ArticlePDF Available

Reassessing taxonomic relationships in the Berkheya clade (Asteraceae, Arctotideae-Gorteriinae): The utility of Achene morphology

Authors:

Abstract

The genus Berkheya is paraphyletic with the related genera Cullumia, Cuspidia, Didelta and Heterorhachis embedded within a broader clade termed the ‘Berkheya clade’. As a contribution towards reassessment of species relationships and delimitation of species groupings within the clade, the utility of external achene morphology for supporting natural species groups within the clade was evaluated. Achenes from 67 species and 10 infraspecific taxa were examined, including representatives of each genus currently recognised in the Berkheya clade. Achene pubescence and pappus characteristics (e.g., scale shape and number of scales within a series) were indicated to be particularly variable and thus less reliable for assessment of species relationships. Of the currently recognised genera, only segregation of Cullumia was supported by achene and pappus characteristics. Species groupings implicit in Roessler’s infrageneric classification of eight series within Berkheya were to a large extent supported. However, in particular, series Speciosae was indicated to be a disparate species grouping and the distinction of series Angustae and Cruciatae is not supported by achene and pappus morphology. Achene and pappus characteristics suggest several novel species groupings (e.g. an affinity between Berkheya cuneata, B. ferox and B. spinosa). Although the taxonomic sampling in published molecular phylogenetic analyses is limited, achene and pappus morphology partially supports clades resolved in prevailing phylogenetic reconstructions for the clade. More complete sampling in future molecular phylogenetic analyses is required to test novel species affinities suggested in the present study and to explore the evolution of the achene and pappus in the Berkheya clade.
Phytotaxa 246 (1): 001–022
http://www.mapress.com/j/pt/
Copyright © 2016 Magnolia Press Article PHYTOTAXA
ISSN 1179-3155 (print edition)
ISSN 1179-3163 (online edition)
Accepted by Alexander Sennikov: 13 Apr. 2015; published: 5 Feb. 2016
http://dx.doi.org/10.11646/phytotaxa.246.1.1
1
Reassessing taxonomic relationships in the Berkheya clade (Asteraceae,
Arctotideae–Gorteriinae): the utility of achene morphology
NTOMBIFIKILE PHALISO1,2, ROBERT J. MCKENZIE1,*, NOLUTHANDO C. NETNOU-NKOANA1,3, PER OLA
KARIS4 & NIGEL P. BARKER1,5
1Department of Botany, Rhodes University, P.O. Box 94, Grahamstown, 6140, South Africa
2Current address: National Herbarium, South African National Biodiversity Institute, Private Bag X101, Pretoria 0001, South Africa
3Current address: Department of Agriculture, Forestry and Fisheries, Private Bag X250, Pretoria 0001, South Africa
4Department of Ecology, Environment and Plant Sciences, Stockholm University, Stockholm, SE-106 91, Sweden
5Current address: School of Integrated Plant and Soil Sciences, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa
*Corresponding author; tel. +27 46 6229698; fax +27 46 6229550, cell +27 764616509, email r.mckenzie@ru.ac.za
Abstract
The genus Berkheya is paraphyletic with the related genera Cullumia, Cuspidia, Didelta and Heterorhachis embedded within
a broader clade termed the Berkheya clade’. As a contribution towards reassessment of species relationships and delimita-
tion of species groupings within the clade, the utility of external achene morphology for supporting natural species groups
within the clade was evaluated. Achenes from 67 species and 10 infraspecific taxa were examined, including representatives
of each genus currently recognised in the Berkheya clade. Achene pubescence and pappus characteristics (e.g., scale shape
and number of scales within a series) were indicated to be particularly variable and thus less reliable for assessment of spe-
cies relationships. Of the currently recognised genera, only segregation of Cullumia was supported by achene and pappus
characteristics. Species groupings implicit in Roessler’s infrageneric classification of eight series within Berkheya were
to a large extent supported. However, in particular, series Speciosae was indicated to be a disparate species grouping and
the distinction of series Angustae and Cruciatae is not supported by achene and pappus morphology. Achene and pappus
characteristics suggest several novel species groupings (e.g. an affinity between Berkheya cuneata, B. ferox and B. spinosa).
Although the taxonomic sampling in published molecular phylogenetic analyses is limited, achene and pappus morphology
partially supports clades resolved in prevailing phylogenetic reconstructions for the clade. More complete sampling in future
molecular phylogenetic analyses is required to test novel species affinities suggested in the present study and to explore the
evolution of the achene and pappus in the Berkheya clade.
Key words: African flora, Compositae, fruit, taxonomy
Introduction
The tribe Arctotideae contains approximately 220 species with a predominantly southern African distribution. Two
subtribes are recognized, namely Arctotidinae (ca. 80–90 species) and Gorteriinae (ca. 135 species) (Karis et al. 2009).
The last full revision of the Gorteriinae was by Roessler (1959), whose generic treatment is still followed (Karis 2007;
Karis et al. 2009). Recent molecular phylogenetic studies (Funk et al. 2004; Funk & Chan 2008) and a morphological
cladistic analysis (Karis 2006) indicate that the Gorteriinae consists of two sister groups. The Berkheya clade’ consists
of Berkheya Ehrhart (1784: 303) (ca. 80 species), Cullumia R.Br. in Aiton (1813: 137) (15 species), Cuspidia Gaertner
(1791: 454) (one species), Didelta L’Héritier de Brutelle (1786: 55) (two species) and Heterorhachis Walpers (1847:
278) (two species). The Gorteria clade’ currently comprises the genera Gorteria Linnaeus (1759: 1229), Gazania
Gaertner (1791: 451) and Hirpicium Cassini (1820: 27) (Karis 2007; Karis et al. 2009).
In the most recent monograph of Berkheya, Roessler (1959) accepted 72 species divided into eight series. Seven
species and three infraspecific taxa of Berkheya and one Heterorhachis species have been described subsequently
(Roessler 1960, 1973; Hilliard & Burtt 1975, 1985, 1989; Manning et al. 2010; Manning & Goldblatt 2012; Bergh
& Helme 2014). Recent molecular phylogenetic studies indicate that Berkheya is paraphyletic as it is currently
circumscribed with Cullumia, Cuspidia, Didelta and Heterorhachis nested within Berkheya (Funk et al. 2004; Funk &
PHALISO ET AL.
2 Phytotaxa 246 (1) © 2016 Magnolia Press
Chan 2008; Phaliso 2013). These studies also indicate that Roessler’s infrageneric classification of Berkheya requires
reassessment, although interpretation is hindered by insufficient sampling for this purpose. Therefore, a thorough
reappraisal of species relationships and generic circumscription within the Berkheya clade is required.
Berkheya was established by Ehrhart (1784), based on B. fruticosa (Linnaeus 1753: 829) Ehrhart (1784: 138)
(Atractylis fruticosa L.) as the type. Subsequently, among major botanical works, L’Héritier de Brutelle (1786)
established Didelta based in large part on capitulum morphology. Under the Caracter essentialis for the genus L’Héritier
de Brutelle noted the differences in floret type and sex functionality within a capitulum (neuter ligulate ray florets,
sterile central disc florets, and perfect lateral disc florets), the seeds are embedded, the receptacle is separable, and
that the receptacle segments are hardened at fruiting. Gaertner (1791) established Cuspidia based on the morphology
of the involucral bracts and receptacle, floret types and their sex functionality, monomorphic achenes, and ‘plumose’
pappus. Rohria Vahl (1791: 15) was established for two species (those currently known as Berkheya carlinoides (Vahl
1791: 17) Willdenow (1803: 2275) and B. herbacea (Linnaeus 1782: 381) Druce (1917: 608)) with a honeycombed
receptacle surface, multimeric pappus, and sterile ray florets. Stobaea Thunberg (1800: 141) was erected for nine
newly described species, but the characters by which the genus was distinguished were not stated. A glabrous achene
and absence of a pappus were emphasised to support the segregation of Cullumia (Aiton 1813). Stephanocoma
Lessing (1832: 56) was segregated for two species (those currently known as Berkheya carduoides (Lessing 1832:
56) Hutch. in Fourcade (1932: 86) and B. decurrens (Thunberg 1793: 104) Willdenow (1803: 2276)) with achenes
bearing a uniseriate, coronate pappus. Heterorhachis was established on the basis of receptacle morphology (Schultz
Bipontinus 1844; Walpers 1847). Harvey (1865) maintained Berkheya and Stobaea as distinct genera, distinguished by
the acuminate, acute or taper-pointed pappus versus ‘blunt’ pappus, respectively. Harvey also retained Stephanocoma
but noted that it was distinguished only in possessing a coronate pappus.
Lessing (1832) was the first author to propose an infrageneric classification of Berkheya, dividing the genus
into four subgenera: Evopis, Agriphyllum, Basteria and Berkheya. Candolle (1838) partially followed Lessing’s
classification, subdividing the genus into six sections. Lessing’s subgenera Evopis, Agriphyllum and Basteria were
retained as sections; subgenus Berkheya was renamed sect. Euberkheya, and two new sections, sectt. Trichodes and
Trichocoma, were established. Harvey (1865) slightly modified Candolle’s classification by merging sectt. Evopis and
Agriphyllum and erecting a new section, Pseudostobaea. Each of these infrageneric classifications were based on a
combination of morphological characters, but mainly on life form and leaf morphology.
In his revision of Gorteriinae, Roessler (1959) sunk Stobaea, Rohria and Stephanocoma into Berkheya, and divided
the expanded Berkheya into eight series: Fruticosae (13 species), Angustae (monotypic), Cruciatae (monotypic),
Armatae (6 species), Speciosae (9 species), Rigidae (14 species), Subulatae (11 species) and Decurrentes (17 species).
A combination of vegetative and reproductive morphological characters was used to distinguish each series. Some
character states were shared among multiple series, and some series had multiple states for a specific character.
In particular, Roessler emphasized growth form, leaf shape and arrangement, involucral bract morphology (series
Angustae only), receptacle alveoles, achene length and pubescence, pappus series, pappus scale shape and apex shape,
and discoid versus radiate capitulum type.
As a contribution towards reassessment of species relationships and delimitation of species groupings within the
Berkheya clade, the utility of external achene morphology for indicating natural species groups within the clade was
evaluated. Achene morphological characters have proved valuable to aid elucidation of taxonomic relationships in
diverse Asteraceae groups (e.g. Grau 1980; Tadesse et al. 1995; Dittrich 1996; Vincent & Wilson 1997; Zhang et al.
2013), including the sister subtribe Arctotidinae (McKenzie et al. 2005). As is evident in the preceding discussion,
achene and pappus characters have been emphasised previously to support, either entirely or partially, circumscription
of genera in the Gorteriinae. However, a detailed assessment of the taxonomic value of achene and pappus characters
in the Berkheya clade has not been undertaken previously. In addition, achene and pappus morphology is compared
with existing phylogenetic knowledge.
