The following text is from: E. Eric Grissell and M. E. Schauff. 1997. A Handbook of the Families of Nearctic Chalcidoidea (Hymenoptera): Second Edition, Revised. Entomological Society of Washington. 87pp.

Literature

All references in these files are contained in the Chalcid Literature document.

Systematic Overview

The Chalcidoidea eventually may be recognized as numerically among the largest and most biologically diverse of all insect groups. Recently published figures indicate that chalcidoids already equal the number of Ichneumonidae in described species (ca. 19,000) and estimates by current workers in the group agree that between 60,000 to 100,000 species is not unreasonable (Gordh 1979; Noyes 1990). Although the numbers alone are somewhat daunting, they only hint at some of the many problems, both discovered and yet to be discovered, that await us in the Chalcidoidea. Because of the relative abundance of species (mostly undescribed), the worldwide geographic range, the complex biological and often baffling morphological diversity, and the few taxonomists who have studied the superfamily, much remains to be accomplished if ever we are to understand the Chalcidoidea in a comprehensive way.

There is a great amount of confusion even among specialists about the definitions of families, genera, and species of Chalcidoidea, and a handbook such as this one cannot possibly solve the problems that even now puzzle the experts. Therefore we have taken a practical, middle of the road approach in writing this handbook. For those readers interested in a general discussion of some aspects of chalcidoid taxonomy we provide a discussion in the remainder of this section. It is an overview of chalcidoids with reference especially to biological and morphological diversity and its effects on taxonomy and identification. However if you are simply interested in identifying specimens you may wish to proceed directly to the identification key on p. 20.

Our knowledge of the evolutionary relationships of the major groups of chalcidoids is still in its infancy. Recent work suggests that the Mymarommatidae are the probable sister group to the rest of the superfamily and, as mentioned elsewhere, there is still some debate as to whether or not mymarommatids are true chalcidoids. There also seems to be evidence to indicate that the Mymaridae are the likely sister group to the remaining families of Chalcidoidea. Within the bulk of the chalcidoids, however, interrelationships are not well understood, and workers often refer to "lines" of families to indicate groupings of taxa which seem to be related to one another. One group that has been investigated to some degree is the Encyrtidae/ Eupelmidae/ Tanaostigmatidae complex which, evidence suggests, does represent a cohesive and closely related group linked by shared derived characters (LaSalle 1987, Gibson 1989). Other workers refer to the "eulophid line" (Eulophidae, Signiphoridae, Aphelinidae, Elasmidae) or "pteromalid line" (Pteromalidae, Eucharitidae, Perilampidae), but these other lineages have not been shown to be natural groups. Much work remains to be done and it will likely be many years before higher level relationships are clearly understood.

Family Classification: The number, definition, and limits of chalcidoid families have been in a state of flux since the early part of the 1800's. The number and components of families have varied from 1 (Westwood 1840, Howard 1885-86, Handlirsch 1925) to 9 (Riek 1970), 11 (Burks 1979), 14 (Ashmead 1904), 16 (Yoshimoto 1984), and 20 and over (Foerster 1856, Nikol'skaya 1952). The most recent, comprehensive treatment lists 21 families (Boucek 1988), but this number was reduced to 20 by Gibson (1993) who removed Mymarommatidae to its own superfamily.

Given this diversity of opinion one might assume that the family and subfamily classifications are not yet the most stable of systems. One would be correct. There is still no consensus among workers as to exact family limits, and we are convinced there will be none for some time to come. We attempt to strike a compromise position and rely on recent works by specialists in the various groups as well as discussions with them to indicate what suprageneric names to employ. We treat 19 families but exclude the Rotoitidae known from New Zealand and Chile (Boucek and Noyes 1987, Boucek 1988, Gibson 1989). We treat Mymarommatidae as its own superfamily, realizing that for most users of this handbook its technical placement will not matter.

