Pleurandra diamesogenos Steud. was described by Steudel (1845) based on Preiss 2141, collected near Busselton in south-west Western Australia. The specific epithet is from the Greek dia- (across), meso- (middle) and genos (a kind, in the sense of genus; Sharr & George 2019), indicating Steudel's uncertainty as to whether the species should be placed in Pleurandra Labill. or Hibbertia Andrews, which at that time were recognised as separate genera following De Candolle (1817), Meisner (1837) and Endlicher (1839). Bentham (1863) subsequently treated Pleurandra as a section of Hibbertia, but placed Steudel's P. diamesogenos, as a variety (var. parviflora Benth.) of Hibbertia lineata Steud., in his sect. Hemipleurandra Benth. rather than in sect. Pleurandra (Labill.) Benth.

Neither of these names (Pleurandra diamesogenos or Hibbertia lineata var. parviflora) were taken up in Western Australia, with the later name Hibbertia rhadinopoda F.Muell. being used instead for the species by Green (1985), Wheeler (1987) and Paczkowska & Chapman (2000). Wheeler (2004) recognised that the earlier name has priority and made the combination Hibbertia diamesogenos (Steud.) J.R.Wheeler. She described H. diamesogenos as variable in habit, leaf and flower size, indumentum, and in the presence or absence of staminodes. The latter is surprising, as staminodes have long been regarded as important characters in Hibbertia; indeed, it was the presence of staminodes on the type that led Bentham (1863) to place the species in his sect. Hemipleurandra (which has staminodes) rather than sect. Pleurandra (which does not).

Examination of all material of H. diamesogenos at PERTH and in the field in spring 2020 indicated that the species as currently recognised comprised two distinct and broadly sympatric species (one of which matches the type and the other of which is segregated here as H. parvula K.R.Thiele), each of which was morphologically relatively consistent throughout its range. The recognition of these two species explains the unusual degree of variability of H. diamesogenos as circumscribed by Wheeler (2004), with each species now substantially more morphologically consistent than was H. diamesogenos sens. lat.

While there are two heterotypic synonyms of H. diamesogenos (H. rhadinopoda and H. kochii Maiden & E.Betche), both are good matches for its type and are retained as synonyms of that species.

Subsequent to the segregation of H. parvula, further significant variation within its initial circumscription became apparent when our colleagues Fred & Jean Hort provided photographs of plants from the Darling Range east of Perth, which showed that the species as originally segregated from herbarium specimens had significant and striking variation in androecium and gynoecium morphology, readily observed in their fresh flowers. Plants from the southern part of the range (of H. parvula as initially segregated) have stamens closely parallel and curved over the carpels like a hand of bananas, with the styles short and curving forwards beneath them. However, a photograph of a plant from the northern part of the range showed a flower with much longer stamens, filaments that are free to the base, anthers that are erect and divergent and form an open cluster on one side of the flower, and styles that are long and curve upwards from the carpels on one side of the stamen group. Such differences are highly significant in Hibbertia.

Before seeing the photographs, we had not recognised the striking difference in androecium and gynoecium morphology when examining herbarium material of the 'H. parvula' segregate because it is not readily assessed on dried material. However, closer examination showed that separation between the two morphologies is straightforward if style and stamen lengths are measured: the species with 'hand-of-banana' stamens has distinctly shorter stamens, anthers and styles than the species with divergent, erect stamens. This closer examination also showed that the latter species comprises the northern part of the range of the 'H. parvula' segregate, between approx. York and Bindoon. With this clarification, Hibbertia parvula is here restricted to the widespread southern species with 'hand-of-banana' stamens, while the northern species with an open, erect androecium is described as H. xenandra K.R.Thiele.

In this paper we also describe two new species segregated from Hibbertia avonensis J.R.Wheeler. As circumscribed (Wheeler 2002), H. avonensis is reasonably widespread in the Avon Wheatbelt IBRA bioregion in Western Australia (hence the name). It is an erect shrub to 0.6 m, with narrow, ericoid leaves with the margins recurved tightly to the abaxial midrib so that the abaxial lamina is hidden within lacunae formed between the margins and the midrib. The indumentum of the true abaxial leaf lamina is a taxonomically very significant character. In describing the leaves of H. avonensis Wheeler noted 'the margin of the midrib apparently papillose' but did not describe the abaxial leaf indumentum further. Close examination shows that, in all but one of the specimens at PERTH determined as this species by Wheeler, the true abaxial leaf lamina is densely pubescent with fine, straight, white hairs; these are quite distinct from, and do not appear to grade into, the marginal papillae described by Wheeler, which are the only abaxial leaf surface feature usually visible without dissection. Hibbertia avonensis also has a distinctive, broadly ovate primary bract immediately beneath the calyx, with 1–3 secondary bracts at the base of the pedicel that are similar in size and shape to the primary bract.

Subsequent to Wheeler's description of H. avonensis, two specimens collected near Wandering in the Jarrah Forest IBRA bioregion were determined as that species in PERTH, largely because they have similar floral bracts. These, however, are low, almost prostrate shrubs, and dissection shows that their leaves lack pubescence on the abaxial surface, being instead sparsely papillate. They also consistently have 3–6 stamens, rather than the 10(15) stamens described for H. avonensis. The Wandering specimens are described here as the new species H. micrantha K.R.Thiele & T.Hammer. Likewise, a small group of specimens from west and south-west of Katanning included in H avonensis at PERTH (one of which was determined in that species by Wheeler) are similar in overall morphology to H. micrantha and differ from H. avonensis in lacking pubescence on the abaxial leaf lamina. They differ from H. micrantha in having ten stamens, and are described here as H. humilis K.R.Thiele.

Open-access, online, rapid taxonomy ISSN: 2653-4649 (Online)

Australian Journal of

Taxonomy

Open-access, online, rapid taxonomy
ISSN: 2653-4649 (Online)