The genus Idertia (Ochnaceae)

and deemed continuous. Only a single, though variable species distributed from Guinea to Uganda is recognized. A full taxonomic treatment is provided and two lectotypes are designated. Conclusions – Idertia is a well-defined and sufficiently distinct monotypic genus.


INTRODUCTION
The genus Idertia was created by Farron (1963: 211-212) to accommodate three tropical African species previously treated as members of the genus Ouratea Aubl.s.l.Later, Farron (1968: 227;1985: 60, 66) added a fourth species, based on a single specimen from Sao Tomé, but left it undescribed.Amaral & Bittrich (accepted) followed Farron in his concept of this genus.
Following the subdivision of the family by Kanis (1968), Idertia belongs to the subfamily Ochnoideae, tribe Ochnea e, subtribe Ouratinae (Engl.)Kanis.This subtribe contains three other genera: Rhabdophyllum Tiegh.(Africa, eight species; Sosef 2008), Ouratea Aubl.(Neotropics, c. 140 species; Sastre 1988) and Campylospermum Tiegh.(some 55 species in Africa and Madagascar, and two in tropical Asia; Farron 1985, Kanis 1968).It is characterized by having 10 stamens with sessile or subsessile anthers and a gynobasic style in between a ring of almost free carpels.These develop into an apocarpous fruit with the mericarps sitting on a swollen re-ceptacle.The interesting group of African Ouratinae are the subject of on-going research efforts at Wageningen, in cooperation with the Senckenberg Research Institute at Frankfurt.Rhabdophyllum has been revised some years ago (Sosef 2008) while the notoriously problematic Campylospermum is subject to a revision of the continental African species by Bissiengou (Bissiengou & Sosef 2008;Bissiengou et al. 2013).Ouratea has been studied by Sastre (1988).Although Farron (1963Farron ( , 1985) ) and Bamps & Farron (1967) represent useful contributions, Idertia has never been subject to a thorough taxonomic revision, a situation we hope to improve now.

MATERIAL AND METHODS
We carefully studied the variation within dried material of the genus Idertia from the following herbaria: B, BM, BR, BRLU, COI, EA, G, K, MO, P, WAG.Standard herbarium techniques were used.In all, this material represents 73 different collections.Many of these (50) were collected after Pl.Ecol. Evol. 146 (3), 2013 Farron's 1985 paper or were never seen by him and hence we are now better equipped to make the right taxonomic decisions.All specimen data have been entered into the BRAHMS (Botanical Research and Herbarium Management System, http://herbaria.plants.ox.ac.uk) database of the National Herbarium of the Netherlands (now Naturalis Biodiversity Center) and will soon be available on-line through http://vstbol.leidenuniv.nlor upon request to the author.