It may seem odd that a relatively important group such as chalcidoids remains so poorly understood, but, as we have previously noted, the number of taxonomists working in the group has been underwhelming and the number of taxa conversely large. Added to this is the fact that much of the family classification was hypothesized by European workers of the 19th century working primarily on the European fauna. We are beginning to realize that Europe is not necessarily the center of chalcidoid diversity! In addition, the small size of most chalcidoids made it difficult for early workers to study many characters. Thus, relatively easily observed and countable characters such as antennal segmentation, wing venation, and tarsal numbers became of prime importance in defining taxa without much regard for evolutionary implications. These factors have resulted in chalcidoid workers basing classifications upon inadequate samples of the available morphological characters and upon only a small portion of the extant fauna. Present workers may resample the same faunas (or explore new ones) and find material that does not precisely fit the system previously followed. They then either try to force the new material to fit into the old system or they try to construct a new system. It is this striving for the "new system" which is reflected in the numerous instances of family classification cited above. The more new sets of data that are added, the more difficult becomes the assessment of higher classification.

This is not to say that no progress is being made towards a better understanding of the evolutionary relationships of the many groups that make up the superfamily. For example, recent studies (Gibson 1986a) have shown that the superfamily itself is, indeed, a natural group. This had not been well established previously and resulted in some workers moving groups like the Mymaridae back and forth between superfamilies. As a result, we are steadily gaining a better handle on some higher level relationships within the chalcidoids as well as creating a basis from which to start inferring the relationships of others. Nevertheless an enormous amount of work remains to be done.

Ultimately, to help resolve the problems of higher classification, we will have to accept that most of the families we currently recognize are simply "convenience groups" that allow some ease of communication about large assemblages of taxa and also make it easier to provide names for identification. In the original version of this handbook (1980) we stated that the only method of solving the problems of the higher classification of Chalcidoidea would be to rebuild its structure from a foundation of monophyletic groups. Although this would mean some temporary uncertainty, it would ultimately lend stability to a revised classification (Heraty and Darling 1984; Gibson 1989). Until this goal is attained, however, it seems most conservative to emphasize the family groupings as visualized at present. By operational definition they will be approximately correct for our purposes.

Generic classification: The number of valid world genera of Chalcidoidea was most recently estimated at about 2000 (Noyes 1990). Recent estimates for the Nearctic place the number of genera at over 700 (Gibson, et al. 1997), an increase of nearly 40% from the figures cited less than 20 years ago (Krombein, et al. 1979). In the Nearctic Region at least half the chalcidoid genera contain only 1 species each. We are not certain if this is true also for the rest of the world, but we suspect so. This results in a few genera with many species and many genera with only 1 or 2 species each. We feel that these figures indicate that there are a large number of poorly or artificially defined genera. (This critical lack of generic definition also makes it difficult to define subfamilies and/or families.) Complicating the factor of large numbers of genera is the lack of written information with which to identify them. For example, the family Pteromalidae has about 230 genera in the Nearctic region but we have no keys with which to identify specimens to genus. Often there are keys to species, but because there is no way to identify genera these keys are not accessible. In the case of Pteromalidae the best we can do is run specimens through Palearctic keys (Graham 1969, Boucek and Rasplus 1991), realizing that one third of our genera are not recognized in that key. The same is true for Eulophidae with about 120 Nearctic genera. The good news, however, is that this situation will change shortly because there is a multinational, cooperative project underway to produce a key to all nearctic genera of Chalcidoidea (see Literature Resources, p. 9).

Species classification: While identification at the generic level is useful, what most researchers want is a species name. Unfortunately this level of comprehension is well beyond our abilities in the present handbook. Indeed, there are literally thousands of pages of literature devoted to the descriptions and identifications of chalcidoid species, but most of it is diffuse and difficult to assimilate. In spite of the seemingly large amount of literature, much of it is of little use because the results were based upon limited population sample sizes, limited understanding of population variation, inadequate geographic samples and analyses, inferior microscopes, and poor microscopic techniques.

