Introduction
When Brand (Reference Brand1915a) first published the genus Adelocaryum Brand it comprised five species, A. anchusoides (Lindl.) Brand, A. capusii (Franch.) Brand, A. coelestinum (Lindl.) Brand, A. malabaricum (C.B.Clarke) Brand and A. schlagintweitii Brand. Four of these had previously been grouped in Lindelofia Lehm., Cynoglossum L. or Paracaryum (A.DC.) Boiss., while Adelocaryum schlagintweitii was new. Later Brand described two further species in the genus, namely Adelocaryum flexuosum Brand (Brand, Reference Brand1915b) and A. erythraeum Brand (Brand, Reference Brand and Engler1921). Adelocaryum flexuosum and A. schlagintweitii, both described from Kashmir, are still very poorly known; more recent authorities (e.g. Kazmi, Reference Kazmi1971; Nasir, Reference Nasir1989) place one or both of them in Cynoglossum although either or both may be better placed in Paracaryum (see my comment in litt. concerning the possible conspecificity of Paracaryum trinervium Duthie with Adelocaryum/Cynoglossum schlagintweitii: Zhu et al., Reference Zhu, Riedl and Kamelin1995: 424).
Adelocaryum erythraeum was made the type of the new genus Brandella R.R.Mill by Mill (Reference Mill1986). Taxonomic and nomenclatural reasons for its segregation were set out in that paper. However, when I wrote my 1986 paper I had mistakenly (see Riedl, Reference Riedl1992) believed that Riedl (Reference Riedl1962) had chosen Lindelofia anchusoides (Lindl.) Lehm. (based on Cynoglossum anchusoides Lindl.) as ‘provisional’ lectotype of Lindelofia, and Riedl (Reference Riedl1971) had reclassified all Brand’s original five species of Adelocaryum in either Cynoglossum or Lindelofia, leaving Adelocaryum at that time unavailable for Brandella. Since then, Riedl (Reference Riedl1992) has revived the genus Adelocaryum and included Brandella erythraea (Brand) R.R.Mill within it. However, I disagree with that treatment; Adelocaryum as recognised by Riedl (Reference Riedl1992), and by myself (Mill, Reference Mill1991) under the name Paracaryopsis (H.Riedl) R.R.Mill, chiefly (in Riedl’s circumscription) or wholly comprises tall woodland mesophytic herbs with large leaves from rain- and fog-forests of western India and Oman. Brandella erythraea on the other hand is an annual or possibly short-lived biennial herb of the deserts of the Horn of Africa and Arabia. Riedl (Reference Riedl1992) maintained that its nutlet morphology was no different to the other species he included in Adelocaryum and that I ‘obviously … did not know Cynoglossum coelestinum and C. malabaricum’. Neither of those statements is correct and there are considerable differences in the nutlet morphologies of those two members of Adelocaryum compared with Brandella. Brandella is, in its habit, distribution and general morphology (including its nutlets), more similar to Microparacaryum (Popov ex H.Riedl) Hilger & Podlech than it is to any of the other members of Adelocaryum. Indeed, like Microparacaryum, Brandella exhibits heteromericarpy (Riedl, Reference Riedl1992) which the mesophytic perennials of Adelocaryum sensu stricto do not. Riedl himself commented that if one recognises Microparacaryum ‘it is necessary also to separate Brandella Mill from Adelocaryum’. For that reason I continue to exclude Brandella erythraea from Adelocaryum and will be recognising both genera in my treatment of Boraginaceae tribe Cynoglosseae for Flora of Arabia. Riedl (Reference Riedl1992) regarded the heteromericarpic form with a wide, incurved nutlet margin as Adelocaryum erythraeum forma erythraeum because that was the state found on the lectotype. He described the other form, with narrow marginal wing and long flattened glochids, as Adelocaryum erythraeum forma subexalata H.Riedl. This is here recombined under Brandella.Footnote 1
