Acrolejeunea (Spruce) Schiffn., 1893

Renner, Matt A. M., 2013, A new subspecies of Acrolejeunea arcuata, and notes on typification, synonymy, and distribution of other Australasian Lejeuneaceae, Phytotaxa 83 (1), pp. 39-53 : 40-42

Acrolejeunea (Spruce) Schiffn. in Engler & Prantl (1893: 119, 128), nom. cons.

The genus Acrolejeunea was validly published in 1893 based on the occurrence of a preprint (see Gradstein 1974: 329), and was circumscribed to include species with entire underleaves, cortical stem cells larger than medulla stem cells, cordate trigones on leaf lobe cell walls, perianths with 2–5 ventral plicae, and lacking subfloral innovations. There is evidence that plants possessing this combination of characters are monophyletic ( Wilson et al. 2007). Acrolejeunea was monographed on a worldwide basis by Gradstein (1975), who recognized two subgenera, 15 species, six subspecies, and four varieties, of which two species, one subspecies, and one variety were described as new. No new taxa have been proposed in Acrolejeunea since Gradstein’s monograph. Acrolejeunea species are predominantly tropical, with subgenera distributed on either side of the Pacific Ocean, subg. Acrolejeunea occurring in the Afro-American tropics, and subg. Isolejeunea Gradstein (1975: 83) distributed through south-east Asia, Melanesia, Australasia and the Pacific. Two species were reported for Australia by Gradstein (1975), A. aulacophora ( Montagne 1843: 259) Stephani (1895: 317) from Tasmania and Norfolk Island, and A. securifolia (Nees in Endlicher 1833: 5) Steph. ex Watts (1901: 215) subsp. securifolia from north eastern New South Wales, south eastern Queensland, and Norfolk Island.

However, the total distribution of Acrolejeunea presented by Gradstein (1975) did not include the Wet Tropics Bioregion of North-east Queensland, presumably because no specimens from this diverse but undercollected region existed, or were available for study. As a result of subsequent collecting effort, three species of Acrolejeunea are currently recognized for Australia and all are known to occur in the Wet Tropics Bioregion, including A. pycnoclada ( Taylor 1846: 385) Schiffner (1893: 126) subsp. pycnoclada , first reported for Australia by Thiers & Gradstein (1989).

During fieldwork within the Wet Tropics Bioregion in July 2005 a fourth Acrolejeunea was collected from two montane sites. This entity is distinctive in its reddish-brown coloration, long sparingly branched shoots, long narrow lobules with two teeth, oblong underleaves, autoicy, male bracteoles smaller than vegetative leaves and produced on short lateral branches, rounded female bracts in one gyre, and 8–10 isoplicate perianths, but is difficult to assign with confidence to any recognized species. It has affinities to A. arcuata in its reddish brown colouration, and lobule shape, and is vegetatively most similar to the type material of Ptychocoleus hians Stephani (1912: 45) in its large size, its rounded leaf lobes, its narrow-oblong underleaves and the second lobule tooth being longer than the first. However, unlike this plant Australian material is autoicous, and has a single gyre of female bracts, whose lobe apices are rounded. This combination of character states is intermediate between A. arcuata and A. pycnoclada . Although similar to A. arcuata in overall morphology, including the distinctive colouration, Australian plants differ from A. arcuata as defined by Gradstein (1975) in their rounded female bract apices, autoicy, larger stature, the oblong-elliptic underleaves, the first lobule tooth is smaller than the second lobule tooth, and geographic range.

The taxonomic value of character differences

Lobule shape and the number and shape of the lobule teeth; the shape of the underleaf; and the shape of the female involucre were all regarded as taxonomically informative characters by Gradstein (1975). Sexuality was regarded as providing supporting evidence for the separation of taxonomic entities, for instance A. securifolia subsp. securifolia (autoicious) and A. securifolia subsp. hartmannii ( Stephani 1889: 164) Gradstein (1975: 99) (dioicous), however several species are apparently both autoicious and dioicous, including A. emergens ( Mitten 1879: 397) Stephani (1895: 65) , A. securifolia , and A. torulosa ( Lehmann 1834: 41) Schiffner (1893: 128) . The Australian plants differ from overseas plants of A. arcuata in these four character systems.

What to do with Australian Plants?

