4. The taxonomic position of the genus Fistulinella.

The genus Fistulinella was based on F. staudtii Henn. from Africa. The type of this genus is not, as was generally assumed (Guzman 1974), lost but has been rediscovered and restudied (Singer 1973) - I agree with Horak (1968) and Guzman (1974) who consider Ixechinus Heim identical with Fistulinella. I have shown that all pickled material of boletes has separable (`individual') tubes if the lateral stratum swells up to push the tubes away from each other and if the mediostratum is sufficiently gelatinized or thin to permit the fluid to dissolve enough of the gelatinous mass and thus sever the few thin-walled hyphae connecting the two sides of the lateral strata. Even while growing, some tubes may slide downwards a fraction of a millimeter in relation to others surrounding them, and thus create an uneven pore level, which emphasizes the false impression of separate individual `pores' as known in Fistulina. However, a section through the tubes of dried Fistulinella campinaranae Sing., F. minor (Heim) Guzman, F. venezuelae (Sing. & Digilio) Sing. and F. violaceiporus (Stevenson) Sing., comb. ined., as studied by me show that the hymenophoral trama, aside from being somewhat more gelatinized, is not different from other corresponding structures in Boletaceae and Strobilomycetaceae. The `outer' layer of separated tubes as removed from hymenophores in liquid preparations consists of a naked lateral stratum with free hyphae dangling in the medium.

I have (Singer 1975) inserted Fistulinella as a section in Tylopilus, but have insisted that this is a temporary and tentative solution, waiting for more studies of fresh material. Such fresh material has been obtained recently from the campinarana vegetation in Brazil. These fresh collections were prepared as dried herbarium material and thus a continuous tube-layer was maintained. This species is considered new and most closely related to F. mexicana Guzman. A fraction of the formalin material on which this latter species, recently published (Guzman, 1974), was based, was kindly sent to the present author for comparison.

What do the newly discovered species of Fistulinella reveal about the position of this genus, its limits, and distribution?

In the first place, it appears that all have numerous characters in common and that their distribution and ecology is totally different from that of Tylopilus as well as the smooth-spored species of Porphyrellus. Fistulinella is tropical, extending to New Zealand, and non-mycorrhizal, whereas Tylopilus - although including some tropical species - is predominantly north-temperate and consistently ectomycorrhizal. Porphyrellus on the other hand, is, as far as smooth-spored species are concerned, north-temperate but reaching the palaeotropics as well as the Nothofagus zone of Australia and New Zealand.

Being non-mycorrhizal and narrow-stemmed, with generally very long spores, these often being pseudoamyloid and/or with apical germ pore, with a spore print in color much like that of Porphyrellus gracilis, it becomes obvious that Fistulinella is much closer to Porphyrellus than to Tylopilus. It differs nevertheless by the spores which are smooth and by the strong gelatinization not only of the hymenophoral trama but also of the surface layers, at least of the pileus, and the relatively slender stipe.

If we now exclude all species of Porphyrellus which agree in their characteristics with Fistulinella, we have to transfer the subsection Viscidini and with it the type of section Pseudotylopili Sing. to Fistulinella and transfer the subsection Niveini to section Graciles of Porphyrellus.

The genus Fistulinella will then be placed side by side with Porphyrellus in the Strobilomycetaceae, and will contain the following species

F. staudtii Henn.; F. minor (Heim) Guzman, F. major (Heim) Guzman, F. campinaranae Sing. sp. nov., F. mexicana Guzman; F. viscida (McNabb) Sing., comb. nov. [Basionym: Porphyrellus viscidus McNabb in New Zealand J. Bot. 5: 547. 1967]; F. venezuelae (Sing. & Digilio) Sing., comb. nov. [Basionym: Tylopilus venezuelae Sing. & Digilio in Lilloa 30: 163. 1960], and obviously also Boletellus violaceiporus Stevenson (type revised).

Observations. This specimen is very similar to P. viscidus var. viscidus McNabb by the presence of the ixotrichodermium, basidial and cystidial characters. Spores of this specimen are longer by 5 µm (at the upper range limit) and by 2.5 µm at the L^m value that var. viscidus. The spore E^m value is also greater than that of var. viscidus by 0.3. The spore character differences appear to be sufficiently different for varietal recognition. More studies of general collections will be necessary for confirmation of this conclusion.

Because this taxon belongs in the genus Mucilopilus the following new combination is proposed.
Mucilopilus viscidus var. macrosporus (McNabb) Wolfe, comb. nov.
Basionym: Porphyrellus viscidus var. macrosporus McNabb. 1967. New Zealand J. Bot. 5: 546.