Adelocaryum anchusoides was originally one of the two species treated in the protologue of the genus Lindelofia Lehm. and is correctly classified in that genus (cf. Riedl, Reference Riedl and Rechinger1967, Reference Riedl1971; Kazmi, Reference Kazmi1971; Nasir, Reference Nasir1989; Sadat, Reference Sadat1989). Lindelofia is a genus of about 12 species distributed in the Irano-Turanian semi-deserts and mountains, from Iran through Afghanistan, Pakistan, Russia and Middle Asia to China (Riedl, Reference Riedl and Rechinger1967; Nasir, Reference Nasir1989), and is both very geographically disjunct and ecologically distinct from Adelocaryum which, as defined here and as Paracaryopsis by Mill (Reference Mill1991), is endemic to the subtropical rain- and fog-forests of the eastern and northern shores of the Arabian Sea. The morphological differences between Adelocaryum and Lindelofia are set out in Riedl’s emended Latin diagnosis of Adelocaryum (Riedl, Reference Riedl1992: 21), which also gives its differences from Paracaryum (A.DC.) Boiss. and Cynoglossum L. The typification of Lindelofia has for a long time been a minefield into which I myself (Mill, Reference Mill1991) once stumbled and it is safest to agree (contrary to what Riedl, Reference Riedl1992 has written) with the statement in Index Nominum Genericorum (Farr et al., Reference Farr, Leussink and Stafleu1979 and http://botany.si.edu/ing/ accessed 11 August 2009) that the type remains ‘non designatus’ and to address that matter here. When first described by Lehmann (Reference Lehmann1850), Lindelofia comprised two species, L. spectabilis Lehm. and L. anchusoides (Lindl.) Lehm.; the correct name for the former species is L. longiflora (Benth.) Baill., based on Cynoglossum longiflorum Benth. which Lehmann included in his synonymy of what was in effect a new name for that species rather than a species new to science. A few later authors have indicated no definite type species for Lindelofia (e.g. Popov, Reference Popov and Shiskin1953 who seemed to argue a case for both but designated neither). Most workers have indicated that Lindelofia spectabilis, or its correct name L. longiflora, is the type of Lindelofia (e.g. Brand, Reference Brand and Engler1921; Riedl, Reference Riedl1962, Reference Riedl and Rechinger1967, Reference Riedl1992; Kazmi, Reference Kazmi1971). Others have stated, or at least suggested, that Lindelofia anchusoides is its type species (e.g. Sadat, Reference Sadat1989; Mill, Reference Mill1991). Riedl (Reference Riedl1992) considered Sadat’s choice of Lindelofia anchusoides to be a lectotypification and rejected it on the basis that he had earlier (Riedl, Reference Riedl and Rechinger1967: 137) lectotypified the genus by L. longiflora. However, in reality neither of these actions constituted lectotypifications; both were informal statements in floristic treatments. Both are pre-dated by a statement by Riedl (Reference Riedl1962: 385 – ‘Die erstgenannte Art, die wohl als Typus der Gattung zu betrachten ist’) in which he appeared to consider that Lindelofia anchusoides was the type, which would have serious consequences if it were to be regarded as a formal typification statement (which in his 1992 paper Riedl emphasised it was not intended to be). In order to fix the application of the genus name, Lindelofia spectabilis Lehm. is therefore here designated lectotype of Lindelofia Lehm.Footnote 2 Either it, or Lindelofia longiflora of which it is a synonym, have most frequently been indicated as type (see citations above). The alternative choice of Lindelofia anchusoides (Riedl, Reference Riedl1962; Sadat, Reference Sadat1989) is here formally rejected since it would upset infrageneric nomenclature within Lindelofia (the name is the type of Lindelofia sect. Brandia M.Popov) and would also cause Adelocaryum to become a synonym of Lindelofia because Brand (Reference Brand1915a, Reference Brand and Engler1921) had included L. anchusoides within Adelocaryum. Of the original five species of Adelocaryum, A. capusii also belongs in Lindelofia (Popov, Reference Popov and Shiskin1953; Riedl, Reference Riedl1992).