All of the available literature at the species level relies upon two basic methods of making identifications. The first and most commonly used method is the morphological approach wherein taxa are defined by stable and consistent differences in appearance. Most chalcidoid taxonomy is based upon external morphology, but internal characters may be used as well. The second approach is to test the reproductive isolation of taxa by hybridization studies. Either method may initially rely heavily on host and habitat data to suggest that more than one taxon is involved. Additional information used to identify species may include aspects of life history such as placement and shape of emergence holes and meconia, pupal coloration and its duration over time, courtship and mating behavior, and karyological characteristics.

Hybridization studies mostly have been applied in two genera, namely scale parasites of the genus Aphytis (Aphelinidae) and egg parasites of the genus Trichogramma (Trichogrammatidae). In these groups the term "cryptic species" has been applied to morphologically indistinguishable but reproductively isolated populations. Knowing that cryptic or sibling species occur in several families of Chalcidoidea, and suspecting that it may be a common phenomenon, one wonders if one shouldn't consider the majority of chalcidoid species as "morpho-species", with each morphologically distinct taxon possibly representing dozens of biologically distinct species (variously called cryptic species, sibling species, semi-species, races, strains, or biotypes).

A factor which greatly complicates any approach to studying chalcidoids is that we do not yet fully understand the degree of morphological and biological plasticity inherent in a single species. Some aspects of variation known to occur within species of chalcidoids are as follows:

Different hosts may cause progeny of the same female wasp to vary morphologically. For example, in one species of Trichogramma (Trichogrammatidae), eggs deposited in one host will produce normal male offspring but when placed in another host will produce a male with no wings and reduced antennal segments (see Trichogrammatidae section).

Different parts of the same host may cause morphological variation; e.g. in one species of eulophid the progeny will vary depending on whether they feed on body tissue of the host or haemolymph of the host (see Eulophidae section).

Different sizes of the same host may cause 3 or 4-fold differences in size of progeny from the same female; this may result in changing ratios of measured body parts such as antennal segments, ocellar distances, ovipositor lengths, and numbers and sizes of setae and sensilla. Additionally, sculpture and carinae on specimens from small hosts may be obscure or absent when compared to larger specimens, color of small specimens is usually darker than large specimens, and finally, progeny from the same female may even vary biologically when reared from different sizes of the same host. For example, in a species of Trichogramma, large progeny from large hosts oviposit faster than small progeny, they also live longer, they have a higher reproductive potential and attack mostly large hosts (see Trichogrammatidae section).

In addition to host induced variation, seasonal dimorphism is also known in chalcidoids. This has been shown in several families (e. g. Torymidae, Eulophidae) and usually takes the form of darker individuals produced in the fall. Ovipositor and wing lengths have also been shown to vary seasonally for these families.

Concerning our knowledge of chalcidoid species in general, we find two rather opposing problems. As hybridization studies have shown, there can be complexes of morphologically identical but non-interbreeding "species." On the other hand, as rearing studies have shown, the progeny of a single female may vary so greatly in morphology and even biology that their progeny do not even appear to be the same species. Delucchi, Rosen, and Schlinger (1976) summed up the species problem this way:

"The identification of a species according to the morphological concept ... represents a necessary step to the final determination of a true specific entity. Whether or not the morpho-logical criteria are sufficient...will be decided case after case by taxonomist-biological control specialists."

In summary, then, we believe that the problems of interpreting families, genera, and species are not easily solved even with the most sophisticated of study methods. The museum taxonomist, the identifier, the biologist, the ecologist, and the biocontrol worker must all be aware of both biological and procedural problems if we expect ever to achieve a workable concept of taxa in the Chalcidoidea.

