The Braconidae constitute one of the most species-rich families of insects. Although tropical faunas are still relatively poorly understood at the species level, most taxonomists in this group would agree that a rough, probably highly conservative, estimate of 40-50,000 species worldwide is reasonable as an extrapolation from the current described number of roughly 12,000 species. Among extant groups, the sister group of the Braconidae is the Ichneumonidae, an equally enormous group (Sharkey and Wahl, 1992; Quicke et al. 1999).

The family appears to date from early Cretaceous (assuming Eobracon is properly assigned to family – Rasnitsyn, 1983; Whitfield, 2002), diversifying extensively in the mid to late Cretaceous and early Tertiary, when flowering plants and their associated holometabolous herbivores, the main hosts for braconid parasitoids, radiated (Basibuyuk et al., 1999; Quicke et al., 1999; Belshaw et al., 2000; Whitfield, 2002). The species richness of the family is matched by a morphological diversity virtually unrivalled among the Hymenoptera. They range in size from approximately 1 mm in length to 3-4 cm (not counting the ovipositor, which in some species can be several times as long as the body).

Some of the groups are parts of extensive Müllerian mimicry complexes (Quicke, 1986), and exhibit striking color patterns (some of which are recurrent within regions), while others are among the most inconspicuous of Hymenoptera. The braconids display a bewildering array of wing venation patterns and body forms to stymie the beginning student. Female external genitalia (ovipositor mechanisms) vary considerably intraspecifically and are widely used for species discrimination and identification, while the male genital capsules tend to be somewhat more conservative and have been underutilized relative to other insect groups.

The vast majority of braconids are primary parasitoids of other insects, especially upon the larval stages of Coleoptera, Diptera and Lepidoptera but also including some hemimetabolus insects (aphids, Heteroptera, Embiidina). As parasitoids they almost invariably kill their hosts, although a few only cause their hosts to become sterile and less active. Both external and internal parasitoids are common in the family, and the latter forms often display elaborate physiological adaptations for enhancement of larval survival within host insects, including the co-option of endosymbiotic viruses for compromising host immune defenses (Stoltz and Vinson, 1979; Stoltz, 1986; Whitfield, 1990; Beckage, 1993, Stoltz and Whitfield, 1992; Whitfield, 2002; Whitfield and Asgari, 2003).

Early larval development in braconids has also yielded surprises, such as the discovery of relatively closely related genera that differ in such import aspects as syncitial versus holoblastic cleavage, normally characterizing major animal phyla (Grbic and Strand, 1998; Grbic, 2000)! Parasitism of adult insects (especially of Hemiptera and Coleoptera) is also known, and members of two subfamilies (Mesostoinae and Doryctinae) form galls on plants (Infante et al., 1995; Austin and Dangerfield, 1998). Several excellent general reviews of braconid biology are available (Matthews, 1974; Shaw and Huddleston, 1991; Shaw, 1995; Wharton, 1993a).

Description & Statistics

Braconidae are a large cosmopolitan family with circa 10,125 described species world wide as of 1993. Important morphological characters are antennae with 17 or more segments; 1st M-2 cell absent; costal cell absent; basal part of media complete, dividing area behind stigma into two cells. The body is elongate and slender. The ovipositor is usually not longer than the overall body length. Most braconids are primary endoparasitoids of Lepidoptera larvae, although most holometabolous groups may be attacked, e.g., Diptera, Coleoptera and other Hymenoptera. Some species attack spiders, and some are hyperparasitic. There are both solitary and gregarious species in the family, and some are ectoparasitic. Braconids have been used extensively in biological control with much success.

This is one of the major groups of insect parasitoids that includes a large number of species that are effective enough to exert a definite regulatory impact on the increase of numerous important plant pests. Hyperparasitism is less developed than in the Ichneumonidae and is of rare occurrence. Braconids are thus almost wholly beneficial. Principal exceptions are species of Perilitus (Dynocampus) that parasitize adults of entomophagous Coccinellidae (Clausen 1940/1962).

Sharkey (1993) noted that in the Braconidae, vein 2m-cu of the forewing is absent, except in specimens of Apozyx penyai Mason from Chile (present in 95% of Ichneumonidae). Vein 1/Rs+M of the forewing is present ca. 85% of the time (absent in all Ichneumonidae). Vein 1r-m of the hind wing is usually (95%) basal to separation of R1 and Rs (this is opposite or apical in Ichneumonidae). Metasomal tergum 2 is fused with 3, though it is secondarily flexible in Aphidiinae (90% of Ichneumonidae have a flexible suture).

Braconidae is the second largest family of Hymenoptera, with over 40,000 species (Sharkey 1993). The family is cosmopolitan and diverse in all areas, with no strong preference for tropical or temperate regions or for wet or dry habitats. Van Achterberg (1976) gave a summary of the taxonomic history of the family. Shaw & Huddleston (1991) discussed the family’s classification and biology.

Braconidae show a variety of biologies. Hosts are usually the larvae of Holometabola, although nymphs of Hemimetabola and adults of both Holometabola and Hemimetabola are also parasitized. Two major lineages occur within this family, the cyclostome and non-cyclostome braconids. Most species are endoparasitic koinobionts, although a large number are idiobiont ectoparasitoids. idiobionts generally paralyze their hosts, lay an egg on or near the host, and begin consuming it immediately after the egg hatches. Most idiobionts are ectoparasitoids (Sharkey 1993). Koinobionts usually do not paralyze their prey, and typically an egg is laid inside the host. The egg hatches immediately but undergoes a quiescent period while the host grows to an appropriate size and stage. Koinobionts usually exercise some control over the development of their hosts (Vinson & Iwantsch 1980), and because they are closely associated with the life cycles of their hosts they have limited host ranges. On the other hand, idiobionts are usually not closely synchronized with their hosts, and host ranges are generally quite large (Askew & Shaw 1986). Ectoparasitism and the idiobiont development are ground-plan attributes of Braconidae (Sharkey 1993). Nevertheless, both endoparasitism and koinobiosis appear to have developed a few times within the family.

There is no consensus on the number of braconid subfamilies, but Sharkey (1993) proposed 29 as follows: Adeliinae, Agathidinae, Alysiinae, Amicrocentrinae, Aphidiinae, Apozyginae, Braconinae, Cardiochilinae, Cheloninae, Doryctinae, Dirrhopinae, Euphorinae, Gnamptodontinae, Helconinae, Homolobinae, Ichneutinae, Khoikhoiinae, Macrocentrinae, Meteoridiinae, Meteorinae, Microgastrinae, Miracinae, Neoneurinae, Opiinae, Orgilinae, Rogadinae, Sigalphinae, Trachypetinae and Xiphozelinae.

Key references are Shenefelt 91965) who published a bibliography on Braconidae, Shenefelt (1969, 1970a, 1970b, 1972, 1973a, 1973b, 1974, 1975, 1978), Shenefelt & Marsh 91976), Fischer (1971), Mackauer & Stary (1967) and Mackauer (1968) catalogued the described species of Braconidae. Shenefelt (1980) indexed his catalogs. Tobias, Belokobylskij & Kotenko (1986) and Tobias, Yakimavichus & Kirijak (1986), building on the historic works of Telenga (1936, 1941, 1955) and Tobias (1975), keyed all the described species of Braconidae of European USSR. Marsh (1979) cataloged the described species of Braconidae of North America. marsh et al. (1987) keyed the genera of Braconidae occurring in North America. Van Achterberg (1990) gave a key to the Holarctic subfamilies and (1984a) discussed their phylogenetic relationships. „apek (1970) investigated larval braconids and discussed the phylogeny of Braconidae in light of larval morphology. „apek (1973) gave a key to the larvae of braconid subfamilies. Huddleston (1988) discussed Braconidae of Great Britain at the subfamily level. Shaw & Huddleston (1991) provided a key to the subfamilies of Britain and discussed in great detail their biological attributes.

Sharkey (1993) discussed the subfamilies of Braconidae by first dividing the family into its two major groups, noncyclostome and cyclostome.

Non‑cyclostome Braconidae

Labrum not concave, usually sculptured, and often concealed beneath mandibles; vein m‑cu of hind wing absent, except for a very small stub in specimens of Dirrhope and Acampsis; spiracle of metasomal tergum 2 usually (99%) on laterotergite; metasoma not bent between segments 2 and 3; species not parasitic on Aphidae (Homoptera).
All are endoparasitoids, most are koinobionts, though secondarily some have reverted to the idiobiont mode of parasitism. Female venom glands usually have the musculature reduced, suggesting little venom is used. This seems probable in light of the fact that most hosts are parasitized when they are small compared with the parasitoid.
Interestingly, members of Ichneutinae have heavily muscled venom glands, and they parasitize the egg stage of their host. When this occurs, the muscles may be necessary to counter the pressure in the egg.

The most common hosts are Lepidoptera larvae, followed by Coleoptera larvae. Other hosts include Hemiptera nymphs and adults; Orthoptera and Psocoptera nymphs; adult Hymenoptera, Coleoptera and Chrysopidae (Neuroptera); and Symphyta larvae. Typically, an early instar of the host is parasitized, the parasitoid egg hatches, and its development is arrested until the host is nearly full grown, at which time the parasitoid rapidly develops and consumes the host. The parasitoid often oviposits into a host egg and delays development as described above. The development of the host is often regulated so that the parasitoid emerges in synchrony with the next generation of hosts. Members of Euphorinae and several other smaller groups oviposit into adult or nymphal insects, and some are idiobionts.

Adeliinae

Have vein Rs of forewing not tubular to wing margin; forewing without vein r; hind wing with vein A not tubular; transscutal suture present; metasomal terga 1-3 fused but not forming carapace over remaining terga. They are solitary endoparasitoids of leafmining Lepidoptera larvae. Because of their close phylogenetic relationship with Cheloninae, members are possibly egg-larval koinobiont parasitoids.

Distribution is worldwide, although not recorded in the Neotropical or Oriental regions. However, several species are in the Canadian National Collection from South America (Bolivia, Ecuador), and undoubtedly members will be found in the Oriental region as well; two genera.

Agathidinae (including Mesocoelus)

Have the forewing with cell lRs small or absent, with last abscissa of vein Rs close to stigma such that cell 2R1 narrow, and with wing fold between prestigma and vein 1/Rs; gena and mouthparts sometimes (25~o) elongate; occipital carina absent.

They are solitary koinobiont endoparasitoids of Lepidoptera larvae. Most are diurnal, but about 1O% to 20% are nocturnal. The last instar parasitoid larva leaves the body of the host and consumes theremains externally, except for the head capsule. Agathidines spin a cocoon, which is often inside the cocoon spun by the final instar of the host.
Distribution is worldwide; 54 genera. Sharkey (1986) discussed the rationale for including Mesocoelus in Agathidinae.

Amicrocentrinae

Have their body length more than 8 mm; occipital carina absent; epicnemial carina absent; forewing with cell lRs with five bordering veins; ovipositor about as long as forewing. They are parasitoids of large, stem boring Lepidoptera larvae. The pale yellowish brown color of adults suggests that they are nocturnal.
Distribution includes Madagascar and continental Africa; one genus (Amicrocentrum).

Cardiochilinae

Have the forewing with last abscissa of vein Rs weakly sclerotized and decurved, and with cell lRs present and distinctly wider than long; antenna with more than 16 flagellomeres; metasomal tergum 1 with spiracle on laterotergite; occipital carina absent; transscutal suture inflexible, superficial.
They are solitary koinobiont endoparasitoids of Lepidoptera larvae. Distribution is worldwide; five genera.

Cheloninae

have metasomal terga 1-3 fused, forming a carapace covering remaining terga; forewing with veins present though not always tubular; postpectal carina present anterior to mesocoxa; epicnemial carina absent.
They are solitary egg, larval koinobiont endoparasitoids of Lepidoptera. The egg is laid in the host egg, and development of the parasitoid is arrested at the first instar larval stage until the host larva has prepared a pupation retreat (Shaw and Huddleston 1991). Distribution is worldwide; seven genera.

Dirrhopinae

Have forewing with vein Rs not tubular to wing margin, and with vein r present; metatarsomere 1 with longitudinal comb of closely appressed setae; transscutal articulation groove like; metasoma not greatly constricted anteriorly, not petiolate; metasomal terga 1 – 3 not forming a carapace.
They are solitary endoparasitoids of leaf mining Lepidoptera larvae. Because of their close phylogenetic relationship with Cheloninae they may be egg and larval koinobionts. Described species are from the Palaearctic and Nearctic regions; undescribed specimens from South Africa and the Solomon Islands (Australian region) are in the Canadian National Collection; more collecting will likely extend the distribution to other regions; one genus (Dirrhope).

Euphorinae

(including Centistini, and Ecnomios) is a rather diverse and likely polyphyletic assemblage that is difficult to diagnose; metasoma usually (85%) petiolate; forewing with cell 2Cu open apically and with vein r – m usually (85%) absent; maxillary palpus usually (85%) 5 segmented 
They are solitary, rarely gregarious, usually koinobiont endoparasitoids of several different orders of insects including the following, in order of frequency: Coleoptera, Hemiptera, Neuroptera, Psocoptera, Orthoptera and Hymenoptera. Adult and nymphal stages are usually parasitized, although the larvae of Coleoptera are parasitized by a few members. Distribution is worldwide; 63 genera.