Materials and Methods
Achenes from 67 species and 10 infraspecific taxa were examined, including representatives of each genus currently
recognised in the Berkheya clade (Karis 2007; Karis et al. 2009) and seven of the eight Berkheya series circumscribed
by Roessler (1959) (Table 1). Achenes were examined from herbarium specimens lodged in the Selmar Schonland
Herbarium, Grahamstown (GRA), the Compton Herbarium, Cape Town (NBG), and the National Herbarium, Pretoria
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 3
(PRE), and from fresh collections. The taxa examined and voucher details are listed in Appendix 1. For comparability,
achenes of outer disc florets were examined; the ray florets are sterile in all Gorteriinae that bear this floret type, and
some taxa have discoid capitula (Karis 2007; Karis et al. 2009). Wherever possible, mature achenes were examined;
immature achenes were studied only when mature achenes were unavailable and are indicated as such in Appendix 1.
Achenes from multiple specimens of a taxon were examined wherever possible to obtain an indication of intraspecific
variability. However, sample sizes were usually small (for most taxa <10 specimens per taxon), which reflects the
scarcity of herbarium specimens bearing mature achenes that are readily extractable without causing damage to the
specimen. As the achenes were dry, no treatment was necessary prior to direct mounting on stubs without pretreatment,
sputter-coated with gold-palladium and observation with a JEOL JSM 480 scanning electron microscope (SEM) at
the Electron Microscope Unit, Rhodes University, Grahamstown, and a Leica Stereoscan 420 SEM at the Electron
Microscope Unit, University of Limpopo, Medunsa campus. Features of external achene and pappus morphology
were recorded. Where necessary, supplementary observations were made with a compound microscope. Features
that exhibited variability between taxa were assessed for their taxonomic utility. Achene morphogenesis was not
investigated. Continuous quantitative characters were not recorded, as these might be influenced by growth conditions,
stage of achene development, and the proportion of florets within a capitulum that develop a mature achene containing
a fertile seed (Breitwieser & Ward 2003).
TABLE 1. Summary of the taxa sampled in this study. The classification follows Roessler (1959) and incorporates taxa
described subsequently (Roessler 1960, 1973; Hilliard & Burtt 1975, 1985, 1989; Manning et al. 2010; Manning & Goldblatt
2012; Bergh & Helme 2014). Infraspecific taxa are indicated in brackets.
Taxon Total number of species Number of species sampled
Berkheya series Fruticosae 13(3) 11
series Angustae 1 0
series Cruciatae 1(1) 1(1)
series Armatae 6 5
series Speciosae 10(6) 7(2)
series Rigidae 17(8) 13(3)
series Subulatae 11(3) 4(1)
series Decurrentes 19(1) 15(1)
Cullumia 15(7) 7(1)
Cuspidia 1(1) 1
Didelta 2(1) 2(1)
Heterorhachis 2 1
Total 98(31) 67(10)
Results and Discussion
Description of morphological characters
Achene
Achene shape was recorded reliably for mature filled achenes. The mature achene shape was turbinate (e.g., Berkheya
zeyheri (Harvey 1865: 496) Oliver & Hiern (1877: 429)), obovoid-turbinate (e.g., B. montana J.M.Wood & M.S.Evans
in Medley Wood (1897: 351)), cylindric-turbinate (e.g., B. heterophylla (Thunberg 1800: 141) Hoffman (1910: 314)),
attenuate-turbinate (e.g., B. cirsiifolia (Candolle 1838: 519) Roessler (1959: 250)), cylindric-obovoid to cylindric-
ellipsoid (Cullumia spp.), pyriform (e.g., B. cuneata (Thunberg 1793: 105) Willdenow (1803: 2270)), or obovoid-
unguiculate (Didelta spinosa (Linnaeus 1782: 384) Aiton (1789: 256)).
In many taxa ribs (which coincide with vascular traces) were conspicuous on the achene surface. In certain taxa,
notably Cullumia species, ribs were not obvious on the achene surface. In taxa with a dense covering of twin hairs on
the achene surface, the presence of ribs was not always discernible.
In Cullumia species, a ring of pale, fleshy tissue was observable at the base of freshly collected achenes. In
achenes from older herbarium specimens, this tissue is desiccated and inconspicuous and the achene base appears to
have a minute stalk. This fleshy tissue may represent an elaiosome, which refers to an external tissue rich in lipids,
proteins and sugars that is attractive to ants (Bennett & Krebs 1987). Elaiosome-bearing diaspores are inferred to be
dispersed by ants (myrmecochory), which is recorded in about 20% of plant species in the Cape Floristic Region (Bond
PHALISO ET AL.
4 Phytotaxa 246 (1) © 2016 Magnolia Press
& Slingsby 1983) and 17% of all angiosperm families worldwide, including the Asteraceae (Lengyel et al. 2010). Ants
take elaiosome-bearing diaspores to their nest, where the elaiosome is consumed and the remaining part of the seed is
discarded. Further studies of Cullumia achenes are needed to elucidate the biochemical composition of the tissue and
its attractiveness to ants. In certain Berkheya species, e.g., B. chamaepeuce (Moore in Schinz 1904: 1024) Roessler
(1959: 141) and B. schinzii O.Hoffm. in Schinz (1894: 213), a narrow ring of bare, seemingly non-cellular tissue, from
which no twin hairs arise, was present at the base of the achene. This tissue might be comparable to a carpopodium,
but further examination of freshly collected achenes is needed to clarify the nature of this tissue.
Differences in the sculpturing patterns on the achene surface, based on the external morphology of the epidermal
cells, have been noted previously (Phaliso 2013). In the present study, due to unavailability of mature achenes, only
semi-mature achenes were observed for some taxa. The degree of achene maturity may affect the external morphology
of the epidermal cells. In addition, for taxa with a dense covering of twin hairs on the achene surface, removal of the
twin hairs by physical or chemical means without potentially causing artefacts or damage to the epidermal cells could
not be guaranteed. Consequently, sculpturing pattern on the achene surface was not recorded in the present study.
Trichomes
In some taxa, the achene surface was completely glabrous, e.g. Berkheya cirsiifolia (Fig. 1A), B. ferox O.Hoffm. in
Schinz (1894: 214) (Fig. 3B), and Cullumia reticulata (Linnaeus 1753: 170) Greuter et al. (2005: 155) (Fig. 3F). In the
remaining taxa three trichome types were observed on the achene surface. The most common trichome type was the
twin hair or ‘Zwillingshaare’ (Hess 1938; Herman 2001; Fig. 1B–E), which consists of two parallel, elongate terminal
cells, sometimes of unequal length and often separated at their apices, and an extremely short basal cell. The basal
cell may be swollen (Fig. 1D) or not visibly swollen relative to the terminal cells. In many species, the terminal cells
are spirally twisted (Fig. 1C, E). Biseriate glandular hairs about 30 µm long, comprising a short stalk and a swollen
terminal cell, were observed on the achenes of four species, namely Berkheya latifolia J.M.Wood & M.S.Evans
in Medley Wood (1897: 351), B. speciosa (Candolle 1838: 518) O.Hoffm. in Zahlbruckner (1910: 314) (Fig. 1F),
Cuspidia cernua (Linnaeus 1782: 382) Burtt (1948: 316) and Heterorhachis aculeata (Burman 1768: 23) Roessler
(1959: 313). Achenial glandular hairs are also reported in Heterorhachis hysterix Manning et al. (2010: 188). The
possibility that glandular hairs are present on the ovary of additional taxa but are lost during achene development was
not examined and cannot be discounted. Uniseriate multicellular trichomes, similar to those observed in Arctotheca
calendula (Linnaeus 1753: 922) Levyns (1942: 284) in the Arctotidinae (McKenzie et al. 2005), were observed on the
surface of Cuspidia cernua achenes.
Twin hairs and uniseriate multicellular trichomes (when present) are most commonly distributed over the entire
achene surface, but may be concentrated in the apical portion of the achene, e.g. Berkheya annectens Harvey (1865:
509) (Fig. 2K), or at the base of the achene as in Cullumia patula subsp. uncinata Roessler (1959: 300) (Fig. 3G). In
some taxa, the twin hairs occurred predominantly between the ribs, although owing to variability this was not recorded
separately. The density of twin hairs ranged from a dense covering to a sparse covering on the achene surface.
The length of twin hairs relative to the achene length exhibited a continuum among taxa from individual twin hairs
greatly exceeding the achene in length (to ca. 4 mm long in Berkheya canescens Candolle (1838: 507) [Fig. 2C] and
B. schinzii) to extremely reduced twin hairs as short as 10 µm in length (Berkheya robusta Bohnen ex Roessler (1959:
260)).
Pappus
A pappus was present in all taxa, except all but one Cullumia species. In all pappose taxa, the pappus consisted of
a single or multiple series of scales at the apical end of the achene (Fig. 1G–N). Close examination of the insertion
of the scales on the achene was required to determine the number of series reliably. In most taxa one or two series
were present, but the pappus of Berkheya canescens was sometimes indeterminate (here termed ‘pluriseriate’) in
part because of the high number of scales present in this species. In some taxa the scales were partially connate only
at the base, e.g. Didelta carnosa (Linnaeus 1782: 384) Aiton (1789: 256) (Fig. 1M) and B. cirsiifolia (Fig. 1N), and
in occasional taxa the scales were completely connate so that discrimination of individual scales was subjective or
impossible, e.g. Berkheya carduoides. The scales were either morphologically uniform (monomorphic) or showed
obvious dimorphism between the scale series, e.g. alternate broad and narrow scales on an achene, such as in Berkheya
robusta (Fig. 1J) and Cuspidia cernua (Fig. 3I).
The scale shape ranged from linear-subulate (Fig. 1G) to lanceolate (Fig. 1H), ovate, oblong (Fig. 1I), obovate
or suborbicular. The lateral margin of the scale was entire, serrate (±regularly dentate; Fig. 1G, H), spinescent (deeply
serrated with spine-like protrusions; Fig. 1K) or fimbriate-spinescent (with long cilia-like protusions; Fig. 1L). The apex
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 5
shape ranged from truncate to acuminate. The apical margin was entire, erose (irregularly indented; Fig. 1I, J), serrate,
crenate (with rounded teeth), lacerate (deeply and irregularly serrated; Fig. 1N), spinescent or fimbriate. In some taxa the
scale apex was truncate and either erose, serrate or lacerate (here collectively termed ‘praemorse’; Fig. 1J).
AB C
D E F
G H I
JKL
NM O
FIGURE 1. Achene and pappus morphological features. A. Achene surface glabrous (Berkheya cirsiifolia). B. Dense twin hairs (Berkheya
francisci). C. Twisted twin hairs (Berkheya carlinopsis). D. Swollen basal cell of twin hair (Berkheya pinnatifida subsp. pinnatifida).
E. Twin hair (Cullumia aculeata). F. Glandular hairs (Berkheya speciosa subsp. speciosa). G. Linear-subulate pappus scales (Berkheya
umbellata). H. Lanceolate pappus scales (Berkheya pinnatifida subsp. pinnatifida). I. Oblong pappus scales (Berkheya bipinnatifida subsp.
bipinnatifida). J. Dimorphic, uniseriate pappus scales (Berkheya robusta). K. Spinescent pappus scales (Berkheya ferox). L. Fimbriate-
spinescent pappus scales (Berkheya chamaepeuce). M. Partially connate, spinescent pappus scales (Didelta carnosa var. tomentosa). N.
Partially connate pappus scales (Berkheya cirsiifolia). O. Pappus absent (Cullumia decurrens). Scale bars: A, K, M, 300 μm; B, H, O, 200
μm; C, E, 30 μm; D, L, N, 100 μm; F, 20 μm; G, 1 mm; I, J, 500 μm.
