Russula atroviridis Buyck, Bull. Jard. Bot. Nat. Belg. 60(1-2): 202 (1990)

{zoologyFigureCitations}

MB127270

Holotype: New Zealand: Prov. Mid-Canterbury, Craigieburn Range, Kowai Bush, on soil under Nothofagus solandri, 5 Febr. 1969, leg. E. Horak (PDD 30655 holotype, ZT69.305 isotype)

= R. atrovirens R.F.R. McNabb New Zealand journal of botany 11: 703, nom. illeg., non Beeli (1928)

{botanyFigureCitations}

Figure 8. Russula atroviridis (isotype). a. Hyphal terminations of the pileipellis constituting the pseudoparenchymatic layer. b. Emerging ‘hairs’ from the pseudoparenchyma containing refringent bands that are clearly visible in Congo red mounts. Scale bar = 10 µm. Drawings B. Buyck

Figure 9. Russula atroviridis (isotype). Elements of the hymenium. a. Spores as observed in Melzer’s reagent. b. Basidia. c. Basidiola. d. Gloeocystidia on sides of lamellae with indication of contents as observed in Congo red in two cells. e. Gloeocystidia on edge of lamellae. Scale bar = 10 µm, but only 5 µm for spores. Drawings B. Buyck.

Spores broadly ellipsoid, (7.9–)8.0–8.45–8.9(–9.4) x (5.8–)6.5–6.98–7.5(–7.7) µm, Q = (1.08–)1.14–1.21–1.29(–1.36), sparsely to typically very densely ornamented with strongly amyloid, cylindrical to conical or smaller, obtuse-rounded warts; the warts isolated or more often grouped with two or more in dense clusters or aligned into short crests that never encircle the spore, low and short linear interconnections are sometimes present but never to such a degree that the spore could be considered subrecticulate; suprahilar spot large and mostly distinct, often even strongly amyloid, but sometimes also just greying in Melzer's reagent or simply verruculose. Basidia strongly clavate, not pedicellate, 35–40(–45) x 11–14 µm, 4–spored; sterigmata robust, ca 8 x 2 µm. Hymenial gloeocystidia on sides of lamellae clavate to almost fusiformous, 40–60 x 10–14 µm, mucronate or more frequently appendiculate; the appendix measuring up to 25 µm long, not thick–walled, filled with abundant, refringent, granular-crystalline contents. Marginal cells not observed; the lamella edge has basidia and cheilocystidia similar to pleurocystidia, but (much) smaller, sometimes as small as 20 x 7 µm. Pileipellis two-layered, a gelified subpellis of intertwined hyphae is present beneath the pseudoparenchyma at the surface; the pseudoparenchyma composed of hyphal terminations that start with basal, strongly inflated, globose to ellipsoid cells, sometimes up to 25 µm diam, from which arise short chains of 3–6 gradually smaller cells, rarely ramified, that are mostly barrel–shaped to almost cylindrical with the terminal cell typically short and slightly narrowing, utriform to even clavate, measuring 8–20 x 4–8 µm; these terminations mixed with much longer, more slender ones that are composed of 4 to 8 subcylindrical cells, forming ‘hairs’ that extend up to more than 100 µm and are strongly emergent, the uppermost 1-2 cells banded with refringent contents (pigments ?) particularly near the septa; the terminal cell mostly narrowing, up to 40 µm long, sometimes subcapitate (and reminiscent of mucronate pileocystidia); toward the pileus margin parts of the suprapellis has seceded leaving the narrow, thin-walled hyphae of the subpellis exposed; these hyphae individually distanced from each other by gelification and intermixed with oleiferous fragments that are more strongly undulate-tortuous. Typical pileocystidia and cystidioid hyphae not observed. Clamp connections absent.

Notes: The original description of R. atroviridis pictures a mild, medium-sized (up to 65 mm diam.) species with a multicoloured pileus, predominantly tinged with dull greyish magenta, greyish green, greenish-yellow and dark brown, a pinkish stipe, 40 x 15 mm, and pale coloured lamellae (McNabb 1973: as R. atrovirens). It was originally collected under Nothofagus solandri, but is also listed with several other Nothofagus.

Our examination of the isotype of Russula atroviridis revealed a pileipellis that could be described as a pseudoparenchyma. It differs from a typical ‘virescens’- structure in the presence of dispersed, septate, cylindrical, emerging ‘hairs’ (fig. 8b). These ‘hairs’ are easily located in microscopic preparations as their terminal and sometimes subterminal cell are filled with distinct, mostly banded or agglutinated, refringent contents. Also our observations on the spores of the isotype (fig. 9a) correspond very well to McNabb’s description which mentions a ‘densely verrucose’ ornamentation.

While neither the type nor any of the original material of R. atroviridis has been sequenced, all of the recently published ITS and LSU sequences deposited under that name place McNabb’s species in the /wielangtae-lineage (Adamčík et al. 2019, Ghosh et al. 2023) and the most disconcerting fact resulting from our microscopic examination of the type collection, is that the pileipellis of R. atroviridis is very different from the only two other known species in the /wielangtae-lineage. Indeed, the type has exactly the same pileipellis structure as found in Oceanian species that are part of the unrelated subsection Tricholomopsidum Buyck & V. Hofst. (Buyck in Rossi et al. 2020). The latter subsection is by far the most common and diverse Russula lineage in New Zealand and harbours several species with very similar overall colouration and equally associated with Nothofagus trees.

When looking at images of the microscopic features of these sequenced “R. atroviridis” collections (available on https://scd.landcareresearch.co.nz/), it is impossible not to notice a major discrepancy been these specimens and the type specimen. These sequenced collections all possess strongly ornamented subreticulate spores, which is in complete disagreement both with McNabb’s original description (1973: 703-704) and our own observations on the type. It is therefore obvious that the modern interpretation of R. atroviridis in New Zealand does not conform to the type.

The complexity of the situation does not end here, however, because a comparison between the type collection of R. atroviridis and the other species described by McNabb (1973) reveals near-identical features for R. roseostipitata McNabb. The similarities are such that we suspect that McNabb’s holotype of R. atroviridis is a specimen of R. roseostipitata. Whether or not R. atroviridis becomes indeed a synonym of R. roseostipitata will need a more in-depth study of this species complex as there appear to exist other closely related but still undescribed taxa based on deposited sequences in GenBank. This possible synonymy is not without interest in the context of our original quest as R. roseostipitata was suggested by McNabb to be closely related to Virescentinae (exactly the subsection where the Malagasy R. prolifica is placed). However, McNabb ignored the importance of the strongly amyloid suprahilar spot on the spores of his species, which excludes such affinity. Available sequences for R. roseostipitata place it in subsect. Tricholomopsidum and this placement conforms to our observations on the R. atroviridis type specimen. It is also interesting that the two earliest deposited sequences for R. atroviridis (GU222285 and MW296245) are identical to the more recently deposited sequences for R. roseostipitata.

The recent misinterpretation of “R. atroviridis”, however, corresponds to the best of our knowledge to a still undescribed species. This misinterpreted “R. atroviridis“, which is part of the /wielangtae lineage, does indeed possess the typical features of this lineage and is described below as R. cooperiana sp. nov.