LITERATURE RESOURCES

New World (western Hemisphere)

The New World Chalcidoidea are well cataloged so that references are available to list the names (and sources of descriptions) of almost all the described New World species. Note that we used the term "list." We might know what the names are, but that doesn't mean we know how to identify or apply those names.

The Nearctic Chalcidoidea were cataloged by Peck (1963) who gave complete bibliographic references to all included species. This listing was updated (in a simplified version) by Burks, Gordh, and Grissell (all 1979) in the "Catalog of Hymenoptera in America North of Mexico." Unfortunately some parts of the catalog were completed in the early 1970's and thus were outdated before publication. Because of this, and also because no update of the catalog has been published since 1979, it is necessary to consult the Nearctic literature from the period of about 1972 to the present to have an accurate list of species for this region.

The Neotropical species are nearly completely cataloged to date thanks to a series of papers by L. DeSantis. In 1967 he cataloged the species of Argentina and in l980 the species of Brazil. In 1979 he cataloged all other regions of the Neotropics. He published a first supplement to these catalogs in 1983, a second in 1989, and a third in 1994 (with P. Fidalgo). A recent monograph on the Hymenoptera of Costa Rica (Hanson and Gauld 1995) contains considerable information on the chalcidoid families including some keys to subfamilies and/or genera.

Probably the most used of all Nearctic keys to families of chalcidoids are the various editions of "An Introduction to the Study of Insects," now in its 6th edition (Borror, Triplehorn, and Johnson 1989). Recently Yoshimoto (l984) wrote a guide to the Canadian families and subfamilies of chalcidoid wasps. This is the first effort of its kind in the New World and contains keys to families and subfamilies, 95 illustrations (many habitus), and a write-up of each group. Both of the above works rely on rather traditional keys that can be difficult for student or professional alike. In addition, the keys are not entirely accurate in technical detail. Our handbook is broader in geographic scope than is Yoshimoto's and more up-to-date. We have striven to produce a workable, easy to use, but technically correct, key to families and subfamilies.

The only key available to the genera of Nearctic Chalcidoidea is the recently published one by Gibson, et al. (1997). This key was produced by the joint efforts of most of the taxonomists working today on Nearctic chalcidoids. Ashmead (1904) produced a key to world genera, a monumental undertaking at any time, but it is of little use today.

Old World (eastern Hemisphere)
There is no single, comprehensive catalog for Old World species. There are taxon specific catalogs, however, some of which are world wide, and we mention these under the various family discussions. Relatively broad-based coverage of Old World Chalcidoidea in general may be found in the following sources: PALEARCTIC: Boucek (in Peck, Boucek, and Hoffer 1964; keys to families and genera of Chalcidoidea of Czechoslovakia, but a useful place to start even for Nearctic workers); Nikol'skaya 1952 (English translation in 1963 of Russian work that keyed genera and many species of European USSR); Tryapitzyn 1978 (English translation in 1987 that keys all known genera and species of European part of former Soviet Republic). AFROTROPICAL: Prinsloo 1980 (a concise guide to the families of sub-Saharan Africa). AUSTRALIAN: Riek 1970 (a key to families of Australia, a new classification, and general comments about the families); Noyes and Valentine 1989b (key to numerically smaller families and genera of New Zealand). AUSTRAL-ASIAN: Boucek 1988 (a review of 14 families of Australasian chalcidoids with keys to 550 genera and lists of over 2360 valid species). ORIENTAL: Mani 1938 (catalog of chalcidoids of India); Subba Rao and Hayat l985 (keys to families and genera of Indian chalcidoids) and l986 (catalog of Indian chalcidoids); Mani 1989 (keys to genera of Indian chalcidoids). There are additional, regional sources of literature, but they are fairly widely scattered.

World
Recently Gibson (1993) produced an illustrated key to world families of Chalcidoidea that attempted to key nearly all known forms regardless of sex or polymorphic states. Gibson also reviewed subfamilies, biology, taxonomy, and literature at the world level.