PHALISO ET AL.
6 Phytotaxa 246 (1) © 2016 Magnolia Press
ABC
DEF
G H I
JKL
FIGURE 2. Achenes of selected species of Berkheya. A. Berkheya fruticosa. B. Berkheya spinosissima. C. Berkheya canescens. D.
Berkheya cruciata. E. Berkheya herbacea. F. Berkheya speciosa subsp. speciosa. G. Berkheya purpurea. H. Berkheya carduoides. I.
Berkheya echinacea subsp. echinacea. J. Berkheya umbellata. K. Berkheya annectens. L. Berkheya heterophylla var. radiata. Scale bars:
A, B, D, E–H, L, 500 μm; C, J, 1 mm; I, 300 μm; K, 200 μm.
On the outer surface of the pappus scale, cell apices projected to varying degrees (termed papillate; Fig. 1H, I)
or did not project (smooth; Fig. 1N). In some species the projections were short, rounded and papilla-like, whereas in
some species the projections were longer, acute and spine-like.
The total number of scales per achene was highly variable among taxa and often within a taxon. In many taxa the
number of scales per achene varied within a limited range, usually centred around 10, 20 or 30. Overall, the number
of individual scales per achene (excluding taxa with completely connate scales) ranged from 6–9 (partially connate) in
Berkheya mackenii (Harvey 1865: 494) Roessler (1959: 261) to 40–50 in B. canescens. In his taxonomic descriptions
Roessler (1959) cited the predominant pappus scale number per achene (e.g. ‘10 + 10’, ‘cr. 20’) and rarely indicated
the range of variation within a taxon.
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 7
ABC
DEF
L
K
J
G H I
FIGURE 3. Achenes of selected species of Berkheya, Cullumia, Cuspidia, Didelta and Heterorhachis. A. Berkheya cardopatifolia.
B. Berkheya ferox var. ferox. C. Berkheya acanthopoda. D. Berkheya maritima. E. Cullumia decurrens. F. Cullumia reticulata subsp.
reticulata. G. Cullumia patula subsp. uncinata (immature). H. Cullumia aculeata var. aculeata (immature). I. Cuspidia cernua. J.
Heterorachis aculeata. K. Didelta carnosa var. tomentosa. L. Didelta spinosa. Scale bars: A, C, D, G, H, 500 μm; B, E, F, I, J, K, 1 mm.
Infraspecific variation in achene and pappus morphology
Acknowledging that sample sizes per taxon were small in the present study (achenes were examined from mostly <10
specimens per taxon), infraspecific variation was limited and predominantly observed in achene pubescence, number
of pappus scales per achene, and indentation of the pappus scale margin (see Table 2). Pappus scale shape commonly
varied on an individual achene, but in comparison showed limited variation between specimens within a taxon. Achene
shape was normally consistent within a taxon. However, slight curvature of the achene from the outermost disc florets
within a capitulum was observed in some Berkheya species and was especially marked in Didelta spinosa.
PHALISO ET AL.
8 Phytotaxa 246 (1) © 2016 Magnolia Press
TABLE 2. Achene and pappus morphological characters for the Berkheya clade’ taxa studied. ‘Group’ refers to species
groupings discriminated in this study exclusively on achene and pappus features (see Results and Discussion for brief
descriptions of each grouping).
Taxon Group I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII XIX XX XXI
Berkheya
Series Armatae
B. armata I 0 1 1 0 1 0 0 1 0 1 1 0 1 2,3 2 0 1 20–25 0 1,3 1,2
B. francisci I 0 1 1 0 1 0 0 1 0 1 0 0 0,1 1,2 2,4 0 1 20–22 0 1 0
B. herbacea I 0 1 1 0 1 0 0 1 0 1 1 0 1 2,3 2 0 1 20–30 0 1,3 1
B. macrocephala I 0 1 1 0 1 0 0 1 0 1 1 0 1 2,3 2 0 2 24 0 3 2
B. rosulata I 2 1 2 0 1 0 0,1 1 0 1 0 0 1 2 2 0 2 20 0 3 0,1
Series Cruciatae
B. cruciata IV 0 1 2 1 1 0 0 1 0 1 1 0 0,1 1 1,2 0, 1 1 10–18 0 0,1 0
Series
Decurrentes
B. acanthopoda VIII 1 0 0 NA NA NA NA 1 0 1 0 1 0 0 4 0 1 10 (5+5) 1 0,1 0
B. cf. caffra VIII 3 0 0 NA NA NA NA 1 0 1 0 1 0 0,1 4 1 1 10(5+5) 1 1 0
B. cirsiifolia VIII 3 0 0 NA NA NA NA 1 0 1 1 0 0 0,1,2 1 0,1 1 16–21 0 0,1,
3
0
B. decurrens VIII 1,3 0 0 NA NA NA NA 1 0 1 0 1 0 0 1,4 0 1 10 (5+5) 1 0,1 0
B. discolor VIII 3 0,1 0,2 0 0 1 1 1 0 1 0 1 0 0 1,4 0,1 1 10
(5+5)
1 1 0
B. griquana VIII 3 0, 1 0 NA NA NA NA 1 0 1 0 1 1 0 1 0 1 10 (5+5) 1 1 0
B. latifolia VIII 3 0,2 0 NA NA NA NA 1 0 1 0 1 0 1 1 0 1 10(5+5)–
13
1 0 0
B. sp. cf. latifolia VIII 3 0 0 NA NA NA NA 1 0 1 0 1 0 1 2,4 0 1 17–22 1 1 0
B. mackenii VIII 3 0 0 NA NA NA NA 1 0 1 0 1 0 0,1 1,4 0,1 1 6–9, 10
(5+5)
0, 1 1 0
B. maritima VIIIA 3 0 0 NA NA NA NA 1 0 1 0 1 1 1 1 0 2 10 1 1 0
B. montana VIIIA 1 0 0 NA NA NA NA 1 0 1 0 0 0 0, 2 1,4 0 1 10
(5+5)
1 1 0
B. onopordifolia
glabra
VIII 3 0 0 NA NA NA NA 1 0 1 0 1 0 0, 2 1 1 1 8 1 NA 0
B. onopordifolia
onopordifolia
VIII 3 0 0 NA NA NA NA 1 0 1 0 1 0 0, 2 1 1 1 12–14 1 NA 0
B. pauciflora VIII 3 1 2 0 0 1 1 1 0 1 0 0 0 1,2 1,4 0 2 20–21 0 1,3 0
B. radula VIII 3 0 0 NA NA NA NA 1 0 1 0 0 0 0 1,2 0,1 1 10
(5+5)–13
1 1 0
B. robusta VIIIA 3 0,1 2 0 0 0 1 0 0 1 0 1 1 0 1 1 1 10(5+5)–
14
1 1 0
B.
sphaerocephala
VIII 3 0 0 NA NA NA NA 1 0 1 0 1 0 2 1 1 1 10–13 0 1,3 0
Series
Fruticosae
B. angustifolia II 0,1 1 1 0 1 0 0 1 0 1 1 0 2 3 5 0 2 16–23 0 3 1
B. barbata II 0,1 1 1 0 1 0 0 1 0 1 0 0 2 3 5 0 2 20
(10+10)
0 1,3 0
B. canescens IIA 0,1 1 1 0 1 0 1 1 0 1 0 0 1 2 2 0 2,3 40–50 0 4 2
B. carlinopsis
magalismontana
II 0,1 1 1 0 1 0 0 1 0 1 0 0 1 0,1(2) 4,5 0 2 18–21 0 1 0
B. chamaepeuce II 0,1 1 1 1 1 0 0 1 0 1 0 0 0 2 5 0 2 20–21 0 1,3,
4, 5
0
B. coriacea II 0,1 1 1 0 1 0 0 1 0 1 1 0 1 2 1,4 0 2 18–24 0 1,3 1
B. cuneata IX 4 1 1,2 0 1 0 0,1 1 0 1 1 0 2 3 0,2 0 1 16–20 0 3,4 0
B. fruticosa II 0,1 1 1 0 1 0 0 1 0 1 1 0 0,1 2,3 1,2 0 2 20 (10 +
10)
1 3,5 1
B. schinzii II 2 1 1 0 1 0 0 1 0 1 1 0 1 2,3 1,4 0 2 18–20 1 1,3 0
B. spinosa IX 4 0,1 1 0 1 0 0 1 0 1 1 0 1 2,3 0,2 0 2 18–20 0 4 2
B. spinosissima
spinosissima
II 1,6 1 2 0 1 0 1 1 0 1 1 0 2,3 3 1 0 2 20
(10+10)
1 3,4 0
...continued on the next page
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 9
TABLE 2. (Continued)
Taxon Group I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII XIX XX XXI
Series Rigidae
B. annectens VIIA 2 1 2 0 0 1 0 1 0 1 1 0 1 2 0,2 0 1 18–20 0 1,3 0
B. bipinnatifida VIIB 3 1 2 0 1 0 1 1 0 1 1 0 0 0,1 1,2 0 1 16–20 0 0,1 0
B.