Helconinae

(comprising Blacini (including Dyscoletes, Brachistini, Brulleiini, Cenocoeliini, Diospilini and Helconini). Brulleiini, Cenocoeliini, Diospilini and Helconini have the occipital carina present; forewing with vein r‑m present and with cell lRs quadrate or pentagonal; metasomal tergum 1 usually (95%) rugose and remaining terga smooth; metasomal tergum 1 not greatly narrowed anteriorly. Blacini: occipital carina present; forewing with vein r‑m usually (99%) absent (except Dyscoletes), and with vein 2cu‑a usually (99%) absent (except members of Blacozona and Stegnocella); metasomal tergum 1 with dorsolateral pits.
Brachistini: occipital carina present; forewing without vein r‑m and with vein 2cu‑a usually (95%) present; metasomal tergum 1 with dorsolateral pits very weak or absent.

They are mostly koinobiont endoparasitoids of Coleoptera larvae. Male mating swarms have been observed in several species of Blacus (Southwood 1957, Haeselbarth 1973). Members of Dyscoletes (Blacini) are parasitic on Boreus larvae (Mecoptera) (Mason 1976). Distribution is worldwide; about 50 genera.

Homolobinae

(comprising Homolobini and Microtypini). Homolobini: forewing with vein r‑m present but not tubular; body usually (98%) yellowish brown; metasomal tergum 1 long, almost parallel‑sided, and usually widest at the spiracles. Microtypini: forewing and cell lRs triangular, vein A with an a’ cross vein; occipital carinae and subpronope present.
They are solitary koinobiont endoparasitoids of Lepidoptera larvae. Most (90%) species are nocturnal. Distribution is worldwide; four genera.

Ichneutinae

have the dorsal longitudinal carina of propleuron usually (95%) absent; occipital carina absent; epicnemial carina often (50%) absent; metasomal terga 1‑3 not forming a carapace.

Members of the tribe Ichneutini and Proteropini are solitary koinobiont larval parasitoids of larval Symphyta; those of Muesebeckiini parasitize leaf mining Lepidoptera. The egg or first instar larva of the host is parasitized, and development is delayed until the host larva has spun its cocoon. Distribution is worldwide; nine genera.

Khoikhoiinae

have metasomal tergum 1 with spiracle on laterotergite; transscutal articulation functional, groove like, not superficially impressed; forewing with apical abscissa of vein Rs spectral or nebulous; epicnemial carina absent.
Their biology is unknown, but based on their close relationship to Cardiochilinae, members of Khoikhoiinae are likely solitary koinobiont endoparasitoids of Lepidoptera larvae. They are found in South Africa; two genera.

Macrocentrinae

(including Chamton). Charmonini: forewing without vein r‑m; occipital carina reduced dorsally; hind wing with vein 2a’. Macrocentrini: metatrochantellus with spines; occipital carina absent; body usually (95%) yellowish brown; sclerotized bridge present between metacoxal cavities and propodeal foramen.
They are solitary or gregarious endoparasitoids of early to late instar Lepidoptera larvae. Some gregarious species of Macrocentrus are known to be polyembryonic; many species are nocturnal. Distribution is worldwide; eight genera.

Meteoridiinae

have the hind wing with vein 2/Cu; forewing with cell lRs quadrate and with vein 2cu‑a present.
They are larval‑pupal endoparasitoids of Lepidoptera. Eggs are laid into the host larvae, but adults emerge from the host pupae. Distribution is worldwide; two genera.

Meteorinae

(including Zele) have metasomal tergum 1 usually (95%) much narrower anteriorly than posteriorly; forewing with vein r‑m present and with cell lRs quadrate; metacoxal cavities not separated from propodeal foramen by sclerotized bridge; metasomal tergum 1 often (80%) with deep dorsal pits. Some researchers consider the two included genera to be members of Euphorinae and others treat them as separate monotypic subfamilies.
They are solitary or gregarious koinobiont endoparasitoids of Coleoptera or Lepidoptera larvae. Some species of Meteorus that parasitize Lepidoptera larvae suspend their cocoons from a line of silk resembling a meteor, hence the generic name. Many species of Meteorus and Zele are nocturnal. Some species are idiobionts. Distribution is worldwide; two genera.

Microgastrinae

have forewing with last abscissa of vein Rs not tubular; metasomal tergum 1 with spiracle on laterotergite; occipital carina absent; antenna with 16 flagellomeres (because of a median constriction in each flagellomere, the flagellum may appear to have 32 articles); apical (ventral) margin of clypeus concave.
They are solitary or gregarious koinobiont endoparasitoids of Lepidoptera larvae (see Shaw and Huddleston, 1990, for a detailed summary). Usually koinobionts, sometimes egg and larval parasitoids. Distribution is worldwide; 52 genera (largest braconid subfamily in terms of number of species).

Miracinae

have the forewing with last abscissa of vein Rs not sclerotized and without vein r‑m; metasomal tergum 1 with spiracle on membranous laterotergite; antenna with 12 flagellomeres.
They are solitary endoparasitoids of leaf mining Lepidoptera larvae. Distribution is worldwide; two genera.

Neoneurinae

have the epicnemial carina absent; occipital carina absent; maxillary palpus with 3 segments; labial palpus with 2 segments; metasomal terga with setae scattered over surface and not restricted to posterior transverse row. This subfamily may be a derived lineage of Centistini (Euphorinae). Some attributes that suggest this supposition are parasitization of adult insects, laterally compressed ovipositor, and specialized pit like antennal sensilla.
They are internal parasitoids of adult Formicidae. The ovipositor is greatly curved and is thrust directly into the gaster of the adult ant. Distribution is Holarctic, including northern Africa; two genera.

Orgilinae

have the forewing with vein r‑m usually absent but, if present, then cell 1‑Rs triangular, and vein A lacking anal crossveins; occipital carina usually (90~o) present; no wing fold between vein 1/Rs and stigma; dorsal pit absent; vein 2cu‑a of forewing present. Antestrix, known only from two Chilean species, is not included in this diagnosis.
They are solitary koinobiont endoparasitoids of Lepidoptera larvae. Distribution is worldwide; seven genera.

Sigalphinae

(including Acampsis, Minanga, Neoacampsis, Pselaphanus and Sigalphus) have the occipital carina absent dorsally; forewing with free apical abscissa of vein Cu present and with vein r‑m present; metacoxal cavities open to propodeal foramen and not separated by sclerotized bridge; metasomal tergum 1 usually (95~O) with pair of percurrent longitudinal carinae.
They are solitary endoparasitoids of Lepidoptera larvae. Distribution is worldwide; five genera.

Trachypetinae

have their forewing with cell lRs pentagonal; ovipositor barely exerted; antennal flagellomere with more than 50 articles.
Biology is unknown. Sharkey (1993) suspects, based on presumed phylogenetic affinites, that members are endoparasitoids of Lepidoptera larvae. More than half of the known specimens have been collected at light in desert habitats (A.D. Austin, pers. commun.). They are found only in the Australian region; three genera.

Xiphozelinae

have the occipital carina absent; sclerotized bridge between metacoxal cavities and propodeal foramen; hind wing with cell lCu much longer on posterior margin than on anterior margin (vein lA much longer than vein M+Cu); metasomal segment 1 with median tergite more than five times as long as apical width.
They are solitary nocturnal endoparasitoids of Noctuidae (Lepidoptera) larvae. Distribution is Palaearctic, Oriental, and Australian regions; two genera.

Cyclostome Braconidae

The labrum is usually (70%) concave, smooth, and often (80%) mostly glabrous (many Doryctinae have microsculpture and many setae on the labrum, but in those forms the labrum is distinctly concave); many members without these attributes have exodont mandibles; hind wing with vein m‑cu often (50%) present; metasomal tergum 2 with spiracle usually (90%) on median tergite. Members of Aphidiinae, which are here associated with the cyclostomes for the first time, do not share some of the more obvious diagnostic attributes of the other cyclostomes. They are best diagnosed by the following: joint between metasomal terga 2 and 3 flexible; metasoma often (50%) bent ventrally between segments 2 and 3; hind wing with vein lA and vein cu‑a absent or spectral, never sclerotized, and with long sensory setae near junction of veins R and lr‑m; parasitoids of Aphidae (Homoptera).

Most are idiobiont ectoparasitoids of Lepidoptera larvae and Coleoptera, although many members are endoparasitoids of Diptera, Aphidae (Homoptera) and Lepidoptera, and many of these are koinobionts. Isoptera and Embioptera are also known hosts and one genus may be phytophagous (Marsh 1991). Most females have well muscled venom glands and use venom to subdue their prey and paralyze it, at least temporarily. Development of the parasitoid usually begins immediately, with little or no visible effect on the development of the host, which is quickly consumed.

Alysiinae

have their mandibles exodont, not touching when closed; epicnemial carina absent; occipital carina absent; hind wing with vein 2m‑cu often (50%) present. Alysiinae is a derived lineage of the paraphyletic subfamily Opiinae (Buckingham and Sharkey 1988).
Most (90%) are solitary koinobionts and all are endoparasitic on Cyclorrhapha (Diptera) larvae. Distribution is worldwide, but much more speciose in temperate regions; 65 genera.

Aphidiinae

have their antenna usually (80%) curved ventrally in dead specimens; flexible joint between metasomal terga 2 and 3 (in dead specimens the metasoma is often (50%) bent at this point); hind wing with veins lA and cu‑a absent or not tubular; parasitic on Aphidae (Homoptera). Aphidiinae is placed with the idiobiont (cyclostome) Braconidae because of the presence of the following apomorphies: hind wing with anterior margin excavated basally and with long sensory setae present near junction of veins R and r‑m; metasomal terga weakly sclerotized, labrum smooth, triangular, and mostly glabrous; and host mummified (as in most Rogadinae).
They are solitary koinobiont endoparasitoids of Aphidae (Homoptera) nymphs and adults. Distribution is worldwide, but more speciose in temperate regions; about 51 genera (P. Stary, pers. commun.).

Apozyginae

have the forewing with vein 2m‑cu; labrum glabrous and concave; hind wing with vein 2/Cu. Apozyginae is treated by Sharkey (1993) as a subfamily of Braconidae rather than as a distinct family as originally proposed by Mason (1978, 1987). Sharkey and Wahl (1992) justified this placement.
Their biology is unknown, but judging from a general similarity to some Doryctinae members are possibly idiobiont ectoparasitoids of xylophagous Coleoptera larvae. They are found in Chile; one genus with one species, Apozyx penai Mason.

Braconinae

(including Vapellina) have the labrum concave; occipital carina absent; epicnemial carina absent; hind wing with vein 1/M at least twice as long as M+Cu.
Most species are idiobiont ectoparasitoids of concealed larvae of xylophagous and stem boring Coleoptera and Lepidoptera larvae, and rarely of Diptera and Symphyta. Several genera are gregarious endoparasitoids of Lepidoptera pupae (van Achterberg 1984b, Quicke 1987a). Distribution is worldwide; 151 genera.

Doryctinae

(including Histeromerus) have the labrum concave; protarsus usually (99%) with spines along anterior margin; occipital carina present but usually (80~) absent ventrally; epicnemial carina present.
They are mostly solitary idiobiont ectoparasitoids of xylophagous and stem boring Coleoptera larvae, though one genus is known to parasitize Embioptera (Shaw and Edgerly 1985) and another may be phytophagous (Marsh 1991). Distribution is worldwide; 75 genera.

Gnamptodontinae

(including Telengaia) have the labrum concave to flat; forewing with cell 2Cu open (vein lA incomplete); metasomal tergum 2 with smooth, anterior, transverse elevation; propodeum without sculpture. Telengai is similar to most gnamptodontines. Members have a modified metasoma, but wing venation and head and mesosomal structures lead me to believe that Gnamptodontinae, including Telengaia is monophyletic.
These are parasitoids of leaf mining larvae of Nepticulidae (Lepidoptera). It is not known if they are endoparasitoids or ectoparasitoids, but based on their possible sister group relationship with Opiinae plus Alysiinae (Buckingham and Sharkey 1988). They are believed to be endoparasitoids. Distribution is worldwide; four genera.

Opiinae

(including Mesostoa) have the epicnemial carina absent (except Ademon); occipital carina often (85%) absent dorsally but usually (98%) present laterally; occipital carina, when present, usually (98%) meeting subgenal carina, not hypostomal carina; hind wing with vein 2m‑cu often (50%) present; clypeus with ventral margin usually (90%) not concave. Van Achterberg (1975) proposed Mesostoa as a separate monotypic subfamily. It is included here in the Opiinae based on the shared possession of the following: labrum not greatly concave; clypeus straight ventrally; epicnemial carina absent; ovipositor bent dorsally. Opiinae is paraphyletic in that Alysiinae is a derived lineage of this assemblage (Buckingham and Sharkey 1988).
They are solitary endoparasitoids of Cyclorrhapha (Diptera) larvae. As with members of Alysiinae, they often parasitize a late larval instar, but they are also known to parasitize eggs and early instar larvae. Distribution is worldwide; 17 genera; most species are in the large genus Opius, which is divided into about 50 sub~enera.

