Description & Statistics
There are circa 86 valid genera and 1,170 species in the cosmopolitan Mymaridae. Important morphological characters include the torulae being closer to the compound eyes than to each other; head bears three sets of sclerotized bars; hind wing petiolate; antenna long in both sexes, female antenna with club (1-3 segmented), male antenna filiform (lacking a club); hind wing spatulate (very narrow), fore- and hind wing with long marginal fringe. Tarsi 4-5 segmented. Mymarids are of minute size; the stigmal vein is reduced and the antennae lack annuli.
All Mymaridae are primary, solitary, internal, egg parasitoids, principally of auchenorrhynchous Homoptera. They will, however, attack other groups of insects. Several species have been used successfully in biological control in Hawaii, New Zealand, Australia, South America and South Africa, primarily against leafhoppers. One species was very effective in the biological control of eucalyptus snout beetle in southern Africa and South America.
Gibson (1993) described Mymaridae as having a body usually less than 1.5 mm, but rarely up to 5 mm long, at most with obscure metallic luster. The antennae are inserted closer to each eye than to each other (except if the head appears wedge-shaped in lateral view). They are almost always much longer than the combined length of the head and mesosoma. The first flagellomere is very rarely ring-like in males. The antennal club is distinct in females, and absent from the male. The head has an H-like, often dark, set of marks. There is a transverse mark across the head above the toruli and a lateral mark along the inner margin of each eye on the face vertex. These parasitoids are usually fully winged. The hind wing usually either has a membrane arising apically from the stalk formed by the submarginal vein, or the wing is composed only of a stalk, but rarely with the membrane narrowly extending to the wing base. The protibial spur is quite long and curved. The tarsi have 4-5 tarsomeres. The metasoma is widely attached to the mesosoma (sessile), or separated by a distinct constriction (subsessile) or tubular petiole (petiolate).
Mymarids were noted as being widely distributed by Gibson (1993), with ca. 100 genera and 1,400 nominal species. The family is usually divided into 2 subfamilies in two different classifications. In one classification the number of tarsomeres is used to characterize the subfamilies Mymarine (with 4 tarsomeres) and Gonatocerinae (with 5 tarsomeres). In the alternative classification, the manner of attachment of the metasoma to the mesosoma is used to distinguish Alaptinae (widely attached) and Mymarinae (subsessile or petiolate). A 3rd subfamily, Eubroncinae, is based mostly on a wedge-like head structure (Yoshimoto et al. 1972). Although not all Eubroncinae have their head conspicuously wedge-shaped, members differ from other mymarids by having the hind wing membrane narrowly extending to the base of the wing so that it is not stalked. Mymaridae is indicated to be a monophyletic group based on the pattern of lines on the head plus other features listed by Schauff (1984). Mymarids are distinct structurally from other chalcidoids, so that they sometimes are considered to be the sister group of the rest of the superfamily. But, Chalcidoidea excluding Mymaridae lack known synapomorphic features, so that it is also likely that Mymaridae represents the sister group of only some part of the superfamily.
Gibson (1993) noted that all mymarids are parasitoids of insect eggs, mostly eggs that are laid in concealed localities, such as plant tissue, under scales, or in the soil. Hosts are mostly Homoptera and Hemiptera, but Psocoptera, Orthoptera, Coleoptera and Diptera may be used.
Key references are Annecke & Doutt (1961) who revised the world genera. Schauff (1984) revised the North American genera, and Noyes & Valentine (1989a) the New Zealand genera. Yoshimoto (1990) revised the New World genera. Huber (1986) provided a historical review of mymarid classification and recorded hosts and included an extended bibliography.
Mymarids are cosmopolitan and consist of species of minute size, often only a fraction of 1 mm in length with exceptionally long antennae, of which the distal segment is much enlarged in the female and the wings are long, narrow, and fringed with long hairs (Clausen 1940/1962). Some species, such as Polynema microptera Bakk., have the wings greatly reduced, although they appear normal in form. There are thought to be three forms or races of Prestwichia aquaticaLubb. (Henriksen 1922), the typical one having fully winged females and males with rudimentary wings. In the second form the wings are rudimentary in both sexes, and in the third the wings of the females are reduced. All mymarids are internal parasitoids in the eggs of other insects, in particular Homoptera, but also of Odonata, Lepidoptera, Coleoptera, Neuroptera, Corrodentia (Psocidae) and Hemiptera. Some have been recorded from Coccidae and Aleyrodidae, but their relationship is uncertain (Clausen 1940/1962).
