Studies on Potheae (Araceae) of Borneo II: Pedicellarum M.Hotta subsumed into Pothos L., and recognition of three new species

Pedicellarum is subsumed into Pothos as Pothos paiei. Four species, three taxonomically novel (Pothos degenerans, ecclesiae and P. fractiflexus) and one pre-existing (Pothos oliganthus) are discussed. The reproductive structures are reinterpreted and floral terminology used in this paper is in line with paving the way to interpreting the ‘inflorescence’ of Pothos, and indeed of all other aroids, as a unique array of structures that are neither flowers nor inflorescences as defined by current prevailing orthodoxy, but something of both.


INTRODUCTION
Pedicellarum paiei M.Hotta (Hotta 1976) is the sole described species of a genus of nomadic vines restricted to upper hill ridgetop kerangas in SW Sarawak. It was originally based on a single collection made close to the border between Sarawak and Kalimantan Barat in 1962, but later Nicolson (1984 drew attention to a previously overlooked second collection made 60 km further north about 35 years before. Pedicellarum has been considered unique amongst bisexual-floreted genera by possessing stipitate florets, with the top of the stipe swollen into a receptacle, an urceolate perigone of fused tepals, and (purportedly) introrse anther dehiscence. Grayum (1984: 634;1992) remarked on a suite fertile and vegetative morphological similarities seemingly connecting Pedicellarum and Pothos species, and proposed a combination in Pothos although without effective publication (Grayum 1984: 634). Later, Grayum (1990: 670) revised his opinion and left the status of Pedicellarum unchanged. In fact, the earlier comparisons were based upon examination of material (P.J. Martin & O.Ismawi S 36660) mis-determined as Pedicellarum and then actually an undescribed species of Pothos. Boyce and Hay (2001: 455) expanded on the similarities of Pedicellarum and Pothos species of the Goniurus supergroup (to which Hotta (1976) had also alluded relationship), noting the presence of a distinct receptacle, and fused tepals, and stating (as did Mayo et al. 1997) that anther dehiscence is latrorse, not introrse. Boyce & Hay also clarified the identity of the P.J.Martin & O.Ismawi S 36660 collection referred to by Grayum (1990), describing it as a new species -Pothos oliganthus P.C.Boyce & A.Hay, and assigned an additional eight collections to Pedicellarum, extending the range of the genus south into Kalimantan Barat and north east to eastern Sabah.
While databasing specimens preparatory for the second author's Master's thesis it was necessary to reexamine the SAN duplicates of the two Sabah collections [Fidilis Krispinus SAN 104271 and Patrick Lassan SAN 107216] assigned to Pedicellarum paiei by Boyce and Hay (2000), and study for the first time a third, later, collection [Suzana S. SAN 154991]. All three collections, although agreeing with Pedicellarum paiei in having stipitate florets on a characteristically zig-zagging axis, lack the characteristic conspicuous floral receptacle, and have the fused tepals forming a shallow saucer; additionally the leaf blades are proportionately narrower and thicker textured than in P. paiei, and more or less straight, lacking the characteristic narrowed falcate terminus, while the mid-rib is abaxially much more pronounced, and drying much paler than the surrounding tissues. It was obvious that these collections, gathered over 800 km to the north east of the Type of Pedicellarum paiei, represented an undescribed species. Thus alerted, we examined the other duplicates assigned to Pedicellarum paiei, starting with three Alison Church collections [A.C. Church et al. 303, A.C. Church et al. 647and A.C. Church et al. 1014] from Kalimantan Barat. While appearing an excellent match for P. paiei by having stipitate florets with a receptacle and chartaceous leaf blades with a conspicuously narrowed falcate terminus, detailed examination, however, revealed that although the perigone lobes are fused below, the tips are evidently free. Examination of two collections previously unseen (P. Wilkie 94175 and Tuke P9 1102) from 75 km further south in Kalimantan Tengah confirmed the same morphologies. These five collections represented a third species distinct from both Pedicellarum paiei and from the Sabahan material.