Rogadinae

(including Exothecini, Hormiini, Lysterimini, Pambolini, Rhyssalini, Rhysipolini, Hydrangiacolini, Rogadini and Ypsistoceratini). Sharkey (1993) commented that this subfamily is certainly not monophyletic. Many of the constituent tribes have been placed in other subfamilies (Doryctinae) or have been treated as independent subfamilies. Because the relationships of these taxa are poorly understood, he adopted a very broad definition of Rogadinae. Most of the tribes comprising the Rogadinae are discussed and diagnosed separately as follows:

Tribe Exothecini

Labrum concave; occipital carina ending ventrally on subgenal carina or absent ventrally but present at least laterally; epicnemial carina absent.These are idiobiont ectoparasitoids of concealed (usually leaf‑mining) Lepidoptera, Diptera, Coleoptera, and Symphyta larvae. Distribution is probably worldwide; five genera.

Tribes Hormiini, Lysterimini and Pambolini

Labrum concave; metasomal tergum 2 with spiracle on median tergite or near margin of median and lateral tergites; occipital carina absent ventrally or meeting hypostomal carina; metasomal tergum 1 without median longitudinal carina, often (80%) with 2 percurrent longitudinal carinae; protibia without pegs or spines; epicnemial carina present. Hormiini: metasomal terga, except first, membranous. Lysterimini: metasomal segments 1‑ 3 with median tergites heavily sclerotized and sculptured and usually covering following terga. Pambolini: propodeum often with posterolateral spine or bump; metasomal terga 2 and 3 not membranous, usually smooth, but if sculptured then not covering following terga.
Members of Hormiini and Lysterimini are usually gregarious ectoparasitoids of concealed Lepidoptera larvae. Members of Pambolini are solitary ectoparasitoids of Coleoptera and Lepidoptera larvae. Distribution is worldwide; about 15 genera.

Tribe Rhyssalini

Labrum concave; metasomal tergum 2 with spiracle on laterotergite, well below margin of median tergite; occipital carina ending ventrally on hypostomal carina; metasomal tergum 1 without median longitudinal carina, or metasoma not coarsely sculptured beyond tergum 1, or both; protibia without spines or pegs on anterior surface.
They are usually gregarious, sometimes solitary, idiobiont ectoparasitoids of Coleoptera and Lepidoptera larvae. Distribution is worldwide (Australian?); five genera.

Tribe Rhysipolini (including Hydrangiacolini)

Labrum concave; occipital carina ending ventrally on subgenal carina; metasomal tergum 1 without median longitudinal carina, or metasoma not coarsely sculptured beyond tergum 1, or both; anterior surface of protibia without pegs and spines.
They are koinobiont ectoparasitoids of Lepidoptera larvae. Distribution is worldwide; seven genera.

Tribe Rogadini (including Betylobracon, Leurinion and Ypsistoceratini)

Occipital carina present; labrum usually (95%) concave; metasomal tergum 1 usually (95%) with sharp median longitudinal carina; metasomal terga 3 and 4 also often (75%) have median longitudinal carina; protibia without pegs and spines on anterior surface.
Leurinion is usually placed in Hormiinae because, as in members of Hormiinae, the median tergites of some metasomal segments are membranous. However, this attribute occurs in several other cyclostome subfamilies including members of Rogadinae, e.g., Aeliodes excavatus (Telenga). Furthermore, the membranous portions of the metasomal terga of Leurinion species are more widespread than the membranous proportions of Hormiinae species and include the posterior part of tergum 1. Members of Leurinion have a sharp median longitudinal carina on the propodeum and metasomal terga 2 and 3. This combination of derived characters is unknown outside Rogadinae

Ypsistoceratini, composed of Ypsistocerus, Terrnitobracon and an undescribed genus from the southeastern USA, are sometimes placed in their own subfamily. They are a derived group morphologically and have lost many attributes that allow for their easy placement in any cyclostome subfamily. They share several derived characters with the rogadine genus Yelicones: expanded apical tarsomeres of all legs and the presence of spines on the anterior surface of the protibia.

Tobias (1979) placed Betylobracon in its own subfamily. However, based on the presence of a strong m‑cu’ vein in the forewing (a derived character within the idiobiont Braconidae) and metasomal spiracles located on the median tergites (a primitive character that excludes it from the koinobiont Braconidae) it appears to belong to the idiobiont lineage. Of the subfamilies in this lineage it appears to be closely related to Yelicones and allies. This is shown by the following: femora swollen, apical tarsomeres swollen and elongate, and ovipositor short (shared by all Rogadini).Betylobracon waterhousei Tobias, the only species described to date, bears a striking resemblance to Yelicones delicatus (Cresson). The principal difference between the two species is that the clypeus of the former is not concave ventrally, but this character state reversal is not uncommon within the idiobiont clade.

These are koinobiont endoparasitoids of Lepidoptera larvae; pupation takes place inside the mummified remains of the host, except for members of Leurinion, which do not mummify their hosts. The biology of members of Ypsistoceratini is unknown, but they have been associated with termite nests (Isoptera). Distribution is worldwide; about 45 genera.

Further Description

Braconidae continue to be of considerable value in biological control. Ischiogonus syagrii Ful. was imported into Hawaii from Australia in 1921 in a successful effort of control against the fern weevil, Syagris fulvitarsis Pasc. Opius fletcheri Silv, a parasitoid of the melon fly, Bactrocera curcurbitae Coq., and O. tryoni Cam., attacking Mediterranean fruit fly, Ceratitis capitata Wied., have been credited with appreciable reduction of these two pests in Hawaii and have allowed the continued growing of certain fruits and vegetables that were previously heavily infested. The reduction in infestation of C. capitata in coffee was particularly important. Apanteles solitarius Ratz and Meteorus versicolor Wesm., both of European origin, have been responsible for adequate control of the satin moth, Stilpnotia salicis L. in some parts of North America (Clausen 1940/1962). Apanteles oenone Nixon and Chelonus sp. nr. curvimaculatus Cameron are considered probable good candidates for importation against the pink bollworm, Pectinophora gossypiella Saunders, attacking cotton (E. F. Legner, unpub. data), as they are associated with low densities of the pest in its endemic range in northwestern Australia.

Host Preferences

Regarding host preferences, because of the large number of species involved and the extensive studies that continue to be made, host preferences will be discussed using the Clausen (1940/1962)format, on the basis of principal subfamilies. There is an exceptional uniformity of habit within these groups, not only in the choice of hosts but in the manner of development (Clausen 1940/1962).

Vipioninae

are principally external parasitoids of the larvae of Lepidoptera, though a considerable number attack coleopterous larvae and a few species are parasitic on those of sawflies and Diptera (Cecidomyiidae). The dominant genus is Microbracon, which is cosmopolitan and attacks a wide range of hosts. Practically all hosts attacked by members of this subfamily are contained in a cell, burrow or cocoon or are protected by a web. Free-living larvae normally are not subject to attack, though an undetermined species of Microbracon parasitizes the uncovered larvae of the teak leaf skeletonizer, Hapalia machaeralis Wlk. in India and M. brevicornis Wesm. attacks the free-living larvae of Heliothis in South Africa (Beeson & Chatterjee (1935). Most species are gregarious in habit, although the number developing on each individual host is small. An occasional species is predaceous rather than parasitic, such as M. lendicivorus Cush. (Williams 1928), which develops at the expense of the larvae of the cecidomyiid, Asphotrophia fici Barnes. These larvae live in the receptacles of the fruit of Ficus nota in the Philippines (Clausen 1940/1962).

Spathiinae and Doryctinae

Few studies have been made on the Spathiinae and Doryctinae, but the species seem to attack principally the larvae of bark and wood-boring Coleoptera. The genera most commonly encountered are Spathius, Doryctes and Dendrosoter, which are externally parasitic.

Rhogadinae

are apparently limited in their host preferences mainly to the larvae of Lepidoptera, on which they develop internally. Rogas is the most common genus of the subfamily. Oncophanes lanceolator Nees differs in habit from others of the subfamily in developing as a gregarious external parasitoid.

Cheloninae

are mainly solitary internal parasitoids of lepidoptera larvae. The genera Ascogaster, represented by the codling moth parasitoid A. quadridentata Wesm. (carpocapsae Vier.), Phanerotoma, and Chelonus, are restricted to lepidopterous hosts and have the habit of ovipositing in the egg and completing their larval development when the host larva is nearly mature. Such behavior makes it difficult to understand how they might be valuable in biological control, since the host continues to cause damage even though parasitized in the larval stage. However, the impact on the population in precluding adult reproduction may be profound, and should not be underestimated over the long term.

Triaspinae

are parasitic in the larvae of Bruchidae and occasionally Curculionidae, but little information is available on the manner of attack and development. Triaspis is frequently reared from weevil-infested beans, peas and other seeds (Clausen 1940/1962).

Neoneurinae

Little information is available on host preferences of the Neoneurinae. Elasmosoma berolinense Ruthe is reported attacking adult ants of Formica fusca var. japonica Motsch (Kariya 1932). The female pounces on the ants at the entrance to the nest and inserts the ovipositor into the abdominal region. Observations on American and European species point also toward a relationship with ants (Clausen 1940/1962).

Microgasterinae

are limited in their host preferences principally to lepidoptera larvae, and many of the hosts are fee-living in habit. Dominant genera that are very common are Apanteles, Microgaster and Microplitis. Development takes place internally, the only apparent exception being A. canarsiae Ashm., which was recorded as a solitary external parasitoid of the larva of Desmia funeralis Hbn. by Strauss (1916). Clausen (1940/1962) considered this record questionable, however, because it may be based on observations made during the very short period of external feeding which occurs in some species between emergence from the host body and spinning of the cocoon.

Braconinae

are most often found as parasitoids of lepidopterous caterpillars, though some species attack trypetid and curculionid larvae. Species of the well known genus Bassus develop internally in lepidopterous larvae of shoot or stem boring habit.

Blacinae

Little is known regarding the host preferences of the Blacinae. Eubadizon and Orgilus are parasitic in lepidopterous larvae, whereas Syrrhizus attacks adult chrysomelid beetles of the genus Diabrotica.

Macrocentrinae

are solitary or gregarious internal parasitoids of lepidopterous larvae. The dominant genus Macrocentrus is well studied by virtue of its attack of the European corn borer and oriental fruit moth. Initial attack is on the young larva in its burrow, and parasitoid development is completed when the host larva is full grown.

Opiinae

Information on the Opiinae is principally from the genus Opius, many species of which are parasitic in dipterous larvae of the family Trypetidae, though a number of species have been reared from Agromyzidae. Oviposition is in the maggot in almost any stage of development, and adult emergence is from the host puparium.

Euphorinae

are distinctive in that the majority of species are internal parasitoids of adult Coleoptera. The well known genus Perilitus (Dinocampus) is confined in its host preferences to Coccinellidae and Curculionidae, while Microctonus attacks mainly the Curculionidae, Chrysomelidae and Tenebrionidae. Perilitus is solitary but in some species of Microctonus a large number may develop in a single host. Aridelus attacks nymphs and adults of Pentatomidae and Euphorus and Euphoriana are parasitic in Miridae (Muesebeck 1936). Exceptions to the above generalization are that M. aethiops Nees has been reared from adults and larvae of Phyllotreta and Sitona, and M. brevicollis Hal. has its first generation in the larvae of Haltica ampelophaga Guer. and the second generation in adult beetles (Kunckel D’Herculais & Langlois 1891).

Meteorinae

are solitary or gregarious internal parasitoids of the larvae of many Lepidoptera and have also been recorded from those of bark and wood-boring Coleoptera. Extensive studies exist on the habits of a number of species of Meteorus, the single genus of the subfamily. Oviposition occurs in young host larvae while they are exposed, and development is completed before the pupal stage is attained. Some species produce two complete generations upon one brood of host larvae, the first being upon very young individuals and the second upon those which are nearly mature (Clausen 1940/1962).

Aphidiinae

are very consistent in their host preferences, and all species that have been studied are solitary internal parasitoids of Aphididae. They can be very effective in reducing infestations of their hosts, even though this usually takes place after the latter have reached a high population density and considerable crop injury has already occurred. Clausen (1940/1962) believed that this was due to the ability of the aphids to reproduce at lower temperatures than the parasitoids, and the pest population attains a relatively high abundance before conditions for parasitoid development become favorable. Common genera are Aphidius, Lysiphlebus, Praon and Ephedrus.

Alysiinae

There have been no extensive studies on Alysiinae, but common members are the genera Alysia and Dacnusa, internal parasitoids of Diptera. Alysia is more particularly a parasitoid of blowflies of the genera Sarcophaga, Lucilia and Calliphora, and the adult emerges from the puparium. Extended studies of A. manducator Panz., have been made by Graham-Smith (1916, 1919), Altson (1920), Myers (1927a), Salt (1932) and Holdaway & Smith (1932). Also some species of Dacnusa attack the larvae and emerge from the puparia, principal hosts being Agromyzidae.

