Three overlooked species of Bacidia from insular Laurimacaronesia

1 –––––––––––––––––––––––––––––––––––––––– © 2021 The Authors. Nordic Journal of Botany published by John Wiley & Sons Ltd on behalf of Nordic Society Oikos This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Subject Editor and Editor-in-Chief: Torbjörn Tyler Accepted 12 January 2021 2021: e03055 doi: 10.1111/njb.03055 00 1–11


Introduction
During ongoing investigations of the lichen flora of Macaronesia (van den Boom and Magain 2020 and references therein), the junior author came across three Bacidia-like species. Further studies of that material as well as a number of types of previously published names indicated that two of them belong in Bacidia s. str. (Ramalinaceae) and correspond to the named but poorly understood species B. amylothelia (Vain.) Vain. and B. endoleucoides (Nyl.) Zahlbr. We provide here the first modern descriptions and illustrations of these species and provide an identification key to all currently known members of Bacidia s. str. in insular Laurimacaronesia. The third species posed more of 2 an enigma, displaying superficial similarities with Bacidia s. str. in the acicular ascospores (Ekman 1996, Kistenich et al. 2018). However, it finally became clear from ascus studies that similarities with the Ramalinaceae were only superficial and that the third species represents an undescribed species that is likely to belong in the Malmideaceae. It is described here but provisionally placed in the genus Bacidia as B. deludens.

Material and methods
Microscopic characters were investigated either in a 10% aqueous solution of KOH (ascospores, paraphyses) or in pure water (all other characters). The nomenclature and identification of apothecial and pycnidial pigments follow Meyer and Printzen (2000). Descriptions are based on the cited Macaronesian material only. Colour reactions of pigments were observed in K (a 10% aqueous solution of KOH), N (a 35% aqueous solution of HNO 3 ) and a 15% aqueous solution of HCl. Measurements of quantitative characters are given either as 'minimum valuemaximum value' or 'minimum value -arithmetic mean value -maximum value (s = sample standard deviation, n = sample size)'. Asci were stained with a 0.3% w/w aqueous solution of IKI for the study of tholus structures. Lichen substances were screened using high performance thin layer chromatography (HPTLC) according to Arup et al. (1993) in system A (Bacidia amylothelia, B. endoleucoides) or thin layer chromatography (TLC) according to Orange et al. (2010) in system A, B and C (Bacidia deludens). Coordinates are provided as latitude-longitude in the WGS84 reference system, either as decimal degrees or degrees + decimal minutes.

Nomenclature
When Vainio (1901) described Lecidea laurocerasi var. amylothelia, he cited the Welwitsch gathering Iter Angolense 184, which may be represented by duplicates in other herbaria. However, the preface by George Murray in Hiern (1896, p. v-vi) makes it clear that the entire catalogue project dealing with the Welwitsch collections, to which Vainio contributed the lichens, is based on the specimens housed in BM. There is no type material in Vainio's lichen herbarium at TUR. Consequently, the name appears to be based on the single specimen in BM, which we refer to as the holotype.

Description
Thallus crustose, thin to medium thick, almost white to light greenish grey, ranging from discontinuous, of scattered, ± convex and often slightly effigurate areoles, to continuous, ± cracked, with warted surface. Prothallus lacking or present, forming thin black lines in competition with other lichens. Photobiont a member of Trebouxiophyceae, unicellular, cells globose or short-ellipsoidal, 4.5-11.0 μm long.

Distribution and habitat
Bacidia amylothelia was described from northwestern Angola by Vainio (1901) and is reported here as new to insular Laurimacaronesia, where it is known from the two Canary Islands Tenerife and La Palma. It has also been reported from South Africa by Vainio (1926), but we were unable to locate that material. As the species was originally discovered on 'young trees', it might turn out to be weedy and substantially overlooked in at least the Old World tropics and subtropics. Two Tenerife collections are from the phorophyte Laurus novocanariensis, growing on branches. One La Palma collection was found on a trunk of a small Laurus tree (9 cm diam.) and one on a medium-sized trunk of Ocotea foetens.