The remaining two species in Brand’s protologue of Adelocaryum are A. coelestinum and A. malabaricum. For these, and a third allied species previously called Paracaryum lambertianum C.B.Clarke, I proposed the new genus Paracaryopsis (H.Riedl) R.R.Mill with P. coelestina (Lindl.) R.R.Mill as its type (Mill, Reference Mill1991). I used the name Paracaryopsis because at that time I believed that the type of Lindelofia was L. anchusoides (see above), and consequently that Adelocaryum could not be used. The following year Riedl (Reference Riedl1992) correctly revived the name Adelocaryum. His generic concept of Adelocaryum was slightly different to my circumscription of Paracaryopsis; it excluded Paracaryopsis lambertiana (C.B.Clarke) R.R.Mill but, as discussed above, included Brandella erythraea as Adelocaryum erythraeum. I restrict Adelocaryum to tall mesophytic perennial herbs but unlike Riedl (Reference Riedl1992) include one (Paracaryopsis lambertiana, for which a new combination is provided in the synopsis) in which the wing-like margin is particularly strongly developed in a similar manner to Paracaryum.
Riedl (Reference Riedl1992) lectotypified Adelocaryum for the first time, choosing Cynoglossum coelestinum Lindl. [Adelocaryum coelestinum (Lindl.) Brand] as lectotype. This has the advantage of retaining Adelocaryum as the name of an accepted genus, whereas if A. anchusoides had been selected Adelocaryum would have become synonymous with Lindelofia (in part). However, Riedl’s choice has the following nomenclatural consequences:
1 Paracaryopsis (H.Riedl) R.R.Mill, based on Cynoglossum L. sect. Paracaryopsis H.Riedl in Österr. Bot. Z. 102: 393 (1962), has the same type as Adelocaryum (as lectotypified by Riedl, Reference Riedl1992) and thus becomes a synonym of the latter generic name.
2 Because Riedl (Reference Riedl1992) restricted Adelocaryum to three species (A. coelestinum, A. malabaricum and A. erythraeum), a new combination is required for Paracaryopsis lambertiana (C.B.Clarke) R.R.Mill, which I still consider to be congeneric with A. coelestinum and A. malabaricum. This combination is made later in this paper.
Two of the three already described species that I recognise in Adelocaryum, A. lambertianum and A. malabaricum, are endemic to western India around Mahabaleshwar (Maharashtra). The third, Adelocaryum coelestinum, has a wider range in India, from Gujarat to Karnataka, but seems to be commonest (or at least most frequently collected) in the same area (Mahabaleshwar) as the other two. In my earlier paper (Mill, Reference Mill1991), I also included within Paracaryopsis coelestina (with some reservations) three specimens from Oman. These obviously belonged to Adelocaryum as now circumscribed (= Paracaryopsis) and had the same general facies as A. coelestinum. Indeed, Johnston (Reference Johnston1957: 283), although he at first ‘believed an undescribed species was involved’, eventually decided that the material was ‘indistinguishable’ from Indian material of the species he called Cynoglossum coelestinum. My earlier examination of the Arabian material revealed that it differed from the Indian material of Adelocaryum coelestinum in having primarily white (not blue) corollas and nutlets with a glabrous, not glochidiate, disc, but at the time I also considered the material to be, probably, conspecific with A. coelestinum.
I have since had the opportunity of examining more material of the Arabian plant, including colour slides taken in the field by Tony Miller (Royal Botanic Garden Edinburgh). Although the Oman plant is certainly a close relative of Adelocaryum coelestinum, the differences between it and A. coelestinum that were mentioned in my earlier paper appear to be constant, with the exception of the flower colour as noted below under A. coelestinum. Additional points of difference not noticed in the course of the earlier study are that the wing of the nutlets of the Oman taxon is always very shallowly crenulate on the inner (incurved) margin, with slender glochids having non-expanded bases, whereas in all the Indian material of Adelocaryum coelestinum the nutlet margin is conspicuously dentate with the teeth broadened at the base and only gradually passing into a glochidiate tip. Further minor differences are that the calyces of the Oman plant are more sparsely hairy and its flowers are slightly larger in diameter. As a result of this re-evaluation of the taxonomic status of the Oman Adelocaryum, which is geographically disjunct from the range of A. coelestinum and consequently reproductively isolated from the latter, the material from Oman is described in this paper as the new species A. nebulicola R.R.Mill.