Biology and Behavior

The position of the host larvae with respect to the plant may influence parasitoid attack (Cushman 1926a). A parasitoid species may attack two or more hosts that are widely separated taxonomically but have the same relationship to the plant, whereas closely related species, that are of different habits, are not subject to attack. Cushman concluded that the relationship of the host insects to the plant, rather than taxonomic relationships, often governs the choice of hosts by the parasitoid. This applies not only to many Braconidae but to various other parasitic groups and even to some predators as is discussed under the section on “Host Selection.” Other early authors recognized the close association of a parasitoid species with a particular plant species or group, which limits their attack to insects infesting those plants. Picard (1919) in studies of insects associated with fig trees, concluded by Sycosoter lavagnei P. & L. is primarily attracted to fig trees rather than to the particular coleopterous species which it parasitizes. Taylor (1932) reported that in South Africa Microbracon brevicornis attacks larvae of Heliothis armigera Hbn. on Anthirrhinum only, even though caterpillars are present on an array of other plants.

The host food plant often has a major influence on the extent of parasitization, which is demonstrated by the case of Apanteles congregatus, parasitoid of Phlegothontius larvae in North America. When larvae feed on wild Solanaceae, parasitization can be very high. Morgan (1910) noted that larvae occurring on tobacco are seldom parasitized, and he thought that parasitized individuals found upon this plant may have moved to it from native vegetation. The difference in degree of successful attack is believed due probably to the toxic effect of the caterpillar’s food on the early stages of the parasitoid (Clausen 1940/1962). Additional observations on Apanteles parasitization of Phlegothontius were made by Gilmore (1938), who found that lack of effective parasitization was most pronounced when the host occurred on dark-fired tobacco. Larvae feeding only on such foliage were frequently parasitized but the Apanteles larvae were unable to develop to maturity. The nicotine content of this tobacco is high, and it was thought that this toxic character may be present in sufficient concentration in the host blood to bring about the death of the parasitic larvae.

The quantity and quality of parasitoid or predator progeny on different hosts seem to vary in the insectary according to evolutionary contact, as mentioned by Legner & Thompson (1977). Their study compared the suitability of the potato tuberworm and the pink bollworm, the original source host, as hosts for a braconid, Chelonus sp. nr. curvimaculatus Cameron. It was found that after being reared for many generations on the potato tuberworm, and then for one generation on pink bollworm, the parasitoid was stimulated to increase its destruction of and fecundity on the factitious host. This group of Chelonus parasitoids responds to kairomones in the body scales of several lepidopterans (Chiri & Legner 1986), and might be characterized as generalists.

Regarding biology and behavior, parasitism by braconids may be either external or internal, and modifications in habit are correlated with habits of the host stages that are subjected to the attack. Generally, internal parasitism occurs if the hosts are free-living as in the case of adult beetles attacked by the Euphorinae and other groups, the foliage-feeding lepidopterous larvae by Microgasterinae, Meteorinae, etc., and the aphids that serve as hosts of the Aphidiinae (Clausen 1940/1962). However, external parasitism is general in hosts that live in confined quarters and thus the Vipioninae, attacking principally caterpillars in tunnels, leaf rolls, etc., and the groups that occur on the larvae of bark and wood-boring Coleoptera, develop externally. The latter groups complete development on the host instar upon which the egg was laid. The Cheloninae, Macrocentrinae and Triaspinae seem to be exceptions to these generalizations, for they are internal parasitoids of hosts occurring in burrows or cavities in plant stems, fruits and seeds. But these parasitoids oviposit either in the egg or the young larva attaining larval maturity in the fully grown host, so that the internal habit is necessary (Clausen 1940/1962).

Referring to adult habits, adults of most species are believed to feed mainly on honeydew and various plant exudations; but females of a large number subsist almost entirely upon the body fluids of the host stages that they attack. Such habits occur generally in the subfamily Vipioninae and occasionally in the Microgasterinae. It is well developed in the genus Microbracon and has been noted in almost all species that have been studied. The host feeding habit was first observed by Trouvelot (1921) in M. gelechiae Ashm. (johannseni Vier.); and he and Genieys (1925) in the case of M. brevicornis Wesm., described the habit in some detail and gave an account of the formation of the feeding tube under the conditions that prevent normal direct feeding. Genieys stated that ordinary laboratory food materials do not fulfill the nutritional requirements of the parasitoid female and that feeding on the body fluids of the host larva was essential before oviposition could take place. Chelonus shoshoeanorum Vier., was found to feed on the fluids that exude from the puncture in the egg of the host. How the embryo survives and the hatched larva fairs after such an attack is worthy of investigation.

A large number of species are able to begin oviposition on the day of emergence of the female from the cocoon, although there is no uniformity here even within genera. Opius fulvicornis Thoms. is ready for immediate oviposition, while O. melleus Gahan requires a gestation period of circa 13 days. Apanteles melanoscelus ratz. (Crossman 1922) and Ascogaster quadridentata are able to oviposit soon after emergence. Some species of Microbracon seem to require as much as 15 days. Genieys noted an unusual condition in M. brevicornis, where unmated females began oviposition 4-5 days after emergence, but mated females required 14-18 days. Some researchers defined the gestation period as the time elapsing between mating and first oviposition but, with the exception of the above instance, this has no bearing on the time at which the first eggs are laid as oogenesis proceeds whether or not mating has taken place. In extended studies on the biology and habits of M. hebetor (= M. brevicornis), Hase (1922) observed that females can be successfully mated and produce female progeny even after 40 days, during which time male progeny are produced at the normal rate. Oviposition is inhibited below 15°C (Clausen 1940/1962).

External braconid parasitoids usually attack host larvae that are half grown or larger which are contained in cells, leaf rolls, or burrows or are beneath a web or other covering. Females of some species of Microbracon penetrate host burrows, attacking larvae directly. Bracon glaphyrus Mshll. burrows in the soil in search of Baris larvae, and Bracon sp. nr. hylobii, scrapes away the frass at the entrance of the burrow of Hylobius, turns about, and inserts the ovipositor into the burrow. The female of Cardiochiles nigriceps Vier. straddles the young host larva and inserts the ovipositor by a downward thrust, while most other species of internal parasitoids oviposit by a forward thrust of the ovipositor between the legs. In Apanteles, several earlier researchers insisted that oviposition was in the host egg, but this has not been proved (Clausen 1940/1962). A. militaris Walsh inserts the ovipositor in the body of the caterpillar and then folds its legs and retains its hold only by the ovipositor. A. glomeratus L. oviposits by preference in the newly hatched cabbage worms (Picard 1922). In the attack by A. machaeralis Wlkn. of young caterpillars of Hapalia, the latter drop from the leaves and suspend themselves by a silken threat, whereupon the parasitoid quickly descends the thread and oviposits. Meteorus hypophloei Cush. attacks bark beetle larvae only when they are crawling about on the surface of the bark. Among the Euphorinae, the species of Perilitus usually oviposit in the host beetles by inserting the ovipositor through the intersegmental membrane in the abdomen. Several researchers believed that P. coccinellae oviposits in larvae and pupae as well as in the adult beetles, which has not been confirmed (Clausen 1940/1962). Microctonus melanopusRuthe was noted by Speyer (1925) to insert the ovipositor through the anal opening of the host or the membrane nearby. Females of Euphorus helopeltidis Ferr. jump on the back of the mirid nymph, curve their abdomen beneath the body and make the insertion in an abdominal suture or at the base of a coxa (Menzel 1926, 1929). Females of Syrrhizus diabroticae Gahan, which oviposit in adult Diabrotica beetles, mount the back of the host, and the ovipositor is inserted dorsally at the base of the elytra. Cosmophorus henscheli Ruschka has a different habit, where the female attacks the host beetle head to head, grasps the back of the thorax between the mandibles and then brings the ovipositor forward and inserts it in one of the thoracic sutures.

Among endoparasitic and gregarious species, the full complement of eggs is usually deposited at one insertion of the ovipositor, and often very quickly. Therefore, the female of Apanteles militaris deposits up to 72 eggs at one insertion in less than one second (Tower 1915). However, Apanteles sagax Wlkn which develops in considerable numbers in the caterpillars of Sylepta derogata F., deposits eggs singly at some distance from each other (Wilkinson 1937).

Host larvae attacked by ectophagous braconids are usually permanently and completely paralyzed. Once exception is Microbracon pini Mues., which inflicts only temporary paralysis, lasting ca. 1-hr, upon the larva of Pissodes strobi Peck. Bracon sp. attacking the larvae of Hylobius abietis L. does not sting the host, which are contained in burrows where they do not move very much during normal feeding (Munro 1917). Donohoe (Clausen 1940/1962) found that some of the larvae of Ephestia figulilella Greg which have been stung by M. hebetor Say are able to effect complete recovery. In the first 15 days of parasitoid activity, 12% of the hosts recovered, while 29% of those attacked during the third 15-day period recovered, showing a progressive lessening in effectiveness of the sting (Clausen 1940/1962). Internal parasitoids seldom paralyze their hosts although larvae of bark beetles attacked by Coeloides dendroctoni Cush become inactive and appear paralyzed within two days after attack and are dead on the third day (De Leon 1935b). Larval hosts of C. pissoidis are permanently paralyzed and sometimes killed by the sting. Adult scolytid beetles attacked by Cosmophorus henscheli are paralyzed for one hour or more but recover completely. Dipterous larvae parasitized by Alysia manducator are paralyzed for a period of 1-2 min, and recovery is accompanied by pronounced writhing movements. Repetition of attack upon the same individual results in its death in one or two days, evidently from an excess of poison.

Among the groups that oviposit in the host egg, such as Ascogaster, Chelonus and Phanerotoma, the stage of development of the egg at the time of parasitoid oviposition has a notable influence on the extent to which attack is successful. In A. quadridentata (= carpocapsae Vier.), Cox (1932) found that the female will oviposit in codling moth eggs in any stage of development, but parasitization is not successful if the “black-spot” stage has been reached. Rosenberg (1934) secured successful parasitization even in that stage, however. The egg is placed at random in the cytoplasm, but in such a position that it is always outside the host embryo when the latter becomes fully formed. Entry into the body of the host embryo is accomplished by the newly hatched larva. Vance (1932b) determined that females of C. annulipes Wesm. oviposit indiscriminately in corn borer eggs of any stage of development, but Wishart & Van Steenburgh (1934) showed that there is a marked difference in the number which attain maturity when oviposition takes place in those of different ages. Maturity was reached by 26.6% of the parasitoid individuals when the host eggs were <24-h old, 36% in those which were 2.5 days old, and only 11.4% in those which were nearly ready to hatch. In fresh eggs there is a good chance that the larva will be found outside the body of the embryo after it has developed, in which case no further growth can take place. However, it may pass between the lateral folds before completion of the dorsal closure and become successful established. Eggs placed in the yolk after dorsal closure are a complete loss. After the embryo rotates, practically all eggs are placed within it.

In the Microgasterinae the only species known to oviposit in the host egg is Microgaster marginatus Nees, a parasitoid of Polia spp. in Russia (Zorin 1930). The egg is laid in the host embryo 24h prior to hatching. There are several parasitoids of Diptera among the Alysiinae which oviposit in host eggs bud do not attain larval maturity until the pupal stage. Examples are Sympha agromyzae Roh. in Agromyza, Coelinidea meromyzae Forbes in Meromyza, and C. niger Nees in Chlorops (Clausen 1940/1962).

The gregarious external parasitoid of the larvae of Hapalia machaeralis Wlk and other Pyralidae in India (Beeson & Chatterjee 1935) and China (Chu 1935), Cedria paradoxa Wlkn. was the only braconid found demonstrating maternal care. The researchers emphasized the persistence of the female with the brood after the host was stung and the eggs laid, which care continued until the progeny attained the adult stage. Another host may then be attacked, and the same events take place. Chu believed that the females did not feed during this period. A maximum of five broods was produced by each female, and the number of individuals in each brood decreased progressively. The averages for a series of females producing four broods was 31.4, 20.3, 11.0 and 3.0 eggs in successive ovipositions. Beeson & Chatterjee (1935) concluded that the total number of eggs deposited on a series of hosts was no greater than may normally be laid on a single one, and implied that a female normally attacks only one host individual during her lifetime. They considered that the main purpose of brood care was for protection from attack by chalcidoid parasitoids.

Immature Stages of Braconidae

The Egg.

The general form of the eggs of the Braconidae is simple, ranging in outline from broadly oval to almost cylindrical but frequently somewhat pear‑shaped, or elongate and tapering at both ends and usually without a stalk or pedicel. The egg of Microbracon lendicivorus (Fig. 19B) differs from those of other Vipioninae in having a slender tapering stalk, slightly longer than the egg body, at what is pre­sumably the anterior end. In Dendrosoter protuberans Nees (Fig. 19G), the stalk is very broad and bent back upon the egg body in a characteristic way, whereas, the stalk of Coeloides subconcolor Russo (Fig. 19F) is long and has a distinctly segmented appearance (Russo, 1938). Most of the species of Apanteles have a short peduncle at the posterior end, and Opius crawfordi Vier. has this peduncle equal to or longer than the egg body. Other species of the latter genus lack the peduncle entirely or have it in a very reduced form (Fig. 19D, E). The egg of O. tryoni is enveloped by a thin, transparent membrane possibly the exochorion, which is broken during the period of incubation owing to increase in size of the egg. This recalls a similar egg envelope found by Dowden in Brachymeria compsilurae Cw.

