Notes:

The description of the new R. discolor depicts a species with a variable pileus colour. In general terms, it could be described as an overall reddish species that discolours easily and rapidly, while developing different tinges. Also the colour of the lamellae changes distinctly between young and mature fruiting bodies. Notwithstanding this important colour variation, the species is considered in Australia to be ‘easily recognizable in the field’ (Richardson 2011, Bougher & Syme 1998). This was confirmed by our sequencing results because all of the received collections (see studied material) produced identical ITS sequences. It seems to be widespread in southwestern Australia where it is exclusively associated with various Eucalyptus species that compose the Karri (Eucalyptus diversicolor), Tingle (E. jacksonii & E. guilfoylei) and Jarrah (E. marginata) - Marri (Corymbia calophylla) forest stands. We can also confirm here that the voucher that served for the illustration and description in Bougher & Syme (1998) of R. aff. cyanoxantha / R. multicolor [as listed on p. 380 of that work] was successfully sequenced and represents indeed R. discolor.

Russula wielangtae, described from Tasmania and equally a eucalypt associate, is definitely close (ITS similarity between 97-98%) to the here newly described R. discolor and shares with it a very similar spore ornamentation, a similar composition of the pileipellis and even a similar overall colour, although less variable. The presence of ‘pileocystidia’ was considered ‘doubtful’ in the original description of R. wielangtae and the legend of fig. 74 in Adamčík et al. (2019), mentioning twice ‘pileocystidia’, probably needs correction as none of the hyphal terminations in the pileipellis of these Oceanian species has the typical contents of the generally accepted definition of gloeocystidia in the genus. Russula wielangtae might also occur on the Australian continent as suggested by the collection MEL 2238232 (see Fig. 3a,b). The ITS sequence for this collection was deposited in GenBank associated with the name R. wollumbina Grgur. and still represents the single deposited ITS sequence under that name (Lebel & Tonkin 2007). Unfortunately this sequence seems of unsufficient quality, lacking more than 70 base pairs at the start of ITS1 and having several ambiguous base pairs further down the sequence. The specimen was identified and collected by J. E. Tonkin in eucalypt woodland in southeastern Australia near Mt Richmond (https://biocache.ala.org.au/occurrences/9063ba82-91d0-4c74-9e1f-329ed608b9db). Tonkin's identification was questioned by Adamčík et al. (2019) who assumed that it rather represents their newly described R. wielangtae (we agree on the basis of the ITS sequence similarity) as the spores of the type specimen of R. wollumbina were described and illustrated as having much more isolated warts and scarcely any ridges (Grgurinovic 1997), a feature that was confirmed for the type in Adamčík et al. (2019).

There are many more collections reported as R. wollumbina in the ‘Atlas of Living Australia’ (https://www.ala.org.au), but none of these have been sequenced. The fact that the type of R. wollumbina is based on a collection by Cleland (1934) originally identified as R. xerampelina, a species often reported from Australia in the past (e.g.: Dickinson & Lucas 1979, Gill & Ashton 1971, Shepherd & Totterdell 1988, Sinnott 1976) seems to be responsible for a misleading shortcut. Indeed, certain websites on Australian mushrooms automatically seem to assume that all mentions of R. xerampelina in Australia correspond now to R. wollumbina, which is absolutely not the case. An interesting and recent find that might correspond to R. wollumbina is a description by P. Leonard for a specimen identified as such and uploaded on https://qldfungi.org.au/wp-content/uploads/2022/02/Russula-wollumbina.pdf. This specimen indeed shows some overall resemblance to close relatives of R. xerampelina, the earlier interpretation of R. wollumbina. Leonard’s description mentions pale gills that produce a whitish or very pale cream spore print (IIa following the Romagnesi 1967 scale), not the ochre or yellow spore print observed for R. wielangtae or R. discolor, respectively, and stresses the often strongly tapering stipe, two characters that match the original interpretation of Cleland (1934, as ‘R. xerampelina’). However, the spores of Leonard’s collection are described as having strongly amyloid ridges, which seems in contradiction with the spore description given by Grgurinovic (1997) for the type specimen of R. wollumbina. Leonard’s interpretation of R. wollumbina suggests a very different species compared to the here newly described R. discolor which has more variegated colours, much darker spore print and a cylindrical to slightly bulbous stipe. The stipe of R. discolor also lacks the pinkish tints observed on nearly all of the other species that are discussed here. A correct interpretation of R. wollumbina remains therefore problematic for the moment.

The only other species that has been placed in the /wielangtae lineage in recently published phylogenies was identified as R. atroviridis Buyck from New Zealand (Adamčík et al. 2019, Cooper 2021). It differs already from both other Oceanian species in this lineage because it associates not with eucalypts, but principally with Nothofagus. Our phylogeny (Fig. 3b) suggests that there are possibly two different, but very close phylogenetic entities. Although considerable effort has been made to obtain sequences for most of the Russula species of New Zealand (Cooper & Leonard 2014, Cooper 2021), there remains an urgent need for better and more detailed descriptions, particularly for the type specimens. We, therefore, provide below our own microscopic examination of the isotype of R. atroviridis. As discussed below, the results of this re-examination raised more questions.