The genus Adelocaryum accordingly now comprises four species, and a revised key is given below. Full descriptions of Adelocaryum coelestinum, A. lambertianum and A. malabaricum were given by Mill (Reference Mill1991, under Paracaryopsis). Because the description of Adelocaryum coelestinum given in my earlier paper (Mill, Reference Mill1991 as Paracaryopsis coelestina) was a composite based on Arabian specimens here assigned to A. nebulicola as well as true Indian A. coelestinum, an amended account of A. coelestinum is also provided below.
Synopsis of the Genus Adelocaryum
Adelocaryum Brand, Repert. Spec. Nov. Regni Veg. 13: 547 (1915). – Cynoglossum L. subgen. Eleutherostylum (Brand) H.Riedl sect. Paracaryopsis H.Riedl, Österr. Bot. Z. 102: 393 (1962). – Paracaryopsis (H.Riedl) R.R.Mill, Edinburgh J. Bot. 48: 56 (1991), nom. superfl. – Lectotype (Riedl, Reference Riedl1992: 21): Adelocaryum coelestinum (Lindl.) Brand based on Cynoglossum coelestinum Lindl., Edwards’ Bot. Reg. 25: t. 36 (1839).
Emended description: as that given for Paracaryopsis (H.Riedl) R.R.Mill by Mill (Reference Mill1991: 56). Differs from Brandella R.R.Mill by tall mesophytic perennial habit; stamens inserted ± at same level as throat scales (not slightly below them); nutlets with few scattered short glochids or none at all on wing-like margin, and few or none on disc (in Brandella with disc densely covered with slender glochids and numerous long glochids at the edge of the wing-like margin).
Revised key to the species of Adelocaryum
1a. Nutlets 10–14 × 8–10.5 mm, smooth, the margin ± flat; radical leaves with cuneate base 4. A. lambertianum
1b. Nutlets 4.5–6.5 × 3–5.5 mm, glochidiate at least on the erect, ± incurving margin; radical leaves truncate to ± deeply cordate at base 2
2a. Flowers 8–11 mm long, uniformly dark blue; anthers with apices just exserted from corolla tube 3. A. malabaricum
2b. Flowers 4–5 mm long, pale blue or white with darker centre; anthers included in corolla tube 3
3a. Disc of nutlet with scattered glochids; incurved margin of nutlet rather coarsely dentate, the teeth triangular, c.0.5 × 0.5 mm excluding their glochids 1. A. coelestinum
3b. Disc of nutlet completely glabrous; incurved margin of nutlet shallowly crenulate or undulate, the undulations 0.1–0.3 × 0.5–0.7 mm excluding their glochids 2. A. nebulicola
1. Adelocaryum coelestinum (Lindl.) Brand, Repert. Spec. Nov. Regni Veg. 13: 549 (1915). – Cynoglossum coelestinum Lindl., Edwards’ Bot. Reg. 25: t. 36 (1839). – Paracaryum coelestinum (Lindl.) Benth. & Hook.f., Gen. Pl. 2: 850 (1876) in adnot.; C.B.Clarke in Hook.f., Fl. Brit. India 4: 160 (1883). – Paracaryopsis coelestina (Lindl.) R.R.Mill, Edinburgh J. Bot. 48: 57 (1991) [q.v. for additional synonymy, citation of Indian specimens and discussion (excluding its first paragraph)]. – Described from cultivated material raised in England by the Horticultural Society from seeds presented by John Nimmo Esq. of Bombay and which flowered for the first time in August 1838. No type indicated.