Additional material from related genera was studied at WAG. Farron (1963) provides several morphological characteristics for Idertia that make it stand out from the rest of the subtribe.He, as well as the earlier Ochnaceae specialist van Tieghem (1902avan Tieghem ( , 1902bvan Tieghem ( , 1907)), attributes great value to the shape and position of the embryo and cotyledons.For Idertia, Farron mentions a straight embryo accompanied by two straight coty ledons of similar size with the radicle in between them (cotyledons accumbent).Within the Ouratinae, this state is only encountered in the New World genus Ouratea from which Idertia sufficiently differs in having persistent and strongly accrescent sepals and intra-axillary and fused stipules.It shares the latter characters with the related Campylospermum and Rhabdophyllum.Moreover, the few-flowered and axillary inflorescence of Idertia also sets it apart from the vast majority of other Ouratinae species while the leaf margins carry unique long setae.Furthermore, Farron (1963) has observed a hypogeal germination (cotyledons below ground) in Idertia morsonii Hutch.& Dalz.while it is epigeal (cotyledons above ground) in two species of Campylospermum, but there are too few observations to give them some taxonomic weight.Finally, Farron (1963) states that the pollen of both Idertia species he studied differs from all other African and American Ouratinae.Unfortunately, he never published his palynological observations, and it seems worthwhile to perform such a study, the one of Amaral (1991) providing only limited data.Wood anatomy -The wood anatomy of Idertia was studied by den Outer (1977) and den Outer & Schutz (1981).The observations place Idertia closest to Campylospermum with the main part of the axial system occupied by parenchyma cells with scattered sieve tubes that are almost rectangular in transverse section.Idertia differs from the three investigated Campylospermum species in having on average 4 cells per phloem-parenchyma strand, and phloem rays with long or short uniseriate tails and a maximum of 5 rows of cells.In Idertia, no crystals were observed, which are usually present in Ochnaceae.Stipules -In Idertia, these are intra-axillary and fused, as in all African Ouratinae [except for Campylospermum lecomtei (Tiegh.)Farron, where they are linear and have secondarily become free again], and the structure regularly carries two apices.The American genus Ouratea stands out in the subtribe by having lateral and free stipules.Leaves -The margin of Idertia leaves is undulate to bluntly serrate, with the actual margin consisting of a slightly thick-ened rim similar to the situation in most species in Campylospermum and Rhabdophyllum.The most striking leaf feature are the long setae that protrude from this thickened rim (fig.2B).Despite of what Farron (1963: 211) says ("sétules ..... qui prolongent le plus souvent les nervures secondaires"), we observed that the setae actually arise in an irregular pattern, and are not associated with the secondary veins but rather with the undulations or blunt teeth.

Diagnostic characters for Idertia
Within the Ouratinae, the leaf venation patterns provide important characteristics (Farron 1968: 188).In Idertia, the main secondary veins protrude in a slightly oblique angle from the midrib to run straight or only with a slight curve to the margin.Just before the margin they curve up and fuse with it (figs 1 & 2B).As such, this venation type is intermediate between the camptodromous type (Ellis et al. 2009; main secondaries curving up to eventually run parallel to the margin, as in most Campylospermum species) and the craspedodromous type (main secondaries running straight to the margin, as in all Rhabdophyllum species).The main secondaries are fairly regularly and rather closely spaced (when compared with Campylospermum), but do not come near the very dense spacing found in Rhabdophyllum.Intersecondary veins (not reaching the margin) are present, 0-2(-3) in between each pair of main secondaries.The intercostal veins are scalariform, at least in the outer half of the leaf blade (fig.2B), similar to the situation in most species of Campylospermum.
The stomata of the twenty species of African Ouratinae investigated by Farron (1963) are always present only at the lower leaf surface.While in Rhabdophyllum the stomata are of the paracytical type (subsidiary cells that flank the stomata parallel to the long axis of the guard cells), he observed stomata of the anomocytic type (no distinct flanking cells, surrounding cells inirregular pattern) in both Idertia and Campylospermum.Inflorescence -In Idertia, they are strictly axillary, a character shared with all species of Rhabdophyllum, a few species of Campylospermum and other genera of the tribe Ochneae like Ochna.They are composed of a single fascicle of only 1-2(-4) flowers, representing a strongly reduced cyme.Occasionally, the inflorescence may seem terminal, because several fascicles are positioned near the tip of a twig.However, upon closer inspection one can observe that these are all arising from the axil of a stipule-like scale and that growth continues apically.Such seemingly terminal inflorescences may occur when the apical growth has almost halted (during the dry season?).Later on, the twig starts growing again, producing leaves etc., and the fascicles are now more obviously in an axillary position.A similar phenomenon can be observed in Rhabdophyllum.