buphthalmoides
VII 2 1 2 0 0 1 0,1 1 0 1 1 0 0 0,1 0,1 0 1 20 0 0,1 0
B. cardopatifolia VIIB 3 1 2 0 0 1 1 1 0 1 1 1 0 0 2 0 1 13–20 0 1 0
B. carlinifolia VII 2 1 2 0 0,1 0,1 0,1 1 0 1 0,1 0 0,1 0,1,2 1,2 0 1,2 14–22 0 1 0
B. eriobasis VIIA 2 1 2 0 1 1,1 1 1 0 1 0,1 0 0,1 1,2 0,1,2 0 1 14–21 0 1,3 0
B. ferox IX 4 0 0 NA NA NA NA 1 0 1 0 0 1,2 3 2 0 1 20 0 4 0,1
B. glabrata VII 2 1 2 0 1 0 0 1 0 1 1 0 0 1 1,2 0 2 19–20 0 0,1 0
B. heterophylla
heterophylla
VII 2 1 2 0 1 0 0,1 1 0 1 1 0 0,1 0,1 1 0 1 17–20 0 1 0
B. heterophylla
radiata
VII 2 0,1 2 0 0,1 0,1 0,1 1 0 1 0,1 0 0 0,1,2 1,2 0 1 16–22 0 0,1 0
B. jardineana VII 2 1 2 0 0 0 0 1 0 1 0 0 1 1 2 0 2 19–22 0 0,1 0
B. onobromoides VII 2 1 2 0 1 1 1 1 0 1 1 1 1 0 2 0 1 20–21 0 0,1,5 0
B. pinnatifida
pinnatifida
VIIA 2 1 2 0 1 0,1 0,1 1 0 1 1 0 1 2 1,2 0 2 16–24 0 1,3,
5
0
B. pinnatifida
ingrata
VIIA 2 0 0 NA NA NA NA 1 0 1 0 0 1 2 1,2 0 2 20 0 0,1,3 0
B. pinnatifida
stobaeoides
VIIA 2 0 0 NA NA NA NA 1 0 1 0 0 1 2,3 2 0 2 18–20 0 3 0
B. rigida VII 2 1 2 0 0 0,1 0,1 1 0 1 0 0 0 0,1 1 0 1 13–20 0 1 0
Series Speciosae
B. carduoides VIII 3 0,1 0,2 0 0 1 1 1 0 1 0 1 0 0 1,4 1 0,1 8–14 0 NA 0
B. echinacea III 0 1 1,2 1 1 0 0 1 0 1 1 0 1,2 0 1,4 0 2 20 0 0,1 0
B. purpurea VIII 1 1 1 0 1 0 1 1 0 1 0 0 0 1,2 1,2 0 2 16–20 0 0,1 0
B. rhapontica
aristosa aristosa
III 0 1 2 1 1 1 0 1 0 1 0 0,1 1 1,2 2,4 0 2 14–20 0 1,3,5 0
B. rhapontica
rhapontica
III 0 1 2 1 1 0 1 1 0 1 0 0 0 1 1,2 0 2 20 0 0,1 0
B. setifera IIIA 0 1 2 1 0 0 0 1 0 1 0 0 1 3 2 0 2 30–36 1 4 2
B. speciosa V 0 1,2 2 0 0 0 1 1 0 1 0 1 1 0,1,2 1,2 0,1 2 14–20 0 0 1,2
B. umbellata IIIA 0 1 2 0 0 0 1 1 0 1 1 0 1 3 2 0 2 29–40 0 4 2
Series Subulatae
B. coddii VI 2 1 2 0 0 0 0,1 1 0 1 0 0 0,1 1 1,2,3 0 2 17–20 0 0,1 2
B. densifolia VI 2 1 2 0 1 0 0 1 0 1 0 0 0 0,1 1,2 0 1 11–15 0 0,
1, 2
0
B. subulata
wilmsiana
VI 2 1 1 0 1 0 0 1 0 1 1 0 1 1 1,4 0 1 18–20 0 0 1
B. zeyheri
rehmannii
VI 0 1 2 0 0 0 1 1 0 1 1 0 1,2 0,1 1,4 0 2 20 0 0 2
B. zeyheri
zeyheri
VI 0 1 2 0 0 0 1 1 0 1 1 0 1,2 0,1 1,4 0 2 20 0 0 2
Cullumia
Cullumia
aculeata
XIVA 5 1 2 0 1 0 1 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
Cullumia bisulca XIV 5 0 0 NA NA NA NA 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
Cullumia
decurrens
XIV 5 0 0 NA NA NA NA 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
Cullumia patula
patula
XIV 5 0 0 NA NA NA NA 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
Cullumia patula
uncinata
XIVB 5 1 2 0 1 2 0 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
Cullumia
reticulata
XIV 5 0 0 NA NA NA NA 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
Cullumia sp. XIV 5 0 0 NA NA NA NA 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
...continued on the next page
PHALISO ET AL.
10 Phytotaxa 246 (1) © 2016 Magnolia Press
TABLE 2. (Continued)
Taxon Group I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII XIX XX XXI
Cullumia sulcata XIV 5 0 0 NA NA NA NA 0 1 0 NA NA NA NA NA NA NA NA NA NA NA
Cuspidia
C. cernua X 4 2,3 2 NA NA 0 0 1 0 1 1 0 2,3 3 1,5,6 0 2 20
(10+10)
1 3,4 0
Didelta
Didelta carnosa XII 4 0 NA NA NA NA NA 0 0 1 0 0 2 3 5 1 1 10–11 0 3 0
Didelta spinosa XIII 6 1 2 1 1 0 1 1 0 1 0 0 2 2,3 5 0 1 14–20 0 3 0
Heterorhachis
H. aculeata XI 2 1,2 2 0 1 0 1 1 0 1 1,2 0 3 2,3 5,6 1 2 14–20 1 3,4 1
Key to characters
Achene: I, Achene shape: 0, turbinate; 1, obovoid-turbinate; 2, cylindric-turbinate; 3, attenuate-turbinate; 4, pyriform; 5, cylindric-
obovoid to cylindric-ellipsoid; 6, obovoid-unguiculate. II, Hairs on achene surface: 0, glabrous; 1, twin hairs present; 2, glandular hairs
present; 3, uniseriate multiseptate trichomes present. III, Twin hair/uniseriate hair length: 0, hairs absent; 1, approximately half or longer
than the cypsela length; 2, less than half the cypsela length. IV, Twin hair basal cell morphology: 0, basal cell not swollen; 1, basal cell
swollen. V, Terminal cell morphology: 0, straight; 1, twisted. VI, Twin hair/uniseriate hair distribution: 0, distributed over entire cypsela
surface; 1, mainly apical; 2, basal only. VII, Twin hair density: 0, dense; 1, sparse. VIII, Ribs: 0, absent; 1, present. IX, Elaiosomes: 0,
absent; 1, present. Pappus: X, Pappus: 0, absent; 1, present. XI, Scale surface papillate: 0, absent; 1, present; 2, glandular. XII, Scale
apex praemorse: 0, absent; 1, present. XIII, Scale lateral margins: 0, entire; 1, serrate; 2, spinescent; 3, fimbriate. XIV, Scale apex shape:
0, truncate; 1, obtuse; 2, acute; 3, acuminate. XV, Scale apex margin: 0, entire; 1, erose; 2, serrate; 3, crenate; 4, lacerate; 5, spinescent;
6, fimbriate. XVI, Scale fusion: 0, scales free; 1, scales at least fused at base; 2, coronate. XVII, Number of scale series: 1, uniseriate;
2, biseriate; 3, pluriseriate. XVIII, Total number of scales. XIX, Pappus dimorphism: 0, absent; 1, present. XX, Pappus scale shape: 0,
obovate; 1, oblong; 2, suborbicular; 3, lanceolate; 4, linear-subulate; 5, ovate. XXI, Pappus scale length relative to achene: 0, shorter; 1,
±equal; 2, longer.
The small sample sizes per taxon in the present study strongly reflects the limited accessibility of mature achenes
from herbarium vouchers, which have been predominantly collected at anthesis and lack capitula containing mature
achenes. Additional sampling is required to determine if the characteristics described for each taxon are representative
of the entire taxon and to better sample the entire range of infraspecific variation. While fresh collections were made
for many taxa during the course of this study, given the wide geographic distribution of the Berkheya clade in sub-
Saharan Africa, and the enormous climatic and ecological diversity under which different species grow, collecting
fresh mature achenes from multiple populations of each taxon is not a simple undertaking.
Of the taxa for which samples from >10 specimens were examined, some variation in achene pubescence and
pappus scale number per achene was observed in Berkheya carlinifolia, B. heterophylla and B. speciosa subsp.
speciosa (see Table 2). Both B. carlinifolia and B. heterophylla are morphologically variable species complexes that
are currently under investigation (PO Karis unpublished data). The variation in achene and pappus morphology in
these two species might be consistent with variability in other vegetative and reproductive organs and thus might be
informative to assist with delimitation of morphology-defined entities within both complexes.
With the above caveats in mind, the following discussion relates what is presently known about the achene and
pappus morphology of the studied taxa to the prevailing generic concepts in the Berkheya clade and Roessler’s (1959)
infrageneric classification of Berkheya.
Achene and pappus morphology in relation to the current taxonomic classification
Generic circumscription
With regard to the current generic circumscription in the Berkheya clade, only segregation of Cullumia is supported
by achene morphological characters considered in isolation from other morphological data. Cullumia possesses an
achene type unique in the clade, namely a cylindric-obovoid or cylindric-ellipsoid, usually glabrous achene with
inconspicuous ribs (Fig. 3E, F), and with an alleged occurrence of elaiosomes. Cullumia might further be diagnosed
as epappose (Fig. 1O), except for the presence of minute scales observed in some exemplars of C. aculeata (Houttuyn
1776: 158) Roessler (1959: 291) (Karis 2006). As molecular phylogenetic reconstructions indicate that Cullumia is
embedded within a paraphyletic Berkheya (Funk & Chan 2008), the Cullumia achene type can be interpreted as an
evolutionary novelty with a single origin within Berkheya s.str.
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 11
Two Cullumia taxa stand apart from the remaining taxa in the genus in producing achenes bearing twin hairs on
the surface as reported by Roessler (1959). Cullumia aculeata achenes have a moderate or dense covering of short
(<0.5 mm long) twin hairs on the achene surface except in the basal portion (Fig. 1E, 3H). In Cullumia patula subsp.
uncinata deciduous twin hairs (ca. 2 mm long) about half the achene length are attached only in the basal portion of the
achene (Fig. 3G). In both Cullumia species, especially C. patula subsp. uncinata, the terminal cells of the twin hairs
differ considerably in length.
Although other morphological characters are held to support the generic segregation of Cuspidia, Didelta and
Heterorhachis (Karis 2006), the achene and pappus of the species presently classified in these genera show few
morphological features that are unique within the Berkheya clade. In Didelta spinosa the lanceolate, fimbriate-spinescent
pappus scales (Fig. 3L) are similar to those of Berkheya cuneata and B. ferox. The present study revealed similarities
in the achene and pappus morphology of Cuspidia cernua (Fig. 3I) with those of Didelta carnosa, Berkheya cuneata,
B. ferox and B. spinosa (Linnaeus 1782: 381) Druce (1917: 609), which has not been noted previously. In these species
the achene is pyriform in shape, with the basal, constricted portion more elongated in Cuspidia cernua (Fig. 3I) and
Didelta carnosa (Fig. 3K). The pappus scales are variable in shape and length but are spinescent to differing degrees.
The presence of glandular trichomes on the pappus scale surface of Cuspidia cernua is a feature shared with the two
Heterorhachis species in the Berkheya clade (Karis 2006; Manning et al. 2010; this study). The presence of glandular
trichomes on the achene of Heterorhachis species (this study; Manning et al. 2010) is shared with Berkheya latifolia,
B. speciosa and Cuspidia cernua, and is indicated to be homoplasious based on the phylogenetic reconstructions
of Phaliso (2013). However, the spinescent-fimbriate pappus scales of Heterorhachis species are unusual and have
their closest resemblance in the spinescent-fimbriate scales of Berkheya chamaepeuce (termed ‘fimbriate’ by Roessler
1959).
Roessler’s Berkheya series
Roessler (1959) divided Berkheya into eight infrageneric series. The achene and pappus features used, in conjunction
with other morphological features, by Roessler to characterize these series are listed in Table 3. Observations in the
present study indicate that achene and pappus morphology partially supports these groupings. However, Roessler
used a combination of vegetative and reproductive morphological characters to distinguish each series. Consequently,
it is not unexpected that certain series are indicated to be unnatural groupings based solely on achene and pappus
features.
TABLE 3. Comparison of achene and pappus morphological features emphasized by Roessler (1959) to delimit series in
Berkheya. Note that Roessler also emphasized other vegetative and reproductive morphological features to distinguish the
series.