BIOLOGICAL OVERVIEW

As with most large groups of organisms a variety of types of exploitation exist and it is not always easy to generalize even at the generic level. However, some overall comments may be made. Hagen (in DeBach 1964:219-220) summarized the types of feeding behavior of all parasitoids under the following specific categories (somewhat modified for the purpose of this handbook):

Of the above fourteen categories Hagen excluded chalcidoids from two, namely categories 2 (eggs hatch after ingestion by host) and 14 (eggs laid in adult insects and emerge from adult insects). To date there is still no evidence to support category 2, but category 14 may be supported by at least the genus Tomicobia (Pteromalidae) which attacks scolytid beetle adults. In essence the chalcidoids are the only group of parasitoids that exploit virtually every type of feeding situation cited above. In addition to Hagen's list, chalcidoids are also obligatorily phytophagous (e.g., seed-feeders) and some are "omnivorous" in either a facultative or obligate manner. No other group of parasitoids (or possibly other group of insects) demonstrates this wide a range of feeding habits.

This diversity of feeding types is reflected in the extensive list of hosts for chalcidoids which includes 13 orders of Insecta and 2 of Arachnida (Araneida, Acarina). Chalcidoids attack the holometabolus orders Neuroptera, Coleoptera, Lepidoptera, Diptera, Siphonaptera, Strepsiptera and Hymenoptera, as well as members of the Odonata, Orthoptera, Thysanoptera, Heteroptera, Homoptera, and Psocoptera. The other groups of parasitic Hymenoptera together attack about the same number of orders.

Those species that are strictly phytophagous either cause galls on plants (especially grasses, e.g., some Eurytomidae) or are seed-feeders (especially in coniferous and rosaceous plants, e.g., Torymidae). Some species of chalcidoids feed initially on the host insect and when this is gone they continue to feed on plant tissue. This is especially true of parasitoids of gall-forming insects (e.g., Eurytomidae, Torymidae). There is at least one report of a eurytomid that feeds initially on plant tissue surrounding a host until the host larva is mature, then the eurytomid consumes the larva.

Families, genera, and species of chalcidoids may be classified as specialists, generalists, or opportunists depending in part upon the heirarchy of the category. For example, the families Mymaridae and Trichogrammatidae are specialists in that all species (so far known) are internal parasites of insect eggs. Other specialist families (or subfamilies, depending upon the taxonomist) include the Agaonidae -- all pollinators of figs, the Eucharitidae -- all parasites of ant larvae and pupae, the Leucospidae -- all parasites of bees and wasps, and the Tanaostigmatidae -- all phytophagous in plant tissue.

Within the Trichogrammatidae and Mymaridae, some genera are also specialists on certain groups of eggs, such as bruchids, bugs, or Lepidoptera. Some species, however, within a specialist genus may display some rather generalist sort of behavior. For example, a given Trichogramma species may be able to oviposit into almost any insect eggs in a particular habitat, such as savannah, arboreal, or aquatic.

The same is true of some genera that attack, for example, leaf mining or gall-forming insects. The insect in the leaf mine might be lepidopterous, dipterous, or even another hymenopteran, but the chalcidoid will be able to parasitize it as long as the habitat is appropriate.

A few chalcidoid species appear to be opportunists in that they are reared from many insect orders in many habitats. This is not particularly common at the species level, but several families demonstrate a very polyphagous nature. The Encyrtidae, for example, parasitize everything from ticks to beetles, aphids, coccids, Lepidoptera, adult Heteroptera, and even themselves.

Additional details are given in the family discussions below.

COLLECTING and MOUNTING

The rather specialized subject of collecting and mounting chalcidoids was discussed thoroughly in a well-illustrated paper by Noyes (1982, updated 1990c). We do not explain all of the equipment or techniques here because this information is available in most generalized or introductory insect books. Instead, we explain how the equipment and techniques pertain specifically to chalcidoids.