Previously, it was reported that Rhabdophyllum and Campylospermum would differ in having monochasial and dichasial fascicles (or cymules), respectively (Farron 1968: 192, Sosef 2008).However, new observations now revealed the presence of dichasial fascicles in at least two species of Rhabdophyllum [R.arnoldianum (De Wild.& T.Durand) Tiegh.and R. calophyllum (Hook.f.) Tiegh.], while in others the true nature of the fascicles was difficult to establish because of the low number of flowers per fascicle.For the Sosef, The genus Idertia same reason, it could not be determined whether the fascicles in Idertia are dichasial or monochasial, and it seems best to refrain from using this character for a distinction at genus level.Flower -In Idertia, the pedicel is articulated at or very close to the base and is accrescent in fruit.The sepals are quincuncial (two outside, two inside, one half outside, half inside).In fruit, they are persistent, turn red or pink and are strongly accrescent (fig.2D), the ultimate size being matched only by species like Campylospermum umbricolum (Tiegh.)Farron or Ochna staudtii Engl.& Gilg.Sastre (1988) claims that the six species of Ouratea sect.Persistens Sastre also have persistent sepals.These do indeed remain attached during the first phase of the fruit development, but fall off before the fruits are fully mature and, moreover, are not accrescent and coloured.The latter feature is a clear element of the bird dispersal syndrome (see below) typical for the African Ouratinae.The sepals lack the character typical for Rhabdophyllum where most carry one or two dorsal strips that interlock with the margin of the adjacent sepal in bud (Sosef 2008: Fig. 3).The petals stretch to become narrowly elliptic, not unguiculate (figs 1, 2A & C).In bud, they are not curved inward to clasp an anther, and so the bud is not 'partitioned' [or "cloisonnée", van Tieghem (1902b: 182)].Hence, they are unlike those in the related genera of the Ouratinae.There are (9-) 10 stamens, the anthers of which sit on a short filament and open by apical pores, which is typical for the subtribe.The anthers are, however, smooth (fig.2A) and linear, not transversely wrinkled and narrowly pear-shaped as in the other three Ouratinae genera.The gynoecium has a singly gynobasic style and is composed of 5-6 almost free carpels which soon become entirely free in fruit.Fruit -In Idertia, the receptacle, and to a lesser extent also the pedicel, becomes inflated and whitish to pinkish or reddish in fruit, contrasting with the black mericarps (fig.2D) and probably favouring bird dispersal (van der Pijl 1982, Steentoft 1988).In Ouratea, which lacks the accrescent sepals in fruit, the swelling of the receptacle and pedicel is often much more pronounced.Idertia seeds have a straight embryo and straight accumbent cotyledons (see above).
In all, Idertia seems to be a clear member of the Ouratinae (fused intra-axillary stipules, 10 stamens on short filaments, persistent and accrescent sepals), but seems to hold an intermediate position between Campylospermum and Rhabdophyllum on the one hand and Ouratea on the other.Moreover, some features, such as the smooth and linear anthers and axillary inflorescences, may form a bridge to other members of the tribe Ochneae outside of the subtribe.Preliminary results of a molecular phylogenetic analysis of the Ochnaceae (Bissiengou, Wageningen University, The Netherlands & CENAREST, Gabon, pers.comm.)show Idertia as sister to Ouratea and Rhabdophyllum with Campylospermum and Ochna s.l.being sister to that, but resolution as well as support for this tree are both still low.In conclusion, there are ample reasons to regard Idertia as being a well-defined and distinct genus of the Ouratinae; the key below providing the most prominent morphological differences between the four genera of this subtribe.

How many species?
As said above, Farron (1968Farron ( , 1985) ) recognized four species within Idertia, one of which remained undescribed.His species are clearly geographically separated: I. morsonii (Hutch.& Dalz.)Farron in the Upper Guinean phytochorion (Guinea to Ghana; White 1979, Linder et al. 2005), I. axillaris (Oliv.)Farron in Lower Guinea (Nigeria to Gabon), I. mildbraedii (Gilg) Farron in Congolia (Democratic Republic of the Congo and Uganda), and his undescribed species on the island of Sao Tomé.In his 1985 publication, Farron gives a key to all four species, providing the characters he uses to distinguish them: leaf consistency, leaf measurements and shape, acumen length, relative abundance of setae.