Mymarids have been successfully utilized for the biological control of crop pests. In 1904 a number of species were introduced to Hawaii from Australia for the control of the sugarcane leafhopper, Perkinsiella saccharicida Kir. Of these Anagyrus frequens Perk., Paranagyrus optabilis Perk., and P. perforator Perk. were established and with other parasitoids, produced partial biological control. Later complete biological control was achieved with the introduction of the mirid egg predator, Cyrtorhinus mundulus Bredd., from Australia and Fiji in 1920. Anaphoidea nitens Gir., parasitic in eggs of the eucalyptus weevil, Gonipterus scutellarus Gyll., in Australia, has been used effectively against the same host in New Zealand and South Africa and against G. gibberus Bsd. in Argentina with some success.
Specific Geographic Areas
NEARCTIC (CANADA).– Yoshimoto (1984) noted that “The family Mymaridae is an unusual group in the superfamily Chalcidoidea. In some fossil species the pronotum reaches the tegulae (as in Proctotrupoidea). In most extant species the antennal toruli are far apart, usually near the eye margin. However, in the subfamily Eubroncinae (Yoshimoto, Kozlov & Trypitsin 1972), which is not present in Canada, the antennal toruli are on a projection from the middle of the face and closer to each other than to the eye margin. The larvae are similar to the scelionid larvae (Nikol’skaya 1952). Yoshimoto (1975) postulated that the mymarids have evolved independently from the primitive eurytomid-torymid ancestor group and before the pteromalid and tetracampid-eulophid lines.””Members of this family are recognized by the following characters: Size usually 0.35-2.00 mm. Body nonmetallic, generally slender, delicate, with long thin antennae and legs. Dorsum of head with 2 sets of sutures and a carina, 1 strong prominent transverse suture (= transverse trabecula) between inner orbits of eyes dorsad of antennal toruli, 1 pair of frontal carinae extending from extremities of transverse suture ventrally along inner orbits of eyes, and 2 supra-orbital sutures (= supra-orbital trabeculae) extending from extremities of tranverse suture dorsally toward posterior part of head. Transverse and supra-orbital sutures of vertex folded along the margins of head sclerites, and sometimes appearing as carinae or striations; folded margin of sclerite often obscuring view of suture from above. Wings usually long and narrow, often with long marginal fringe of setae, and always with greatly reduced venation in fore wings; venation usually restricted to basal third of wing where marginal vein short or absent; stigmal and postmarginal veins usually absent or vestigial; fore and hind wings pedunculate (stalked). Antennae long and slender in both sexes, 8-13-segmented, without anelli; male antennae filamentous; female antennae with distinct club, 1-3-segmented, with longitudinal sensoria (usually obscure unless slide-mounted); toruli much nearer to eyes than to each other. Legs often long and thin in relation to body, with tarsi 4- or 5-segmented; frontal tibial spur long, curved, bifid, forming a well-developed strigil.”There are two main systems of higher classification of the Mymaridae, as exemplified by Annecke and Doutt (1961) and by Peck et al. (1964). In the former system the mymarids are divided into the subfamilies Alaptinae and Mymarinae, based on whether the gaster is sessile (tribes Anagrini, Alaptini) or subsessile (tribes Anaphini, Ooctonini) to petiolate (tribe Mymarini), and on the extent to which the mesophragma projects into the gaster. In the latter system the family is divided into the Gonatocerinae (= Lymaenoninae) and Mymarinae, based on whether the tarsi are 4- or 5-segmented. In both systems tribal categories are used, which are based on the major character for subfamily separation in Annecke and Doutt’s (1961) classification.” Yoshimoto (1984) stated that in his manual he was following the system of Ashmead (1904), Debauche (1948), Nikol’skaya (1952), Bou…ek (in Peck et al. 1964), and Schauff (1983).