COLLECTING
Chalcidoids may be collected by using three broadly defined methods: 1) rearing, 2) active sampling (e.g., sweeping, beating, vacuuming), 3) or passive sampling (e.g., trapping). Each method has its own advantages and special uses, and in general all methods should be used where possible.

Rearing: This is by far the best method for scientific purposes. Not only is something learned about the habitat and host association of the specimens obtained, but males and females may be positively associated and large numbers of specimens may be obtained with which to study variation. This all leads to a better understanding, or characterization, of a species.

As a purely practical matter, knowing something about the host of a specimen helps quite a bit in identification. For example, a chalcidoid reared from a butterfly egg almost certainly will be a trichogrammatid or eupelmid, but a chalcidoid wasp reared from a leafhopper egg will most likely be a mymarid. A chalcidoid reared from ants will likely be a eucharitid, but rarely it might be a eulophid.

True host associations may be more subtle than first observation would indicate. For example, leucospids are parasites of stinging wasps and bees. It is important to know that bee and wasp hosts can nest in old plant galls. Therefore it is possible to rear a leucospid from a gall.

Biological information must be tempered with reason and preciseness. It is difficult to make "absolute" statements in biology, and we are always running across biological surprises. Although it may take some time to become familiar with host associations of chalcidoids, we believe it is well worth the effort.

Exact and precise rearing methodologies go beyond the scope of this handbook, but they really are just matters of degree of refinement. The simplest method is merely gathering up host material and placing it in a plastic bag, jar, or box ... the "host-in-a-box" method. For example, one may gather up 500 rose hips, place them in a container, and see what happens. Sometimes nothing happens and you start again. Sometimes endless numbers of insects of all kinds emerge. Sometimes it is only great numbers of one kind of insect. This is a coarse level of rearing and you really don't know exactly what is going on (but you still have specimens to study!). With this method you may rear parasites from hosts on the surface of the hip (e.g., scales, or insect eggs of any number of orders), from the pulp of the hip (e.g., a boring lepidopterous, dipterous, or coleopterous larva), or from the seeds (e.g., a true seed-chalcid, or parasite of the seed-feeder). This relatively crude rearing allows you to assay the whole fauna.

Studies must become more refined to get at the specifics. To continue with the case of rose hips, for example, the seeds might be extracted and reared as separate units from the pulp. You would then know exactly what came out of a seed as opposed to the surface or pulp. You still would not know what went on inside the seed (there might be an entire community of insects inside), but you would be getting more refined information than the original "host-in-a-box" method. Ultimately you might want to dissect each seed and study the interactions of all the life-forms. Only then will you actually know "what" the parasitic wasp is doing in the host.

Rearing can be as easy or difficult, as gross or precise, as you want to make it. It is somewhat easier to rear parasites from mature or quiescent hosts because you don't have to go through the process of rearing and feeding the host to obtain the parasite. For example, mature scales, seeds, fruits, galls, mines, eggs, or pupae do not require much investment in active maintenance to obtain results. Sometimes parasites emerge the same day the host is collected. On the other hand, seed-feeders may take a year or two to emerge (but at least you don't have to grow the plant the whole time if the seeds are mature). With larvae or other immature hosts, however, you will have to feed and maintain the host long enough for the parasite to develop. Often this requires specialized conditions such as fresh host-plant material or environmental controls to induce or break diapause.

One way to get started rearing chalcidoids is to work from the known host ranges, that is, to collect hosts that are known to harbor parasites in the first place. If you know chalcidoids have been collected from wild rose seeds or gypsy moth pupae it is relatively easy to find and rear these hosts. If, on the other hand, you want to make an entirely new host discovery, you could try confining poison oak seeds, damselfly naiads, or any of a thousand other hosts for which no records are known and hope that something emerges. By consulting the host list in a publication such as the "Catalog of Hymenoptera of America North of Mexico" (Krombein, et al. 1979) you can see what hosts have already been reported. Everything else is fair hunting for new host records and even new genera and species of chalcidoids!

