Alyxia rosmarinifolia (Baill.) Guillaumin (1941: 366)

Alyxia rosmarinifolia (Baill.) Guillaumin (1941: 366) View in CoL .

Gynopogon rosmarinifolius (“rosmarinifolium”) Baillon (1889: 782).

Lectotype (designated by Middleton (2002: 66):— NEW CALEDONIA, Baie de Canala , sur les collines éruptives, 29 June 1869, Balansa 2428: P [ P00156839 !]; isolectotype: P [ P00156840 !], K [ K00894172 !] ( Figs. 14 View FIGURE 14 , 15 View FIGURE 15 )

Species circumscription: — Alyxia rosmarinifolia was considered a synonym of Alyxia tisserantii by both Boiteau (1981) and Middleton (2002). We propose to recognise this name as a valid species. Alyxia rosmarinifolia is close to A. dolioliflora on the basis of sterile material because of its leaf shape. However, A. rosmarinifolia has flowers with a corolla tube 0.9–2 mm long, vs 2.5–4.1 mm long for A. dolioliflora . Its fruit articles are also of spherical shape, while A. dolioliflora has ovoid fruit articles. Finally, A. rosmarinifolia is found between 800 and 1000 m elev. in its only present day known location (Mont Do, south of Grande Terre), while A. dolioliflora grows in low altitude maquis in the Northwestern ultramafic maquis of Grande Terre. The species is also very similar to Alyxia minimiflora described here, but differs from it by being entirely glabrous, and having flower borne on inflorescences with a primary axis ca. 3.7 mm (vs 1 mm in A. minimiflora ), and longer pedicels (up to 2 mm). Since Balansa’s original collection (only fruiting material), no further specimen of this species has been collected in the type locality (Baie de Canala, sur les collines éruptives). We have some doubts because of locality disjunction between Baie de Canala and Mont Do about identifying the specimens from the latter locality under A. rosmarinifolia . But, at this stage, this seems the best option instead of resurrecting a possibly extinct species and describing a new narrow range one from Mont Do.

Distribution and habitat:— The only known population of A. rosmarinifolia occurs on top of Mont Do, in the southern part of Grande Terre. It grows in open herbaceous maquis, on peridotite rocks, between 800 and 1000 m elev.

Phenology:— From herbarium specimens and available pictures, the species is flowering in October and November. Fruits were observed in November, March, April, and June.

Notes:— The isolectotype held in K (K00894172) has “Nouméa” written on the label, as the locality. It carries the same Balansa collection number, however, and regarding shape and drying, is the same plant as the ones at P. Although the original label is Balansa’s one, hand-written entries (name, locality and number) are not from Balansa. In fact, the handwritten number is the same as the P isolectotype and not the same as the P lectotype, the latter being written by Balansa himself. In addition to the name, the locality and number were most likely written by Guillaumin when he published the invalid combination for the first time ( Guillaumin 1911: 194) and before the specimen was sent to K.

Specimens examined:— NEW CALEDONIA; South Province; Mont Do , elev. 950 m, 21°45'28"S, 166°0'06"E, 20 April 2021, Lannuzel 431 ( NOU108249 !) GoogleMaps ; Crête sommet Mt. Do, elev. 900 m, 28 November 1966, MacKee 15975 (L.2699553!, MO1114544 !, NOU058322 !, P04225751!) ; Mt. Do , elev. 900 m, 23 March 1977, MacKee 32938 ( NOU058325 !, P04225542!) ; Mont Do, elev. 800 m, 21 October 1975, Sévenet 1075 ( NOU058320 !, NOU058321 !, P04225557!, P04225558!) ; Mont Do , elev. 800–1000 m, 28 November 1966, Veillon 941 (L.3729837!, NOU058324 !, P04220405!) .