The degree to which trichomycetes have become adapted to specific hosts appears to vary among the taxa. The lists in Appendix A show that many species of trichomycetes are reported from only one species of arthropod. Are these particular fungi species specific? Two major problems become immediately apparent before that question can be answered. First, insufficient collections and lack of studies of related arthropod species may be responsible for the presumed restricted host range. Second, there has been a tendency among some researchers to name new species of trichomycetes largely on the basis of new hosts. Unfortunately, speciation problems are not easily resolved in some instances, even when careful taxonomic judgment is employed; this is because of the variation in morphology that some species exhibit, especially in the Eccrinaceae. This problem is discussed in more detail in Chapter 10.
The ultimate determination of the degree of host specificity should employ experimental methods. Most studies of this kind, out of necessity, have used uncultured fungus inoculum and hosts that are not raised initially under axenic conditions. It is necessary to insure that these hosts are indeed devoid of trichomycetes. This can be done (1) by selecting populations that are probably uninfested, as determined by sufficiently large sample dissections; (2) by using arthropods that have molted and have been immediately separated from their exuviae and other substrates that may contain spores, in order to preclude reinfestation; or (3) by using clean or surface sterilized eggs for hatching and rearing the arthropods.
As early as 1906, Chatton (1906a) tested the host range of the ectocommensal Amoebidium parasiticum and found that spores developing from thalli attached to Daphnia would affix as well to many other arthropods, even to glass, thus confirming his contention that the spores are not particular about the substrates to which they will attach in nature. Whisler and Fuller (1968), using a pure culture of A. parasiticum for the purpose of studying the fine structure of holdfast formation, did not obtain adhesion of spores to chitin or cotton threads when the organism was grown in a liquid nutrient medium; however, spores did attach to cotton fibers, nylon, silk, and hair (in decreasing order of success) when thalli were transferred to a dilute salt medium. Whether or not spores of Amoebidium that become attached to nonarthropod substrates in the natural environment develop to maturity has not been determined.
Manier (1963a) studied the infestation of the terrestrial isopod Armadillidium simoni by Asellaria armadillidii in southern France. Ten other species of Oniscoidea collected in that general region were not infested, but she found that Asellaria could be transmitted to another isopod species, Acaeroplastes melanurus, when individuals were kept in a small container with infested A. simoni. She concluded from the cross-infestation tests that because of habitat preference when in their natural environment, the population of the normally uninfested isopod species did not have sufficient contact with A. simoni to pick up the fungus; however, once transmitted, development of Asellaria in the new hosts was very good. Manier's conclusion fits what we have observed in Kansas, USA, where Armadillidium vulgare and A. nasatum live together and compete for the same resources. Mixed populations of the two isopod species may be infested with Asellaria armadillidii, others with Parataeniella armadillidii (Lichtwardt and Chen, 1964). Neither fungus has been found restricted to only one of the two host species.
Tuzet et al. (1961) did extensive collecting of mosquito larvae in several French Mediterranean departments, and found nine of seventeen species to be infested with Smittium culicis. They performed several experiments to determine whether the lack of infestation in some mosquito species in their natural habitats was due to a lack of receptivity to the fungus or to extrinsic factors. Theobaldia longeareolata is a mosquito species that was consistently devoid of S. culicis in the natural environment, even in the presence of heavily infested larvae of other species. Two tree-hole species without natural infestation, Aedes geniculatus and A. berlandi, could be readily infested experimentally with S. culicis, even in their native water, thus ruling out the possibility that the tree-hole water was in some way fungistatic. They theorized that the lack of natural infestation in the receptive tree-hole larvae of the two species was due primarily to the inability of the fungus inoculum to reach the tree-hole inhabitants. (However, see Chapter 8 where the authors report on several similar kinds of isolated sites with infested mosquito larvae.) Tuzet et al. also concluded that Anopheles claviger was not infested naturally due to the fact that it is a surface feeder and less likely to ingest spores of Smittium, which settle on the bottom substrate. When they kept larvae of this species in very shallow layers of water, thereby forcing the larvae to feed on the bottom substrate, all became infested in vessels containing the fungi, whereas another similar set of larvae of this species kept in water 15 cm deep did not. It seems clear from these experiments that a wide range of mosquito species, although not all, are infestable by S. culicis, but that environmental and behavioral factors may also determine whether the fungus becomes established in a given population.
Coluzzi (1966) in Italy was aware of the experiments by Tuzet et al. and conducted infestation tests with an undetermined species of Smittium (similar to S. culisetae) reported as being common in four local species of mosquito larvae. Using one natural host (Culex pipiens) and two species not normally infested (Aedes aegypti and Anopheles gambiae), he placed laboratory colonies of 2nd instar larvae of each species in 30-cm diameter plastic dishes filled to about 6 cm depth with well water, and added 4th instar larval molts containing the Smittium. On reaching the 4th instar, 93-100% of the larvae were infested. He attributed the high infestation rate in A. gambiae, in contrast to the results of Tuzet et al. with A. claviger kept in deeper water, to the fact that A. gambiae is generally a bottom feeder.
Experimental infestations of a range of hosts by one fungal species were also conducted by Moss (1972) using 19 genera of midge (Chironomidae) larvae collected from an exceptionally rich 21-m stretch of a brook in England. Species of five of the tubiculous midge genera were found to be naturally infested with Stachylina grandispora. Moss kept 4th instar larvae of infested and uninfested species in shallow dishes with pond water for 2 weeks, and was able to infest species of larvae of 7 genera that had not been found to be naturally infested. An interesting result was that none of the species of free-living, carnivorous genera became infested by this procedure. This study provided clear evidence of the wide host range of Stachylina grandispora, and that in this particular site infestation was probably determined by the feeding habits of the midge larvae.
