S.AfrJ.Bot.
, 1993, 59(4)
Short Communications I Kart Mededelings
Morphometric analysis of vegetative
characters of Encephalartos woodii, E.
natalensis and an apparent intermediate
R. Osborne* and R.T. Paschke
Department of Chemistry, University of Natal, King George V
Avenue, Durban, 4001 Republic of South Africa
Received 15 January 1993; revised 11 March 1993
Morphometric analyses were made on quantifiable leaf and
leaflet characters of 5 specimens each of Encephalartos
woodii Sander, E. natalensis Dyer & Verdoorn and of an
apparent intermediate believed to have originated in the
Krantzkloof Nature Reserve near Durban. Principal-
component analyses showed the Krantzkloof plants to be
intermediate between the two species.
Morfometriese analises is gedoen op kantifiseerbare blaar-
en pinnakenmerke van 5 eksemplare elk van Encephalartos
woodii Sander, E. natalensis Dyer & Verdoorn en van 'n
klaarblyklike oorgangsvorm vermoedelik afkomstig van die
Krantzkloof Natuurreservaat naby Durban. Hoofkomponent-
analises het getoon dat die Krantzkloof plante intermedier
tussen die twee spesies is.
Keywords : Encephalartos, Zamiaceae, Cycadales, morpho-
metric analysis.
• To whom correspondence should be addressed.
The account (Dyer 1965) of the discovery in 1895 by John
Medley Wood of Encephalartos woodii Sander (Zamiaceae,
Cycadales), as a single multi-trunked male specimen in the
Ngoye Forest, KwaZulu (28°51'S, 31 0 45'E), initiated the
popular belief that this plant is amongst the rarest in the
world. The report relates how this specimen was removed
from habitat in the form of several large trunks and basal
offsets between 1903 and 1916 - the major trunks thriving
to this day as particularly impressive feature plants in the
Durban Botanic Gardens (Osborne 1986). Despite intermit-
tent searches in the Ngoye Forest and its surroundings, no
other plant of this species was ever located and the species
is technically described as extinct (Lucas & Synge 1978).
It is fortunate that the original material re-established well
in cultivation and that vegetative propagation from basal
suckers and lateral branches has allowed plants of this clone
to be made available to selected botanical gardens and
private collectors. The 1991 - 1992 World Cycad Census
(Osborne 1993a) lists 37 mature E. woodii specimens in
cultivation in six countries (South Africa, Zimbabwe,
England, France, Italy and the USA). Of this number, 32 are
accounted for as originating directly or indirectly from the
Ngoye stock. The remaining 5 plants, all mature male speci-
mens in cultivation in the greater Durban area, could not be
ascribed to the Ngoye clone, but discussions with their
present owners indicated that these plants had their origin in
the Krantzkloof Nature Reserve near Durban (29°45'S,
300 52'E), and some 180 km in distance from the Ngoye
455
Forest site. These plants, four of which are in private
possession and one in the Old Fort Garden in Durban,
appear closely similar to, but not identical to, the plants
from Ngoye.
Individual representative whole-leaf samples were collect-
ed in December 1992 from 5 specimens of E. woodii at the
Durban Botanic Gardens, from 5 specimens of E. natalensis
Dyer & Verdoorn of known origin in a private collection in
Durban, and from each of the 5 specimens thought to have
their origin in the Krantzkloof area. Voucher specimens
(Osborne 2004 - 2020) are filed at the Herbaria of the
University of Natal, Durban, and the University of Durban-
Westville, Durban. After careful inspection, 13 quantifiable
characters were selected for assessment (Table 1). Leaflet
measurements were made on ten median leaflets and aver-
aged. Primary data were standardized and subjected to
principal-component analysis using the STAND, SIMINT,
EIGEN and PROJ routines of NT-SYS.pc (Rohlf et al. 1971).
Results of these analyses are shown in ordination diagrams
of the first (x-axis) and second (y-axis) components (Figure
la) and the first (x-axis) and third (y-axis) components
(Figure Ib). The first component accounts for 44% of the
variation and is influenced mainly by leaflet number, inser-
tion angle, width, area, position of maximum width and
leaflet shape index. The second component, accounting for
20% of the variation, comprises mainly the effect of inter-
leaflet distance and leaflet symmetry. The third component
accounts for 15% of the variation and results chiefly from
leaf length and leaflet length/width ratio data.
The diagrams show clearly that, in terms of the vegetative
characters used in this analysis, the Krantzkloof plants are
intermediate between the two known species. A casual
inspection of the cones over the past five years has indicated
that the reproductive characters of these plants may be
closer to E. woodii than to E. natalensis (Osborne 1993b),
but the use of these characters in morphometric studies was
prohibited by the non-availability of cone material from all
specimens. In order to examine more closely the true genetic
status of the Krantzkloof plants, it is planned to carry out a
restriction fragment length polymorphic analysis using
chloroplast DNA. It is significant that this technique has
recently been used by Italian workers to distinguish between
E. natalensis and E. woodii (Moretti & Norstog 1992).
Furthermore, useful information may arise from a proposed
anatomical investigation of the specimens. We conclude
Table 1 Schedule of characters used in this study
Leaf length
2 Number of leaflets
3 Mean interleaflet distance
4 Mean leaflet length
5 Mean leaflet width
6 Mean leaflet area
7 Mean leaflet length/width ratio
8 Mean leaflet position of maximum width
9 Mean leaflet shape index (Hill 1980)
10 Mean leaflet percentage symmetry (relative areas about a con-
structed longitudinal axis)
11 Mean angle (transverse view) between opposing leaflets
12 Mean angle (planar view) between the leaflet longitudinal axis
and the leaf rachis
13 Mean angle of insertion of the leaflet onto the rachis
456 S.-Afr.Tydskr.Plantk., 1993, 59(4)

