Name changes in the Old World Rhus and recognition of Searsia ( Anacardiaceae )

The background to and status o f the genus Searsia F.A.Barkley (Anacardiaceae) is discussed and reasons given as to why it is the correct name for those Old World species in the Rhus complex fonnely regarded as subgenus Thezera (DC.) K.Koch (section Gerontogeae Engl.). An annotated list o f all the accepted 111 species and 28 fiirther infraspecific taxa in Searsia is presented, and where necessary, new combinations are made and types are designated.


INTRODUCTION
When the author revised the southern African spe cies of Rhus L., Anacardiaceae, for the Flora o f southern Africa (Moffett 1993), he stated that he was retaining the species in Rhus pending the results of ongoing research on the generic status of this heterogeneous genus.Although he was aware that Barkley had previously pub lished the name Searsia for the Old World Rhus species, he preferred to follow Brizicky (1963) in treating this group as subgenus Thezera (DC.) K. Koch.Recent phylogenetic analyses using DNA and gene spacers of the Rhus complex have shown that Searsia is clearly monophyletic and widely separated from Rhus s. str (Miller et al. 2001;Yi et al. 2004) and there is there fore no further reason not to uphold Searsia F.A.Barkley as the correct name for the 'Old World' Rhus species.

BRIEF HISTORY OF THE NOMENCLATURE OF THE 'OLD W ORLD' RHUS SPECIES
The Old World Rhus species are found in the Mediter ranean, Africa and Asia.Linnaeus included three of these from the Cape in his list of 12 species in Species plantarum (Linnaeus 1753).O f the other nine, four have subse quently been retained as Rhus L., three as Toxicodendron Mill., one as Cotinus Mill, and one as AUophylus L., (Sapindaceae) [Index Nominum Genericorum (plantarum) (ING) 2006], epitomizing the heterogeneous nature of Rhus sensu lato.The type species of Rhus and there fore that which governs the application of Rhus sensu stricto is Rhus coriaria L., the sumach of southern Europe, Mediterranean region and the near East.With its pinnately compound leaves, drupes with prominent red glandular hairs and resinous mesocarp, it is outwardly markedly different to the species of subgenus Thezera, which are characterized by temate, rarely simple or pal mate leaves and drupes pale, glabrous or tomentose with a resinous mesocarp adhering to the bony endocarp.
The heterogeneity of Rhus was recognized early, and Bemhardi (1838) remarked at the end of a paper on Laurophyllus Thunb., that the temate species of Rhus from the Cape were different to the true Rhus species and seemed to form a distinct genus which one could name Terminthia Bemh.Despite this name being used by Wu & Ming (1979) in Flora Yunnan, it is illegitimate, as accord ing to Art. 34.1 (b) of the St Louis Code (Greuter et al. 2000), it is a provisional name.McNeill & Greuter (pers. comm. October 2006) are of the opinion that Bemhardi's use of the words 'welche man Terminthia nennen konnte' is a clear indication of provisional status and nonaccep tance o f the name.
The first person to lump all the Old World Rhus spe cies together was Engler (1881), who placed them in his new section Gerontogeae (the old rhusses), and who soon after provided the first detailed account of the genus in A. «fe C. De Candolle's Monographiae phanerogamarum (Engler 1883).His encyclopaedic treatment of Rhus with its four sections, Ttichocarpae, Venenatae, Gerontogeae and Melanocarpae, formed the foundation by which Rhus was measured for the next 60 years until Barkley (1942Barkley ( , 1943Barkley ( , 1950Barkley ( , 1965) ) criticized the traditional concept of the genus.Engler's use of section Gerontogeae was actually illegitimate as it was predated by section Thezera of De Candolle (1825).
The name Searsia first appeared in a footnote to a key to the genera of the Anacardiaceae where Barkley (1942) stated 'Searsia n.gen.= (Rhus) Section Gerontogeae.Named after Paul B. Sears '. Sears (1891'. Sears ( -1990)), who was to become a renowned stratigraphic palaeontologist, eco logist and head of the Yale School of Botany, was one of a number of botanists thanked by Barkley for encour agement and assistance during the early part o f his Ph.D. studies on American Rhus species (Barkley 1937).
The above publication, however, was illegitimate as there was no description of the genus and this was cor rected the following year when Searsia appeared as genus six in Flora o f Texas, with a fiill description and includ ing the cultivated Searsia lancea (L.f.) F.A.Barkley, a South African species (Barkley 1943).In that publication, Barkley also made the new combination of S. tomentosa (L.) F.A.Barkley and designated this species as the type species.
Eight further combinations in Searsia were made by Barkley in 1950, and eleven more in a 1965 Iraq pub lication.In this latter paper, titled 'A criticism o f the traditional concept of the genus Rhus', Barkley rec ognized ten genera in the Rhus complex and in place of Searsia tomentosa, designated Searsia pentaphylla (Jacq.)F.A.Barkley as the type of the genus.In separa ting the genera o f the Rhus complex, he chose not to fol low Brizicky, who two years previously, in a paper on the generic limits of Rhus, preferred to retain the name Rhus at genus level and recognized six subgenera, one of which was subgen.Thezera (DC.) K.Koch and which included Engler's Gerontogeae (Brizicky 1963).Despite Barkley's criticism, the name Rhus has been maintained in all the various Afiican regional floras pub lished subsequently, as well as in the few new species, the latest being Rhus pygmaea Moffett (Moffett 1999) and Rhus gallagheri Ghaz., (Ghazanfur 2002).