Lastly, we turned attention to F.H. Endert 2942 from central Kalimantan Timur and a previously unseen collection (Bernard Lee Meng Hock S 45534) from Belaga, Kapit, Sarawak. Both have overall similarity to the Sabah specimens in possessing straight coriaceous leaf blades with an abaxially pronounced mid-rib drying paler than the surrounding tissues. Examination of the blooms, however, revealed differences in the straight, not zig-zagging spadix axis, almost sessile florets with largely, but not wholly, fused tepals, and blooms all depauperate, each with just two or three florets. Moreover, the blooms are produced in dense fascicles and (presumably) sequentially.
These examinations of material previously associated with Pedicellarum paiei revealed that there are five species involved, with three -one from western Kalimantan, one from central Borneo, and one from Sabah -representing undescribed taxa. Furthermore, taken as a whole these species cover a full range of the combinations of the 'characters' used to define Pedicellarum and those of Pothos such that maintenance of Pedicellarum as genus separate from Pothos is no longer justifiable. We therefore propose that Pedicellarum be merged with Pothos, including transferal of Pedicellarum paiei to Pothos, and that the three taxonomically novel species discussed above be described in Pothos.
Some floral terminology used in this paper differs intentionally from those of previous papers on Pothos and is intended to pave the way to interpreting the 'inflorescence' of Pothos, and indeed of all other aroids, as a unique array of structures that are neither flowers nor inflorescences as defined by current prevailing orthodoxy, but something of both. For the background to this conceptual change see Hay (2019), Hay and Mabberley (1991), Mabberley and Hay (1994). The 'pedicellarums' described in this paper are perhaps particularly germane as they present some of the inflorescences or blooms, and flowers or florets of Araceae that most resemble, at least to first glance, racemose inflorescences of more or less standard pedicellate trimerous monocotyledonous flowers of text books. Nevertheless, this is an illusion.
First, the spadices in these species, where phenology is known, are protogynous as entities, which is the standard behaviour of Araceae. So, despite their strikingly distant positioning, the florets do not behave as individual flowers but act in concert in a protogynous spadix. Second, again a standard feature of Araceae, there are no bracts subtending the florets, in spite of the resemblance of the spadix to a raceme in these species. The use of the term 'pedicel' for the stalks of the indi-vidual florets is prejudiced towards the conventional interpretation of the bloom as an inflorescence, despite the absence of subtending bracts and the phenological behaviour of the florets being subordinate to that of the spadix. We therefore use the more neutral (in this context) term 'stipe' and the corresponding adjective 'stipitate', rather than describing the florets as 'pedicellate'.
Another interesting aspect of the morphology of the florets of this group of species concerns the perigon. That of the sessile florets of Pothos oliganthus consists, conventionally for Pothos, of six free tepals, whereas that of the stipitate florets of Pothos paiei is a truncate cupule set on what appears to be an enlarged receptacle. This cupule is conventionally interpreted as a perigone of connate tepals, though, since there are no apparent tepal lobes at all, it might also be suggested that the tepals have been replaced by this cupular outgrowth of the receptacle. [Sadly, there are no floral developmental studies of this species group, for which it is extremely hard to find suitable quantities of fresh material]. Were one to view floret variation here as a morphocline, of the species newly described in this paper, Pothos degenerans, with sessile florets and a proximally cupular but distally lobed perigone sits nearer to P. oliganthus; Pothos ecclesiae, with stipitate florets and like P. degenerans a proximally cupular and distally lobed perigone is somewhat nearer P. paiei; while P. fractiflexus, with stipitate florets and an entirely cupular (but very shallow) perigone is nearer still to P. paiei.
We cannot speculate on the ecological function of these strangely stipitate, cupular-perigonate florets. However, it is noteworthy that they occur in perhaps the most extreme leptocaul of all aroids, a family mostly with relatively to absolutely massive pachycaul construction (except Lemnoideae, of course). Perhaps extreme leptocauly can sometimes lead to structural decanalisation.