Details on anatomy of Mymaridae is found in Debauche (1948), and a key to genera of the world in Annecke & Doutt (1961). Schauff (1983) included 22 valid genera for the Holarctic, with synonymy, phylogeny, generic diagnoses, morphology, and a key to genera. Graham (1982) examined and identified all of Haliday’s collection of Mymaridae, with redescriptions and comments on each type. A key to genera of the New World is given by Yoshimoto [1984 in preparation].
Yoshimoto (1984) noted that all Mymaridae are internal primary egg parasitoids of a wide variety of insects. Although often poorly represented in collections because of their minute size, mymarids are distributed worldwide and are common. Specimens ought to be mounted on glass slides for study. Yoshimoto (1984) gave a key to subfamilies of Mymaridae: Gonatocerinae and Mymarinae.
All genera within Gonatocerinae have a simple antennal club (except Eustochomorpha Girault, which is not known from North America). There are 9 genera in North America, of which Dicopus Enock, Macrocamptera Girault, LitusHaliday, Alaptus Westwood, Arescon Walker, Camptoptera Förster, Gonatocerus Nees, and Ooctonus Haliday occur in Canada (Yoshimoto 1984).
All Gonatocerus spp. parasitize species of Coleoptera, Hemiptera and Homoptera. The species of Ooctonus parasitize Hymenoptera and members of Alaptus parasitize Psocoptera. The genus Litus is associated with Staphylinidae and species of Camptoptera parasitize Buprestidae, Cicadellidae and Thripidae (Yoshimoto 1984).
The antennal club has 1-3 segments; 14 genera are found in North America, of which Anagrus Haliday, Cleruchus Enock (= Paracleruchus Yoshimoto 1971, sensu Viggiani 1974), Anaphes Haliday, Erythmelus Enock, Omyomymar Schauff, Eustochus Haliday, Chaetomymar Ogloblin, Stephanodes Enock, Acmopolynema Ogloblin, Mymar Curtis, Caraphractus Walker, and Polynema Haliday have been known in Canada (Yoshimoto 1984).
The genus Anagrus was revised by Gordh (1977) with five species for North America.
In Mymarinae, species of Anagrus are parasitoids of Hemiptera, Homoptera, Lepidoptera and Diptera; species of Cleruchus parasitize Coleoptera; species of Anaphes (Patasson) parasitize Coleoptera and Hemiptera; Mymar parasitizes Coleoptera, Hemiptera, Homoptera and Lepidoptera; species of Polynema parasitize Odonata, Coleoptera, Homoptera, Diptera and Hymenoptera; Chaetomymar parasitizes Cicadellidae (Homoptera) and Lyonetiidae (Lepidoptera); Acmopolynema parasitizes Acrididae (Orthoptera); and Erythmelus parasitizes Cicadellidae (Homoptera), and Tingidae and Miridae (Hemiptera) (Yoshimoto 1984).
PALEARCTIC (EUROPEAN former USSR).– Trjapitcyn (1978/1987), as translated from the Russian, described this family as “Minute chalcids, usually less than 1.0 mm long. Antennae long, without rings, in female with 8 to 11 segments with large nonsegmented or two-segmented clava; antennae of male filamentous, with 10 to 13 segments. Shield of mesonotum with complete parapsidal grooves. Wings narrow and long, usually with long marginal fimbria. Legs long and slender. Abdomen often with long petiole. Body black, brown, or yellow, without metallic sheen. Egg parasites (endoparasites of eggs) of various insects, predominantly cicadids (Auchenorrhyncha) and beetles (Coleoptera). Reports of parasitism by mymarids of armored scales (Diaspididae) and other coccids are erroneous. More than 70 genera with more than 1,100 species described in world fauna; 23 genera with more than 200 species known in Europe.”
AFRICA.– Prinsloo (1980) commented that Mymaridae form an extensive family of minute wasps which include some of the smallest insects known, and many species measure only a fraction of a millimeter. The family, also known as the “fairy flies,” comprises more than 100 genera and had been reviewed by Annecke & Doutt (1961). Prinsloo (1980) discusses this family further as follows:
Relationships & Diagnosis.– “Like the trichogrammatids, the mymarids are exclusively egg parasitoids and minute in size, but there are many structural characters, such as number of tarsal segments and shape of the wings and antennae, which separate the two families. The Mymaridae is probably most readily separated from all other chalcidoid groups by the petiolate hind wings.”
