Inventory of the caterpillars, their food plants and parasitoids began in 1978 for today's Area de Conservacion Guanacaste (ACG), in northwestern Costa Rica. This complex mosaic of 120 000 ha of conserved and regenerating dry, cloud and rain forest over 0-2000 m elevation contains at least 10 000 species of non-leaf-mining caterpillars used by more than 5000 species of parasitoids. Several hundred thousand specimens of ACG-reared adult Lepidoptera and parasitoids have been intensively and extensively studied morphologically by many taxonomists, including most of the co-authors. DNA barcoding -the use of a standardized short mitochondrial DNA sequence to identify specimens and flush out undisclosed species -was added to the taxonomic identification process in 2003.

The four major biological strategies of ichneumonoid parasitoids, koinobiont and idiohiont, ecto-and endoparasitism, are discussed and the evolutionary radiations of the two families Ichneumonidae and Braconidae compared in an attempt to relate differences in patterns of host utilization to differences in evolutionary history. The most primitive members of both families are idiobiont ectoparasitoids of hosts concealed in plant tissue. Idiobiont ectoparasitic braconids are all still primarily associated with such hosts, hut idiobiont ectoparasitic ichneumonids have radiated to attack hosts in other situations, such as in aculeate nests or in cocoons. A shift in emphasis between the behavioural steps, host habitat location and host location, is envisaged as being important in such evolutionary change. Idiohiont endoparasitism is postulated as having arisen amongst ectoparasitoids attacking cocooned hosts, as an adaptation that allows them to exploit pupae and puparia in relatively exposed positions; it is a fairly common strategy in the Ichneumonidae, but virtually unknown in the Braconidae. Koinobiosis is perceived as having evolved in association with hosts which feed in a relatively weakly concealed position, hut pupate in a more secluded and safe location. The strategy is advantageous as it allows a parasitoid to oviposit on an easily discoverable host, but to use the host's pupation concealment to complete its own development. The evolution of koinobiosis has allowed parasitoids to exploit hosts that feed in exposed positions, and to attack hosts at a younger and numerically more common stage in the host's life cycle. Koinohiont ectoparasitism is envisaged, in some braconid and ichneumonid groups, to occupy an evolutionary transitional position between idiobiosis and endoparasitic koinohiosis; only in the Ichneumonidae have large radiations of koinohiont ectoparasitoids occurred. Endoparasitic koinobiosis is hypothesized as having arisen in the Braconidae in association with lepidopterous/coleopterous hosts, whilst in the major lineage of endoparasitic koinobiont ichneumonids, this habit is hypothesized as having arisen in association with symphytan hosts. The great majority of braconids are koinobiont endoparasitoids, but only about 50% of the Ichneumonidae have this habit. Very few koinohiont braconids develop as endoparasitoids of hymenopterous hosts, although many endoparasitic ichneumonids attack Hymenoptera. However, lineages of the Braconidae have radiated to exploit adult insects and exoptrrygote nymphs; ichneumonids do not utilize such hosts.

The monophyly of the ichneumonid clade Pimpliformes is established and the phylogenetic relationships of the eight component subfamilies are resolved. The clade (Acaenitinae + (Diacritinae + (Cylloceriinae + (Diplazontinae + Orthocentrinae)))) is the sister‐lineage to the clade (Pimplinae + (Rhyssinae + Poemeniinae)). The Nearctic genus Cressonia Dasch is transferred to the Diacritinae from the Orthocentrinae. Tribes are not recognized in the Acaenitinae as the Coleocentrini (sensu Townes, 1971) is paraphyletic with respect to the Acaenitini. The Cylloceriinae is recognized as comprising three genera, Cylloceria Schiødte, Allomacrus Förster and Sweaterella gen.n. The Orthocentrinae, including the Helictinae of authors, is shown to be monophyletic, but the latter is clearly shown to be paraphyletic if the Orthocentrus genus‐group is excluded. The Pimplinae comprises four monophyletic tribes: the Delomeristini, consisting of Delomerista Förster and Atractogaster Kriechbaumer; the Perithoini trib.n., which includes only Perithous Holmgren (= Hybomischos Baltazar syn.n.); the Pimplini, which includes the Theronia genus‐group as well as the Pimpla genus‐group; and the Ephialtini, which includes the Polysphinctini syn.n., a monophyletic group that previously rendered the restricted Ephialtini paraphyletic. The tribe Delomeristini is the sister‐group to the clade (Ephialtini + (Perithoini + Pimplini)). The subfamily Poemeniinae is recognized as comprising three tribes: the Pseudorhyssini (trib.n.) which includes the single Holarctic genus Pseudorhyssa Merrill; the Rodrigamini (trib.n.) which includes only the Costa Rican genus Rodrigama Gauld; and the Poemeniini. The tribe Pseudorhyssini is the sister‐group to the clade (Rodrigamini + Poemeniini). The phylogenetic inter‐relationships of the genera of Poemeniini are resolved. A new genus from South Africa, Guptella (gen.n.) is described, and Achorocephalus Kriechbaumer is shown to be a synonym of Eugalta Cameron (syn.n.). The evolution of biological traits within the Pimpliformes is discussed with reference to the elucidated phylogeny, and zoogeographic patterns are outlined.