Coste-Mathiez (1970) in Manier's laboratory attempted some more or less reciprocal infestations with two species of Smittium. One, S. mucronatum, has been found in nature only in larvae of Psectrocladius sordidellus (Chironomidae), and is known to produce zygospores in addition to trichospores in that host. Chironomid larvae infested with S. mucronatum were placed in containers with uninfested 4th instar larvae of the mosquito Culex pipiens. At the end of 3 days the molts of the mosquito larvae were all infested with the fungus, and 2 of the 35 molts had some zygospores as well. She found, however, that the reproductive structures were somewhat abnormal in size and shape (Manier and Mathiez, 1965). Coste-Mathiez also took uninfested larvae of Chironomus sp. (Chironomidae) and placed in their containers sporulating thalli from a culture of S. culicis, a species that grows naturally in mosquito larvae, but which is not known to infest chironomids naturally. All Chironomus larvae became infested with S. culicis, but the development was extremely slow for this species. The thalli eventually degenerated, and by the end of 2 months there was no longer infestation in the larvae that remained. Apparently reinfestation by means of trichospores produced from the original growth in the guts did not take place.
There are no records in the literature indicating that strictly carnivorous or predaceous arthropods become infested with trichomycetes. An exception was found in Australia when Lichtwardt was working with Dr. A.W. Sweeney in 1983 near Innisfail, Queensland. Some specimens of the predacious mosquito larva Culex halifaxii were collected, placed individually in containers in the laboratory for 8 days, and, during this period, fed larvae of another, nonpredaceous species of Culex that were infested with Smittium simulii and S. culisetae. One of the seven specimens of C. halifaxii dissected proved to be highly infested with S. simulii. The same site also contained bloodworms (Chironomus sp.), most of which were infested with both Stachylina grandispora and Smittium simulii. Some predaceous midge larvae (Pentaneura sp.) living among them, however, were uninfested. Thus, the infestation of the predaceous mosquito larva appears to be an exception to the rule. Another interesting exception to host specificity was found in rock pools of the Georges River near Cambelltown, New South Wales, Australia. The pools contained mosquito larvae (Aedes rupestris), bloodworms (Chironomus alternans), and ceratopogonid larvae (Dasyhelea sp.), many of which were infested with Smittium culisetae. Among these dipteran larvae were some immature mayfly nymphs (Ephemeroptera), one of which had S. culisetae living in its gut, the first known instance of a nondipteran host for any Smittium species.
Williams and Lichtwardt (1972a) made use of a collection of axenic isolates of four species of Smittium obtained from mosquito, chironomid, and blackfly larvae to test their ability to infest one host species: larvae of Aedes aegypti. The data obtained in this experiment are presented in Chapter 9 (Table 9.3).
The natural infestation by S. simulii of dipteran larvae belonging to different families (most commonly Simuliidae and Chironomidae) may be due to the fact that the chironomid larvae in which this fungus has been found inhabit flowing waters, as do the larvae of simuliids. Perhaps, over time, this permitted adaptation by the fungus to both families of dipteran hosts through repeated transfer of spores.
The degree of specificity of many trichomycetes has been established by many data obtained in the field. For example, in a single stream one may find blackfly larvae infested with species of their particular fungal genera (Harpella, Pennella, Genistellospora, Simuliomyces, etc.), whereas the mayfly nymphs, even on the same rocks or sticks in the water, have their own individual flora (Legeriomyces, Glotzia, Zygopolaris, etc.). Species of these genera are morphologically quite distinct in the sporulating stage, and are not identified on the basis of the host (so circular logic is not a factor here). Furthermore, some of the fungal genera are restricted to mayflies of the family Baetidae, even when other families of mayfly nymphs are present on the same substrates. Refer to the genus Zygopolaris in Chapter 11 for further examples of specificity within families of mayflies.
The invasion of an individual arthropod by one species of trichomycete does not necessarily exclude the establishment of other trichomycetes in the same gut. The most striking example of this phenomenon is found in some populations of blackfly larvae. It is not uncommon to find larvae containing more than one trichomycete species, and frequently two or three species are found in one blackfly larva in good sites. An example of an excellent site is an inconspicuous small stream, Johnson Creek, that drains into Swan Lake in northwestern Montana, U.S.A. In 1975 Lichtwardt found seven species of trichomycetes in three blackfly dissections. One of those larvae contained Harpella leptosa, Genistellospora homothallica, Simuliomyces microsporus, and Paramoebidium curvum. In that same general area of Montana, thalli of H. melusinae and H. leptosa have been found sporulating side by side in the midgut of blackflies (Moss and Lichtwardt, 1980). Also, in several populations of chironomid larvae there and in other regions of the world, the authors have found combinations of two species of Smittium in the hindgut. We can state, therefore, that not only may one species of trichomycete infest many species of hosts, but also that in some habitats one species of host may be infested by two or more trichomycetes.
If any generalization can be made about host specificity in these gut-inhabiting fungi, it is that species of trichomycetes often are capable of infesting more than one species or one genus of hosts, but may be restricted to one arthropod family. The range of hosts can extend beyond a single family in a few trichomycetes (e.g., Enteromyces callianassae, Taeniella carcini, Smittium simulii). The genus of hosts, rather than the family, may delimit the range of infestation by some species of trichomycetes, especially in cases where those genera are phylogenetically distinct and when species of those genera are ecologically isolated from other related genera by virtue of their restricted habitat (e.g., the marine isopod genera Limnoria and Ligia). It is also conceivable that many trichomycetes are, in fact, species specific as the current record indicates, but in all likelihood further studies will show that, in instances where closely related species of arthropods have similar feeding habits and a sympatric distribution, the fungal species will not be found restricted to a single host species.