“Minute species, usually much less than 1 mm in length; body weakly sclerotized; antennal sockets placed far apart, close to the eye margins; antenna long and slender, clavate in the female, filiform in the male; funicle with four to seven segments, the club with one to three segments; fore wing venation reduced, the marginal vein usually short, the stigmal greatly reduced and postmarginal vein absent; fore wing often slender or petiolate with extremely long marginal fringe; hind wing always with a basal stalk which is composed solely of the submarginal vein, the wing disc ribbon-like, not extending to its base; abdomen with gaster broadly sessile or petiolate, the ovipositor sometimes very long; legs usually long and slender, the tarsi with four or five segments.”
Biology.– “Mymarids are all, as far as is known, primary endoparasitoids of insect eggs, and there appears to be a preference for hosts belonging to the Homoptera, although many species of Heteroptera, Lepidoptera, Coleoptera and even Odonata and Psocoptera have been recorded as hosts. One of the best known species from southern Africa is Patasson nitens (Girault) which was introduced into South Africa from Australia against the eucalyptus snout beetle, Gonipterus scutellatus Gyllenhal. In most parts of South Africa complete biological control has been achieved against this weevil by P. nitens, and the biology of the parasitoid has been well documented (Tooke 1955). Species of Anagrus are known to attack a number of injurious leafhoppers, and in South Africa species of this genus develop from the eggs of the maize leafhopper, Cicadulina mbila (Naudé). Two species of Alaptus have been recorded from Africa as parasitic in the eggs of psocids.”
African Mymaridae.— “Some 20 genera of mymarids have been recorded from the Ethiopian region. These may be divided into two subfamilies, namely the Alaptinae, which as the gaster broadly and truncately sessile, and the Mymarinae which has the gaster convexly rounded at its base (Annecke & Doutt 1961). The majority of the genera belong to the latter subfamily. Anagrus, which is cosmopolitan, is probably the best known genus in the Alaptinae, and is distinguished from most other genera in the subfamily by the four-segmented tarsi. Mymar (Mymarinae) is a typical mymarid with slender petiolate fore wings and long, slender, clavate antennae; in Polynema, another common genus of the Mymarinae, the wings are much broader, without a long basal petiole.”
INDIA & ENVIRONS.– Subba-Rao & Hayat (1988) noted that “Species of the family Mymaridae, so far as their biology is known, are parasitic in the eggs of other insects. They also include some of the smallest known insects in nature, though species of some genera (Polynema, Gonatocerus) may exceed 1 mm. in length. The hosts attacked include Hemiptera, Lepidoptera, Coleoptera and also Odonata and Corrodentia. There are records of mymarids having been reared from nymphs of coccids and alyrodid pupae, but these undoubtedly are based on misidentification of the host stage. Mymarids have been utilized in the biological control of insect pests of crops with varying degrees of success.”
“Mymarids are cosmopolitan in distribution and can usually be collected in large numbers. A single sweep on vegetation may sometimes yield representatives of several genera. However, these are easily overlooked by the unfamiliar because of their small size. Therefore, mymarids have received very little attention in the past and, consequently, their taxonomy remained in a confused state till very recently.”
History.– “The family contains about a hundred genera and over 1,100 species. The earlier major contributions to the family were made by Girault (1911, 1912, 1929) on the Australian and American fauna. Other noteworthy contributions were those by Debauche (1948), Soyka (1949, 1956), Hincks (1950, 1952), Kryger (1950) and Ogloblin (1946). The credit, however, goes to Annecke & Doutt (1961) who for the first time provided a comprehensive account of the family and gave keys to the genera. Following this paper, several important papers were published by Viaggiani, Subba Rao, Taguchi, and more recently by Schauff.”
“The mymarid fauna of the Indian subcontinent was studied mainly by Mani (1939), Narayanan, Subba Rao 7 Kaur, Subba Rao and Mani & Saraswat. Recently, Subba Rao and Hayat (1983) catalogued the Oriental species and provided a key to the genera. In the present work we recognize 20 genera and 67 species from the region. As with most of the other chalcidoid families, this number probably represents only a small fraction of the actual number of species present in the region.”