Characteristic Fruticosae Angustae Cruciatae Armatae Speciosae Rigidiae Subulatae Decurrentes
Achene
Hair covering
(when present)
Pilose-
sericeous
Pilose-
sericeous
Pilose-
sericeous
Pilose-
sericeous
Pilose-sericeous Pilose-sericeous Pilose-
sericeous
Pilose
Hair length ca. 1–4 mm 0.5 mm 0.5–1 mm 0.5–3 mm To ca. 0.5 mm To ca. 0.2 mm 0.2–1 mm
Hair density Dense Dense Dense Dense Dense, sparse,
subglabrous or
glabrous
Dense, sparse,
subglabrous or
glabrous
Dense or
sparse
Glabrous or
sparse
Pappus
No. of series Subbiseriate
or pluriseriate
Uniseriate Uniseriate Subbiseriate Subbiseriate or
pluriseriate
Subbiseriate or
uniseriate
Subbiseriate
or ±uniseriate
Uniseriate
No. of scales ca. 10 + 10 or
40-50
ca. 20 ca. 10–15 ca. 10 + 10 ca. 10 + 10 ca. 20 10 + 10 or
10–16
ca. 10 or up
to 20
Scale shape Lanceolate,
oblong,linear-
subulate,
linear-setiform
Linear-oblong Oblong or
obovate-
lanceolate
Lanceolate
or oblong
Obovate, oblong,
lanceolate,
linear-setiform
Subrotund, ovate,
oblong or linear-
lanceolate
Obovate or
lanceolate
Oblong
Scale apex ±Acute Acute Acute or
subobtuse
Acute Acute, subacute
or obtuse
Obtuse or acute Obtise or
acute
Obtuse or
subacute
Scale fusion Free Free Free or connate Free Free Free Free Free or
connate
Scale length 1–7 mm ca. 1 mm 0.5–1 mm 0.5–5 mm ca. 1–8 mm 0.2–3 mm 0.8–6 mm 0.2–2 mm
PHALISO ET AL.
12 Phytotaxa 246 (1) © 2016 Magnolia Press
Berkheya series Fruticosae contains 13 species, including the type for the genus, B. fruticosa. Eleven species
were sampled in the present study (e.g. Fig. 2A–C). The majority of these species had turbinate or obovoid-turbinate
achenes with a dense covering of twin hairs over the entire surface. The pappus scales were biseriate (sometimes
pluriseriate in B. canescens), usually lanceolate, with a serrate margin and an acute or obtuse apex. In some species
the pappus scales may be linear-subulate (B. canescens, B. chamaepeuce and B. spinosissima (Thunberg 1793:
108) Willdenow (1803: 2270)). The scales were somewhat cryptically dimorphic in B. fruticosa, B. schinzii and B.
spinosissima, and monomorphic in the remaining species. Considerable variation in pappus scale number, length,
shape and margin serration was evident among the species. Berkheya cuneata and B. spinosa fundamentally differed
in producing pyriform-shaped achenes. Thus, with the exception of B. cuneata and B. spinosa, the sampled species in
series Fruticosae produced broadly similar achenes but showed considerable diversity in pappus morphology.
Berkheya cruciata (Houttuyn 1779a: 21) Willdenow (1803: 2276) was placed in the monotypic Berkheya series
Cruciatae by Roessler (1959). The achene and pappus of B. cruciata (Fig. 2D) do not exhibit a unique character state
or combination of character states that support its distinction in its own series. The achene was turbinate with a dense
covering of twin hairs, shorter than the achene, over the entire surface. The pappus scales were free, uniseriate, 10–18,
oblong or obovate, with an erose or serrate apex. The pappus scales were markedly shorter than the achene, but exhibited
no distinctive character states not observed in other species. In addition to achene and pappus characteristics, Roessler
(1959) used receptacle alveole features, growth form, leaf characters and radiate capitula to support segregation of
series Cruciatae.
Berkheya series Armatae contains six species, of which five were sampled in the current study (e.g. Fig. 3E).
The achenes were similar to those of the preceding series in being obovoid-turbinate with a dense covering of twin
hairs, shorter or longer than the achene, distributed over the entire surface. The pappus scales were free, uniseriate or
biseriate, 20–30, monomorphic, oblong or lanceolate. The scale margins were usually serrate (rarely entire) and the
apex was usually acute to acuminate (sometimes obtuse). The pappus scale length relative to the achene length varied
considerably between species. The pappus scale outer surface was papillate in B. armata (Vahl 1793: 39) Druce (1917:
608), B. herbacea and B. macrocephala Medley Wood (1907: 50), but smooth in B. francisci Bolus (1906: 396) and
B. rosulata Roessler (1959: 163). Thus the achenes of the sampled species are broadly similar in morphology with
comparatively limited variation in pappus morphology.
Nine species were classified in Berkheya series Speciosae by Roessler (1959) and B. pannosa Hilliard & Burtt (1975:
77) was described subsequently (Hilliard & Burtt 1975). Considerable variation in achene and pappus morphology was
evident among the six species sampled in the present study (e.g. Fig. 3F–J). The achene was turbinate in most species,
but was cylindric-turbinate in B. carduoides and obovoid-turbinate in B. purpurea (Candolle 1838: 518) Masters (1872:
1262). In most species twin hairs were distributed over the entire achene surface. However, in B. speciosa twin hairs
were absent or, when present, extremely sparse and extremely reduced in size, and short glandular trichomes were
usually present on the achene surface. In B. rhapontica subsp. aristosa (Candolle 1838: 518) Roessler (1959: 170) var.
aristosa the twin hairs were only present in the apical portion of the achene. The twin hairs were usually less than half
the achene length, but were longer in B. purpurea and some samples of B. echinacea (Harvey 1865: 495) O.Hoffm. ex
Burtt Davy in Burtt Davy & Pott-Leendertz (1912: 121). The twin hair density and achene surface sculpturing varied
considerably among the sampled species. The pappus scales were free, biseriate and monomorphic. The pappus of B.
setifera Candolle (1838: 507) and B. umbellata Candolle (1838: 507) comprised 29–40 linear-subulate scales; in the
remaining species the pappus consisted of 14–20, usually obovate or oblong scales (but sometimes lanceolate or ovate
in B. rhapontica subsp. aristosa var. aristosa). The pappus showed considerable variation in margin serration and
apex shape among the sampled species. The pappus scale outer surface was sparsely papillate in B. echinacea and B.
umbellata, and smooth in the remaining species. Achene and pappus morphology suggested that B. carduoides shared
a closer affinity with species in series Decurrentes, on account of having a cylindric-turbinate, usually glabrous achene
with a much shorter, variably connate (sometimes entirely coronate) pappus. Thus, in terms of achene and pappus
morphology, series Speciosae was indicated to be among the most variable series and as circumscribed by Roessler
(1959) may be an unnatural assemblage of species.
Roessler (1959) recognised 14 species in Berkheya series Rigidae, and subsequently three additional species
have been described (Manning et al. 2010; Manning & Goldblatt 2012; Bergh & Helme 2014). Among the 13 species
sampled (e.g. Figs. 2K, L, 3A, B), the achenes were obovoid-turbinate or turbinate, glabrous or with a sparse to dense
covering of twin hairs shorter than the achene, over the entire achene surface or mainly apical, and the achene surface
swollen/colliculate, sunken or reticulate. The pappus scales were much shorter than achene, 13–24, free, monomorphic,
oblong, obovate or lanceolate, outer surface rugose and papillate or not, margin entire, erose or serrate, apex obtuse,
truncate or acute.
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 13
Three subgroupings were discernible among species classified in Berkheya series Rigidae. In B. carlinifolia
(Candolle 1838: 516) Roessler (1959: 203), B. heterophylla, B. jardineana Manning & Goldblatt (2012: 57), B.
onobromoides (Candolle 1838: 516) O.Hoffm. & Muschl. in Zahlbruckner (1910: 319), and B. rigida (Thunberg 1800:
141) Ewart et al. (1909: 20), the achenes were obovoid-turbinate, with short pappus scales oblong or obovate, with
an obtuse or truncate apex, and much shorter than the achene (0.2–1.5 mm long). The placement of the distinctive B.
jardineana in series Rigidae by Manning & Goldblatt (2010) is supported by achene and pappus morphology. The
achenes of B. bipinnatifida (Harvey 1865: 499) Roessler (1959: 212) and B. cardopatifolia (Candolle 1838: 514)
Roessler (1959: 215) were similar but differed in being distinctly turbinate and in the former species the pappus was
uniseriate with obvious gaps between the individual scales. In B. annectens, B. buphthalmoides (Candolle 1838: 517)
Schlechter (1897: 343), B. eriobasis (Candolle 1838: 516) Roessler (1959: 205), B. glabrata (Thunberg 1800: 141)
Fourcade (1941: 22) and B. pinnatifida (Thunberg 1800: 141) Thell. in Probst (1928: 77), the pappus scales were
slightly longer (0.3–2 mm long vide Roessler 1959), but still shorter than the achene, and differed in overall shape
and apex shape. In B. annectens and B. pinnatifida subsp. pinnatifida the pappus scales were lanceolate with acute
apices and the scale length within a series on an individual achene was variable, and the twin hairs were concentrated
in the apical half of the achene (especially in B. annectens). The achenes were glabrous in B. pinnatifida subspp.
ingrata (Bolus 1907: 346) Roessler (1959: 210) and stobaeoides (Harvey 1865: 505) Roessler (1959: 209). In B.
buphthalmoides, B. eriobasis and B. glabrata the scales were obovate or oblong with an obtuse apex and ±uniform in
length, and the twin hairs were distributed over the entire achene surface.
On the basis of achene and pappus morphology, Berkheya ferox var. ferox (Fig. 3B) was indicated to be misplaced
in series Rigidae. In this taxon, the achenes were pyriform with a uniseriate pappus of free monomorphic scales, which
are linear-subulate with an acuminate apex and a serrate to spinescent margin (features shared with B. cuneata and B.
spinosa in series Fruticosae). From the latter two species, B. ferox differed primarily in the achenes being glabrous
and the pappus scale outer surface smooth. Roessler (1959) gave no indication of the reasons for his placement of B.
ferox in series Rigidae. The similarity in achene morphology of B. cuneata, B. ferox and B. spinosa suggests that a
reappraisal of their affinities might be warranted.
Berkheya series Subulatae contains 11 species, of which four species were examined in the present study. In these
species the achene was turbinate or obovoid-turbinate, always with a sparse or dense covering of twin hairs over the
entire surface. The pappus comprised monomorphic free scales, either uniseriate or biseriate, usually 17–20 (but 11–15
in B. densifolia Bohnen ex Roessler (1959: 229)). The scales were usually obovate, but were occasionally oblong or
suborbicular. The apex was usually obtuse or truncate. The scale outer surface was smooth or papillate. The scale
length varied from shorter to longer than the achene among different taxa. The scale margins were notably variable in
the type of serration.
Roessler (1959) accepted 17 species in Berkheya series Decurrentes, and subsequently B. pauciflora Roessler
(1960: 519) and B. griquana Hilliard & Burtt (1985: 229) were described (Roessler 1960; Hilliard & Burtt 1985).