Abstract.  A cladistic analysis of the Polysphincta genus‐group (= the ‘Polysphinctini’ of authors), a clade of koinobiont ectoparasitoids of spiders, was undertaken using ninety‐six characters for seventy‐seven taxa (sixty‐five ingroup and twelve outgroup). The genus‐group is monophyletic, nested within the Ephialtini as (Iseropus (Gregopimpla (Tromatobia ((Zaglyptus + Clistopyga) + (Polysphincta genus‐group))))). Within the Polysphincta genus‐group, the clade (Piogaster + Inbioia) is sister‐lineage to all other genera. The cosmopolitan genus Zabrachypus is nonmonophyletic, and has been subdivided into a monophyletic Nearctic/Western Palaearctic Zabrachypus s.str. and an Eastern Palaearctic Brachyzapusgen.n., comprising B. nikkoensis (Uchida) comb.n., B. tenuiabdominalis (Uchida) comb.n. and B. unicarinatus (Uchida & Momoi) comb.n. An Afrotropical species placed in Zabrachypus, Z. curvicauda (Seyrig), belongs to Schizopygacomb.n. The monophyly of the cosmopolitan genus Dreisbachia is equivocal, and we consider that species assigned to it are best placed in an expanded Schizopyga (syn.n.). The monobasic Afrotropical genus Afrosphincta is also a synonym of Schizopyga (syn.n.). The newly delimited Schizopyga is the sister‐lineage of Brachyzapus, and these two genera form the sister‐lineage of Zabrachypus s.str. as the monophyletic clade (Zabrachypus + (Schizopyga + Brachyzapus)). The Holarctic genus Sinarachna is monophyletic if the Palaearctic species S. anomala (Holmgren) is excluded and transferred to Zatypotacomb.n. The European species Polysphincta nielseni Roman belongs to the Palaearctic genus Reclinervellus, and (Reclinervellus + Sinarachna) is the sister‐lineage to a monophyletic group of undescribed Asian species herein assigned to a new genus, Chabliseagen.n. All remaining polysphinctine genera form a strongly monophyletic clade, the Polysphincta clade, although the relationship between this clade and the Chablisea and Zabrachypus clades remains an unresolved trichotomy. The Polysphincta clade comprises three lineages as an unresolved trichotomy, a monophyletic Oxyrrhexis, a major group (‘Polysphincta’ (Ticapimpla (Acrotaphus + Hymenoepimecis))) and an Acrodactyla lineage. In the second group, ‘Polysphincta’ is paraphyletic with respect to the other three genera. However, we retain ‘Polysphincta’ as a genus because few species of this very large genus were included in our analysis. We suspect that, when the tropical fauna is better known, it will be possible to subdivide ‘Polysphincta’ into two or more monophyletic taxa. In the third lineage, the Acrodactyla lineage, the Holarctic genus Acrodactyla is monophyletic if the European species A. madida (Haliday) is excluded. Consequently, we erect a new genus Megaetairagen.n. for this species. The monobasic Afrotropical genus Pterinopus is the sister‐lineage to the cosmopolitan genus Eruga. The very large cosmopolitan genus Zatypota seems to be monophyletic if two New World species, Z. parva (Cresson) and Z. gerardoi Gauld, Ugalde & Hanson, are transferred to...

No abstract