The main difference between Farron's I. morsonii and his undescribed species from Sao Tomé on one side and the other two is the leaf size.Although at first glance, the variation in leaf size is striking (those in Upper Guinea regularly being as small as 5 × 2 cm, those in Lower Guinea not seldom as large as 23 × 7 cm), we observed a clear continuum of variation within the material at hand.There is only a tendency of larger-leaved specimens occurring in Lower Guinea and Congolia and smaller-leaved ones in Upper Guinea.In fact, the same is true for the length of the acumen: in smaller-sized leaves the acumen is generally shorter or even absent, in larger leaves the acumen is generally longer, but we were unable to establish a clear correlation between acumen length and any of the other characters.While I. mildbraedii would stand out because of its chartaceous, instead of papyraceous, leaves, various recent collections from eastern D.R.Congo clearly have less firm leaves.The undescribed species from Sao Tomé (based on the specimen G. Watt 7118) would be distinguished based on its narrowly elliptic leaves, but, again, a continuous variation from leaves with a length/width ratio of as low as 2 to as high as 4.5 could be observed in the large amount of material at hand.The Sao Tomé specimen itself also shows high variation in this respect, the length/width ratio of its leaves being 2.3-4.Similarly, leaves of I. axillaris would be, according to Farron's key, always obovate, which could not be confirmed with the present material.Especially the many specimens collected in 2010 by Ms P. Bissiengou in SW Cameroon proved very useful in understanding the natural variability in leaf shape.Finally, Farron used the relative abundance of the setae along the leaf margin as a diagnostic character.He claimed that I. mildbraedii would have few or no setae, but based this statement on just two specimens from D.R.Congo and one from Uganda (Bamps & Farron 1967: 23).We have several recent collections from eastern D.R.Congo showing abundant setae along the margin.
In conclusion, we recognize a single, though variable, species within Idertia.We studied other characters as well, to see if no infraspecific taxa could be recognized.The variation turning out to be truly continuous, we concluded that even that was not possible.An adjusted species description, synonymy, etc. is given below.
Distribution -Guinea, Sierra Leone, Liberia, Ivory Coast, south-western Ghana, Cameroon, Equatorial Guinea (Rio Muni), Sao Tomé & Príncipe (Sao Tomé), Gabon, Republic of the Congo, Democratic Republic of the Congo and Uganda (fig.3).Habitat -Primary and secondary forest, on dry land to seasonally inundated places, in hill forest and forest on river banks, on clayey to sandy or rocky soil; at 50-800 m altitude.Uses -In Liberia, the wood is used to make axe handles (Voorhoeve 2209).In Ivory Coast, the stems are used for roof construction of rural houses (Tra Bi 1997).IUCN conservation assessment -Least Concern.Due to its comparatively large AOO (area of occupancy, 3,027,350 km 2 ) and EOO (extent of occurrence, 2,818,436 km 2 ), and the fact that the species is hardly used and not commercially exploited, it is best accommodated in the IUCN Red List category of Least Concern.Nomenclatorial notes -Gomphia axillaris was described by Oliver (1868) with two syntypes: Mann 1787 from Gabon and Morson s.n.from Sierra Leone.The latter was taken as the type of Ouratea morsonii by Hutchinson & Dalziel (1927).Therefore, the most obvious choice of a lectotype for G. axillaris is the remaining Mann 1787 syntype, but this was never effected.
Ouratea mildbraedii was described with two syntypes, Mildbraed 2918 and 2932, and an accompanying plate.Both Mildbraed specimens were lost in Berlin and despite a very wide search in all relevant herbaria no duplicate was encountered, neither by us nor by Farron (from a remark in his specimen card system).Therefore, the plate, being the only remaining part of the original material (McNeill 2012: Art.9.3), has to be chosen as lectotype.