Achene and pappus morphology was investigated for 16 species (e.g. Fig. 3C, D). The achene shape of these species
was either obovoid-turbinate (and often somewhat bilaterally flattened) (e.g. B. acanthopoda (Candolle 1838: 521)
Roessler (1959: 255), B. maritima Medley Wood & Evans (1899: 253), B. montana and B. robusta) or cylindric-
turbinate (e.g. B. cirsiifolia and B. sphaerocephala (Candolle 1838: 518) Roessler (1959: 265)). The achene surface
was consistently glabrous in most species, but rare short twin hairs (up to about 25 µm long) were present at the apical
end of some achenes of B. discolor (Candolle 1838: 520) O.Hoffm. in Zahlbruckner (1910: 319), B. griquana and
B. robusta. Extremely short and extremely sparse twin hairs were present, only between the ribs, on all achenes of
B. pauciflora examined. Sparse short glandular hairs were observed on the surface of some B. latifolia achenes. The
achene surface was usually reticulate or sunken, but in some species was swollen/colliculate. Ribs were observed in
most species, although in some species ribs were only conspicuous in the basal portion of the achene. The pappus
was always much shorter than the achene. The form of the pappus ranged from entirely free scales (e.g. B. maritima)
to shortly connate at the base of the scales (e.g. B. griquana and B. sphaerocephala) to almost entirely connate (i.e.
coronate; e.g. B. carduoides). The number of scales was diverse, most commonly 10 (e.g. B. discolor and B. montana),
but rarely fewer (ca. 6–9 with partially connate scales in the Mpumalanga form of B. mackenii) or up to 16–22 (B.
cirsiifolia and B. pauciflora). The species with 10 scales usually exhibited a strongly dimorphic pappus with alternate
broad oblong and narrow lanceolate scales. The pappus was uniseriate in most taxa and only rarely biseriate (B.
maritima and B. pauciflora). The scale apex was usually truncate or obtuse; taxa with dimorphic scales the apices
were usually truncate (broad scales) and acute (narrow scales). The apex margin was usually erose or lacerate, but
occasionally serrate. The scale lateral margins were usually entire but occasionally serrate. Thus, collectively the taxa
in series Decurrentes represented a coherent group with two achene types discernible and notable variation in pappus
PHALISO ET AL.
14 Phytotaxa 246 (1) © 2016 Magnolia Press
scale number and degree of fusion. As mentioned above, the present observations suggested that B. carduoides (placed
in series Speciosae by Roessler) may have a closer affinity with series Decurrentes.
Berkheya series Angustae is monotypic and known only from the type collection of B. angusta Schlechter (1899:
219). Attempts to relocate the species in recent years have proved unsuccessful and other herbarium specimens
identified as B. angusta have proved to be misidentified. It was not possible to extract an immature achene from the
isotype lodged in GRA for examination in the present study. Roessler (1959), who provided greater detail than in the
original description by Schlechter (1899), described the achene as being shortly and densely pilose, with a uniseriate
pappus of about 20 scales, with each scale linear-oblong, acute, minutely denticulate and about 1 mm long. Achene
and pappus morphology on its own provides weak support for segregation of series Angustae. In terms of vegetative
morphology, B. angusta shows affinities with B. barbata (Linnaeus 1782: 382) Hutch. in Fourcade (1932: 86) and
B. angustifolia (Houttuyn 1779b: 518) Merrill (1938: 369). Roessler’s (1959) description of B. angusta suggests the
former two species differ primarily in bearing opposite leaves and a biseriate pappus of lanceolate, slightly longer
scales (2–4 mm and 1.5–3 mm, respectively). Roessler also noted morphological similarities between B. angusta and
Cullumia species, but the achene and pappus of Cullumia species differ notably in morphology (e.g., in being epappose
and glabrous or only sparsely hairy). Rediscovery of B. angusta is of utmost importance to provide fresh material with
which to determine its affinities.
Phylogenetic interpretation of achene and pappus morphology in the Berkheya clade
The most complete published molecular phylogenies currently available for the Berkheya clade are those of Funk
& Chan (2008), which were derived from chloroplast and nuclear DNA sequence data. Their study was focused on
elucidating relationships within the entire subtribe Gorteriinae, hence taxonomic sampling of the Berkheya clade was
limited (30 spp.). Therefore, only a preliminary interpretation of achene and pappus morphology in the Berkheya clade
within a phylogenetic context is currently possible.
The phylogenies of Funk and Chan (2008) consistently place both Didelta species in a clade with Berkheya
spinosissima. Phylogenetic analyses of the cpDNA data placed this clade sister to the Gorteria clade, whereas the ITS
and combined data sets placed the clade sister to the remainder of the Berkheya clade. The achenes of these species
differ considerably in shape, pubescence, and pappus morphology, which collectively do not provide support for a
phylogenetic relationship. The most notable inconsistency is that in B. spinosissima the capitulum does not function as
a diaspore and the mature achenes are released from the capitulum to achieve dispersal. Further sampling of relatives
of B. spinosissima (B. chamaepeuce and B. schinzii) and additional phylogenetic analyses are needed to confirm the
validity of this relationship and to resolve the clade’s affinities within the Gorteriinae.
Cuspidia cernua and Heterorhachis aculeata were placed as monotypic lineages within the Berkheya clade by
the cpDNA and ITS data in the study by Funk & Chan (2008). Both data sets implicated C. cernua to be an early
divergence within the clade, whereas H. aculeata was suggested to be an early divergence in the cpDNA phylogeny but
was embedded within the Berkheya clade in the ITS phylogeny. The achene and pappus of both species show unusual
features, of which some, as discussed already, may be indicative of distant affinities with Berkheya species. Therefore
the suggestion of a close relationship with Berkheya species is not unexpected with consideration of achene and pappus
morphology.
The five sampled Cullumia species were embedded within Berkheya in the phylogenies of Funk & Chan (2008).
However, the precise relationships of these taxa were not resolved. Chloroplast sequence data placed the Cullumia
species in a clade with Berkheya cruciata, and sister to a clade consisting of B. canescens, B. fruticosa and B. spinosa.
The ITS sequence data placed all of these taxa in an unresolved polytomy but indicated that Cullumia patula and C.
rigida are more closely related to B. cruciata than to the other sampled Cullumia species. Cullumia species produce
achenes that are morphologically unique within the Berkheya clade. In all Cullumia taxa the achene is cylindric-
obovoid or cylindric-ellipsoid and epappose. The achene surface is glabrous in most taxa, except for C. aculeata
(minutely pilose) and C. patula subsp. uncinata (with twin hairs about half the achene length and attached only in
the basal portion of the achene). The phylogenetic association of B. cruciata with Cullumia species is not supported
by achene and pappus morphology. In B. cruciata the achene is turbinate with a dense covering of twin hairs shorter
than the achene distributed over the entire achene surface. In addition, a uniseriate pappus consisting of 10–18 free
(or sometimes connate; Roessler 1959) scales is present. The glabrous, epappose achene was emphasized to support
the segregation of Cullumia (Aiton 1813). In addition, possession of an elaiosome may be autapomorphic to Cullumia
species within the Berkheya clade. However, the existing phylogenetic evidence suggests that Cullumia should be
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 15
subsumed into Berkheya. Given the presence of twin hairs on the achene surface, it would be interesting to examine
whether C. aculeata and C. patula represent early divergences in the evolution of a Cullumia lineage from within
Berkheya.
Species currently classified in Berkheya exhibit notable diversity in achene and pappus features, but interpretation
in relation to species diversification and evolutionary history is premature. While Funk & Chan (2008) resolved at
least two major clades among Berkheya spp., relationships among these taxa might alter and additional clades might be
resolved with more complete taxonomic coverage. Additional genetic data might also suggest differing relationships.
Therefore discussion of achene and pappus morphology among Berkheya species in an evolutionary context awaits
further phylogenetic analyses with comprehensive sampling of Berkheya species.
Both the Berkheya and Gorteria clades within the subtribe Gorteriinae are notable for the evolution of the capitulum
or capitular segments as a diaspore. Within the Gorteria clade, limited evolution in achene and pappus morphology
is apparent among species of Gorteria and those species formerly placed in Hirpicium (Stångberg 2014; Stångberg
& Anderberg 2014). In the Berkheya clade, taxa showing retention of the achene and dispersal of the capitulum in
segments or as a single unit (Berkheya cuneata, Jan Vlok pers. comm.; Cuspidia cernua and Didelta spp.) show limited
reduction in the size of the pappus scales, at least compared with other taxa in the Berkheya clade, suggesting that
evolution of the capitulum or capitular segments as the diaspore may have been relatively recent evolutionary events.
In these taxa the pappus scales show a tendency towards spinescence, which might be an autapomorphy conferring
additional protection of the achene that promotes retention of the pappus. However, other taxa placed in the same
subclade as Didelta, as well as B. ferox, show similarly or more spinescent pappus scales, so the trait may have pre-
existed before evolution of the capitulum or capitular segments as the diaspore.
The sister subtribes Arctotidinae and Gorteriinae show interesting differences in diaspore dispersal strategies.
Compared with the greater extent of achene evolution in Arctotidinae (McKenzie et al. 2005), achene morphology in
the Berkheya clade, as well as the sister Gorteria clade (Stångberg 2014), appears to have been under weaker selective
pressure and thus more limited achene evolution is evident. However, in both the Berkheya and Gorteria clades
evolution of the entire capitulum or capitular segments as the diaspore, with retention of the individual achenes, has
occurred independently. Such an evolutionary strategy is unknown among extant Arctotidinae. Geocarpy, which has
evolved independently in three Haplocarpha species in the Arctotidinae (Barker 2005), is unknown in Gorteriinae.
Development of wings on the achene, which characterises the achene of Arctotis Linnaeus (1753: 922) (McKenzie et
al. 2005), is unknown elsewhere in the tribe Arctotideae.
Species groupings based on achene and pappus morphology
In the absence of a well-resolved phylogeny with comprehensive taxonomic coverage, the following broad species
groupings are distinguished based on observations in the present study. These groupings could be tested in future
phylogenetic investigations of the Berkheya clade.
GROUP I: Berkheya armata, B. francisci, B. herbacea, B. macrocephala, B. rosulata—achenes turbinate or cylindric-
turbinate, with dense covering of twin hairs over entire surface, shorter or longer than the achene; pappus uniseriate or
biseriate, scales free, monomorphic, 20–30, shorter, ±equal or longer than achene, oblong or lanceolate, lateral margin
entire or serrate, apex acute or acuminate (sometimes obtuse), apex serrate (sometimes lacerate), outer surface smooth
or papillate (Figs. 1B, 2E).
GROUP II: Berkheya angustifolia, B. barbata, B. carlinopsis subsp. magalismontana, B. chamaepeuce, B. coriacea,
B. fruticosa, B. schinzii, B. spinosissima—achene turbinate or obovoid-turbinate, dense covering of twin hairs over
the entire surface, twin hairs ±equal or longer than achene; pappus biseriate, scales free, monomorphic or cryptically
dimorphic, 16–24, lanceolate, oblong or ovate, scale length shorter or ±equal to achene, lateral margin variable (entire,
serrate, spinescent), apex variable (truncate, obtuse, acute, acuminate), apex margin variable (erose, serrate, lacerate,
spinescent), outer surface smooth or papillate (Fig. 1D, L, 2A, B).
SUBGROUP IIA: Berkheya canescens—twin hairs sparse; pappus scales pluriseriate, ca. 40–50, linear-subulate,
longer than achene (Fig. 2C).
GROUP III: Berkheya echinacea, B. rhapontica—achene turbinate, dense covering of twin hairs over the entire surface,
PHALISO ET AL.