Perennial herb (possibly biennial, as described by Riedl, Reference Riedl1992). Stems (30–)90–150 cm, stout, erect, reddish, lower part appressed-pubescent when young, later glabrate, upper part sparsely retrorse adpressed-pubescent. Radical leaves with long petiole (to 15 cm or longer); lamina broadly ovate, to at least 13–15 × 10.5–15 cm, acute at apex, base distinctly cordate; total number of principal veins (including midvein) 17–23, veins arcuately ascending, their hairs retrorse, dense with some hairs deflexed-patent; upper surface dark green with rather sparse ± antrorse setules arising from calcifying tuberculate bases, lower surface paler, setulose only on veins, the hairs dense and mostly retrorse. Cauline leaves ovate, 4–10.5 × 2–4 cm (possibly larger), lower ones very shortly petiolate, upper ones sessile with cuneate base. Inflorescence a panicle of numerous furcate or simple terminal and subterminal scorpioid cymes, the primary dichotomies occasionally simple but usually with at least secondary and frequently tertiary bifurcations; primary branches of inflorescence 12–20 cm in fruit including cyme; cymes of primary branches up to 30-flowered, secondary and tertiary branches shorter and fewer-flowered. Pedicels 0.5–2 mm in flower, up to c.4 mm in fruit but often hidden by the group of nutlets which thus appears subsessile, somewhat curved downwards in fruit. Flowers all ebracteate. Calyx lobed almost to base; lobes ovate or ovate-elliptic, 2–3 × 1–1.5 mm in flower, 4–5 × 1.8–2.5 mm in fruit, densely hairy at base and with a thin line of hairs along midvein and around margin, otherwise glabrous. Corolla apparently normally pale blue (white with blue centre according to the original description and as photographed at one locality; see note), shortly campanulate, 4–5 mm long, at least 6–8 mm in diameter (according to Riedl, Reference Riedl1992 up to 10–12 mm in diameter), the corolla lobes c.3 mm, the midvein normally not noticeably darker than the rest of the lobe (but contrasting at one locality; see note). Faucal scales semilunar, c.0.5–0.6 × 0.9–1 mm, emarginate. Gynobase in fruit 4–5 mm, narrowly pyramidal, the very short style comprising the uppermost c.0.5 mm and hardly projecting beyond the nutlets; stigma capitate. Nutlets 4 (all maturing), light ivory-grey when mature, triangular-napiform (the three distinct edges almost equilateral), 4–6 × 3.5–4.5 mm; wing c.2 mm high, its inner margin incurving slightly over the disc and coarsely dentate with c.5–6 glochidiate teeth on each of the three edges, the teeth 0.6–1 mm, narrowed from a slightly expanded base into a glochidiate tip; disc with scattered, short, erect glochids mainly in its central part, the midline ± faintly raised; ventral surface glochidiate all over except on the attachment scar.
Distribution and ecology. India (Gujarat, Maharashtra, Karnataka). Wet rainforest on the Western Ghats, up to at least 1370 m. Flowering from August to April. A calcicolous species that occurs in two distinct associations, the Arthraxoetum inermi Bharucha, dominated by Arthraxon inermis Hook.f. and Begonia concanensis A.DC., and the Arthraxoetum purandharensis Bharucha, dominated by Arthraxon purandharensis Bharucha and a species of Nepeta (Bharucha & Satyanarayan, Reference Bharucha and Satyanarayan1954 – these authors mistakenly believed that Paracaryum coelestinum was a different species from Adelocaryum coelestinum whereas they are homotypic synonyms).
Photographs of Adelocaryum coelestinum taken on the Kas Plateau (Maharashtra) by Dinesh Valke show a plant having corollas whose lobes are white distally but are marked with a star-like pattern of blue midvein lines and blue shading in the centre, very similar to the pattern found in A. nebulicola described below. Herbarium material seen all has sky blue corollas with no contrast either between the proximal and distal parts of the lobe or between the midvein and the rest of the lobe. Until I saw these photographs in 2009, I had consequently believed that the corollas of Adelocaryum coelestinum were always wholly sky blue while those of A. nebulicola were white with a blue centre, providing a floral distinction between the two species. That would appear to be not always the case. The protologue of Cynoglossum coelestinum indeed illustrates a plant with bicoloured flowers and calls it the ‘Blue and White Hounds-tongue’, stating in the Latin description ‘Corolla cyanea, margine alba’ (Lindley, Reference Lindley1839). More studies are therefore needed concerning flower colour variation in this species, to understand both whether or not there are colour changes that take place on drying (herbarium material of Adelocaryum nebulicola retains its fresh colour) or with age, and whether there is any geographical variation in flower colour.