The Euphorinae have the stalk at the posterior end, and in some species of Perilitus it is nearly as long as the main body. The same is true in a number of species of Meteorinae. The egg of Alysia manducator (Fig. 19C) bears a pronounced button­like tubercle at its larger, presumably anterior end. Although the evidence is not complete and there are several apparent exceptions, it appears that tho stalk of the braconid egg, when present, is usually situated at the posterior end. In no instance does it serve any definite purpose after deposition.

The eggs of Ascogaster, Phanerotoma and Chelonus, which are dcposited in those of the host, are of minute size, measuring 0.2 mm. or less in length. In all species of the family the chorion is thin and transparent, and usually has no surface sculpturing, though in Meteorus versicolor and Microdus dimidiatus it bears minute hexagonal markings.

As for development of immature stages, following the deposition of the egg, an increase in size occurs during incubation in most species that develop within the host body. This is effected by absorption of host fluids by the developing embryo. This enlargement usually causes the egg to become nearly spherical in form, and the caudal pedicel may disappear almost completely because of the stretching of the chorion. In Apanteles thompsoni Lyle (Vance 1931) the increase is about 3X in width and 8X in length. The greatest increase in size during incubation seemingly occurs in the Euphorinae and Meteorinae. Ogloblin (1913) stated that the egg of Perilitus coccinellae Schr. increased from 0.08 X 0.02 to 0.4 X 0.2 mm, equivalent to circa 1,000 times the volume. In P. rutilus Nees (Jackson 1928), the growth was ca. 1,200X by volume. Balduf (1926b) mentioned the appreciable difference in size of eggs of P. coccinellae while still within the female’s ovaries, some being 0.08 mm in length while others measured up to 0.29 mm. This was considered to be due to the retention of the latter individuals for a longer period in the oviduct and probably to absorption of fluids from the parent. The extreme in egg growth during incubation was recorded by Strickland (1923) in Meteorus dimidiatus Cress, a parasitoid of cutworm larvae in North America. At the time of deposition the egg measured 0.14 X 0.04 mm, while just prior to hatching it was practically spherical, with the pedicel barely showing, and ranged up to 1.5 mm in diam. This growth took place very rapidly and dimensions of 0.28 X 0.12 mm were noted within 30 min after egg deposition.

Eggs of ectophagous braconids are usually lightly adherent to the integument of the host, though some are placed only in the immediate vicinity of the host, while in endophagous species they are free-floating or lodged among the muscles. Faure (1926) stated that the eggs of Apanteles glomeratus were attached by their pedicels to the internal organs of the host caterpillar, but other researchers do not verify this observation. In Opius fletcheri (Willard 1920) they are reported firmly attached at one end to the inner wall of the integument of the host by a mass of dark material which may represent a clot formed at the wound.

Clausen (1940/1962) remarked that little need be said regarding braconid larval habits, for they differ only in relatively minor respects from those of other groups. However, in Ascogaster, the first instar must penetrate the host embryo after spending 10-14 hrs in the cytoplasm surrounding it. This is in contrast to the habit of Chelonus, which places the egg directly within the embryo or in such a position in the cytoplasm that it is enveloped by the growing germ band. The Ascogaster larva is minute at hatching (0.2 mm in length), but increases to 10X that length and 250X in volume before the first molt (Rosenberg 1934). The anal vesicle constitutes 1/6th that volume. In C. annulipes, the development of the larva within the body of its host is appreciably affected and controlled by the physiological condition of the latter (Bradley & Arbuthnot 1938). This is shown by a series of rearings on single and multiple brooded strains of the European corn borer at 26.7°C and 70% RH. In the former, all larvae remained in the first stage for 20-36 days, no 2nd instar larvae appeared before the 20th day, and no 3rd instar larvae before the 24th day. In contrast, development in the multiple brooded strain progressed more rapidly. In no case did the first larval stage extend beyond 15 days, the second stage 17 days or the third stage 19 days after oviposition.

First instar Larvae.

The first instar larvae of the Braconidae represent a con­siderable variety of forms, comprising the hymenopteriform, mandibulate, caudate, vesiculate and polypodeiform. It is often debatable; particularly in reference to this family, as to whether a particular larva should be classified as mandibulate, caudate, vesiculate, or polypodeiform, for it may possess two, or in some cases three, of the characters upon which the grouping is based. The hymenopteriform larva has a medium sized head, 13 body segments, which usually bear transverse bands or rows of setae, and spiracles on the first thoracic and the first eight abdominal segments. This type of larva is representative of the ectophagous forms, comprising the Vipioni­nae and representatives of the Braconinae and Horminae, and will doubtless be found in other groups. Bracon sp., probably B. hylobii Ratz., is distinguished by the lack of spiracles (Munro, 1917).

The mandibulate larvae are found generally in the Opiinae and, in combination with the caudate character, in the Euphorinae, Triaspinae, Alysiinne and Pam­bolinae. Vesicle bearing larvae of this type occur in the Macrocentrinae. The larva of Opius tryoni (Pemberton and Willard, 1918) is typical of the Opiinae and has a large, heavily sclerotized head, large falcate mandibles, and short, blunt anten­nae. A pair of fleshy finger like processes is found ventrally at the anterior margin of the first thoracic segment. A well defined tracheal system, with anterior and posterior commissures, is present and filled with air, but there are no spiracles. The larvae of O. humilis, O. fletcheri (Fig. 20), and O. fullawayi are very similar to that of O. tryoni. Keilin and Picado (l913) have made an extended study of the first instar larva of O. crawfordi (Fig. 21A), which has an almost spherical head, a pair of very large mamma like processes, each surmounted by three sensory papillae, on the first thoracic segment and a smaller conical shaped pair on the third segment. These processes are on the concave side of the body, as is also the anal opening. The authors assert that the concave side of the body, to which the mouth opening is directed, is in reality the dorsum and support their conclusion by demonstrating the presence of the nerve cord along the convex side and of the heart on the concave side. Recognition of the markedly concave side of the first instar larva as dorsal rather than ventral is also reported by Baume-Pluvinel (1914) in Adelura gahani B.‑P. (Fig. 21B), which develops in the larvae of various Phytomyzinae. Further investigation of this interesting point in other species would be desirable.

The larva of Ascogaster quadridentata has an exceedingly large head and 13 body segments of diminishing width, with no tail or other fleshy processes, whereas Chelonus annulipes, which is of modified mandibulate form, has a short tail following the 8 or less distinct body segments. The larva of Macrocentrus ancylivoruss is elongate in form and has a pair of short fleshy processes and a short caudal horn on the last segment. The anal vesicle is relatively small. In the Euphorinae the large, heavily sclerotized head, bearing the falcate mandibles at the front, is followed by 12 or 13 body segments of decreasing width and a rather short tail, which bears setae on the distal half or two‑thirds.

The true vesticulate larvae are found mainly in the subfamily Microgasterinae, of which the principal studies have been made in the genera Apanteles and Microgaster. At the time of hatching, many of these have the general appearance of mandibulate larvae, and they may bear a fleshy tail approaching half the length of the body proper. Usually only 10 or 11 ring-like body segments are distinguishable, the last segment apparently representing several that have fused. Each of the segments usually bear a transverse row of setae dorsally. In A. tasmanica Cam. (Dumbleton 1935) and Miscogaster tibialis (Fig. 22a) (Vance, 1932a), the rows of setae are lacking on the first two segments, while in other species they are missing on only the first segment. The vesicle appears shortly after feeding begins, and its width is then equal to or greater than that of the preceding segments (Fig. 22C). The body at this time is somewhat cylindrical, and the tail, which previously was prominent, now appears as only a small ventrally directed “horn” beneath the vesicle. In A. militaris (Tower 1915), A. hyphantriae Riley, and A. thompsoni, there is no indication of a tail structure at any time, and the bulb-like vesicle is well developed even before hatching.

The wall of the proctodeum of the first instar larva of Orgilus obscurator Nees is relatively thin (Fig. 23), but it increases greatly in thickness in the second instar (Thorpe 1932).

The simple caudate type of larva, without other adaptive modifications, is found principally in the Meteorinae and Aphidiinae, whereas the tail in some form is present in practically all groups which develop inter­nally. In Meteorus, there are 12 or 13 segments exclusive of the tail, and the latter may exceed the body in length, though in some species it is only one eighth as long. Each segment usually bears a transverse row of setae on the dorsum, and the tail may also bear setae. The first instar larvae of many of the Aphidiinae are recognizable principally by the possession of a row or comb of heavy setae at the posterior margins of each body segment and by the two ventrally directed lobes of the last segment. The tail in this subfamily is usually somewhat tubular in form with the distal end rounded. There are 13 distinct body segments. The fringe of spines at the posterior margin of each segment dorsally and extending to the lateroventral margins is, so far as known, found only in Praon, while in P. simulans, studied by Timberlake (1910), they occur only on the third thoracic segment and on all abdominal segments except the last. Janiszewska (1933) describes the larva of an undetermined Aphidiine believed to be Aphidius, in which this row or comb of spines is present on each body segment. In Ephedrus incompletus Prov. (Wheeler 1923) and other species of the genus, the larva (Fig. 24) bears on each segment a median transverse ridge which is more pronounced on the dorsum and sides and is strongly serrate, with the teeth directed caudad. The tail also is heavily and completely spined, with the spines arranged in rings about it. These adaptations are possibly locomotory in function. The ventrally directed bilobed process of the caudal segment is found in Praon, in Ephedrus, and in some species of Aphidius, and consists of two conical or finger like processes, about the length of one segment, situated ventrally at the base of the tail. The majority of the species of the genus Aphidius have simple caudate larvae, which lack entirely the integumentary spines and the paired caudal process, and the tail is only lightly spined on its distal portion. Larvae of the Aphidiinae have the anal opening ventrally at the base of the tail rather than dorsally; in the species having the lobed processes, it is situated between the bases of the lobes and the tail.

The polypodeiform larva is found in isolated species in a number of subfamilies. That of Dacnusa navicularis var. cynaraphila Ric. (Ricchello 1928) (Fig. 25A) is, except for its paired ventral processes, typically caudate, with a transverse row of setae dorsally on each abdominal segment. The paired ventral processes occur upon each of the 12 body segments and are surmounted by a group or row of setae. The larva of D. areolaris (Fig. 25D), on the other hand, lacks the tail and the paired ventral processes (Haviland, 1922a). Bassus dimidiator (Fig. 25) (Silvestri, 1923a), B. pumilus Ratz. (Thorpe, 1933), B. stigmaterus Cress, and Macrocentrus gifuensis are distinguished by having two pairs of fleshy processes on each segment. In M. gifuensis, these are present on the first 12 segments and are of uniform size, whereas in the first two species named they are lacking on the first segment and are of slightly greater size on the abdomen.

The respiratory system of first instar larvae of certain of the endoparasitic species, such as the Opiinae, consists of the two lateral trunks with branches at the various segments and an anterior dorsal and a posterior ventral commissure. In a considera­ble number of species, however, there is a complete lack of the tracheal system in this instar.

In the great majority of species, the mandibles are simple, though several excep­tions occur. Those of Microbracon brevicornis are dentate on the lower border, whereas in Bracon tachardiae they are dentate, the teeth being long and spine like, and in Heterospsilus cephi (Hill and Smith, 1931) the main tooth is followed by five or six elongate teeth in comb like arrangement along the inner edge. It will be noted that these departures from the normal are in species which feed externally.

Many first instar Apanteles larvae have a tail, or “caudal horn,” situated beneath the vesicle, which in some species disappears entirely at the first molt and in others per­sists in reduced size in the second instar larva. In contrast to this, the anal vesicle of the larvae that possess it increases in size with each molt but is absent in the final instar. In A. thompsoni, it is said to be present for only a short time after the second molt. The paired ventral processes on the last segment of many aphidiine larvae do not persist beyond the first instar. Many species reveal an increasing number of small teeth on the inner margin of the mandibles in the intermediate instars. In Bracon tachardiae Cam., the four teeth of the first instar are succeeded by five in the second and third, whereas the mandibles of Microbracon mellitor are simple in all instars.

The paired ventral processes that occur on the bodies of first instar polypodeiform larvae, such as those of some species of Bassus and of Macrocentrus gifuensis, persist in much reduced form in the second instar. According to Parker, the larva of the latter species lacks mandibles in this instar.

Intermediate and Final instar Larvae.

The intermediate larval instars of the ectoparasitic species do not differ in any essential character from the first instar. Among the internal parasites, the mandibulate type larva loses the large, heavily sclerotized head, with its long falcate mandibles, at the first molt, and in the caudate forms the tail is reduced in size with each succeeding molt and is entirely lacking in the last instar. In some species having five instars, it disappears after the second molt, and in Cosmophorus henscheli it is entirely absent in the second and following instars.

