Description of Gaertnera luteocarpa (Gentianales: Rubiaceae), with two subspecies, a new forest shrub species from Liberia, Ivory Coast and Ghana

. Gaertnera luteocarpa sp. nov., with two subspecies, is described and illustrated. The new species resembles G. spicata K.Schum. but has yellow, not reddish, fruits. With G. cooperi and G. aurea it shares the ruminate endosperm.


Introduction
The Paleotropical genus Gaertnera Lam.comprises 69 species of shrubs and small trees according to the taxonomical revision by Malcomber & Taylor (2009).The 12 species that they recognize for continental Africa are all endemic.Only five of these grow in the forests west of Nigeria, G. aurea Malcomber, G. cooperi Hutch.& M.B.Moss, G. liberiensis E.M.A.Petit, G. longivaginalis (Schweinf.ex Hiern) E.M.A.Petit and G. paniculata Benth., with the new species described here this number will be six.
Almost 20 years ago I found in Ankasa Forest in south-west Ghana a yellow fruiting Gaertnera shrub.It was thought at that time to be G. cooperi with immature fruits, because of its robust leaves and branchlets, and because the mature fruits of G. cooperi are known to be dark purple.In the revision the fruits of all species of Gaertnera are described as "(as far as observed) all black or violet-black at maturity" (2009: 587).The plant from Ankasa was re-identified by Taylor for the revision as G. aurea (on MO herbarium sheet), probably by eliminating other possible species (the key in the revision only works with flowering plants).Recently, I collected another yellow fruiting Gaertnera, now in Sapo National Park in Liberia, and realised that the large sweet tasting yellow fruits of both plants (Figs 1-2) were in fact mature and not similar to the fruits of G. cooperi (Fig. 3B) or G. aurea.The collection and photos of a red fruiting G. spicata K.Schum.from Gabon by O. Lachenaud (Fig. 3A) confirms that, at least in continental Africa, the colour of the fruits in Gaertnera is indeed more variable than concluded in the revision.The colour of the fruit is not the only character that separates the abovementioned species from G. cooperi and G. aurea.The compact, almost spicate inflorescence and the large size of the mature fruits, in which it resembles G. spicata, are other characters.With G. cooperi and G. aurea it shares the ruminate endosperm, a rare character in African Gaertnera species (Malcomber & Taylor 2009: 581) and not present in G. spicata.There are, in my opinion, clearly 2 new and closely related taxa involved here but, because there are very few open flowers and mature fruits, it is hard to be sure whether they are different species or subspecies.More flowering and fruiting material is needed to get a better idea of their differences.For the moment, I think it is best to describe them as subspecies, G. luteocarpa sp.nov.subsp.luteocarpa and G. luteocarpa sp.nov.subsp.sinoensis subsp.nov.This new species adds a bit more variation to an already very variable genus (Malcomber & Taylor 2009: 590).
The new species was already mentioned a few years ago by William Hawthorne as Gaertnera sp.A (Hawthorne & Jongkind 2006: 646).

Materials and Methods
The available material of Gaertnera from the BR, K, P, and WAG herbaria was studied.The herbarium, where the specimens are located, is indicated by the international code (herbarium acronym) registered in Index Herbariorum (Thiers, continuously updated).Photos were seen from plants in the FHO herbarium cited here.Gaertnera luteocarpa spec.nov.has recently been studied in the field on several occasions.Preliminary assessments of the IUCN Red List categories of threat were performed using the IUCN criteria through the RBG Kew website http://geocat.kew.org.

Diagnosis
Resembling G. spicata K.Schum., with almost similar spicate inflorescences, but differing by its yellow, not red, fruits, ruminate endosperm and by the absence of setae on the edge of the stipular tube.

Etymology
The species is named after its yellow fruits., c. 1 mm high, glabrous or with small hairs on the edge; corolla white, glabrous outside, with acute apex in bud, tube 9 × 2-3 mm, white hairy inside above the insertion of the stamens, glabrous below this point, hairs exserted from the mouth, lobes 5 mm long, glabrous; stamens completely included in the corolla tube; ovary superior.Drupes subglobose with flattened apex, smooth, 2-2.5 cm in diameter, yellow, pulp whitish and sweet, with 2 pyrenes; pyrenes more or less plano-convex, 10 mm in diameter and 5 mm in cross-section, rugose, endosperm ruminated.

Distribution and habitat
Undergrowth of evergreen lowland forest.Known from the south-east of Liberia to the south-west of Ghana.

Conservation status
The "Extent of Occurrence" (EOO) is 20 649 km 2 and the "Area of Occupancy" (AOO) is 36 km 2 , the first counts as "Near Threathened" and the second as "Endangered".The AOO is based on a cell width of 2 km.The calculation of the EOO as one continuous area is (much) too high looking at the big gap between the western occurrences and the eastern ones.Most forest that could be found inside the EOO a century ago has been removed since.All eastern specimens were collected in the Ankasa Reserve and that is the only place where the species is not rare.Because G. luteocarpa sp.nov.is, when flowering or fruiting, a conspicuous plant it should not be easy overlooked in the forest and it can be expected that it is really a rare species that is only becoming more rare because of the ongoing deforestation.Therefore I propose "Endangered" for this new species (B1 & B2 ab(iii) IUCN 2015).

Note
The plagiotrophic branching with distichous leaves has been seen by the author in the field for all Gaertnera species he encountered in West Africa.

Conservation status
The EOO is 1398 km 2 and the AOO is 20 km 2 , both count as "Endangered".The AOO is based on a cell width of 2 km.This calculation of the EOO as one continuous area is much too high looking at the big gap between the most western occurrence and all other ones.Most specimens were collected in the Ankasa Reserve and that is the only forest area where this subspecies is found more than once.Most forests connecting the two locations of the subspecies has disappeared in recent years.Like for the species itself I propose "Endangered" for this subspecies (B1 & B2 ab(iii) IUCN 2015).

Remark
The holotype of G. luteocarpa sp.nov.was not recognized before as a Gaertnera species and for years misplaced in the herbarium as Bertiera racemosa (G.Don) K.Schum.

Diagnosis
Fertile branches half the size as of the other subspecies (2 mm wide in dry state).Stipular tube glabrous, not densely hairy.

Etymology
The subspecies is named after "Sinoe", the river close to which it was seen for the first time.

Distribution and habitat
Known only from the south-east of Liberia.

Conservation status
The EOO is 54 km 2 and the AOO is 12 km 2 , both count as "Endangered".The AOO is based on a cell width of 2 km.Only one of the specimens was collected in a protected area.Considering all the economical development planned in this part of Liberia, I propose "Endangered" for this subspecies (B1 & B2 ab(iii) IUCN 2015).