Biology & Behavior
Mymaridae are true egg parasitoids in that they normally attack the host eggs before an appreciable development of the embryo has occurred. In eggs of Perkinsiella attacked by Paranagrus development of the embryo is inhibited. Prestwichia aquatica is believed to oviposit in eggs of any developmental stage, but it is not able to mature in those with advanced embryos. Anaphes nipponicus Kuway. confines its attack to Lema eggs that are only 1-3 days old. Polynema striaticorne Gir., which produces three successive generations in a single brood of Ceresa eggs, is an exception to the above and is able to attack successfully even those eggs containing well developed embryos (Balduf 1928b).
The majority of species of mymarids are solitary. This applies in particular to the dominant genus Anagrus, and the only exception in this genus is A. atomus L., which, while solitary in the eggs of Erythroneura pallidifrons Edw. in England (MacGill 1934), develops regularly with two individuals in those of Tettigonia viridis L. (Pierre 1906). From 1-7 Anaphes nipponicus emerge from each parasitized egg of Lema oryzae Kuway., although this parasitoid is more often solitary, and an average of seven Anaphoidea calendrae Gahan develop in the eggs of Calendra spp. A maximum of 50 Prestwichia aquatica have been secured from single eggs of the larger species of Dytiscus.
Adult Behavior.–The genera Anagrus, Caraphractus and Prestwichia include species in which the adults are of aquatic habit; this is made necessary by the occurrence of the host eggs in which they develop beneath the water surface in streams and ponds. Several researchers have described the manner of swimming of adult P. aquatica and its varieties. According to Enock (1898) the middle legs only are used, while Henriksen found that the hind legs serve this purpose. Both sexes are able to swim, and they are also able to walk on the surface of the water. Heymons (1908) believed that adults remained under water for five days without being harmed. A. brocheri Schulz is unable to swim but moves about the water by walking on foliage, plant stems and other objects (Clausen 1940/1962).(Enock 1989) discussed the mating of P. aquatic inside dytiscid eggs in which they had developed prior to emergence. Rimsky-Korsakov ((Clausen 1940/1962) confirmed this observation and believed that one male may mate with all the females in turn in the egg. However, Heymons & Henriksen (1922) were unable to verify this habit and saw no indications of mating activity on the part of the males, for all individuals in the egg remained quiescent until one had made an emergence hole.
Life span of the adult is very short, and in a number of species females are able to oviposit immediately after emergence. The reproductive capacity is low, although the P. striaticorne female is thought to be able to deposit 18-20 eggs in quick succession, for each of that number of ovarioles in the reproductive system contained one mature egg (Clausen 1940/1962). Kuwayama (1935) located an average of 26 eggs in the ovaries of gravid Anaphes nipponicus females. Egg production in most species probably does not exceed 100 (Clausen 1940/1962).
Preparatory to oviposition, the female Polynema striaticorne searches rapidly over the bark of twigs infested with Ceresa eggs, tapping the surface constantly with the antennae. When an oviposition scar containing living eggs is encountered, the female inserts her antennae into the crevice and taps the eggs to determine their suitability. She then stands astride the crevice, retaining the tips of antennae in contact with eggs, brings her ovipositor forward to a perpendicular position and thrusts it into the top of an egg. Usually all eggs in a packet, numbering 7-12, are attacked at this time. Mymarids which attack insect eggs deposited upon the surface of foliage find no difficulty in oviposition, but where these are inserted in leaf tissue or in plant stems a portion may be protected from attack. Female Anaphoidea luna Gir. is able to reach only the few eggs of H. variabilis Hbst. that are near the oviposition puncture in fresh alfalfa stems; but she is able to enter dried stems, and all the eggs of the mass are thus exposed to attack. Among egg parasitoids of the sugarcane leafhopper, Paranagyrus optabilis parasitizes principally the eggs at the base of the mid-rib, while Anagrus frequens Perk. limits itself mainly to those at the edge of the distal portions of the leaf (Perkins 1905e).