The mature larvae of the Braconidae are of normal form and have few characters that distinguish them, aside from the tracheal system. In many species, the mandibles have minute teeth, often slender and spine like, on the inner margin, approaching 30 in number in some species, giving a comb­like appearance. Voukassovitch, in describing the mature larva of M. abdominalis mentions a bilobed chitinous “anal capsule,” of which the ventral lobe is more heavily sclerotized and bears a small ventrally directed process. The anal opening is between the two lobes of the capsule. Beeson and Chatterjee refer to a prominent “process” ventrally on the fifth and sixth abdominal segments of the larva of Perilitus mylloceri Wlkn. but do not otherwise describe it.

The mature larvae of many of the ectophagous species bear a dense coating of fine hairs; in some instances, this is uniform over the body, and in others it occurs in a transverse band on each segment and may be absent ventrally.

It has already been pointed out that the respiratory system of the ectoparasitic first instar braconid larva has normally nine pairs of spiracles, situated on the first thoracic and the first eight abdominal segments In these species, largely included in the Vipioninae, this number and arrangement persist through all the following instar. The early instar larvae of the species that develop internally lack the open tracheal system; and, in species that are known to have five instars, the spiracles first appear on the fourth. The species in which only three or four larval instars have been distinguished reveal the spiracles only on the last instar.

De Leon (1934) has summarized the information available regarding the respira­tory system of mature braconid larvae and has attempted to group the subfamilies on the basis of spiracle number and position and on the presence or absence of certain commissures. The information available is sufficient for only a very few subfamilies to permit of generalizations in this respect. It appears, however, that the Vipioninae quite consistently have the number and arrangement given above, and limited information indicates that this is true of the Braconinae also. The most common spiracular arrangement, however, has the same number, but the thoracic pair is situated on the second segment rather than the first. This order appears to pre­dominate in the Macrocentrinae, Meteorinae, Euphorinae, Opiinae and Alysiinae. Macrocentrus ancylivorus is said to have the spiracles on the second and third thoracic and the second to eighth abdominal segments, whereas M. abdominalis has 10 pairs, the additional one being upon the ninth abdominal segment. In the Microgasterinae, the occurrence of eight pairs is quite general, and in most species the single thoracic pair is on the second segment, whereas in a smaller number it is upon the third. Microgaster connexus, however, has only six abdominal pairs rather than seven, and Apanteles lictorius Rein. is said to have nine pairs, though their position is not given.

On the basis of information regarding a limited number of species, it seems that the greatest variation in spiracle arrangement occurs in the Aphidiinae. Aphidius granarius L. has spiracles on the first thoracic and eight abdominal segments, and Ephedrus plagiator Nees (Skriptshinskij, 1930) on the second and third thoracic and seven abdominal segments. Wheeler (1923) states that tracheal system and spiracles are absent in the aphidiine species studied, representing three genera but this is so unusual as to require verification.

The tracheal system of the mature braconid larvae is distinguished from that of the Ichneumonidae chiefly by the absence of the secondary lateral commissures in the thorax, which connect with the main trunks by three branches. The anterior dorsal commissure is present in all species, but the absence of the posterior ventral commis­sure has been noted in species of Chelonus, Apanteles, Microplitis and Meteorus. Ventral abdominal commissures occur in the first eight segments in several species of Vipioninae and in Doryctes gallicus Rh.

Among the species that develop within the host, many have an internal tracheal system but no spiracles in the early instars, and the spiracles appear only in the last larval instar. In Bracon sp. (probably B. hylobii) studied by Munro (1917), which is the single species of external habit that lacks spiracles in the early instars, they appear first in the fourth (penultimate) instar.

According to Glover (1934), who has studied the immature forms of B. tachardiae, the head widths of the five larval instars conform to Dyar’s principle, though the extremes overlap, whereas mandible length shows no overlapping.

De Stephani Perez (1902) has described the chrysalis of Giardinaia urinator Perez, found upon the stems of Potamogeton, which he considers to be the last larval exuviae and within which pupation is said to take place. The last body segment is bifurcate, and the terminal “hooks” are embedded in the stem of the plant. They may bear spiracles, and the supply of air would thus be derived from the plant. The figure of this chrysalis shows 14 segments. The parasite pupa illustrated within it is slightly more than half its length and one sixth to one eighth its volume. At emer­gence, the wasp breaks through the dorsum of the chrysalis and climbs to the surface of the water. It is extremely improbable that this chrysalis is the last larval skin of Giardinaia; it is much more likely to be that of its host. This parasite has been recorded from Hydrellia sp. in Europe and the “chrysalis” described by Perez may be the empty larval skin of a species of Ephydridae. The aquatic larvae of some representatives of this family are known to possess caudal spiracles upon a bifurcate process and to derive their air supply from plants (Clausen 1940).

By 1940 there were no braconids known to retain a connection with the egg shell after hatching, and the larval exuviae are usually entirely discarded by the succeeding instars. However, in Macrocentrus abdominalis F. the 2nd instar larva retains the first exuviae in ribbon-like form at the posterior end of the body, and some species of Apanteles retain it as a band about the abdomen immediately in front of the anal vesicle (Clausen 1940/1962). The species of Opiinae and Alysiinae attacking dipterous larvae, but emerge from the puparia, very consistently delay the first larval molt until host pupation, or lat least until the puparium is formed, after which development is very rapid, and the perfect pupal stage is usually not attained by the host.

In gregarious species of Microgasterinae the emergence of the brood of larvae from a host caterpillar takes place almost simultaneously. Gatenby, referring to Apanteles glomeratus, concluded that the outward movement of the more advanced individuals is due to a physiological stimulus and that their movements influence the remainder to do likewise. In this subfamily the 3rd instar is usually external and nonfeeding, the second molt being coincident with emergence from the host, and the anal vesicle is retracted at emergence or soon thereafter. In A. militaris, its retraction takes place late in the 2nd instar, while in A. sesamiae Cam. the vesicle is believed to persist through the third stage (Ullyett 1935). Third instar larvae of A. lacteicolor Vier. have been found still within the host body where they apparently do some feeding. In those species which form the cocoon on the body of the host, the anal vesicle usually remains embedded in the wound until the preliminary phases of cocoon formation are completed. Many species leave exuviae in the wound at emergence, which partially explains the lack of bleeding from the relatively large perforations in the body wall.

In gregarious species of Macrocentrus, the larvae, while within the host, are all oriented with the head toward that of the host, but during the following period of external feeding they lie in the opposite direction. After spinning of the cocoon, there is still another reversal of position so that the pupae are always oriented in the same way as the host. The third larval molt takes place at the time of emergence form the host (Clausen 1940/1962).

Mature larvae of Perilitus, after completing feeding in the body of the beetle host, accomplish emergence by cutting a hole through the membrane between two segments of the abdomen, usually near the posterior body end. P. coccinellae does so dorsally between the 5th and 6th abdominal segments. Sometimes it has been reported that emergence is through the anal opening, but it was more probably through an incision in the membranae near by (Clausen 1940/1962). In species such as Syrrhizus, that kill the adult host prior to emergence of the larvae, the point of exit is between the abdomen and thorax, or head and thorax, rather than at the posterior end of the body.

The persistence of the embryonic membrane, or trophamnion, as an envelope partly or completely enclosing the 1st instar larva has been noted in many species of braconids, particularly among Microgasterinae, Meteorinae, Euphorinae and Opiinae. Tower stated that the larva of Apanteles militaris was still enclosed in the membrane after escape from the egg and that it soon cut through the membrane in the mouth region and began feeding. As growth proceeded, the membrane was further disrupted and fell away, though portions of it remained about the body until the first molt. In A. thompsoni, the same situation exists and several weeks elapse before the larval frees itself completely from the membrane (Vance 1931). The membranae of Microgaster connexus persists until the larva is circa 2.5 mm in length, indicating that a considerable period is passed within it, and very appreciable growth takes place due to the food derived from it (Gatenby 1919). This is also found in Dacnusa areolaris Nees. The membrane of Meteorus nigricollis Thoms. remains in the egg shell after hatching, and on the 17th day after oviposition, which is about 10 days after hatching, it becomes dissociated into individual isolated cells, which float out into the blood of the host and increase considerably in size (Parker 1931a).

Pemberton & Willard (1918) observed in Opius tryoni, O. humilis Silv. and O. fullawayi Silv, that a gelatinous mass of cells clings to the ventral side of the larva and usually persists until the first molt when it is cast with the exuviae. In O. fletcheri, the newly hatched larva is completely enveloped by this mass of cells, though it does not remain attached to the exuviae after the molt. The 1st instar larva of O. melleus bears a large rounded mass ventrally on the thorax, which may be of the same origin (Lathrop & Newton 1933).

A detailed study of the membrane that surrounds the embryo of P. coccinellae was made by Ogloblin (1924). Through this membrane the embryo secures the food materials required for its development and the oxygen to fill its respiratory demands. At hatching, the cells of the membrane are more or less cubical in form and measure 16 to 19 microns in diameter. Disintegration takes place at this time, and the cells float free in the host blood. They finally attain a diameter of 200-250 microns, which is ca. a 4,000-8,000X increase in volume. In flowing with the blood, they tend to congregate in the posterior region of the abdomen of the host beetle, and the intermediate and mature larvae of the parasitoid, which feeds upon these cells, usually lie with their heads at this point. Jackson (1928, 1935) did extended studies of this embryonic membrane on P. rutilus Nees in adult beetles of the genus Sitona. Sometimes it still encloses the larva after hatching, but dissociation occurs very soon thereafter, and the cells maintain an independent existence in the host body. They store up fats from the body fluids of the host and attain a max. diam. of 0.1 mm. Eventually they are consumed by the parasitoid larva in its last stage. Paillot found them in aphids which contained no parasitoid larvae but thought that in such instances the parasitoid larva itself had died and been completely absorbed.

Polyembryony in Braconidae.

This mode of reproduction was first suspected in Braconidae by Voukassovitch (1927b) in Macrocentrus abdominalis, parasitic in caterpillars of Psammotis and Tortrix in Europe. In each of these hosts, 17-41 parasitoids reached maturity, and it was noted among 25 broods that 13 comprised females only, 8 were males only and 4 broods were mixed. Actual proof of polyembryonic reproduction was given by Parker (1931b) in the case of M. gifuensis, a parasitoid of the European corn borer, which for a time was confused with M. abdominalis. Broods ranged in size from an average of 16 in pure broods of females to 24 if all were males. Among 200 broods for which records were taken, 71 consisted exclusively of males, 54 of females and 75 were mixed. The mixed broods were considered to be the result of duplicate oviposition.

The egg of M. gifuensis transformed to a spherical pregerm, which was usually lodged in a fat cell, and this grew to a primary germ, which divided by fission. The secondary germ continued division, forming tertiary germs, eventually resulting in morulae which finally developed into embryos and into 1st instar larvae. There was no blastula stage, and no proliferation of host tissue to form a cyst about the parasitoid body such as is found in the development of some polyembryonic chalcidoid and serphoid Hymenoptera.

Extensive studies of the solitary internal parasitoid, M. ancylivorus, Daniel (1932) found that development to the embryo stage was similar to that described by Parker for M. gifuensis. However, the first parasitoid body to attain the larval stage exerted an inhibitory influence on the remaining embryos, morulae, germs and pregerms that were present in the host body, and no further development of these took place. The action was not mechanical but was considered to be due to the secretion of some substance by the larva or by the host through the influence of the parasitoid larva. Where two or more larvae developed simultaneously, the death of all but one was brought about by combat. In this species polyembryonic reproduction does not achieve its purpose, for only a single individual attains maturity from any one deposited egg. It is probable that M. ancylivorus was at one time parasitic on a much larger host in which a number of progeny could develop in each, but in its present hosts it seems consistently solitary (Clausen 1940/1962).

Amicroplus collaris Spin, attacking larvae of Euxoa segetum Schiff. in Europe, reproduces polyembryonically (Paillot 1937). Two or three eggs are deposited at each insertion, and 40-50 individuals develop to maturity in each host.

Among endoparasitic species, the habit of completing larval feeding after emergence from the host body often has been noted, in particular among Cheloninae and Macrocentrinae. Ascogaster quadridentata emerges from codling moth larvae after which it entirely consumes the remaining body contents. In Chelonus annulipes this external feeding phase is extensive and is essential to further development, for the cocoon cannot be spun unless feeding has occurred. The larva of C. blackburni Cam. emerges from its host when only half grown. All species of Macrocentrus that have been studies have the external feeding phase, and emergence from the host body coincides with the 3rd molt. The same habit was observed in Bassus hawaiicola Ashm. In the Microgaterinae only a few species are known to feed after emergence from the host body. Among these are Apanteles sp. on Nacoleia octasema Meyr. in Java, A. aristoteliae Vier. in Argyrotaenia citrana Fern., and Microgaster tibialis Nees (Clausen 1940/1962). Solitary Apanteles larvae completely consume the body contents of the host within a very short period after emergence.

Pupation.

Pupation habit of several species of Rogas differs from that of other members of Braconidae which attack lepidopterous larvae, in that the cocoon is spun within the empty skin of the host. Prior to the formation of the cocoon, the mature larva breaks the host skin ventrally and pushes out that portion of the body contents which has not been consumed, and this material dries and fastens the host remains securely to the substratum. However, some species do not make this ventral break in the host skin.

