Around 400 species of mantispids, placed in around 44 genera, are known (Ohl and Oswald 2004). Diversity is higher at lower latitudes. For example, according to Reynoso-Velasco and Contreras-Ramos (2008), the known mantispid fauna of Mexico includes around two dozen species, whereas the (surely more complete) species list for all of the United States and Canada includes just half this number, nearly all of which are present in Mexico as well (and according to Cannings and Cannings , only four species occur in Canada). Caribbean island diversity is equivalent to that of the entire Nearctic (Canada and the United States), although just a few species are not shared with continental areas. A summary of species richness by zoogeographic region reinforces this pattern, with the numbers of species known from the Nearctic or Palearctic regions being many times fewer than for other regions (Ohl and Oswald 2004), despite the fact that these two regions are presumably the most thoroughly known.
With the exception of the subfamily Mantispinae, little is known of the biology and ecology of mantispids. The Mantispinae, however, are relatively well known. In contrast to the scattered data available for the several other subfamilies, which suggest that a wide range of insects may be used as food by developing individuals of various mantispid species, known larval developmental associations of species in the Mantispinae are all with the egg sacs of spiders and the larvae are predators on spider eggs. Larvae of some mantispine species are obligate egg-sac penetrators. These larvae locate and penetrate spider egg sacs that have already been produced and deposited in the environment. Other mantispine species locate and board spiders, then enter the egg sac during its construction. Larvae of still other species use either or both techniques to get inside spider egg sacs. If larvae enter an egg sac containing nearly hatched spiderlings, they may board these. It is not clear whether larvae are able to search out spiders from a distance. Given that adults produce egg clutches containing from several hundred to several thousand eggs, it is possible that searching is purely random and that encounters with spiders are simply fortuitous. Although larvae would presumably benefit from boarding only female spiders, in all species that have been studied larvae board male and female spiders with equal frequency. However, in at least one species, larvae that find themselves on a male will transfer to the male's mate when he copulates; in at least one other species, larvae on a male spider will transfer to a female who cannibalizes the male. (Redborg 1998 and references therein)
Although it has long been known that mantispine larvae feed on spider eggs, only in the 1980s was it recognized that some (possibly all) spider boarders feed on spider hemolymph ("blood") while aboard adult spiders. Spider-boarding mantispids overwinter on their host spiders. The overwintering behavior of egg-sac penetrators is more varied. In some species, individuals overwinter as a mass of unfed first instars and search for egg sacs the following spring. Others may overwinter as larvae or pupae within spider egg sacs. (Redborg 1998 and references therein)
Once inside a spider egg sac, a mantispid larva pierces and drains eggs with its modified mandibles and maxillae. As a third instar larva, it constructs a pupal cocoon within the spider egg sac. In the laboratory at 25º C, the entire life cycle for the well studied Mantispa uhleri takes around 28 days from first instar to adult. The incipient adult bites its way out of both the cocoon and egg sac and walks some distance away before undergoing the final molt, presumably because egg sacs may be located in concealed or awkward places such as under tree bark or within a silken retreat. (Redborg 1998 and references therein)
Although data on which spider species are preferentially attacked are very incomplete (reviewed by Redborg 1998), some tentative generalizations are possible. Host range tends to be fairly broad, with few mantispine species believed to have a host range limited even to a single family. Egg sac-penetrating mantispids probably tend to have broader host ranges and to be more often associated with web-building spiders than are spider-boarding mantispids. Spider-boarding mantispids may be more likely to be associated with wandering spiders, while mantispids associated with the egg sacs of web-building spiders may be more likely to be egg-sac penetrators. At best, however, these are rough generalizations. (Redborg 1998 and references therein)
Little is known about the nonreproductive habits of adult mantispines. Several species appear to be wasp mimics. In the few mantispine species in which mating behavior has been observed, courtship begins with male and female facing each other. Both male and female slowly extend forward and flex the coxa and femur of one raptorial leg followed by identical movements of the other leg. Males possess a glandular abdominal epithelium that produces a suspected pheromone, although this requires further investigation. Copulation occurs venter to venter, which necessitates twisting of the male’s abdomen through a 180º angle to achieve the final orientation of male and female facing away from each other. Copulation duration ranges from 1 hour to 24 hours among species that have been studied. Postcopulatory sparring may occur. A white opalescent sperm-containing spermatophore, present at the tip of the female abdomen following mating, is generally absorbed within 24 to 48 hours. Mantispids lay discrete clutches of very large numbers of eggs, each of which is attached to the substratum by a short stalk produced by secretions from the female's colleterial ("glue") gland. Depending on the species and the size of an individual female, clutch size may vary from a couple of hundred to several thousand eggs, with one to two clutches produced per week for several weeks or even months. (Redborg 1998 and references therein)
Ohl and Oswald (2004) catalogued the world mantispid fauna, including all fossil and extant species known. They present data on original and current generic placements and synonymies and include literature sources, type status, type locality, type depository, and taxon distributions.