shell and the cast skins as an envelope about the caudal end of the body (Clausen 1940/1962). This habit is associated with larvae of both the hymenopteriform and caudate types, which are free living in the host body, and serves no apparent purpose. In C. compressicornis, the successive exuviae cover a considerable portion of the body, and the mandibles of the firs two can be easily distinguished on the mid-ventral area of the 3rd instar larva.

In Encyrtidae the number of larval instars is variable, ranging from 2-5. Anarhopus sydneyensis Timb. is the single species known to have only two (Compere & Flanders 1934). Most species are thought to have three instars, although several have four or five. However, the great majority probably have five and the lesser number recorded in many cases is due to having overlooked early exuviae. An exact number can be determined only by clearing and staining the entire host contents, and in this way the mandibles become recognizable and can be measured.

Modifications in larval development that are most striking relate to respiration of encyrtiform larvae during later stages. At this time a functional relationship is made by some species with the host's tracheal system. This phenomenon was demonstrated in Encyrtus infelix Embl. (Embleton 1904, Thorpe 1936). E. infidus Rossi (Clausen 1932b), Aphycus melanostomatus Timb. (Imms 1918) and Carabunia myersi Waterst. (Myers 1930). Carabunia is parasitic in the nymphs of the froghopper, Clastoptera undulata Uhler, and the remaining species attack lecaniine Coccidae. All these, with the possible exception of A. melanostomatus, pupate and emerge while the host is still alive. Clausen (1940/1962) stated the course of events as follows: "When approaching maturity, but before the last molt, the parasite larva becomes invested with a membranous sheath. At approximately the same time, the tracheal branches of the host fuse with, or become attached to it, in the immediate vicinity of each of the four parasite spiracles. At this time, the functional connection of the larva with the egg stalk is broken. The sheath, which surrounds the larva and later the pupa..., becomes filled with air, and the oxygen supply of the parasite is derived therefrom. Miss Embleton, who was first to observe this remarkable adaptation, surmised that the sheath was probably a cast larval skin, and this interpretation was followed by several later authors. Imms concluded that it arises as a chitinous proliferation of the host tracheae, whereas Thorpe, after a detailed study, has recently stated that it is of host origin but produced by phagocytic action, in the building up of which the find tracheal branches play a part."

Flanders (1938a) arrived at what seems to be the correct conclusion as to the origin of this sheath. He found that the ileac and labial glands of the larvae are apparently identical in function and that they produce a viscid material which exudes from both ends of the body and spreads to form a thin protective covering. The sheath of Encyrtus is consequently a cocoon, in film form rather than composed of strands, and is identical in origin with the common spun cocoon. The sheaths enveloping the larvae and pupae of other endoparasitic Chalcidoidea, in particular the Encyrtidae and Aphelinidae, are developed in the same manner. Although Thrope's interpretation of the origin of the sheath of E. infelix is seemingly not valid, yet his explanation of the manner in which the connections between the sheath and the host tracheae are brought about is of interest. As the sheath develops, the adjacent tracheal trunks of the host form a union with it in the immediate vicinity of the larval spiracles. This is stated to be the result of a physiological rather than a mechanical reaction. The tracheal epithelium is activated by a sudden change in respiratory activity, such as a lowering of the oxygen tension or an increase in carbon dioxide concentration incident to the approach of the pupal stage and to the stoppage of an adequate air supply through the egg stalk. Such a stimulus would naturally be most strongly felt in the areas surrounding the open spiracles. The sharp bending of the tracheal branch, which is evident at the point of attachment, results in most cases in a definite fracture of the tracheal lining (Clausen 1940/1962).

Intermediate instar and Mature Larvae.

The great­est diversification in form occurs in the first instar, and the succeeding forms tend to become more uniform as the final instar is reached. The hymenopteriform larva, which lies free in the body cavity of the host, progresses through the series of molts without appreciable change in its essential characters. In the caudate forms, the tail becomes considerably reduced in size in the second instar and practically disappears in the third. The vesiculate forms, on the other hand, show an enlargement of the vesicle in the second and third instars. In Anarhopus and Tetracnemus, the ring of fleshy protuberances on each body segment of the first instar larva is lacking after the first molt.