Vipioninae usually spin individual cocoons within the cell, cocoon, or leaf roll inhabited by the host, and in particularly confined quarters, as in a tunnel in wood or bark, they may be formed in line (Clausen 1940/1962). Microbracon sp. on the larva of Hapalia in India does not form individual cocoons; instead, the group of larvae spin a thin common web and then pupate in a row beneath it. Aphidiinae, with some exceptions, pupate within the empty skin of the aphid hosts, which is first lined with silk. Emergence is effected by cutting away a circular cap in the dorsum of the abdominal wall. One exception to this occurs in Praon simulans Prov. The mature larva leaves the body by an opening in the venter of the abdomen and spins a tent-like covering beneath it, firmly attaching it to the leaf surface, and the compact spherical cocoon is then spun beneath the host remains and within this shelter (Wheeler 1923).

Considerable variation is shown in the position taken by the larvae of the Microgaterinae at the time of pupation. The solitary species of Apanteles may spin the cocoon beneath the body of the host or apart from it, whereas an Apanteles sp. attacking the larva of Nacoleia in Java spins it within that of the host (Clausen 1940/1962). Some gregarious species, such as A. congregatus, spin the cocoons separately upon the body of the host at the point of emergence from the body, and standing perpendicularly, whereas those of A. glomeratus and A. sagax are found in irregular masses upon the leaf about the host body. Green (1925) observed that the cocoon mass of A. acherontiae may be 3 inches diam. and the 1,200 or more pupae contained within it do not have definite individual cocoons. Microgaster alvearias F. has the unusual habit of forming its cocoons in orderly horizontal tiers, one above the other, like a pile of bricks; those of A. militaris and Microplitis ceratomiae Riley are in compact masses, standing vertically upon the substratum. The solitary cocoon of M. maculipennis Szep. is formed beneath the caudal portion of the host body (Clausen 1940/1962).

Many species of Meteorus suspend the cocoon by a slender thread from the leaf or twig, and in M. pulchricornis this thread may be 8 in. long. Immediately after its emergence from the host body, the larva spins a small mat of silk upon the substratum and then lets itself down by a thread from its mouth. After this the larva reverses its position, grasps the thread with the tip of the abdomen and then completes spinning the cocoon. Because of this manner of spinning, which is common to all species forming such cocoons, the head of the pupa is always downward, at the end of the cocoon opposite the point of attachment of the thread (Clausen 1940/1962). Cedrai paradoxa has an unusual pupation habit. The group of larvae that develop on a host build a common sheet-like cover of circular outline on the leaf, and the individual cocoons are formed beneath this cover. There is usually a central row of cocoons encircled by others, which are arranged radially, with the heads directed outward.

The most frequently encountered species of Euphorinae, Perilitus coccinellae, forms its spindle shaped cocoon longitudinally beneath the body of the living host beetle, with the legs of the latter entangled in the outer strands. Parasitized Hippodamia beetles when freed from the cocoon rush about frantically, apparently in search of the cocoon, and when it is found they settle down upon it and attempt to entangle their legs once again in the web of loose silk surround it (Davis 1928). Captivity is apparently voluntary, even at the time the cocoon is spun. Microctonus melanopus spins its cocoon beneath the host beetle in the same way, though several gregarious species of this genus form them en masse in the soil (Clausen 1940/1962).

Chelonus and Ascogaster spin their cocoons within those of the hosts or in the cells occupied by the latter. Those of Sigalphus bicolor are closely packed in the host cocoon and are oriented so that their axes are nearly perpendicular to that of the host. The cocoons of Macrocentrus abdominalisare formed side by side, all oriented in the same direction. The Opiinae and Alysiinae, which reach larval maturity in the host puparia, do not spin cocoons. Among the Opius and Alysia it seems that the meconium, which is normally cast by the mature larva or prepupa in the pupation cell or cocoon, is retained and voided only by the adult. This is found also in Alysia, which similarly emerges from the puparium of its dipterous host. The last larval exuviae of Opius are consistently found upon the dorsum of the pupa rather than pushed to the posterior end of the puparium. Jackson noted that in the mature larva of P. rutilus before pupation, the meconium is enclosed in a long membranous sac which previously had lined the mid intestines. This is seemingly the “peritrophic sac” that was studied in some detail by Clancy in Helorus paradoxus Prov. A few solitary species of the family which pupate within the skin of the larval host usually orient themselves with the head toward the posterior end, while those which transform within dipterous puparia lie in the reverse position (Clausen 1940/1962).

Life Cycle.

Most Braconidae have relatively short life cycles, particularly among the external parasitoids, for their development is not correlated with that of the host. Information is available regarding a large number of species of Vipioninae, mostly of the genus Microbracon, which have many generations each year. The incubation period covers 1-2 days, the larval feeding period 1.5-7 days, and the cocoon stage 4-8 days. The minimum cycle from egg to adult is 7-9 days, recorded for M. lefroyi Ashm. and M. serinopae Ramkr. in India. M. sordidator Ratz. has the especially long cycle of ca. 56 days. There are marked differences in the cycle of M. hebetor under different temperature conditions, it being complete in 8 days at 32°C but 39 days at 16°C (Harries 1937).

Euphorinae usually have several generations each year, though several species of Microctonus are thought to have only one. The egg stage covers 4-10 days in Perilitus, the larval period 14-20 days and the cocoon stage 6-20 days, with a total of 24-50 days from egg to adult (Clausen 1940/1962). The Meteorinae have a little longer cycle, Meteorus cinctellus Nees having only one generations per year, although M. nigricollis reaches maturity in 45 days after oviposition.

Judging from studies of Apanteles, there are in the Microgasterinae usually several generations each year, with the egg, larval and cocoon stages covering ca. 2-5, 6-15 and 5-10 days, respectively. The minimum cycle is 10.5 days recorded for A. marginiventris Cress. The larval period of Microgaster is a bit longer than that of Apanteles. Among Aphidiinae, Lysiphlebus testaceipes Cress. and Aphidius gomezi Qules complete their life cycles in 7 days, but others of the subfamily take 10-30 days.

Apanteles solitarius Ratz. has one generation and a partial second each year on its host the satin moth. But the species has two distinct methods of passing the winter (Parker 1935). A portion of the brood hibernates in the mature larval stage in the cocoon, with adults appearing in early May. They attack host larvae coming out of hibernation at that time, and the second brood of adults appears circa one month later. A portion of the overwintering population remains in the first larval stage within the living host caterpillars. Their development in the spring is rapid, and the adults appear only slightly later than those which develop from larvae that have hibernated in the mature form in cocoons.

In Cheloninae, Macrocentrinae, Opiinae and Alysiinae, the life cycles are largely dependent on the development of the host, insofar as larval feeding is completed only when a certain stage of the host is attained. Therefore, Chelonus may have a single generation each year on one host, which itself has an annual cycle, while it may have several generations on a multibrooded host. Development of Macrocentrus ancylivorus Roh. is delayed in small host individuals and this effect extends even to a prolongation of the incubation period (Daniel 1932, Fink 1926, 1932). The Opiinae and Alysiinae complete their larval development only after their dipterous hosts have attained the pupal stage. Some of the parasitoid individuals may as a result attain the pupal stage at a time when others of the same age are still 1st instar larvae. Alysia manducator has a much longer life cycle than its host, resulting in the production of only two generations each year as compared with four for the host (Clausen 1940/1962).

Most Braconidae pass the winter as 1st instar larvae within the living hosts. This applies to most species of Apanteles, though a few of these, such as A. glomeratus and A. solitarius, sometimes may be in the mature larval stage in the cocoon, and to the Euphorinae, Triaspinae and some Meteorinae and Braconinae. Polyembryonic Macrocentrus gifuensis hibernates in the morula stage within the living host, while Aphidius fabarum Marsh hibernates in the pupal stage within the cocoon as does Cardiochiles nigriceps. Cedria paradoxa is the only representative of the family known to hibernate as an adult female. These females mate in the autumn, and the males do not survive the winter (Clausen 1940/1962).

Opiinae that pass the winter in the mature larval stage in host puparia, exhibit a pronounced tendency in that stage to go into diapause. In Opius tryoni and O. fullawayi a bit more than 1% of the larvae may persist for several months at a mean temperature of ca. 24°C, and in exceptional cases the diapause may extend to nearly one year. This tendency is even more noticeable in O. melleus, in which a portion of the brood normally carries over to the second year or later (Lathrop & Newton 1933). From a quantity of parasitized host puparia collected in 1925, 78.6% of the parasitoids emerged the following year, 20.3% the 2nd year, and 1.1% the 3rd year (Clausen 1940/1962).

Obligatory alternation of hosts occurs in several species of Apanteles and in occasional species of other groups. The stages of each host that are subject to attack are present in the field for only a small portion of the active season, and thus they can support only one parasitoid generation. The second must occur on one or more alternate hosts having a different seasonal cycle. Parasitoid species that are incapable of passing the winter in the adult or cocoon stage must have one or more hosts that persist as larvae during this period. Therefore A. lacteicolor, which was imported into the northern America as a parasite of the brown-tail moth, passes the winter in young caterpillars of that host, but its summer generations are in young gypsy moth caterpillars and in those of several native species (Muesebeck 1918). In Japan A. liparidis Bouché goes through two generations in gypsy moth larvae during summer, and the overwintering generations carries over in Dendrolimus larvae (Burgess & Crossman 1929). Muesebeck (1918) regarded that Meteorus versicolor had a seasonal cycle and alternation of hosts similar to that given for A. lacteicolor.

The alternation of hosts is much more striking in Microctonus brevicollis than that described above, for the alternate broods are in different stages of the same host rather than in the same stage of different hosts. Kunckel D-Herculais & Langlois (1891) reported the overwintering generations is contained in adult beetles of Haltica ampelophaga Guer., while the summer generations develops exclusively in the larvae of that species. This is comparable to the cycle of the tachinid, Erynnia nitida, R.D., a parasitoid of the elm leaf beetle, Galerucella luteola Mull., in Europe.

The reproductive capacity is relatively high in many Braconidae, in particular those which develop internally in the host. The ovaries of a gravid females of Apanteles golmeratus are believed to contain >2,000 eggs, and 15-35 are deposited at one insertion of the ovipositor. A single female of A. melanoscelus laid 535 eggs in six days, and Crossman (1922) estimated that the capacity of this species under field conditions was circa 1,000. Microgaster marginatus has 900-1000 mature eggs in the ovaries within two weeks after emergence. Among Aphidiinae, the egg production is also high. Females of Lysiphlebus testaceipes may contain as many as 430 eggs at one time, and Perez (1930) reported that Aphidius gomezi had a capacity of 1,500. Nearly 600 aphids were parasitized in one day, and circa 1,000 progeny were secured from a single female. The egg-ovipositing species of the family usually have a high reproductive capacity, as shown by the 655 eggs laid by a female Chelonus annulipes in 22 days, with 156 remaining in the ovaries at death; and another female deposited 165 eggs in a single day (Vance 1932b).

Polyembryonic species of Macrocentrinae, which produce 50 individuals in each host, deposit 200-300 eggs, and they therefore have a reproductive potential of several thousand.

Ectoparasitic Braconidae, of which a number of species of the genus Microbracon have been intensely studied, range from a minimum of 80 eggs for M. greeni Ashm. to the 678 recorded by Taylor (1932) for M. brevicornis upon Heliothis armigera Hbn. in South Africa. An unusually low reproductive capacity is indicated for Ischiogonus syagrii, in which the ovaries were never found to contain more than 12 mature eggs at one time (Clausen 1940/1962).

Regarding development of parasitioids in a single host, Green (1925) reported that a total of 1,226 adults of Apanteles acherontiae Cam. emerged from a single cocoon mass and that several hundred still remained. All these had developed in a larva of the hawk moth, Acherontia lachensis F. A maximum of 129 Apanteles liparidis was recorded from a larva of Dentrolimus (Burgess & Crossman 1929). Euphorine parasitoids of adult beetles are usually solitary, but McColloch recorded the development to maturity of 124 larvae of Microctonus eleodis Vier. in one Eleodes tricostata.

The influence of temperature on the reproductive capacity was studied by Payne (1933, 1934) in the case of M. hebetor, and the reproductive potential was found to be 67.5 at 36°C and 90 at 27°C. and it declined to 28 at 15°C.

The effect of temperature and food conditions upon parasitoid adults varies. Donohoe found that the development of Microbracon hebetor at temperatures fluctuating between 10-26.7°C. resulted in adults which were glossy black in color, while at 26.7°C they were mostly dark brown to black and at 32°C the color was pale brownish yellow to brownish orange. In Sycosoter lavagnei, according to Lichtenstein & Picard (1917) and Picard (1919), there was a seasonal dimorphism in both sexes, in which only the apterous forms were found early in spring, winged females and a preponderance of winged males in midsummer, and the apterous forms of both sexes in great abundance during autumn. Temperature was considered to be the most important factor influencing this cycle, though the quantity and quality of food at different seasons contributed to the change.

As for the number and sex of the broods of other species of Braconidae, there is the possibility of polyembryonic reproduction in M. crambivorus Vier., in which the broods are reported to be of one sex only, and in Sigalphus bicolor Cress. studied by Cushman (1913a), of which as many as 30 may reach maturity in a single caterpillar of Apatela and nearly half of the broods are of one sex only.