Adelocaryo coelestino (Lindl.) Brand habitu et forma foliorum valde similis, a quo nuculae disco glaberrimo (haud glochidiato) et ala levissime tantum crenulata vel undulata (haud argutiuscule dentata) constanter differt, ut videtur venis foliorum radicalium paucioribus (lateralibus in toto 9–15 non 18–22) pilis antrorsis non retrorsis tectis etiam recedit. – Type: Sultanate of Oman, Dhofar, Jebel Qara, Salalah to Ashanhaib road, c.10 km S of Ashanhaib, 800 m, 6 ix 1989, A.G. Miller & J.A. Nyberg M. 9131 (holo E; iso K, KTUH n.v., ON n.v.).
Adelocaryum nebulicola R.R.Mill. A, upper part of flowering stem with inflorescences; B, single cyme with lowest flower open; C, flower viewed from above; D, anthers: left, adaxial view showing filament attachment; right, abaxial view; E, corolla opened out; F, throat scales and stamens. Drawn by Işik Güner.
Adelocaryum nebulicola. A, basal leaf and petiole; B, detail of basal leaf venation and indumentum; C, upper cauline leaf, abaxial surface: detail of venation and indumentum; D, upper cauline leaf, adaxial surface: detail of venation and indumentum; E, fruiting calyx, group of four nutlets and style; F, single nutlet, abaxial view; G, detail of marginal glochids of nutlet. Drawn by Işik Güner.
Stout-stemmed, erect perennial or biennial herb 1–1.5 m tall. Stem 5–10 mm thick or more, hollow, glabrous in lower part, appressed crispate-pubescent above and more densely so in inflorescence. Radical leaves very long-petiolate; petiole to at least 30 cm, channelled, sparsely pubescent; lamina broadly ovate-cordate, 11–25 × 11–30 cm (the larger dimensions when in fruit), the apex acute or shortly acuminate, the base shallowly cordate or cordate-truncate, entire or the margin occasionally appearing 3-lobed (e.g. Collenette 8387), the upper surface deep green with scattered short ± antrorse setules arising from large many-celled calcifying bases, the lower surface much paler green and totally glabrous except for rather dense antrorsely appressed non-setiform hairs on veins and margin; total number of principal veins (including midvein) 9–15, veins arcuately ascending, their hairs antrorse, short, fairly dense but less so than Adelocaryum coelestinum and all ± appressed. Lower cauline leaves similar to radical but more shortly petiolate, middle and upper ones elliptic, 3–22 × 0.8–12 cm, the apex acute, the base cuneate, subsessile or sessile; indumentum as radical; veins c.6 each side of midrib. Inflorescence a panicle of furcate or simple terminal scorpioid cymes, the primary dichotomies often then simple but sometimes with secondary and occasionally tertiary bifurcations; primary branches of inflorescence 9–23 cm in fruit including cyme; cymes of primary branches 12–21-flowered, 7–17 cm in fruit, those of secondary and tertiary branches shorter and fewer-flowered. Pedicels not more than 1.5–2 mm in flower, 2–4.5 mm in fruit, erecto-patent but finally becoming hamately curved downwards in fruit. Flowers all ebracteate. Calyx lobed almost to base; lobes rather broadly elliptic or elliptic-obovate, 2.5–4 × 1.5–2.5 mm in flower, 4.5–6 × 2.5–4 mm in fruit, with a thin line of short appressed hairs along the midvein and around the margin, the rest of the calyx lobe glabrous or with very few scattered hairs. Corolla white with blue or pinkish-blue centre, subrotate or shortly campanulate, 7–9.5 mm in diameter, the corolla lobes suborbicular-ovate, 3.5–4 × 3.5–4 mm, each (when dry) with a narrow blue midvein. Faucal scales transversely subquadrate, c.0.5 × 1 mm, with emarginate apex, densely papillate. Anthers c.0.8 mm, ellipsoid. Gynobase in fruit c.2.5 mm, narrowly pyramidal, the short style comprising the uppermost 0.7–1 mm; stigma capitate. Nutlets 4 (all maturing), greenish becoming olive-brown, ovate-napiform, 4.5–6 × 3–4 mm; wing erect, 2–2.5 mm high, its inner margin incurving over the disc and shallowly undulate or crenulate with slender marginal glochids 0.3–0.4 mm not broadened at their base, and with similar slightly longer (0.3–0.5 mm) glochids around the upper half of the outer surface; disc completely glabrous, with or without a narrow, slightly raised midline; ventral surface with slender glochids scattered over whole surface except for attachment scar.
