These are tiny wasps (1 mm or less) all of which are egg parasites. One genus, Megaphragma, which parasitizes thrips eggs, has species 180 microns (0.18 mm) in length. These are among the smallest known insects (Doutt and Viggiani 1968). The genus Trichogramma has probably had more papers written about it than any other chalcidoid, which is especially regrettable when considering that few, if any, of these papers have used a correct species name and therefore are of dubious value. A notable exception is more recent work by Pinto, Platner, and Oatman (1978). George Salt (1935, 1937, 1940) published a series of papers on the behavior of Trichogramma and other parasites. More recently, Pinto, et al. (1989) and Southamer, et al. (1990) have investigated important biological characteristics such as phenotypic plasticity and thelytokous forms (respectively). The genera of Trichogrammatidae were monographed at a world level by Doutt and Viggiani (1968) who also provided lists of all known species. In 1971 Viggiani presented a key to 27 (of 70) genera based on male genitalia. Yousuf and Shaffee (1986) gave a checklist and bibliography of world species. Pinto (in Gibson, et al. 1997) keyed all the genera for North America.

Of special interest to students of chalcidoid taxonomy is the great difference between the genus Trichogramma and the genus Aphytis (Aphelinidae). Both have been studied fairly well, from a biosystematic standpoint, with completely different results. For example, Aphytis species are believed to be very host specific, with about a 1 host, 1 parasite ratio (Rosen and DeBach 1979). Trichogramma, however, tends to show "habitat specificity" in some species, while others seem to show little specificity of any kind (Nagarkatti and Nagaraja 1977). Salt (1935) reviewed records of 1 species of Trichogramma ovipositing and emerging from eggs of 53 species of Lepidoptera, 6 Diptera, 3 Coleoptera, and Hymenoptera, Hemiptera, and Neuroptera. In fact, Salt showed that a species of Trichogramma attempted oviposition into any object of a certain configuration including sand, glass beads, lumps of flour, seeds, and mercury globules. Given a choice between large sand particles or seeds and a true host egg, Trichogramma actually preferred the false host to the true one! Salt concluded that the Trichogramma female was more impressed with "larger size" than with actual host suitability and that "size is the principal criterion used by Trichogramma in its choice of hosts." A second difference is that the taxonomy of Aphytis and most other chalcidoids is based on female specimens and the male genitalia have had little taxonomic significance. Trichogramma taxonomy, however, is based largely on male genitalia.

From the 2 examples cited above it would be safe to say that each family and probably each genus of chalcidoids needs to be studied on its own merits as well as in comparison to our overall knowledge of Chalcidoidea.

Several authors (Doutt and Viggiani 1968, Nagarkatti and Nagaraja 1977) stated that trichogrammatids are related to eulophids, and in fact, it seems likely that one of the major groups of eulophids (possibly euderines or tetrastichines) gave rise to this group. However they are distinct morphologically and are not readily confused with eulophids. In addition, recent molecular work could change our ideas of relationships. Viggiani (1971) divided the family into 2 subfamilies based upon male genitalia (Trichogrammatinae, Oligosetinae). Because of the difficulty in seeing adult trichogrammatids (let alone their genitalia), their relative uncommonness (at least in general collecting), and the degree of specialization required to work with them, we will treat these parasites only at the family level.

STATISTICS: Number of world species: about 650 (50 Nearctic); number of world genera: 76 (35 Nearctic).

BIOLOGY: These wasps are strictly egg parasites and have been reared from the holometabolus orders as well as Homoptera, Heteroptera, Orthoptera, and Thysanoptera. This family also contains some of the few truly aquatic species of Hymenoptera known. Several species in several genera parasitize eggs of Odonata, Dytiscidae, and other aquatic Heteroptera. Propulsion underwater may be by swimming with the legs or with the wings depending upon the species. As many as 70 individuals may emerge from a single dytiscid egg.

As was mentioned previously, Salt wrote several papers on the ethology of Trichogramma (Clausen 1940 and Askew 1971 list most of these papers). In addition to the example cited above for host selection, Salt has reported on at least 2 other aspects of Trichogramma biology of interest to the study of both taxonomy and biocontrol. These aspects concern the effect that a host has upon its parasite. The first finding was that hosts of T. semblidis caused male sexual dimorphism (much as in Melittobia-Eulophidae). Males reared from Sialis (Neuroptera) eggs were apterous and had different numbers of antennal segments and enlarged legs compared to winged males reared from Lepidoptera. Females were the same from both hosts (Salt 1937). Of particular interest is that males did not differ in size and that the differences in morphology were truly dimorphic, there being no intermediates. Sialis was the "normal" host, but Salt suggested that a "variety of hosts" might be attacked in the summer. The second finding of interest (Salt 1940) was the fact that progeny from large hosts generally were larger in size than from small hosts and that this could effect both morphology and behavior. Morphologically it could result in presence or absence of wings and the relative numbers of setae on wings and antennae being correlated to length. Biologically he found that progeny reared from large hosts behaved differently than progeny from small hosts. For example, large progeny parasitized hosts nearly twice as fast as small progeny, they possibly lived longer (though there was little data to make this statement), and they had a higher reproductive potential. Also large progeny could attack large hosts (which they preferred) or small hosts (reluctantly) but small progeny could successfully attack only small hosts even though they also preferred large ones.

DISTINGUISHING CHARACTERS: This family is distinct from all other chalcidoids by having 3-segmented tarsi. Specimens are nearly all less than 1 mm and are very fragile so that they shrivel up when dead (similar to eulophids and some aphelinids in that respect). The antenna contains 2 or less funicle segments (i.e. antenna appearing 5-segmented or less with the club counted as one). Quite often the wing setae are arranged in rows.

COLLECTING: Usually overlooked when sweeping because of small size, these wasps are fairly common, but in small numbers. Sweeping meadows and herbaceous vegetation is often productive, however, collecting and rearing eggs, especially Lepidoptera, will usually produce the best results.

DISTRIBUTION: Trichogrammatids are distributed worldwide and are part of the "usual" fauna of small, remote islands by virtue of their easy dispersal by wind (Doutt and Viggiani 1968).