Remarks
Bacidia amylothelia is a member of Bacidia in a strict sense on account of the heavily gelatinized excipular cells with long and narrow cell lumina (Kistenich et al. 2018). It can be identified by its distinctly orange-brown hypothecium, superficial white pruina, presence of minute crystals predominantly in the lower part of the exciple near the edge, no radiating clusters of large crystals in the proper exciple, and an indistinctly delimited 5-6-layer zone of enlarged cell lumina along the excipular edge. Apothecia in the Macaronesian specimens are, however, on average somewhat paler than in the type material, although apothecial colour in the holotype varies from grey-brown to orange-brown to purplish black.   (1868)

Distribution and habitat
Bacidia endoleucoides was described from Madeira and has so far never been reported from anywhere else (Carvalho et al. 2008). It appears to be quite widespread in insular Macaronesia, as we have seen specimens from the Canary Islands (Tenerife and La Palma) and the Azores (São Jorge), in addition to Madeira

Remarks
Bacidia endoleucoides is a member of Bacidia in a strict sense on account of the heavily gelatinized excipular cells with long and narrow cell lumina (Kistenich et al. 2018). Apart from the original and quite vague diagnosis and brief mentions in checklists (Tavares 1952, Hafellner 1992, this species has not been discussed in the literature and no modern description is available.  (Ekman 2004). Bacidia caesiovirens, a European oceanic species not known from Macaronesia, can be recognized 6 by its pale yellowish hypothecium and granular thallus with blue-green pigment (Ekman and Holien 1995

Etymology
The epithet deludens means 'deceiving', the present active participle of deludo, and refers to the fact that, morphologically, B. deludens may be taken for a member of the Ramalinaceae even though it is a member of another family.

Description
Thallus crustose, thin, yellow-grey to pale brown-grey, sorediate, composed of scattered, irregular, convex areoles that coalesce to form a continuous and ± cracked crust with smooth or rough surface, on furrowed bark forming small, indeterminate patches of areoles up to 0.08 mm diam. with rough surface, on smooth bark usually well delimited, to 20 mm diam., areolate (areoles up to 0.16 mm diam.) or more often continuous, then cracked but with otherwise ± smooth surface. Soralia ± tinged pale brownish in outer part, whitish in the centre where soredia have been shed, on furrowed bark efflorescent, irregularly rounded or (following the ridges of the bark) ellipsoidal, flat to convex, to 0.8 mm when elongate; on smooth bark bursting through the thallus, mostly rounded and to 0.5 mm diam., flat and level with the surrounding thallus, often forming a ± vertical, discontinuous rim along the edge of the soralia. Soredia globose to ellipsoidal, 20-28-41 μm long (s = 7, n = 50). Prothallus, when present, blackish, particularly prominent where several specimens form a mosaic. Photobiont a member of Trebouxiophyceae, unicellular, cells globose or short-ellipsoidal, 5-8 μm long.

Distribution and habitat
Bacidia deludens is currently known from the Canary Islands (Tenerife and La Gomera) as well as Madeira at altitudes between 600 and 1200 m a.s.l. We have observed it on bark of Erica, Vaccinium, an unidentified shrub in laurisilva and on Cupressus in a mixed forest. Localities vary from disturbed and poor in lichens to fairly species-rich. In the type locality, where B. deludens inhabits bark of Erica, accompanying microlichens include Byssoloma marginatum, Coenogonium luteum, Endohyalina ericina, Fellhaneropsis vezdae, Jamesiella anastomosans, Micarea alabastrites, M. doliiformis, M. pycnidiophora, Porina coralloidea and Scoliciosporum pruinosum. One specimen lacks apothecia, which opens the possibility that B. deludens may occur as an overlooked 'sterile, sorediate crust' in insular Laurimacaronesia.