Distribution and ecology. Sultanate of Oman; endemic to Jebel Qara, Dhofar. Escarpment woodland dominated by Anogeissus dhofarica A.J.Scott and in thickets and scrub, typically associated with Ficus vasta Forssk.; sometimes in the contact zone between this vegetation type and adjacent drier Euphorbia balsamifera Aiton scrub but never in the scrub proper; 360–850 m. Collenette’s collection is labelled as being collected in ‘open grassland’. This is apparently atypical (although Vesey-FitzGerald’s specimen, the earliest known, was also collected in ‘upland grassland’ the plant was apparently shaded by a fig tree). Her gathering was also made at the lowest altitude; other collections with altitude indicated have all been made above 600 m. Flowering commences at the beginning of August, for c.5 weeks until mid-September; fruits ripe September–mid-October.
Etymology. The specific epithet nebulicola (fog- or mist-inhabiting, from Latin nebula, fog or mist, and indeclinable suffix -cola, dweller) alludes to the plant’s ecological requirement, it being confined to escarpment woodlands which, during its growing season (see above), are subjected to constant mist drip.
Additional specimens examined. Sultanate of Oman. Dhofar, Jebel Qara, under shade of Ficus only, 4 x 1943, D. Vesey-FitzGerald 12448/4 (BM, 2 sheets); Dhofar, Jebel Qara, Salalah to Thumrait road, 5 km S of Ayun turnoff, 600 m, 13 x 1979, A.G. Miller 2709 (E); Jebel Qara, nr Kaftawt, 790 m, 1 viii 1985, A.G. Miller 7202 (E); Thumrait to Salalah road, by Hagaif turnoff, 850 m, 7 ix 1985, A.G. Miller 7509 (E, K); Nashib road, east of Salalah, open grassland, 1200 ft, 3 x 1992, I.S. Collenette 8387 (E).
Adelocaryum nebulicola has so far only been collected from Jebel Qara, the central of the three ranges of fog-affected escarpment mountains that make up the so-called ‘Dhofar fog oasis’ (Miller, Reference Miller, Davis, Heywood and Hamilton1994). Here, dense fogs caused by upwelling of cold water off the coast build up on the escarpments during the southwest monsoon (mid-June to mid-September). (In Mill, Reference Mill1991, the collection date of Miller 7509 was incorrectly given as ‘7 xi 1985’, an error which led to the flowering time of Arabian ‘A. coelestinum’ being wrongly given in that paper as ‘August–November’.) The life cycle of Adelocaryum nebulicola would appear to coincide almost exactly with this monsoon; germination, or regrowth of the underground parts, presumably commences with the breaking of the monsoon, with flowering beginning about a month later in late July/early August and being completed about 5 weeks afterwards in mid-September, and the fruits being fully ripe from about the end of the monsoon (mid-September) onwards. In its habit and life cycle Adelocaryum nebulicola thus resembles the other species of Adelocaryum as here circumscribed, all of which are tall mesophytic herbs of the rainforests and forest margins on the western seaboard of India. Dhofar is the only part of Arabia with a comparable climate to this area of India. The occurrence of the genus Adelocaryum in Dhofar is thus of particular interest; the apparent disjunction between Dhofar and western India is a natural and not unexpected one despite the considerable distance. The genus is very unlikely to occur anywhere else in Arabia although it should be looked out for on the other ranges of the Dhofar fog oasis. These, especially Jebel Qamar to the west, have similar Anogeissus dhofarica escarpment woodlands to those on Jebel Qara in which Adelocaryum nebulicola has been collected and thus the occurrence of Adelocaryum in similar habitats elsewhere in Dhofar is not unlikely. Jebel Samhan, to the east of Jebel Qara, is somewhat drier and conditions may not be ideal for Adelocaryum nebulicola to thrive. The Dhofar mountain ranges are a local centre of endemism: of c.900 species, c.60 are endemic, including two genera (Cibirhiza P.Bruyns and Dhofaria A.G.Mill.: Miller, Reference Miller, Davis, Heywood and Hamilton1994). To this total must now be added this new species of Adelocaryum. Its existence on Jebel Qara was first noted by I. M. Johnston (Johnston, Reference Johnston1957: 283), under the name Cynoglossum coelestinum. Johnston (annotated drawing in sched.) also examined the pollen of Vesey-FitzGerald 12448/4 and found it to be 9–12 × 6–8 μm with a very shallow equatorial constriction.