The number and position of the spiracles of the larvae are an exceedingly variable character in Encyrtidae. In the hymenopteriform larva, the spiracles are lacking in the first and second instars, but they appear on the second to the tenth body segments in the third or a later instar. Among the species having caudate larvae, Cerapterocerus mirabilis Westw. is stated to lack spiracles until the fourth instar, at which time the nine pairs appear in the position already mentioned. In Carabunia myersi (Fig. 76) they are first found on what is stated to be the third and final instar, and only three in number, the anterior pair being on one of the thoracic segments and the remaining two pairs on the abdominal segments immediately preceding the caudal appendage.

Information regarding the spiracle arrangement of vesiculate larvae is available only for Anarhopus sydneyensis. In this species, they are lacking on the first instar and occur on the second to tenth body segments of the second and last instar. Tetrac­ncmus pretiosus has no open spiracles until the final instar, when the full complement appears.

The first instar encyrtiform larvae possess a single pair of spiracles on the last apparent abdominal segment. This arrangement persists in the following two instars, and the nine pairs of spiracles then appear on the fourth instar. In Microterys speciosus (Ishii, 1923), they are stated to appear on the third instar. Clancy mentioned that the second instar of Isodromus is readily distin­guished from the first by the presence of the spiracular spurs in the second to ninth body segments, and this character may be common to many second instar larvae of the encyrtiform type. A marked departure from the normal for the family occurs in Metaphycus lounsburyi (Smith and Compere, 1920), in which the single caudal pair of spiracles of the first instar is followed by three additional pairs, situated on the second to fourth body segments, on the second instar, and by the usual nine on the third instar.

A further modification in spiracular arrangement is found in certain species of Encyrtus having encyrtiform larvae, which acquire in their later stages an intimate connection with the host respiratory system. In E. infelix (Thorpe, 1936), the fourth instar has 3 pairs of spiracles, one of which is on the prothorax and the remaining two at the posterior end of the abdomen. The caudal spiracles are borne at the end of a pair of slender tube like processes, merely enclosing the tracheal tubes, half to two thirds the length of the body proper. The fifth instar larva bears only two pairs of spiracles, one at each end of the body. This spiracular modification occurs also in E. infidus, though one instar apparently was overlooked, and the described third (Fig. 72B) is identical with the fourth of E. infelix. Ishii (1932a) described the supposed first instar larva of E. barbatus Timb. which has the three pairs of spiracles arranged in identically the same manner as is given above for the fourth instar. The large size of this larva indicates that it may be a later instar than that stated (Clausen 1940).

In considering the various adaptations, it is seen that the characters mentioned are common to a number of genera and that, in some instances, all species of a given genus do not reveal the same modifications.

Also of interest in the biology of Encyrtus is the habit of the mature larva of reversing its position within the sheath prior to pupation. This occurs whether the parasitoid is solitary in young scales, as E. infelix in Saissetia hemisphaerica Targ., or gregarious, as E. infidus in Lecanium kunoensis Kuw. In the latter, an average of 6.4 Encyrtus individuals reach maturity in each full grown female scale. Without a reversal in position, the pupae and the newly transformed adults would lie with their heads directed downward toward the venter of the scale, and emergence from the living host would be encumbered. However, the head of the pupa is directed outward near the point of insertion of the egg, and the instincts of the adult to move directly forward bring it quickly to the body wall of the host where emergence may occur.

Why E. infelix changes its position is not so clear; for the solitary parasitoid is oriented along the longitudinal axis of the host, and emergence of the adult would seem to be as readily accomplished from one end or the other. Thorpe stated "That this turning movement is the result of an innate instinct and is not dependent on some stimulus provided by the tissues of the host is shown by the fact that even in those rare cases where the egg has been deposited anteriorly the tendency to turn is still manifest."

Pupation usually takes place inside the body of the host. In some species the host is not killed until after the adult encyrtid emerges. In these species the mature larva makes a pupation chamber in the form of a membranous envelope which becomes confluent with the tracheal system of the host. The envelope then becomes filled with air thus enabling the pupa to respire (Embleton, 1904: Clausen, 1932; Thorpe, 1936). Overwintering is generally as a mature larva or pupa within the body of the host.

Regarding pupal respiration, the change in position has little effect, for in either case the two points of fusion of the sheath and the host tracheae would overlie the two pairs of functional spiracles of the pupa.

The way of formation of the sheath and the pupation habit of Carabunia myersi are apparently identical with those of Encyrtus, but it is interesting that the early larval instars are of the caudate type rather than encyrtiform. Tracheal attachment may occur at several or all of the six larval spiracles. The sheath does not become filled with air until sometime after pupation, while in Encyrtus the connection is functional during the last larval stage and air bubbles surround the pupa at its formation.

top of page