Most species have only a single generation each year, which is correlated with the annual host cycle. P. striaticorne, which has been studied by Balduf, is parasitic in the eggs of various Membracidae, in particular the buffalo treehopper, Ceresa bubalus F., and this host has a single annual generation, with the winter passed in the egg stage. However, the parasitoid passes through three generations in this one brood of eggs. Other species, having multibrooded hosts, breed without interruption throughout the season, so long as temperature conditions are favorable. Pananagyrus optabilis produces a generation every three weeks under subtropical conditions, Anaphes nipponicus completes its cycle in 8-13 days and has 5-6 generations each year, Anaphoidea nitens requires 17-22 days and Anagrus atomus L. 16 days (Clausen 1940/1962).
The egg stage is of short duration, as the larva of Anaphoidea calendrae was found less than six hours after oviposition, and in A. nitens hatching occurs in 1-2 days. In the latter, larval and pupal stages require 6-8 and 10-12 days, respectively. Hibernation is known for several species of Polynema, Ooctonus and Lymnaenon, and in each case the 1st instar larva is found in host eggs during winter. Anagrus armatus var. nigriventris Gir. hibernates in the half grown “histriobdellid” form (Armstrong 1936).
Sex Ratio & Parthenogenesis
Females of P. aquatica outnumber males 20:1, and in Anaphes nipponicus, Anagrus incarnatus and A. armatus var. nigriventris by 2- or 3:1. Mossop (1929) found the ratio in Anaphoidea nitens Gir. to vary under field conditions, but females consistently predominated, the highest ratio being 3:1. In A. calendrae (Satterthwait 1931), that produces an average of seven parasitoids in each host egg, 72.5% of the broods contain only a single male and only 15% contain more than one. In contrast to this males predominate in the ratio of more than 3:1 in Polynema striaticorne. Collections of the latter covering an entire season gave a ratio of 3/1.
Anagrus atomus, A. frequens, Paranagrus optabilis, P. perforator and Polynema euchariformis Hal. reproduce unisexually. However, occasional males are found in each species except P. euchariformis.
Immature Stages of Mymaridae
Ganin (1869) reviewed work on embryological development in mymarids. In Anagrus and Prestwichia aquatica var. solitaria R. & T., there is an appreciable growth of the egg during incubation, which in the latter species results in an increase in length from 0.22 to 0.38 mm within 24h after being laid.
There is little variation in egg form within the family. The main body of the egg is ellipsoidal, ovoid, or spindle shaped, with a slender tapering peduncle at the anterior end ranging in length from one tenth that of the egg body in Anaphoidea nitens and Caraphractus to equal its length in Polynema striaticorne (Fig. 54A). The ovarian egg is practically identical with the oviposited egg. These eggs are exceedingly minute, ranging from O.06 mm. in length in Anagrus atomus to 0.25 mm. in P. striaticorne.
Development of the sacciform 1st instar larvae of Prestwichia and several species of Anagrus is little known because they lie in the egg fluids and are entirely incapable of movement. Respiration, and probably a certain intake of food also is by diffusion through the integument. The mymariform 1st instar larvae, which occur in most other genera, are well developed and capable of considerable activity. The tail is utilized for moving about the host egg, this movement being facilitated by the use of the several rows of long spines on the body segments. Such activity disorganizes the egg contents.
A great deal of confusion exists regarding the larval forms of the Mymaridae. This is primarily due to the exceedingly minute size of the early stages combined with the lack, in the instars following the first, of heavily sclerotized or indurated structures of fixed form. Often several species have been involved in the descriptions of the instars of what was supposedly a single species. This was the case in the account given by Ayers (1884) of Teleas sp., a scelionid parasite in the eggs of Oecanthus in the United States. His figures 2, 3, 8, 12, and 13, of plate 24, are undoubtedly of Polynema (Fig. 55A), which is stated by E. W. Wheeler (1923) to be P. bifasciatipenna Gir. The Polynema sp. figured by Ganin is stated by Bakkendorf to be Anagrus, probably A. incarnatus Hal., though Henriksen believed it to be A. subfuscus Foerst.