Adelocaryum nebulicola is closest in its morphology to A. coelestinum, which has the widest range of habitats and geographical distribution of the three Indian species. It is more tolerant of drier habitats than Adelocaryum lambertianum and A. malabaricum, both of which seem to be restricted to rainforest around Mahabaleshwar in Maharashtra. Adelocaryum coelestinum extends as far north as Gujarat (Mill, Reference Mill1991). Adelocaryum nebulicola has presumably evolved from an A. coelestinum-like ancestor and has diverged sufficiently due to its geographical isolation for its differential characters to have become constant. Specific rank therefore seems fully justified, although more study on flower colour variation in Indian Adelocaryum coelestinum is clearly needed, given the existence of plants with A. nebulicola-like corolla colouring on the Kas Plateau in Maharashtra (see note under A. coelestinum). Interestingly, the Kas Plateau is on almost exactly the same latitude as the locality of Adelocaryum nebulicola, but 2000 km to the east on the opposite side of the Arabian Sea and on the inland (leeward) side of the Western Ghats.
3. Adelocaryum malabaricum (C.B.Clarke) Brand, Repert. Spec. Nov. Regni Veg. 13: 549 (1915). – Paracaryum malabaricum C.B.Clarke in Hook.f., Fl. Brit. India 4: 160 (1883). – Cynoglossum malabaricum (C.B.Clarke) H.Riedl, Österr. Bot. Z. 119: 71 (1971). – Paracaryopsis malabarica (C.B.Clarke) R.R.Mill, Edinburgh J. Bot. 48: 59 (1991). – Type: [India] Western Deccan Peninsula, Canala and Mysore, Law (holo K, as ‘Cynoglossum no. 13’; iso W n.v.).
For description and other information see Mill (Reference Mill1991: 59, under Paracaryopsis) and Riedl (Reference Riedl1992: 24–25).
4. Adelocaryum lambertianum (C.B.Clarke) R.R.Mill, comb. nov.
Paracaryum lambertianum C.B.Clarke in Hook.f., Fl. Brit. India 4: 161 (1883). – Mattiastrum lambertianum (C.B.Clarke) Brand in Engler, Pflanzenr. IV, 252 (Heft 78): 61 (1921). – Paracaryopsis lambertiana (C.B.Clarke) R.R.Mill, Edinburgh J. Bot. 48: 59 (1991). – Type: [India] Bombay, Lambert s.n. as ‘Omphalodes’ (holo K; photo E).
For description and full taxonomic account see Mill (Reference Mill1991: 59–61, under Paracaryopsis).
Acknowledgements
I thank A. G. Miller (Royal Botanic Garden Edinburgh) for making available the colour slides of Adelocaryum nebulicola which he took during his expedition to Oman. Işik Güner drew the illustrations. John McNeill and an anonymous referee are thanked for their comments that have helped to improve this paper. The Royal Botanic Garden Edinburgh is supported by The Scottish Government Rural and Environment Research and Analysis Directorate (RERAD). I also thank the Centre of Excellence in Biodiversity Research, King Saud University, Riyadh, Saudi Arabia for funding the illustrations.