Active Sampling: By active sampling we mean "collecting" in the traditional sense of the stereotypical entomologist, net-in-hand, wildly swinging away. This is a wholesale approach to gathering specimens because you often don't know what you've collected until you get back to the lab and sort things out. Sometimes it is possible to infer some biological information about a chalcidoid when collecting in this fashion, but not very often.

With the net method, chalcidoids are rarely "chased-down" like butterflies. Instead the net is used as an imaginary broom to sweep habitats of interest. The simplest method is to sweep everything in sight, choke the net bag off, and put it temporarily into a large killing bottle. When everything is dead, sort through the sweepings (often contaminated with sticks, stems, leaves, seeds) and pick out the target specimens. The resultant material is relatively dry so care must be taken in sorting. Generally it must be sorted the same day as collected. A variation on this technique, and the one we use most often, is to dump the net-bag contents into a gallon-size jar containing a pint or two of alcohol. This must be done quickly, not only to prevent specimens from flying out, but also to avoid stinging wasps that might have accidentally been caught in the sweeping process. With material in alcohol you can sort at almost any time, even weeks after collecting. This wet method is generally better for the specimens than the dry method, as will be explained later under the "Mounting" section.

Variations on the above techniques are certainly possible. For example, instead of an insect net you could use a beating sheet or vacuum device. Also, specimens may be removed from the net by several means. One is to use an aspirator (suction device) to pick out just the specimens of interest. Because the specimens are still alive, this requires holding the net so material does not escape, finding what you want, getting the aspirator to it, and sucking up the specimen. Some dexterity, agility, and coordination are required. Another method is to quickly invert the net bag into a trap or light box of some type in which the specimens work their way from the dark interior of the box (trap) to a light place (usually a collecting tube of some sort).

Refinements of the general sweeping technique are possible. We prefer to begin sweeping in as narrow a habitat-type as possible. For example, in a woodland-savannah situation, we would first sweep one species of tree. If we found nothing we would switch to another species and so on. Then we would sweep shrub species, individual species of herbaceous plants, or species of grasses. In doing so, it is possible to gain some potential biological data otherwise overlooked by broad-scale sweeping. For example, we have found that sweeping a single species of composite in bloom yielded quantities of tephritid or cecidomyiid parasites (often the potential host is swept at the same time), sweeping an oak produced numerous cynipid parasites, and sweeping holly produced huge numbers of holly seed-feeding chalcidoids.

As often as not, nothing is collected by sweeping in this narrowly defined manner. If it does not work we then try a broader approach. For example, at some point it pays to sweep habitat types. That is, we sweep foliage of trees and shrubs, pond-side grass-es and sedges, upland grasses, herbaceous fields, or grazed pastures. By doing this we can often obtain numerous species of chalcidoids that we define only as "grass inhabiting" species or "woodland" species. It is also known that different strata of the same habitat produce different species of wasps. For example, wasps found on ground-level grasses (grazed or mown) may be different than wasps found at the tops of tall grasses in the same habitat.

There is one special sort of collecting site which is often the very best in terms of quantity and diversity. Although it is not commonly found, and is not particularly biologically informative, these areas may yield specimens of nearly every family of chalcidoid. The collecting site is one that provides abundant nectar or honeydew in a restricted habitat. A small patch of flowers blooming in the middle of an otherwise dry field may attract wasps from miles around. Small-flowered plants such as Baccharis, Eriogonum, Euphorbia, or umbellifers are particulary good for small wasps. A spring, seep, leaking horse trough, or cloudburst puddle in the middle of a desert or along the drainage of a dry road is often rich in material if plants have sprung up in response to the moisture. A tree or shrub full of honeydew-producing aphids can be exceedingly rich in a diversity of wasps which have no interest in the aphids themselves. Some plants, having extrafloral nectaries that attract throngs of wasps, are easily overlooked because there is no showy flower to be seen from a distance. A highly disturbed site, full of introduced, blooming "weeds," can be an attraction to wasps, even if few other native plants are blooming in the area. These types of sites are a mecca for small Hymenoptera, and finding them is often difficult. The results can be overwhelming.