There are two general types of first instar larvae. The first of these is oblong or flask shaped, of the sacciform type, and occurs in the several species of Anagrus that have been studied and in Prestwichia aquatica. The body is merely a bag, without segmentation, and lacks any distinguishing characters. Bakkendorf (1934) figures a somewhat intermediate form between this and the next, in what are considered to be first instar larvae of Allaptus minimus Hal. and Lymnaenon effusi Bakk. (Fig. 55B). These are rather spindle shaped, with indications of segmentation, and have the caudal segment attenuated and tapering to a point. There is no distinct tail structure, nor are there spines or setae.
The second and more common form of the first instar larva is designated as mymariform and occurs in the described species of Polynema, Araphes, Ooctonus and Anaphoidea. The head is large, drawn out into a curved median conical process, with a second smaller process beneath it, representing the mouth. The thorax and abdomen consist of six to eight segments, often indistinguishable, with transverse rings of long hairs, which are most numerous upon the dorsum. The dorsal spines of P. euchariformis (Fig. 56A) and Anaphoidea nitens are exceptionally long and heavy. The caudal end of the body bears a long, curved or abruptly bent process, often equal to the body in length, which, in some species, is compressed laterally into blade like form and bears a single large tooth or one or more smaller paired teeth on the ventral margin. Clark (1931) described two distinct mymariform instars in A. nitens; the first of these has the tail slender and bent at right angles twice, first ventrally and then dorsally, whereas in the second form it is constricted and toothed at several points (Fig. 55C). Clausen (1940) believed this to be highly improbable, and the two forms doubtless represent either two species or different ages of the first instar of the same species. In size the first instar larvae of the Mymaridae are small and range from 0.1 to 0.3 mm. in total length at the time of hatching.
The number of larval instars following the first is very uncertain. Several authors asserted that there are only the first and the mature forms, though in most species there are said to be three. Balduf described four in Polynema striaticorne.
The second instar larva of Anagrus is of distinctive form and has been designated as “histriobdellid” by Ganin, who first observed and described it. This larva (Fig. 56C) is cylindrical in form and is divided by constrictions into six segments, of which the first and last are largest. The head bears a pair of large, conical or cylindrical fleshy processes lateroventrally, which are said to be the antennae, and the extruded mandibles, which are long, slender, and curved, lie parallel to each other. The last segment bears a pair of large ear like organs, of unknown function, lateroventrally. This type of larva has thus far been associated only with the sacciform first instar larva and is not known in any species having mymariform larvae.
The second instar larvae of Prestwichia, Anaphoidea and Polynema (Fig. 56B, D) have few digtinguishing characters except for the relatively large extruded mandibles, which are somewhat fleshy. The body is bag like, without segmentation, and lacks appendages, spines, or setae.
Clausen (1940) noted a conspicuous feature, observable in insect eggs that contain intermediate stage and nearly mature mymarid larvae, where the egg contents are in almost constant agitation as a result of the writhing or rotary movements of the parasitoid larvae. The heavier semisolid contents are in this way prevented from settling to the bottom of the egg, and a constant supply of food material is within reach of the mouth. In Anaphoidea nitens, these movements persist for 3-4 days, during which time one end of the egg becomes filled with a thick yellowish material, presumed to be an excretion of the parasitoid larva. In some chalcidoid and serphoid egg parasitoids, the eggs containing developing parasitoid larvae are also recognizable by a distinctive coloration. This is true of at least some species of Mymaridae, and in several species of Anagrus the host eggs containing advanced larvae and pupae are bright red or yellow. The body color of larvae and the remaining portions of the egg fluids are visible through the delicate chorion. Parasitism of Polynema striaticorne in the eggs of Ceresa results in the deposition of a black pigment in the vitelline membrane, which is a reaction more generally associated with Trichogramma parasitism (Clausen 1940/1962).
The mature larvae of Anagrus and Paranagrus are similar to the histriobdellid form except that the ear like processes of the last segment are lacking and the mandibles and antennal processes are much reduced in size. Those of Polynema (Fig. 54C), Prestwichia, Anaphes and Anaphoidia are indistinctly segmented, and, aside from the large extruded mandibles, have no recognizable characters. The larva of Erythmelus goochi Enock appears to be intermediate in form between Anagrus and the above genera.
In no species is there any indication of an internal tracheal system or spiracles in any of the larval instars.Information courtesy of www.faculty.ucr.edu