Remarks
In a fertile state, B. deludens may remind of a member of Bacidia, Bacidina or Toniniopsis on account of the combination of acicular and transversely septate ascospores, biatorine apothecia and the Trebouxiophyceae photobiont (Kistenich et al. 2018). The prominent, whitish or pale brownish soralia, combined with the somewhat Cliostomumlike apothecia with sparse amounts of pigment and presence of minute crystals only along the edge of the proper exciple is unprecedented in the Ramalinaceae, however. In a sterile state, on the other hand, Bacidia deludens bears a superficial resemblance to Biatora britannica Printzen et al., B. efflorescens (Hedl.) Räsänen and Lecanora jamesii J. R. Laundon, in which argopsin is dominant in the two first and atranorin, usnic acid and 2-O-methylsulphurellin is usually present in the latter (Tønsberg 1992, Lumbsch et al. 1995, Printzen et al. 2001. Bacidia deludens does not fit in any known genus in the Ramalinaceae. The ascus structure instead suggests affinities to the Malmideaceae or Pilocarpaceae. Based on the well developed and strongly gelatinized proper exciple with radiating hyphae, lack of constrictions at the ascospore septa and the abundance of crystals in the hymenium, we suggest that B. deludens is a member of the Malmideaceae. Among the genera currently classified in the Malmideaceae (Spribille et al. 2020, Wijayawardene et al. 2020, Malmidea, Sprucidea and Zhurbenkoa possess mainly non-septate ascospores and a brown hypothecium, Malmidea also having halonate ascospores, Sprucidea producing sporodochia and Zhurbenkoa having a parasitic life-style on other lichens (Kalb et al. 2011, Cáceres et al. 2017, Flakus et al. 2019. Puttea includes species with non-septate ascospores, minute apothecia with a poorly developed proper exciple, and (when pigmented) brown pigment caps on the terminal cells of the excipular hyphae (Dillman et al. 2012). Ascomata are unknown in Cheiromycina and Savoronala, which are instead recognised by their prominent sporodochia (Ertz et al. 2013, Muggia et al. 2017. Ascospores with up to 3 septa are known in Crustospathula and Kalbionora, but Crustospathula is also characterised by stalked soralia, Kalbionora by a brown hypothecium, and both genera by a variety of lichen substances not involving fumarprotocetraric acid (Kalb et al. 2012, Sodamuk et al. 2017. All genera except Malmidea currently include 1-4 species, while Malmidea has 52 species (Wijayawardene et al. 2020). The circumscription and generic classification of the Malmideaceae is poorly known and it has been suggested that several additional taxa may belong in that family, e.g. the genus Porpidinia as well as couple of species formerly referred to Phyllopsora in the Ramalinaceae or the large and distantly related Lecidea, the type of which belongs in the Lecideaceae (Breuss and Lücking 2015, Kistenich et al. 2018, Palice et al. 2018, Flakus et al. 2019. At the moment, we see three options to classify our new species: 1) a new genus in the Malmideaceae could have been erected. We do not favour this option, partly because the genus to which our species belongs may turn out to be already described but misclassified in another family, partly because we would prematurely create a monotypic genus based on fragmentary data in a phylogenetically and taxonomically poorly known family. In the latter case, typification of the genus may turn out to be suboptimal if several species are later shown to belong to the genus. 2) We could have provisionally recognised the species in an already described genus classified in the Malmideaceae. As outlined above, all currently recognised genera are morphologically homogeneous and match poorly with our species. The two genera Crustospathula and Kalbionora may seem as the least bad alternatives because of the presence of 3-septate ascospores, but our species deviates substantially from these genera in characters outlined above. Adding our species to any of the genera would extend the morphological variation and make the genus problematic to characterise, which is why we decided to avoid this option.
3) The third option, which we settled for, is to provisionally classify our new species in the genus Bacidia awaiting a reasonable overview of the phylogeny and classification of the Malmideaceae. Bacidia has historically been the home to basically all crustose lichens with a chlorococcoid photobiont, biatorine apothecia and ascospores with three or more transverse but no longitudinal septa (Ekman 1996). Most of those species are not congeneric with the type species B. rosella (Pers.) De Not., but a more natural classification was recently proposed (Kistenich et al. 2018). Provisionally accepting another species not congeneric with the type in an already heterogeneous assemblage is in line with recent history and would minimally impact the endeavour to achieve a monophyletic Bacidia. In addition, there is precedent for provisionally treating species as members of historically heterogeneous genera awaiting phylogenetic and taxonomic clarification, recent examples being, e.g. Lecidea coriacea Holien & Palice (Holien et al. 2016), Bacidia gullahgeechee Lendemer (Lendemer 2018), Bacidia pruinata Fryday (Fryday 2019) and Lecidea streveleri T. Sprib. (Spribille et al. 2020). The latter was explicitly placed in the Malmideaceae, yet described in Lecidea for the same reason we described our species in Bacidia.