Passive sampling: This form of collecting relies on devices, or traps, that collect continuously while the collector is doing something else. It is not a lazy-man's way of collecting, however, because the traps must be transported, set up, regularly serviced, and repaired. Traps may be stolen, overrun by cattle, run over by off-road vehicles, vandalized, flooded, or simply fall apart. Still, they can be incredibly productive because they allow the "collector" to be in many places at once and to work 24 hours a day.

There are numerous types of insect-collecting traps and a discussion of them would take up quite a bit of space. The book by Steyskal, et al. (1986) (available from the U. S. Government Printing Office, Washington, D. C. 20402) lists and describes many such traps. Noyes (1989) discussed the merits of five different sampling methods (including sweeping), and cited several papers concerning specific traps and their values. Although he concluded that "... for most groups of Hymenoptera, sweeping was the most effective single method of sampling," he also stated that sampling should be "... conducted using as wide a variety of methods as possible because every method undoubtedly has at least one advantage over any other."

The traps that seem to work best for chalcidoids are the flight-interception type (e.g., Malaise trap, windowpane trap) and the yellow-pan or Moericke trap. Other traps may be used, but we have found them to be less productive than the ones just cited. Light traps, for example, generally are not too useful, nor are pitfall traps. Berlese funnels might prove useful but are underused. Technically they might not be considered a trap, however, because you have to bring material to them in the form of soil litter, duff, moss, etc.

MOUNTING
For the most part, chalcidoids are too small to be mounted directly on insect pins. A few of the larger Chalcididae, Leucospidae, or Pteromalidae may be directly pinned, but most chalcidoids are glued on the right side of the thorax to small card triangles called points through which an insect pin is placed. Some workers use card rectangles instead of triangles, and glue the specimen nearly flat upon the card. This is a matter of preference. Extremely small specimens (e.g., Trichogrammatidae, Aphelinidae, Mymaridae) are usually mounted on slides, though some workers glue the specimen to a hair taken from an artist's brush which is, in turn, glued to a point. Mounting techniques may be improvised to the point of extreme, but the result should be to secure the specimen to a substrate that allows complete examination from all angles. This is often difficult at best, but is impossible in the case of slide-mounted material which requires a two-dimensional plane for best visibility.

Slide-making is a long, involved process often approaching an art form. In the past it was the only useful method to mount and examine tiny, weak-bodied specimens that would shrink and distort as soon as dehydration took place. Once the water has left the body the only way to recover a specimen's original shape is to soak it in chemicals and slide-mount it. Now, however, there are drying techniques that allow such small wasps to be handled more easily. (These methods work only if the original specimens are collected directly into alcohol or are freshly killed, that is, they cannot already be shrunken.) The most important drying method is called critical point drying (see discussion by Gordh and Hall 1979), but this requires the use of expensive equipment and is far beyond the scope of home use. A reportedly good, cheap method is to place the specimens in an open, shallow container with a weak solution of alcohol which is then placed in the freezing section of a frost-free refrigerator. In several days, or weeks, the alcohol-water mixture is removed (sublimated) from the specimens while the body remains turgid. When removed to an ambient temperature, the body does not collapse because there is no mositure left in it to evaporate. These dehydrated specimens may be mounted on a small bristle or hair as explained above. They also may be placed in gelatin capsules that are pierced with a pin, labeled and treated much like the more hard-bodied specimens. To examine such specimens requires specialized equipment such as compound or scanning electron microscopes. Admittedly this is not for home use. In general it might be said that studying the very tiny chalcidoids is beyond the scope of a hobby or home-time activity.