Abstract
M.Sc.
Africa's cycads (66 species and 2 subspecies in two endemic genera: Encephalartos and
Stangeria) are extremely endangered with four species extinct in the wild and 80% threatened (CR,
EN, or VU) with all included in CITES Appendix 1. Although South Africa has some of the world’s
strictest cycad legislation, these plants are still under threat from illegal collection for horticulture
and medicine especially where they are seized in an unidentifiable condition. Currently developed
legislation demands accurate identification for permit issue. Ex situ conservation of genetic and
locality based diversity is paramount. Furthermore, taxonomically many species of unknown origin
are difficult to identify especially when diagnostic characters are absent. Species delimitation and
numbers are uncertain with field observations often contradicting current understanding. DNA
barcoding can assist with all the above-mentioned scenarios. In the current study all proposed
DNA barcoding regions (matK, rbcLa, psbA-trnH, and nrITS) along with several additional regions
were tested on ~350 samples from which a phylogeny of 63 of the 65 Encephalartos species was
also constructed. Results show general good amplification and sequencing success of proposed
barcoding regions, although a shift to specialist primers was made in several cases. Genetic
variation however was extremely low as is resolution at species level, even when multi-locus
barcodes were employed. Results obtained from the phylogenetic analyses show an increase in
resolution at both species and higher levels compared to previous work and as such several new
groupings are delimitated. Each species grouping is characterised by shared, derived
morphological, ecological, and geographic characters and when compared to previous
phylogenetic studies are supported to some extent. The current study provides the first step
towards a much-needed monograph and revision of the entire genus Encephalartos.