Chapter 13

Biogeography

Biogeographic theories are only as robust as the knowledge of the actual distribution of genera and species under consideration. In the case of Trichomycetes, it is probable that only a fraction of extant taxa are known. Nonetheless, Trichomycetes have been collected widely and, in some regions of the world, intensively (Fig. 13.1). Many trichomycete genera are almost cosmopolitan wherever their particular kinds of hosts occur, but many species appear to have more limited distributions and may be locally endemic. Where trichomycete species are located is dependent on the habitats of their specific hosts, and, if their host range is narrow, this may account for the limited distribution of some gut fungi. On the other hand, species of Trichomycetes with a broader range of hosts may have a distribution that exceeds that of individual host species. Many examples of this are known (Chapter 11), and a few are cited below.

Identifying patterns of fungal distribution, incomplete as these may be in many instances, is a means to develop theories regarding their historical biogeography. These theories can then be tested by field work directed specifically toward seeking particular kinds of fungus-infested arthropods in particular regions of the world. Current views on trichomycete biogeography are based on several hypotheses:

Hypothesis 1. Ancestral Trichomycetes began to adapt to arthropod gut habitation shortly after insects began to evolve in aquatic habitats some 250 to 190 million years BP.

Evidence that supports this hypothesis is that (a) Harpellales show closer ancestral affinities to the Zygomycetes than do the other trichomycete orders, and are believed to be the first Trichomycetes to have evolved; and (b) all species of Harpellales (with two known exceptions) live in the most primitive aquatic insects, namely stoneflies (Plecoptera), mayflies (Ephemeroptera), and Nematocera (the most primitive suborder of Diptera).

Hypothesis 2. Among plants and animals that inhabited different land masses as they drifted apart during the breakup of Pangea were insects carrying Trichomycetes within their guts. These tectonic events account at least in part for the widespread distribution of many current genera and species of Trichomycetes.

Hypothesis 3. Dispersal of Trichomycetes over distances beyond their immediate host populations normally occurs only in consort with their dispersing hosts.

Propagules of Trichomycetes (trichospores, zygospores, sporangiospores, arthrospores) are adapted primarily to transmit the fungi from one host individual to another over relatively short distances, although downstream drifting of aquatic fungal propagules may occur and ovarian cysts of Harpellales may be carried to new sites by adults. The number of spores produced by individual thalli of Trichomycetes are very few as compared with most other fungi. While phoretic and wind dispersal of Trichomycetes may occur occasionally and randomly, these mechanisms are not sufficiently reliable to have evolved as the primary means of dispersing spores over long distances, because they would have to be deposited by chance close enough to a suitable host to be ingested.

Given the above hypotheses, one can comment on a few distributions based on the types of host habitats.

Marine habitats.

Marine Eccrinales and Asellariales are symbionts of various kinds of crustaceans, namely isopods, amphipods, crabs, and anomurids. The majority of the crustacean hosts are intertidal. Though ocean shorelines may be continuous along a continent or around an island, habitats of the crustaceans are not. They may be restricted, for example, to mud flats, sandy shores, or rocky shores. Successful dispersal of fungus-infested crustaceans to new habitats may have occurred from time to time, for instance by rafting on flotsam.

In the past, crustaceans carrying eccrinids in their guts dispersed widely as they radiated. The most notable generalization is that in many instances the gut fungi did not speciate during much of this time. A good example is Asellaria ligiae (Asellariales) that has been found in the hindgut of various species of the isopod genus Ligia that inhabit upper shore lines or splash zones in the Pacific, Atlantic, Caribbean and Mediterranean shores (Fig. 13.2). A similar situation has been reported in Palavascia sphaeromae (Eccrinales) that lives in several species of the intertidal isopod family Sphaeromidae in France, Japan, and the eastern coast of the USA (Lichtwardt, 1995). Other Eccrinales have a wider host range, and have been found on shores in several oceans (e.g. Enteromyces callianassae and Taeniella carcini in both true crabs and anomurids). It thus appears that some Trichomycetes are more evolutionarily stable than are their hosts, and one might conclude that the fungal species may have established a symbiotic relationship early in the history of host radiations.

Terrestrial habitats.

Sporulating Eccrinales and Asellariales, in contrast to Harpellales, are found most often in mature arthropods, though immature stages may also contain gut fungi. In some cases human activities, such as transport of agricultural products, may have been responsible for spreading fungus-infested arthropods. A few examples include Enterobryus oxidi (Eccrinales) in the common greenhouse millipede, Oxidus gracilis; various species of springtails (Collembola) carrying Orchesellaria mauguioi (Asellariales); and pill bugs and sow bugs (Isopoda) infested with species of Parataeniella (Eccrinales). Lichtwardt (unpublished) found that millipedes (Diplopoda) and terrestrial isopods collected on the Hawaiian island of Oahu-all of them having been introduced and not native to the island-contained Eccrinales from other geographic areas.