16 Phytotaxa 246 (1) © 2016 Magnolia Press
twin hairs shorter than achene; pappus uniseriate or biseriate, scales free, monomorphic, 14–20, shape variable (obovate,
oblong, lanceolate, ovate), scale length shorter than achene, lateral margin variable (entire, serrate, spinescent), apex
variable (truncate, obtuse, acute), apex margin variable (erose, serrate, lacerate), outer surface smooth or papillate (Fig.
2I).
SUBGROUP IIIA: Berkheya setifera, B. umbellata—pappus biseriate, scales monomorphic or cryptically dimorphic,
29–40, linear-subulate, longer than achene, apex acuminate (Fig. 1G, 2J).
GROUP IV: Berkheya cruciata—achene turbinate, with dense covering of twin hairs over entire surface, twin hairs
shorter than achene; pappus uniseriate, markedly shorter than achene, scales free or sometimes connate, monomorphic,
10–18, oblong or obovate, lateral margins entire or serrate, and apex obtuse, apex margin erose or serrate, outer surface
papillate (Fig. 2D).
GROUP V: Berkheya speciosa—achene turbinate, with glandular hairs and sparse twin hairs over entire surface, twin
hairs much shorter than the achene; pappus biseriate, ±equal or longer than achene, scales free or shortly connate at
base, monomorphic, 14–20, obovate, lateral margins serrate, apex truncate, obtuse or acute, apex margin erose or
serrate, outer surface smooth (Fig. 1E, 2F).
GROUP VI: Berkheya coddii, B. densifolia, B. subulata var. wilmsiana, B. zeyheri—achene turbinate or cylindric-
turbinate, with sparse or dense covering of twin hairs over entire surface, twin hairs shorter or longer than achene;
pappus uniseriate or biseriate, shorter to longer than achene, scales free, monomorphic, 11–20, obovate, oblong or
suborbicular, lateral margins entire, serrate or spinescent, apex truncate or obtuse, apex margin variable (erose, serrate,
crenate, lacerate), outer surface smooth or papillate.
GROUP VII: Berkheya buphthalmoides, B. carlinifolia, B. glabrata, B. heterophylla, B. jardineana, B. onobromoides,
B. rigida—achene cylindric-turbinate, with sparse or dense covering of twin hairs over entire surface or in apical
portion of achene, twin hairs much shorter than achene; pappus uniseriate or biseriate, shorter than achene, scales free,
monomorphic, 13–22, obovate or oblong (sometimes ovate), lateral margin entire or serrate, apex obtuse or truncate,
apex margin variable (entire, erose, serrate), outer surface smooth (sometimes papillate) (Fig. 2L).
SUBGROUP VIIA: Berkheya annectens, B. eriobasis, B. pinnatifida—pappus scales lanceolate or oblong (sometimes
ovate), lateral margin serrate (sometimes entire), apex acute (sometimes obtuse or acuminate) (Fig. 1C, H, 2K).
SUBGROUP VIIB: Berkheya bipinnatifida, B. cardopatifolia—achene attenuate-turbinate (Fig. 1I, 3A).
GROUP VIII: Berkheya acanthopoda, B. cf. caffra, B. carduoides, B. cirsiifolia, B. decurrens, B. discolor, B. griquana,
B. latifolia, B. mackenii, B. multijuga, B. onopordifolia, B. pauciflora, B. radula, B. sphaerocephala—achene attenuate-
turbinate, glabrous or with sparse twin hairs in upper portion of achene (sometimes over entire surface), twin hairs
minute, rarely with glandular hairs; pappus uniseriate or biseriate, scales free or connate at base (sometimes more
than half of pappus length), monomorphic or dimorphic, 6–21, commonly 10 (5+5), shorter than achene, oblong or
lanceolate, sometimes obovate, lateral margin entire or serrate, apex truncate, obtuse or acute, apex margin erose or
lacerate (sometimes serrate), outer surface smooth (rarely papillate) (Fig. 1A, N, O, 2H, 3C).
SUBGROUP VIIIA: Berkheya maritima, B. montana, B. purpurea, B. robusta—achene obovoid-turbinate (Fig. 2G,
3D).
GROUP IX: Berkheya cuneata, B. ferox, B. spinosa—achene pyriform, glabrous or with sparse or dense covering
of twin hairs over entire surface, twin hairs shorter or longer than achene; pappus uniseriate or biseriate, scales free,
monomorphic, 16–20, shorter to longer than achene, lanceolate or linear-subulate, lateral margin serrate to spinescent,
apex acute or acuminate, apex margin entire or serrate, outer surface smooth or papillate (Fig. 1K, 3B).
GROUP X: Cuspidia cernua—achene obovoid-pyriform, glandular and dense uniseriate multiseptate hairs over
entire surface, multiseptate hairs shorter than achene; pappus biseriate, scales free, dimorphic, 20 (10+10), shorter
than achene, lanceolate or linear-subulate, lateral margin fimbriate-spinescent, apex acuminate, apex margin erose or
fimbriate-spinescent, outer surface papillate (Fig. 3I).
GROUP XI: Heterorhachis aculeata—achene cylindrical-turbinate, glandular hairs and sparse twin hairs over entire
surface, twin hairs shorter than achene; pappus biseriate, scales free, dimorphic, 14–20, ±equal to achene length,
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 17
lanceolate or linear-subulate, lateral margin fimbriate, apex acute or acuminate, apex margin spinescent-fimbriate,
outer surface papillate and glandular (Fig. 1F, 3J).
GROUP XII: Didelta carnosa—achene pyriform, glabrous; pappus uniseriate, scales free, monomorphic, 10–11,
shorter than achene, lanceolate, lateral margin spinescent, apex acuminate, apex margin spinescent, outer surface
smooth (Fig. 1M, 3K).
GROUP XIII: Didelta spinosa—achene obovoid-unguiculate, sparse twin hairs over entire surface, twin hairs much
shorter than achene; pappus uniseriate, scales free, monomorphic, 14–20, shorter than achene, lanceolate, lateral
margin spinescent, apex acuminate or acute, apex margin spinescent, outer surface smooth (Fig. 3L).
GROUP XIV: Cullumia bisulca, C. decurrens, C. patula subsp. patula, C. reticulata, C. sp., C. sulcata—achene
cylindric-obovoid to cylindric-ellipsoid, glabrous, ribs inconspicuous; pappus absent (Fig. 3E, F).
SUBGROUP XIVA: Cullumia aculeata—sparse twin hairs over entire achene surface, twin hairs much shorter than
achene (Fig. 3H).
SUBGROUP XIVB: Cullumia patula subsp. uncinata—dense deciduous twin hairs only in basal portion of achene,
twin hairs about half achene length (Fig. 3G).
Conclusion
The present study provides provides data on achene and pappus morphology for approximately two-thirds of the
species of the Berkheya clade. Variation in achene shape, pubescence, and pappus form and micromorphology was
recorded, which only partially supports the prevailing taxonomic classification at the generic and infrageneric levels.
Segregation of Cullumia with emphasis on achene characteristics and the absence of a pappus is not supported, however,
by existing molecular data. Achene and pappus morphology generally supports Roessler’s species groupings as series
in Berkheya but with some notable exceptions. In particular, heterogeneity of series Speciosae and misplacement of
a number of species in series Decurrentes and Rigidae are indicated. Detailed reappraisal of other morphological
characters and reconstruction of molecular phylogenies with complete taxonomic coverage are required as part of
the ongoing taxonomic reassessment of the Berkheya clade. Future reconstruction of taxonomically comprehensive
phylogenies also will permit formulation of hypotheses for achene and pappus evolution in the Berkheya clade.
Acknowledgements
This work is based on research supported in part by the National Research Foundation of South Africa (grant numbers
2042600, 2046932 and 2053645 to N.P.B.). We thank the Rhodes University Research Committee for financial support;
the curators of the Selmar Schonland Herbarium, Grahamstown, and the National Herbarium, South African National
Biodiversity Institute, Pretoria, for access to specimens and permission to sample achenes; CapeNature, Northern
Cape Department of Environment and Nature Conservation, Ezemvelo KZN Wildlife, and SANParks for permission
to collect plant specimens; Michael Turner (Electron Microscopy Unit, University of Limpopo, Medunsa campus) and
Shirley Pinchuck and Marvin Randall (Electron Microscopy Unit, Rhodes University) for technical assistance with
electron microscopy; Marinda Koekemoer for generously allowing access to her personal collections; Laco Mucina,
Syd Ramdhani and Hedi Stummer for collection of material.
References
Aiton, W. (1789) Hortus kewensis; or, a catalogue of the plants cultivated in the Royal Botanic Garden at Kew 3. George Nicol, London,
547 pp.
http://dx.doi.org/10.5962/bhl.title.4504
Aiton, W.T. (1813) Hortus kewensis; or, a catalogue of the plants cultivated in the Royal Botanic Garden at Kew by the late William Aiton
(2nd ed.) 5. Longman, Hurst, Rees, Orme, and Brown, London, 568 pp.
PHALISO ET AL.
18 Phytotaxa 246 (1) © 2016 Magnolia Press
Barker, N.P. (2005) A review and survey of basicarpy, geocarpy, and amphicarpy in the African and Madagascan flora. Annals of the
Missouri Botanical Garden 92: 445–462.
Bennett, A. & Krebs, J. (1987) Seed dispersal by ants. Trends in Ecology and Evolution 2: 291–292.
http://dx.doi.org/10.1016/0169-5347(87)90078-4
Bergh, N.G. & Helme, N.A. (2014) A new species of Berkheya (Asteraceae, Arctotideae) from the Northern Cape, South Africa. Bothalia
44(1), Art. #123, 5 pp.
Bolus, H. (1906) Contributions to the African flora. Transactions of the South African Philosophical Society 16: 381–400.
http://dx.doi.org/10.1080/21560382.1905.9526076
Bolus, H. (1907) Contributions to the African flora. Transactions of the South African Philosophical Society 18: 379–399.
Bond, W.J. & Slingsby, P. (1983) Seed dispersal by ants in shrublands of the Cape Province and its evolutionary implications. South
African Journal of Science 79: 231–233.
Breitwieser, I. & Ward, J.M. (2003) Phylogenetic relationships and character evolution in New Zealand and selected Australian Gnaphalieae
(Compositae) inferred from morphological and anatomical data. Botanical Journal of the Linnean Society 141: 183–203.
http://dx.doi.org/10.1046/j.1095-8339.2003.00141.x
Burman, N.L. (1768) Florae capensis prodromus. In: Flora indica: cui accedit series zoophytorum indicorum, nec non prodromus florae
capensis. Cornelium Haek, Lugduni batavorum & Johannem Schreuderum, Amstelaedami, 32 pp.
http://dx.doi.org/10.5962/bhl.title.60581
Burtt, B.L. (1948) Didelta and Cuspidia. Kew Bulletin 1948: 69–73.
http://dx.doi.org/10.2307/4118925
Burtt Davy, J. (1935) New or noteworthy South African plants.—VII. Bulletin of Miscellaneous Information (Royal Gardens, Kew) 1935:
568–572.
http://dx.doi.org/10.2307/4113517
Burtt Davy, J. & Pott-Leendertz, R. (1912) A first check-list of the flowering plants and ferns of the Transvaal and Swaziland. Annals of
the Transvaal Museum 3: 119–182.