CONFIRMATION OF THE STATUS OF SEARSIA
Recent research in the USA on the generic status of Rhus has provided conclusive evidence that the 'Old World' species of Rhus are sufficiently different to warrant generic status of their own, thus vindicating Barkley. Miller et al. (2001) examined the sequences of the internal transcribed spacer region (ITS) of the nuclear ribosomal DNA of six genera in the Rhus sensu lato complex in order to determine the monophyly of Rhus sensu stricto and to provide insight into the phy logenetic and biogeographical history of the genus.Two species from the Old World, viz.Searsia ciliata and S. quartiniana, both ex hort., were included.As outgroups in their analysis, Pistacia vera and Schinus molle were used.The results showed that Rhus sensu stricto is monophyletic and that the other genera, Actinocheita F.A. Barkley, Cotinus, Malosma (Nutt.)Raff., Searsia and Toxicodendron were distinct from Rhus sensu stricto, but the relationships between these other genera were not well resolved.Interestingly Schinus L., Searsia and Toxicodendron formed a clade (bootstrap value of 91 %).The authors suggested that the use of additional charac ters such as chloroplast genes, should help to resolve the intergeneric relationships.This suggestion was followed by Yi et al. (2004), who carried out a phylogenetic analysis of the Rhus complex using ITS of nuclear ribosomal DNA and chloroplast {ndhY and trnL-Y).Among the species included, but not used in the Miller et al. (2001) study, were the southern African Searsia lancea, S. leptodictya, S. pyroides and S. undulata, all ex hort.The phylogenetic analysis of the ITS, the chloroplast and the combined ITS and chloro plast datasets confirmed the monophyly of Rhus sensu Barkley (1937) and that Searsia from Africa was mono phyletic and distinct from the Rhus clade.The ITS data also showed that Searsia lancea, S. leptodicty'a and S. undulata formed a clade, which was sister to the clade composed of S. ciliata, S. pyroides and S. quartiniana.A chronogram indicating diversion times compiled by these authors based on the maximum likelihood tree of the combined ITS and cpDNA data showed that Searsia undulata diverged 55 Ma, Rhus 49 Ma and R. coriaria 24 Ma.SEARSIA F.A.BARKLEY: SPECIES AND TYPES Barkley's original designation of Searsia tomentosa as the type species of the genus (Barkley 1943), and sub sequent replacement by S. pentaphylla (Barkley 1965) requires some explanation, and I am indebted to Prof. John McNeill of Ontario and Edinburgh for clarifying the issue.
Article 7.4 of the Code states 'A new name formed from a previously published legitimate name (stat. nov., comb, nov.) is, in all circumstances, typified by the type o f the basionym, even though it may have been applied erroneously to a taxon now considered not to include the type'.As Barkley, 1943, cited Rhus sect.Gerontogeae Engl.( 1881) as a synonym of his new genus, his des ignation of 5".tomentosa as the type was superfluous, as the type of Searsia is the type of sect.Gerontogeae.Section Gerontogeae is, however, illegitimate as Engler acknowledged in his publication that it was based on the earlier sect.Thezera DC. ( 1825).As Engler did not indicate a type for sect.Gerontogeae, according to Art. 7.5, Searsia is automatically typified by the type of sect.Thezera.Neither De Candolle nor Koch (1853), who treated Thezera as a subgenus, designated types and it was lef^ to Brizicky (1963) to choose a lectotype, viz.Rhus pentaphylla (Jacq.)Desf., which explains why Barkley accepted this in 1965. Despite this, ING (2006), however, still cites R. tomentosa L. as the type o f Searsia.
The following list is based on research undertaken on the taxonomy of all the Old World Rhus species when revising Rhus in southern Africa between 1976 and 1992 (Moffett 1993), and on the relevant literature since then.The few differences between this list and that of the accepted names for sub-Saharan Afnca (Klopper et al. 2006) are intentional.Based on the notes made when the author visited many European herbaria in 1982 and cor respondence with respected taxonomists, types have been designated for those taxa as yet untypified.

*
No specimen annotated by G.Don could be found, thus requiring the designation o f a neotype.Both a and b on the above sheet are almost identical and any one could have been chosen.** Bemhardi is sometimes cited as the author o f the new names based on Krauss's specimens.In this particular case, however, Krauss him self should be recognized as he used the words 'in specimen m eo' when citing Cissus natalensis Bemh.olim in sched.