Diagnosis
Pothos degenerans is most similar to Pothos fractiflexus by the straight coriaceous leaf blades with an abaxially pronounced mid-rib drying paler than the surrounding tissues. However, Pothos degenerans has up to 3 florets per spadix (vs 7 or more in P. fractiflexus), an almost straight or only very weakly (not strongly) zig-zagging spadix axis, sessile (vs stipitate) florets with a receptacle of partially (not wholly) fused tepals.

Etymology
From Latin degenerans, reduced, in allusion to the blooms, the smallest yet recorded for the genus, each with at most three florets.

Distribution
Central Borneo, known from two localities approximately 300 km apart.

Ecology
Lowland forested ridges and riverine forest, to about 200 m asl. The Kapit collection is from Eocene sedimentaries, the Kalimantan one from Cretaceous-Jurassic deepwater sediments.

Notes
The remarkably slender and diminutive blooms of Pothos degenerans are both easily overlooked and readily detached from specimens. Each flowering node has numerous denuded pedicels indicating blooms produced in fascicles, but it is not possible from the available specimens to determine if the blooms mature in a sequential series (as in all known Pothos producing fas-cicles of blooms), or if all the blooms in a fascicle reach anthesis simultaneously. The absence of any fruiting seems to favour the second scenario except that the large number of missing blooms indicated by the barren pedicels might have taken with them any developing fruitsfavouring the first scenario.
Other specimen examined

Diagnosis
Pothos ecclesiae is most easily confounded with Pothos paiei in having a conspicuous receptacle, stipitate florets, and chartaceous leaf blades with a conspicuously narrowed falcate terminus. However, the free perigone lobes of Pothos ecclesiae are immediately diagnostic. The two species are also ecologically dissimilar, with P. ecclesiae occurring in lowland dipterocarp-dominated forest on red clay soils whereas P. paiei occurs on upper hill kerangas ridges.

Description
Probably heterophyllous, root-climbing nomadic leptocaul vines. Adult shoot system not completely known, from available material differentiated into non-flowering orthotropic adherent stems, free lateral fertile stems, and much abbreviated cataphyll-encased (foliage-) leafless flowering shoots arising from below the petiole insertion; eocaul, seedling, and flagellate foraging shoots unknown. Leaves chartaceous, drying dull pale brown; petiole 1-3 × 0.1-0.2 cm, slender, very narrowly canaliculate, apex geniculate, sheath indistinct, no sign of it clasping the stem in material available; blade 5-17 × 1.5-5 cm, elliptic or ovate-elliptic, with one side unequal to the other, apex long-acuminate, markedly falcate, apiculate, base subacute to obtuse; primary lateral veins arising at 60-70°, 2 intramarginal veins per side, 2-15 mm from blade margin, arising from just above base of the midrib, remaining ± parallel to margin, terminating at the tip of the blade. Blooms solitary to rarely paired;    (Figures 4 and 5).

Diagnosis
Pothos fractiflexus is similar to Pothos paiei in having stipitate florets on a characteristically zig-zagging spadix axis but differs in the fused tepals forming a shallow saucer (vs florets with an enlarged floral receptacle and tepals fused into a cupular structure), leaf blades proportionately narrower and thicker textured, and more or less straight (vs leaf blades chartaceous with a characteristic narrowed falcate terminus), and with the midrib abaxially much more pronounced, and drying much paler than the surrounding tissues.

Etymology
From fractiflex -zig-zagging -used to describe the spadix axis.

Distribution
East Sabah.

Ecology
Forested hillsides on Neogene or Quaternary sediments between 300 and 500 m asl.

Notes
In general appearance very close to Pothos degenerans, but readily differentiated by stipitate florets and shallow saucer-like perigone.
Other specimens examined

Distribution
SW Sarawak.

Habitat
Primary to disturbed secondary upper hill ridge kerangas forest between 600-950 m altitude.