AFLP Phylogeny of 36 Erythroxylum Species

Abstract

Four taxa of the plant genus Erythroxylum; Erythroxylum coca var. coca (Ecc), Erythroxylum coca var. ipadu (Eci), Erythroxylum novogranatense var. novogranatense (Enn) and Erythroxylum novogranatense var. truxillense (Ent) are cultivated primarily for the illicit extraction and processing of cocaine. Despite their economic and medical importance, the evolutionary history of these species remains unknown in a modern phylogenetic framework. The aims of this study were to: (a) investigate the relationship among the cultivated and a select number of non-cultivated taxa, and (b) test Plowman's (Journal of Psychodelic Drugs 11:103-117, 1979b) linear progression hypothesis of the cultivated Erythroxylum taxa versus Johnson's et al. (Annals of Botany 95:601-608, 2005) hypothesis that Ec and En are sister species. AFLP phylogeny was used to compare the relationships among 36 Erythroxylum species (133 accessions) spanning the geographic distribution of the genus. A Maximum Parsimony tree revealed both geographic and taxonomic partitioning into clades representing species from Africa, Asia-Pacific and the New World (Tropical Americas). Ec and En formed distinct clades, indicating they are sister species and a cluster of non-cultivated species were the most closely related to the cultivated species. Multivariate ordination analysis was used to evaluate the relationship between cultivated and non-cultivated Erythroxylum taxa from the Tropical Americas. Our results support the hypothesis that the cultivated species are more closely related to each other than to any other species of Erythroxylum, but refute the hypothesis that Ent (and Enn) descended from Ecc. Instead our data suggest an independent, non-linear evolutionary relationship between Ec and En. Finally, the AFLP analyses identified significantly different genetic groups within Erythroxylum suggesting that the current intrageneric classification of this genus be revised.

Phylogenetic inference in section Archerythroxylum informs taxonomy, biogeography, and the domestication of coca (Erythroxylum species)

“…For the sections from which we sampled multiple species, our phylogeny confirms previous suspicions that they do not represent exclusive evolutionary lineages (Rury, ; Plowman and Rivier, ; Emche et al., ; Islam, ). Instead, Archerythroxylum forms at least five major clades, all but one of which (clade II, with three species) contain members of other sections.…”

“…However, several morphological synapomorphies have been revealed to be more variable, and subsequent researchers have cast doubt on the monophyly of these sections (Schultes et al., ; Rury, ; Plowman and Rivier, ). Preliminary phylogenetic analyses have also supported this later view (Emche et al., ; Islam, ). Thus, the first purpose of this analysis is to test the monophyly of Schulz’ sections with the goal of eventually improving upon this intrageneric classification scheme.…”

“…There are two standing hypotheses for the domestication of coca: the linear evolutionary series hypothesis, where E. coca was the progenitor of the other cultivars (Bohm et al., ; Plowman, ), and the sister species hypothesis, where the two cultivated species form monophyletic groups sister to one another with a distinct common ancestor (Johnson et al., ; Emche et al., ). The nonmonophyly of the cultivated coca species in our analysis refutes both of these hypotheses and suggests a more complicated history of domestication.…”

“…Chemotaxonomic studies, however, have found distinct secondary metabolite profiles between the two varieties of E. coca and the varieties of E. novogranatense , suggesting they shared a common ancestor instead of progenitor–derivative relationship (but Amazonian coca and Colombian coca could still be derived from their respective conspecifics; Johnson et al., , ). Phylogenetic and cluster analyses using AFLPs also supported the two species being sister lineages (Johnson et al., ; Emche et al., ). The only other putative ancestor in the AFLP analysis, E. hondense , was not immediately similar to the cocas.…”

“…The only other putative ancestor in the AFLP analysis, E. hondense , was not immediately similar to the cocas. Our phylogenetic analysis permits testing of Plowman's linear evolutionary series hypothesis (Bohm et al., ) as well as Johnson and Emche's sister species hypothesis (Johnson et al., ; Emche et al., ) by sampling from all the cultivated coca taxa as well as thorough sampling of possible wild progenitors.…”