In the Acrolejeunea monograph, Gradstein (1975) discussed species recognition and the treatment of intraspecific variation within Acrolejeunea , and his concepts have been widely applied to liverwort taxonomy. For Gradstein, species were morphological entities diagnosable by the presence of at least two correlated qualitative characters. For variation within species, Gradstein distinguished between continuous and discontinuous intra-specific variation, and only formalized instances of the latter. When clear-cut intraspecific differences, comprising one or two character states, were found in allopatric populations, these were recognized as subspecies. Varietal rank was employed for forms that were sympatric, usually each variety exhibited a limited distribution. Gradstein did not believe that liverwort taxonomists should worry about subtle morphological variants, because the taxonomic interpretation of morphological variation in liverworts is both difficult and subjective. A recent plethora of papers identifying ‘cryptic’ morphological species ( Heinrichs et al. 2009, 2010, 2011, Bombosch et al. 2010, Kriers et al. 2010, Pätsch et al. 2010) suggests the opposite, that liverwort taxonomists must reconcile morphological variation with predictive explanatory hypotheses at every step, especially given studies demonstrating that subtle morphological differences can indicate phylogenetically distinct, indeed isolated, lineages ( Renner et al. 2010, 2011).

While the differences that exist between the Australian plants and the various forms of A. arcuata found elsewhere may seem insignificant, that they are constant suggests origination and fixation of character state differences in an independent, reproductively isolated lineage. Commensurate with established standards for recognition of subtly differentiated morphological forms as infraspecific taxa within Acrolejeunea , I propose a new subspecies of A. arcuata as a formal placeholder for the hypothesis that Australian plants comprise a reproductively isolated, morphologically distinct, lineage.

Key to Acrolejeunea arcuata and taxa in Australia

1a. Plants large, shoots to 1.5 mm wide, reddish brown; leaf lobe cells arranged in radiating rows; lobules narrowly pyriform, keel straight most its length, inclined at 30–45° to stem (measured on shoot apex side of angle), lobule bearing two teeth........................................................................................................................................................... 2

1b. Plants medium to small, shoots 0.8–1.3 mm wide, mid-green, yellow-green or grey-green, never reddish; leaf lobe cells not arranged in radiating rows; lobules narrowly obpyriform to trapeziform, keel straight or arched, inclined at 60–80° to the stem, lobule bearing three or four teeth.................................................................................................. 3

2a. Dioicous; androecia intercalary on leading shoots; female bracts in two gyres, bract lobe apices obtuse-acute, bract lobes particularly those of inner gyre often reflexed; perianths exserted above bracts at maturity, underleaves ovoid to longitudinally oblong; lobule second tooth shorter or longer than lobule first tooth ................................................. ............................................................................................................................. Acrolejeunea arcuata subsp. arcuata

2b. Autoicous; androecia on short lateral branches; female bracts in one gyre, bract lobe apex rounded, bract lobes not reflexed; perianths immersed within bracts at maturity; underleaves longitudinally oblong c. 0.2× the shoot width; lobule second tooth longer than lobule first tooth .......................................... Acrolejeunea arcuata subsp. gradsteinii

3a. Lobule with four teeth, the first of one or two cells, the remainder unicellular, evenly distributed along the broadly rounded lobule free antical margin; underleaf weakly obovate; female bract margins entire ........................................ ............................................................................................................................................ Acrolejeunea aulacophora

3b. Lobule with three teeth, variously one to three celled, unevenly distributed along the lobule free antical margin; underleaf ovate or obovate; female bract margins scalloped or entire ......................................................................... 4

4a. Lobule arch one or two cells, (first lobule tooth one or two cells from the junction of the lobe and lobule), first and/ or second lobule tooth hooked, lobule carinal region broadly pyriform, keel curved; leaf lobe cells finely punctulate; female bracts cucullate, margins entire; perianth rostrum 8–12 cell tiers high .............................................................. ................................................................................................................. Acrolejeunea pycnoclada subsp. pycnoclada

4b. Lobule arch 4–5 cells (first lobule tooth four or five cells from the junction of the lobe and lobule), lobule teeth not hooked, lobule carinal region narrowly pyriform, keel straight; leaf lobe cells unornamented; female bracts plane, margins scalloped; perianth rostrum 3–7 cell tiers high ............................ Acrolejeunea securifolia subsp. securifolia .