0.8
(a)
0.6
E. natalensis
0.4 .2017

0.2 • 2020
.2012

0+-----~------------r-~~~------r-----------~~--~~--------_1

.2013
·0.2

·0.4
. 2016

-0.6
E. sp. 'Krantzkloof'
-0.8
2014

-1+------------,------------,------------,------------,-----------_1
·1 ·0.5 o 0.5 1.5

0.8
(b)
0.6

.2020
0.4

.2017
0.2
. 2007

-0.2

·0.4
'rro"
E. natalensis
.2009
E. woodii

-0.6
.2012

-0.8

-1
-1 -0.5 0 0.5 1.5
Figure 1 Ordination diagrams of the principal-component analysis: (a) for the first two components, and (b) for the fIrst and third
components. Nos. 2004 - 2008, Encephalartos woodii (Durban Botanic Gardens, ex Ngoye For~st); Nos. 2009 - 2013, E. sp. from the
Krantzkloofarea; Nos . 2014, 2016,2017,2019 and 2020, E. natalensisofknown origin.

with the speculation that the Krantzkloof plants either
represent a second population of E. woodii or that these
plants have experienced a significant genetic contribution
from that species.
References
DYER, RA 1965. The cycads of southern Africa Bothalia 8: 405 -
515.
HILL, R.S . 1980. A numerical taxonomic approach to the study of
angiosperm leaves. Bot. Gaz . 141 : 213 - 229.
LUCAS, G.L. & SYNGE, H. 1978. The mCN Plant Red Data
Book. Threatened Plants Committee, Royal Botanic Gardens,
Kew, England.
MORETTI, A. & NORSTOG, K. 1992. Eredita' uniparentale dei
plastidi in Encephalartos Lehm. (Zamiaceae, Cycadales). G.
Bot. Ital. 126: 15.
OSBORNE, R. 1986. Focus on ... Encephalartos woodii.
Encephalartos (Journal of the Cycad Society of South Africa)
5: 4 - 10.
OSBORNE, R. 1993a. The World Cycad Census. Proc. CYCAD 93,
Third Int Conf. on Cycad BioI., Pretoria, July 1993 (in pr~.) .
OSBORNE, R. 1993b. The other Woodii's. Veld & Flora (in
print).
ROHLF, Fl., KISHPAUGH, J. & KIRK, D. 1971. NT-SYS.
Numerical taxonomy system of multivariate statistical pro-
grammes. Tech. R~. State Univ. of New York, Stony Brook,
New York, USA.