Caroxylon today

Mucina, Ladislav, 2017, Caroxylon (Chenopodiaceae s. str.) in continental southern Africa and Madagascar: a preliminary nomenclatural synopsis and biogeographical considerations, Phytotaxa 312 (2), pp. 151-178 : 152-154

Caroxylon today

Caroxylon (tentatively incl. Nitrosalsola ) appears to have two centres of biodiversity. The northern centre includes the arid and semiarid regions of the Mediterranean, North Africa and Asia Minor, the Arabian Peninsula and the neighbouring Horn of Africa region, Iraq, Iran, Afghanistan, Pakistan and the neighbouring Central Asian countries of the latter three, and China (especially the eastern desert regions). The southern centre that covers the arid regions of Angola (probably only one species), Namibia , South Africa and Botswana (seven species according to Setshogo 2005), spanning the climatically harsh hyper-arid Moçâmedes, Namib and Gariep Deserts, semi-deserts of the Karoo and strongly (austral) winter-arid savannas of the Kalahari. In regions such as the western Free State, Caroxylon is found to prefer edges of saltpans. One taxon of Caroxylon occurs in Madagascar.

Several Caroxylon taxa from southern Africa have already been included in molecular phylogenetic studies (as Salsola spp. ). Akhani et al. (2007, but see also Wen et al. 2010) found that in the combined maximum likelihood (ML) analysis of ITS and psbB-psbH markers, southern African S. araneosa Botsch. (1973: 818) , S. zeyheri ( Moquin-Tandon 1849: 176) Bunge (1880: 13) and S. glabrescens Burtt Davy (1926: 177) formed a well-supported clade with Iranian S. abarguensis Assadi (1984: 136) . This clade was found to be sister to almost all analysed Northern Hemisphere Caroxylon accessions.

Caroxylon appears to be one of the flagship genera pointing to the existence of a migrational link or pathway between (semi-)deserts of the Arabian Peninsula, Horn of Africa and southern Africa. This link has been termed the ‘Arid Corridor’ ( Balinsky 1962, Verdcourt 1969, Thulin 1994, Caujapé-Castells et al. 2001, Bellstedt et al. 2008 etc.). It is also presumed to have facilitated the exchange of floras through repeated openings since the Miocene (Caujapé-Castells et al. 2001) and especially during the recurring dry periods of the Pleistocene glacial/inter-glacial cycles. Slightly outlandish early theories were invoking a larger extent of Africa than known today, including putative desert areas in the western Indian Ocean or suggested existence of a hypothetical continent Lemuria ( Suess 1912) in basically the same region. The Soviet botanist Popov (1927, 1963; reprinted in 1983a, 1983b) even suggested that accepting the existence of Lemuria would explain the shared similarities betwen the desert floras of Asia, North Africa, Socotra with those of southern Africa. Popov (1963; 1983b) also suggested that the main elements of almost all desert floras of the world originated from a hypothetical ancestral Gondwanan flora, which he called the “ Welwitschia flora”, as opposed to his equally hypothetical ancestral temperate “ Ginkgo flora”.

Botschantzev (1969b: 992–993), leaning on the work of Soviet paleogeographers Leonov (1956) and Sinitsyn (1967), presented a model attempting to explain the geographic disjunction in Salsola (hence in Caroxylon , the dominating salsoloids in southern Africa). He suggested that representatives of section Caroxylon (within the genus Salsola ) originated in southern Africa as early as the Oligocene-Miocene and then migrated during the Pliocene to the Red Sea region where these representatives supposedly initiated speciation events resulting in the emergence of other sections of Salsola . The disjunction itself (according to Botschantzev 1969b) was a result of the separation of the southern African region from the Northern Hemisphere range of Salsola by presumed formation of a ‘tropical belt’ during the Pliocene. This hypothesis invokes the origin of Caroxylon in southern Africa and therefore I prefer to call it the ‘Ex Africa Australis Hypothesis’. This hypothesis has not been corroborated by the modern palaeoclimatic data that aided the dating of aridification events on our planet (e.g. Zachos et al. 2001). Botschantzev’s (1969b) assertion that his scenario corroborated Popov’s (1927, 1963) hypothesis regarding the Gondwanan origin of desert floras also appears unwarranted.

According to an alternative scenario suggested by Il’in (1937, 1946, 1958;), the halophytic and desert floras have common origins and the littoral habitats of the coasts of the Tethys (or, better to say, the Tethyan region) served most probably as the cradle of many groups of chenopods (incl. Salsola s.l.), wherefrom these ancestors then migrated to southern Africa. This scenario, which I call the ‘Ex-Tethys Hypothesis’, has thus far received greater support ( Kühn 1993, Zhu 1996: Fig. 10, Akhani et al. 1997, Pyankov et al. 2002), than the one suggested by Botschantzev (1969b). Mosyakin (2002) and Feodorova (2009) reviewed these ideas in historical and biogeographical contexts and also came to the conclusions supporting the ‘Ex-Tethys’ concept; however, Mosyakin (l.c.) suggested at least two (or more) migration events, while Feodorova (l.c.) postulated just one. The ‘Ex-Tethys’ scenario is also emerging as a core element of the leading paradigm on the global dispersal of halophytic ( Kadereit et al. 2005, 2006, 2010, Kadereit & Freitag 2011) and other, arid-adapted non-halophytic clades (e.g. Schrire et al. 2005). Molecular phylogenetic analyses (cited above) also clearly suggest that early-branching taxa in Caroxylon (and other genera of the subfamily Salsoloideae ) occur in the Northern Hemisphere.

In light of emerging knowledge concerning the nature of links between arid regions of southern Africa (see above), scanty molecular phylogenetic data, mounting evidence on the nature of radiations in insular habitats (including subcontinental scale, yet isolated regions; Linder 2003) and often-encountered discrepancies between morphological and molecular data in species-rich clades, I predict that:

1) Traditionally used morphological characters will poorly reflect molecular phylogenetic patterns in Caroxylon ( Aizoaceae model; Klak et al. 2004, 2007);

2) Molecular phylogenetic studies will encounter problems with resolution probably due to poor power of conventional markers and/or the possible effect of ancient hybridisation in Caroxylon ;

3) Molecular dating will probably reveal that Caroxylon is a Pliocene or Late Miocene taxon, represented in southern Africa by only a few evolutionarily old (relict) species while most southern African species of Caroxylon will result from recent rapid radiation(s) ( Linder 2008);

4) Tree dating will further reveal that both the species richness and genetic diversity resulted from several recent rapid radiations that followed repeated migrations through the Arid Corridor; and

5) Migrations through the Arid Corridor could have been happening in both directions (Zygophyllym/Roepera model: Bellstedt et al. 2008).

These predictions are testable and make further enquiry into the evolution and current taxonomic structure of the genus Caroxylon an exciting scientific journey.