All genera and species of passalid beetles (Passalidae) collected in North, Central, and South America (Fig. 13.3) were infested with the eccrinid Leidyomyces attenuatus (Lichtwardt et al. 1999). As in the example of Asellaria ligiae cited above, this suggests that the fungal species is very old and was already present in passalid beetles as they began to spread and speciate through the Americas.

Freshwater habitats.

Trichomycetes in streams, lakes, and ponds--as well as in smaller discrete habitats such as water-holding containers and rock pools where mosquitoes (Culicidae), bloodworms (Chironomidae), and biting midges (Ceratopogonidae) breed--are, with few exceptions, Harpellales living in larvae of Nematocera (lower dipterans), or mayflies (Ephemeroptera), and stoneflies (Plecoptera). Harpellales have been collected and studied more intensively in many regions of the world than have Eccrinales and Asellariales. The two largest harpellid genera, Smittium and Stachylina are truly cosmopolitan. Whereas some species of those two genera may have sufficient host specificity to be limited in their geographic occurrence, other species have a wider range of hosts and are common on all continents. Examples include Smittium simulii, S. culicis, and S. culisetae, which have been found in larvae belonging to four to six families of Diptera.

Several different patterns of distribution have been discerned in some genera of Harpellales in species that are limited to living in one family of insects. Some selected examples are described below.

Harpella. Harpella melusinae, which lives attached to the peritrophic membrane in the midgut of numerous genera and species of blackflies (Simuliidae), is widespread on all continents in the Northern Hemisphere. Few populations of simuliids are devoid of this gut symbiont. Harpella melusinae is equally common in endemic species of blackflies in New Zealand and Australia. To date, however, it has not been found in Central or South America. In Costa Rica, over 90% of simuliid larvae were found to be infested with another species, H. tica (Lichtwardt, 1997). Harpella tica has also been found in Puerto Rico and northwestern Argentina (White et al., 2000; Lichtwardt et al., 2000). A third species, H. meridianalis, currently is known from southern Chile (Lichtwardt and Arenas, 1996) and Argentina (Lichtwardt et al., 1999). In the region of Manaus in the central Brazilian Amazon, an undescribed species of Harpella occurs. The apparent absence of H. melusinae in Central and South America, according to current knowledge, cannot at present be correlated with ecological or historical factors, thus its distribution remains unexplainable.

Smittium morbosum. This species is lethal to mosquitoes, and therefore one would expect it to be noticed wherever it occurs, given the importance of mosquitoes as vectors of human diseases. First reported in mosquitoes from laboratory cultures in Italy and Russia, it was described by Sweeney (1981a) as killing up to 90% of larvae in rearing trays. Infections have now been reported from Japan (Sato et al., 1998) and Argentina (López Lastra, 1990) in natural populations of several genera and species of mosquitoes. It appears that the incidence of this species is much more sporadic than the two common nonpathogenic symbionts of mosquito larvae, S. culisetae and S. culicis. Given the host range of S. morbosum, it is not likely that host specificity accounts for its more limited distribution.

Carouxella. Two-possibly three-species of the genus Carouxella are symbionts of different species of Dasyhelea in the dipteran family Ceratopogonidae (biting midges) collected in France, Australia, and Argentina (Fig. 13.4). Almost certainly this widely disjunct distribution on three different continents in one host genus is partly attributable to incomplete knowledge of the actual distribution of Carouxella spp. A supplementary hypothesis is that species of Carouxella were once more widely distributed, but that past extinctions occurred such that only remnants of the original populations are extant.

Plecopteromyces. Stoneflies are particularly interesting for biogeographical studies because of the low vagility of adults. One family, the Gripopterygidae (Suborder Antarctoperlaria), has a circumantarctic distribution attributable to tectonic events during the breakup of Gondwana. A new genus/species of Harpellales, Plecopteromyces patagoniensis, was discovered in Argentina (Lichtwardt et al., 1999). It is now known that eight other (currently unnamed) species of Plecopteromyces exist, four in New Zealand and four in Tasmania, in species of Gripopterygidae endemic to those land masses, respectively. It would appear that the fungi, like their hosts, underwent vicariant speciation after continental drift separated the ancestral species.

Harpellales in Arctoperlaria. Three families of stoneflies (Arctoperlaria), Capniidae, Nemouridae, and Leuctridae, are restricted to the Northern Hemisphere, and are known hosts of a number of harpellid genera. Current data indicate that species of these fungal genera are found either in North America or Europe, but not on both continents. Orphella, has three species in North America associated with Capniidae, Nemouridae, or Leuctridae, and there are three species in Europe in Nemouridae or Leuctridae. Lancisporomyces and Genistelloides likewise have different species on both continents. Capniidae in North America are hosts to two monotypic genera, Ejectosporus and Capniomyces, not yet discovered in Europe. The stonefly hosts species likewise are different on both continents. Implications are that stoneflies of these three families were separated after the land bridge disappeared during formation of the North Atlantic some 20-35 million years BP, and that vacariant speciation occurred subsequently in both the stoneflies and their fungal symbionts.