Regarding sex ratios and parthenogenesis, Braconidae show an exceptional number of species that yield a preponderance of male progeny and only a small number that have a large majority of female progeny (Clausen 1940/1962). Researchers have reported the female:male ratio in Microbracon hebetor, to range from 3:1 to 1:2. Payne (1934) found that the proportion of males was higher under low temperatures, and she ascribed this to a decrease in mating activity. In M. terebella Wesm., the normal ratio is ca. 2:1, with 62.6% of the colonies comprising females only, 26% males only and 11.4% of both sexes (Salt 1931b). The colonies comprise an average 3.3 individuals, those of the males numbering 3.9 and of the females 2.9. Cedria paradoxa was found to have a sex ratio of 5.5 to 1 in both China and India and on different hosts. A preponderance of females is shown also in Lysiphlebus testaceipes and Sigalphus bicolor, the figures being circa 2:1 and 3.4:1, respectively. Heterospilus cephi Roh. and Macrocentrus gifuensis show the males slightly in majority, while in M. ancylivorus females predominate in the ratio of 3:2. Opius fletcheri has females slightly in majority, while other species of that genus have that sex in the minority, the extreme being 1:2 in O. melleus (Clausen 1940/1962).

A marked variation in the sex ratio of Alysia manducator was reported by Holdaway & Smith (1932). It varied with the size of hosts. The small puparia of Lucilia sericata Meig. yield only males, while the largest hosts of Calliphora vomitaria L. gave mostly females. Within each host species, as well as between species, the sex ratio varies consistently, the larger puparia yielding a greater proportion of female parasitoids than do those of smaller size. Fink (1926) and (Stearns (1928) pointed out that in regard to the male progeny of unmated females of M. ancylivorus, these are only 1/2 as large by weight as the males produced by mated females. This is suggestive of diploid males produced by mated females.

Unisexual reproduction is normal in a number of Braconidae, among which are Apanteles thompsoni, Meteorus japonicus Ashm., P. coccinellae, Microctonus brevicollis and Rogas unicolor Wesm. No males of A. thompsoni have ever been found even during large scale rearings. Researchers in European and North America have collected and reared P. coccinellae; yet it was only around 1940 that the first authentic male was recorded. Extended rearings of field material of R. unicolor showed that the males consistently represent less than 1% of the population (Dowden 1938). Mating takes place readily, though this is not essential, as virgin females produce the same sex. Only a single male was reared experimentally, and this individual was the progeny of an unmated female.

The capacity to reproduce parthenogenetically seems to occur generally throughout the family, but sometimes mating seems essential to the production of progeny. MacGill (1923) found that eggs produced by virgin females of Aphidius avenae Hal. consistently failed to hatch. M. brevicollis is thought to produce a summer generation that is exclusively females in larvae of Haltica, and the overwintering generation in adult beetles gives rise to adult parasitoids of both sexes. In Lysiphlebus testaceipes, although males are normally produced by virgin females, an occasional female appears (Hunter 1909, Webster & Phillips 1912). Hunter (1909) stated that these represent 1.7% of the progeny, while the latter researchers recorded that 4 in 48 unmated females produced some female progeny. In both cases these represented only a small portion of the brood, and exclusively male progeny resulted in the second generation. Whiting (1921) has shown that unmated females of Microbracon hebetor similarly produce occasional females.

Effect of Parasitism on the Host.

A number of changes occur in hosts that sustain parasitism by braconids. In Ascogaster quadridentata, the parasitized larvae of the codling moth found in the cocoons and that have completed their feeding are only 1/4th to 1/3rd of normal size and are whitish rather than pink in color. Hosts of numerous Microgasterinae show a similar difference in size as compared with healthy larvae. In contrast, Beeson & Chatterjee (1935a) noted that the larvae of Sphingidae parasitized by Megarhogas theretrae Vier. become nearly double their normal size before death. More than 50 parasitoid larvae develop in each host. This inflation of the host recalls the similar condition occurring in caterpillars parasitized by species that reproduce polyembryonically. European cornborer larvae containing advanced larvae of Meteorus nigricollis are a dirty brown color, and an examination of the internal organs reveals that the fat body bears numerous wounds, apparently inflicted by the parasitoid larva. Haeussler (1932) mentioned that larvae of Grapholitha molesta Busck parasitized by Macrocentrus ancylivorus are retarded in their development and that the parasitoid emerged from the body an average of three days later than the time at which healthy larvae pupate. The dipterous hosts of the Opiinae complete larval development and usually form a normal puparium, though the pupal stage is never attained (Clausen 1940/1962).

As with other groups of parasitoids, death of the host may not take place until some time after emergence of the parasitoid larva from the body. This is true also in the case of many lepidopterous larvae attacked by internal parasitoids of the subfamily Microgasterinae. Gatenby (1919) found the larvae of Pieris brassicae accomplish a partial recovery after the emergence of the brood of Apanteles glomeratus, and some have been known to survive for one month. An examination of haemocoelic fluid of parasitized larvae failed to reveal any very obvious differences when compared with the blood of nonparasitized caterpillars, however. In case only a few parasitoid larvae develop in the body, the host may attain the pupal stage after their emergence, and possibly even the adult stage. Grandori (1911) and Faure (1926) mentioned that attainment of the pupal stage by some individuals. In contrast, the brown tail caterpillar host dies prior to emergence of the larva of A. lacteicolor (Muesebeck 1918). This is due to the destruction of a portion of the central nervous system. A solitary parasitoid such as this, developing in a very young host, naturally consumes a greater portion of the body contents and therefore causes greater injury than do a number of individuals in a nearly mature caterpillar.

In those cases where the adult stage of the host is attacked, or is attained before death, the effect of parasitism on the host is very interesting. The parasitic groups involved are largely subfamilies Euphorinae and Aphidinae, the former being known principally for its attack on adult beetles and the latter on aphids. Observations on the effect of Perilitus coccinellae parasitism on coccinellid adults have been made by Balduf (1926b). The host usually does not die for some time after the emergence of the larva from its body. Parasitized females very seldom show the ovaries to be in a functional condition, and in females that are gravid at the time of attack they are disrupted. No fatal injury is inflicted by the parasitoid larva, either through its feeding or by mechanical injury at the time of emergence, but the host is considerably weakened and dies from starvation because of entanglement in the outer strands of the parasitoid cocoon, which is formed beneath its body. Timberlake (1916) found that the more vigorous beetles of Olla abdominalis recovered completely if freed from entanglement with the cocoon and resumed feeding and oviposition. Sometimes these beetles were again parasitized experimentally and yielded mature larvae. Full recovery from the effects of parasitism requires considerable time, and one female began depositing eggs after 22 days from the time the parasitoid larva emerged from her body (Clausen 1940/1962). atrophied, while the male organs are still functional at the time of emergence of the parasitoid larva.

Studying the effect of Microctonus on its hosts, Kurdjumov & Znamenski (1917), McColloch (1918) and Speyer (1925) found that sterility was produced in the overwintering flea beetles, but those which are attacked in the early spring deposit the greater portion of their eggs before their reproductive activities are affected by the parasitoid. McColloch mentioned that one female of Eleodes tricostata deposited three eggs on the same day that 124 mature larvae of M. elodis emerged from her body. Speyer’s observations on what was probably a species of Microctonus, in Ceutorrhynchus, indicate that host oviposition takes place normally until an advanced stage of development is attained by the parasitoid larva and that the principal effect on the internal organs is the degeneration of the fat body. In Cosmophorus henscheli, Seitner & Notzl (1925) found that mating by the adult beetles of Pityophthorus was not inhibited, though the female organs were atrophied. No eggs or larvae were found in the brood chambers of parasitized beetles.

Adult beetles of Sitona spp, parasitized by Perilitus rutilus show the female reproductive system to be The effect of parasitism on aphids by the Aphidiinae is greater than that on beetles by the Euphorinae, because oviposition by the parasitoid frequently takes place before the adult stage is reached. Hunter (1909) found nymphs of Toxoptera graminum Rond.., if parasitized by Lysiphlebus testaceipes Cress. (tritici Ashm.) during the second stage, attain maturity before death but do not reproduce. When they are stung after the 4th molt, 2-6 offspring may be produced. In observations on the same parasitoid, Webster & Phillips (1912) found that stinging in the first two nymphal stages resulted later in death of both the host and the parasitoid larva, and they note results similar to those given above for T. graminum in case of attack at a later stage. In Aphidius rapae Curt., it was found that irrespective of the stage of development of the embryos in the body of the host Myzus and Aphis aphids, their growth is arrested at the time the parasitoid egg hatches and they finally disintegrate (Spencer 1926). The development of winter eggs is similarly inhibited. The production of eggs or nymphs ceases on the third day after oviposition by the parasitoid, this period representing the time required for incubation of the egg. Great physical changes take place in the host body before death. The parasitoid embryo is thought to secrete a cytolytic enzyme which aids in breaking up the serosa; it is also thought to be responsible for the death of younger larvae of its own species, for the abstraction of food materials from the fat body, and for the stoppage of development of the host eggs or embryos. In the final larval stage of the parasitoid, the embryos and eggs are first consumed, and then the remaining body contents (Clausen 1940/1962).

In some dipterous hosts of Alysia manducator, such as Lucilia, it was found that parasitism has a pronounced influence on the physiological processes involved in pupation. This is not a result of the activities of the parasitoid larva in the body but to a direct consequence of stinging at the time of oviposition. The hosts normally pass winter as mature larvae, yet those which are parasitized always pupate in the autumn. Some toxic substance injected at the time of stinging apparently provides the stimulus for premature pupation. The occurrence of unparasitized puparia among host material collected during the winter is explained by their having been stung but not successfully parasitized.

In a study of the effect of parasitism of caterpillars of Cirphis unipuncta How. by Apanteles militaris, Tower (1916) found that individuals parasitized shortly after the 4th molt ate less than half as much foliage during the period between attack and emergence of the parasitoid larvae from the body as did unparasitized individuals. An incidental effect of parasitization of Ephestia by Microbracon hebetor is infection by Thelohania ephestiae Mattes, wherein the female parasitoid serves as the vector (Payne 1932). The first foci of the disease occur in the thoracic ganglia that have been pierced by the sting of the parasitoid. Later the sporozoan is found throughout the nervous system and the fat body. The disease cannot be transmitted per os and no infection results from contact of healthy larvae with those that are diseased.

Discussion of Phylogenetic Relationships

Early attempts to explore the phylogeny of the major groups of Braconidae (e. g., Tobias, 1967; van Achterberg, 1984) were largely intuitive. Quicke and van Achterberg (1990) made the first large-scale effort to estimate braconid phylogeny using a large data set of morphological characters, and computer-assisted parsimony analysis. Although many aspects of this study have been in dispute (Wharton et al., 1992; van Achterberg and Quicke, 1992), it has stimulated a number of further attempts at resolving the phylogeny of subset groups of the Braconidae, using morphology as well as molecular data (e. g. Belshaw and Quicke, 1997; Whitfield, 1997; Belshaw et al., 1998; Dangerfield et al., 1999; Dowton et al., 1998; Dowton, 1999; Mardulyn and Whitfield, 1999; Quicke and Belshaw, 1999; Quicke et al., 1999; Belshaw et al., 2000; Kambhampati et al., 2000; Sanchis et al., 2000; Belshaw et al., 2001; Dowton et al., 2002; Whitfield et al., 2002; Chen et al. 2003).

The phylogeny used as the backbone here for accessing the subfamilies is adapted and pasted together, with a great deal of conservative poetic license, from these recent studies. It should not be taken all that seriously, as help is on the way in the form of continuing morphological and molecular phylogenetic studies. The reader is encouraged to consult the literature cited above for more in-depth analysis.

Classification

The higher classification of the Braconidae has been a matter of much dispute. Approximately 40 subfamilies are generally recognized, several of them newly discovered within the last 15 years (Mason, 1983; Quicke, 1987; Whitfield and Mason, 1994). Several subfamilies (e.g., Aphidiinae, Alysiinae, Apozyginae) have at one time or another been recognized as separate families. In general there is agreement on the basic subfamily or tribal groupings, with the exception of the “hormiine” and “exothecine” groups of genera (Whitfield, 1992; Quicke, 1993; Wharton, 1993b), but disagreement on the ranking or inclusiveness of some groups, since our understanding of the phylogeny of the family is not yet robust. There is a general perception that the number of recognized subfamilies has become inflated, while the tribal rank has been underutilized. An attempt has been made to address this problem (Wharton, 2001), but requires further phylogenetic testing.

Identification Guides

Keys to the world subfamilies and general introductions to the literature are available in van Achterberg (1993) and Sharkey (1993). A new manual to the genera of the New World has been produced (Wharton et al., 1997 – Spanish version 1998), including individual subfamily chapters contributed by a number of the world’s braconid specialists. Interactive versions of the keys in this manual are available online. van Achterberg (1997) has published the CD-ROM Braconidae. An illustrated Key to all Subfamilies. Information courtesy of www.faculty.ucr.edu