Candolle, A.P. de (1838) Prodromus systematis naturalis regni vegetabilis, sive enumeratio contracta ordinum, generum, specierumque
plantarum huc usque cognitarum, juxta methodi naturalis normas digesta 6. Treuttel and Würtz, Paris, 687 pp.
Cassini, H. (1820) Description d’un noveau genre de plantes (Hirpicium), précédée d’observations sur l’Œdera alienata de Thunberg, et
sur l’Œdera aliena de Jacquin. Bulletin des sciences par la Société Philomatique de Paris 1820: 26–27.
Dittrich, M. (1996) Die Bedeutung morphologischer und anatomischer Achänen–Merkmale für die Systematik der Tribus Echinopeae
Cass. und Carlineae Cass. Boissiera 51: 5–102.
Druce, G.C. (1917) Nomenclatural notes: chiefly African and Australian. Report for 1916 of the Botanical Exchange Club (second
supplement): 601–653.
Ehrhart, F. (1784) Berkheya, eine Pflanzengattung. Neues Magazin für Aerzte 6: 303–307.
Ewart, A.J., White, J. & Rees, B. (1909) Contributions to the flora of Australia, no. 11. Proceedings of the Royal Society of Victoria (New
Series) 22: 6–23.
Fourcade, H.G. (1932) Contributions to the flora of the Knysna and neighbouring divisions. Transactions of the Royal Society of South
Africa 21: 75–102.
http://dx.doi.org/10.1080/00359193209518876
Fourcade, H.G. (1941) Check-list of flowering plants of the divisions of George, Knysna, Humansdorp, and Uniondale. Memoirs of the
Botanical Survey of South Africa no. 20. Government Printer, Pretoria, 127 pp.
Funk, V.A. & Chan, R. (2008) Phylogeny of the spiny African daisies (Compositae, tribe Arctotideae, subtribe Gorteriinae) based on trnL-
F, ndhF, and ITS sequence data. Molecular Phylogenetics and Evolution 48: 47–60.
http://dx.doi.org/10.1016/j.ympev.2008.03.035
Funk, V.A., Chan, R. & Keeley, S.C. (2004) Insights into the evolution of the tribe Arctoteae (Compositae: subfamily Cichorioideae s.s)
using trnL-F, ndhF, and ITS. Taxon 53: 637–655.
http://dx.doi.org/10.2307/4135440
Gaertner, J. (1791) Fructibus et seminibus plantarum 2. Typis Guilielmi Henrici Schrammii, Tubingae, 520 pp.
Grau, J. (1980) Die Testa der Mutisieae und ihre systematische Bedeutung. Mitteilungen der Botanischen Staatssammlung München 16:
269–332.
Greuter, W., Aghababian, M. & Wagenitz, G. (2005) Vaillant on Compositae—systematic concepts and nomenclatural impact. Taxon 54:
149–174.
http://dx.doi.org/10.2307/25065315
Harvey, W.H. (1865) Compositae. In: Harvey, W.H. & Sonder, O.W. (Eds.) Flora capensis, being a systematic description of the plants of
the Cape Colony, Caffraria and Port Natal 3. L. Reeve, London, pp. 44–530.
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 19
Herman, P.P.J. (2001) Observations on the hairs in the capitula of some southern African Asteraceae genera. South African Journal of
Botany 67: 65–68.
http://dx.doi.org/10.1016/S0254-6299(15)31093-0
Hess, R. (1938) Vergleichende Untersuchungen über die Zwillingshaare der Kompositen. Botanische Jahrbücher für Systematik 68: 435–
496.
Hilliard, O.M. & Burtt, B.L. (1975) Notes on some plants of southern Africa chiefly from Natal. IV. Notes from the Royal Botanic Garden,
Edinburgh 34: 73–100.
Hilliard, O.M. & Burtt, B.L. (1985) Notes on some plants of southern Africa chiefly from Natal. XI. Notes from the Royal Botanic Garden,
Edinburgh 42: 227–260.
Hilliard, O.M. & Burtt, B.L. (1989) Notes on some plants of southern Africa chiefly from Natal. XV. Notes from the Royal Botanic Garden,
Edinburgh 45: 179–223.
Houttuyn, M. (1776) Natuurlyke historie of uitvoerige beschriving der dieren, plantenen en mineraalen, volgens het samenstel van den
heer Linnaeus 2(6). Houttuyn, Amsterdam, 468 pp.
Houttuyn, M. (1779a) Natuurlyke historie of uitvoerige beschriving der dieren, plantenen en mineraalen, volgens het samenstel van den
heer Linnaeus 2 (11). Houttuyn, Amsterdam, 456 pp.
Houttuyn, M. (1779b) Natuurlyke historie of uitvoerige beschriving der dieren, plantenen en mineraalen, volgens het samenstel van den
heer Linnaeus 2(10). Houttuyn, Amsterdam, 828 pp.
Karis, P.O. (2006) Morphological data indicates two major clades of the subtribe Gorteriinae (Asteraceae–Arctotideae). Cladistics 22:
199–221.
http://dx.doi.org/10.1111/j.1096-0031.2006.00109.x
Karis, P.O. (2007) Tribe Arctotideae Cass. In: Kadereit, J.W. & Jeffrey, C. (Eds.) The families and genera of vascular plants 8. Springer-
Verlag, Berlin, pp. 200–206.
Karis, P.O., Funk, V.A., McKenzie, R.J., Barker, N.P. & Chan, R. (2009) Arctotideae. In: Funk, V.A., Susanna, A., Stuessy, T. & Bayer,
R. (Eds.) Systematics, evolution and biogeography of the Compositae. International Association of Plant Taxonomists, Vienna, pp.
407–432.
Lengyel, S., Gove, A.D., Latimer, A.M., Majer, J.D. & Dunn, R.R. (2010) Convergent evolution of seed dispersal by ants, and phylogeny
and biogeography in flowering plants: a global survey. Perspectives in Plant Ecology, Evolution and Systematics 12: 43–55.
http://dx.doi.org/10.1016/j.ppees.2009.08.001
Lessing, C.F. (1832) Synopsis generum compositarum earumque dispositionis novae tentamen, monographiis multarum capensium
interjectis. Duncker & Humblot, Berlin, 473 pp.
Levyns, M.R. (1942) Some changes in nomenclature. III. Journal of South African Botany 8: 282–284.
L’Héritier de Brutelle, C.-L. (1786) Stirpes novae aut minus cognitae. Philippi-Dionysii Pierres, Parisii, 184 pp.
Linnaeus, C. (1753) Species plantarum exhibentes plantas rite cognitas, ad genera relatas, cum differentiis specificis, nominibus trivialibus,
synonymis selectis, locis natalibus, secundum systema sexuale digestas. Laurentii Salvii, Holmiæ, 1200 pp.
Linnaeus, C. (1759) Systema naturæ (ed. 10). Laurentii Salvii, Holmiae, 1384 pp.
Linnaeus, C. (1782) Supplementum plantarum systematis vegetabilium editionis decimae tertiae, generum plantarum editiones sextae, et
specierum plantarum editionis secundae. Orphanotrophei, Brunsvigæ, 468 pp.
Manning, J.C. & Goldblatt, P. (2012) Berkheya jardineana (Arctotideae–Gorteriinae), a new dwarf perennial from the Swartruggens,
Western Cape. Bothalia 42: 57–59.
Manning, J.C., Karis, P.O., Goldblatt, P. & Helme, N.A. (2010) Berkheya chrysanthemoides and Heterorhachis hystrix, two new species
of Arctotideae–Gorteriinae from the Southwestern Cape. Bothalia 40: 185–190.
http://dx.doi.org/10.4102/abc.v40i2.217
Masters, M.T. (1872) Berkheya (Stobœa) purpurea. The Gardeners’ Chronicle and Agricultural Gazette 1872: 1262.
McKenzie, R.J., Samuel, J. Muller, E.M., Skinner, A.K.W. & Barker, N.P. (2005) Morphology of cypselae in subtribe Arctotidinae
(Compositae–Arctotideae) and its taxonomic implications. Annals of the Missouri Botanical Garden 92: 569–594.
Medley Wood, J. (1897) New Natal plants. Journal of Botany, British and Foreign 35: 350–353.
Medley Wood, J. & Evans, M.S. (1899) New Natal plants. Journal of Botany, British and Foreign 37: 251–255.
Merrill, E.D. (1938) A critical consideration of Houttuyn’s new genera and new species of plants, 1773–1783. Journal of the Arnold
Arboretum 19: 291–375.
Oliver, D. & Hiern, W.P. (1877) Compositae. In: Oliver, D. (Ed.) Flora of tropical Africa 3. L. Reeve & Co., London, pp. 253–461.
Phaliso, N. (2013) A systematic study of Berkheya and allies (Compositae). MSc thesis, Rhodes University, Grahamstown, South Africa,
122 pp.
Probst, R. (1928) Dritter beitrag zur adventivflora von Solothurn und umgebung. Mitteilungen der Naturforschenden Gesellschaft in
Solothurn 8 (20): 40–82.
PHALISO ET AL.
20 Phytotaxa 246 (1) © 2016 Magnolia Press
Roessler, H. (1959) Revision der Arctotideae–Gorteriinae (Compositae). Mitteilungen der Botanische Staatssammlung München 3: 71–
500.
Roessler, H. (1960) Eine neue Berkheya aus Transvaal. Mitteilungen der Botanischen Staatssammlung München 3: 519–520.
Roessler, H. (1973) Nachträge zur Bearbeitung der Arctotideae–Gorteriinae (Compositae). Mitteilungen der Botanischen Staatssammlung
München 11: 91–99.
Schinz, H. (1894) Beiträge zur Kenntnis der Afrikanischen flora (neue Folge). II. Bulletin de l’Herbier Boissier 2: 180–228.
Schinz, H. (1904) Beiträge zur kenntnis der Afrikanischen-Flora. (Neue Folge.) XVI. Bulletin de l’Herbier Boissier, séries 2 4: 995–
1025.
Schlechter, R. (1897) Decades plantarum novarum austro-africanarum. Decas V. Journal of Botany, British and Foreign 35: 340–345.
Schlechter, R. (1899) Plantae Schlechterianae novae vel minus cognitae describuntur. II. Botanische Jahrbücher für Systematik,
Pflanzengeschichte und Pflanzengeographie 27: 86–220.
Schultz Bipontinus, C.H. (1844) Enumeratio Compositarum a cl. Dr. Krauss annis 1838—40 in Capite bonae spei et ad portum Natalensem
lectarum. Flora 27: 767–783.
Stångberg, F. (2014) Phylogenetics and taxonomy of the Gorteria clade (Asteraceae-Arctotideae). PhD thesis, Stockholm University,
Sweden, 34 pp.
Stångberg, F. & Anderberg, A.A. (2014) Morphology and taxonomic reclassification of Gorteria (Asteraceae). Willdenowia 44: 97–120.
http://dx.doi.org/10.3372/wi.44.44112
Tadesse, M., Crawford, D.J. & Smith, E.B. (1995) Comparative capitular morphology and anatomy of Coreopsis L. and Bidens L.
(Compositae), including a review of generic boundaries. Brittonia 47: 61–91.
http://dx.doi.org/10.2307/2807249
Thunberg, C.P. (1793) Afhandling og Beskrivelse over nogle tilforn ukiendte Arter af Rohria. Skrivter af Naturhistorie-Selskabet 3 (1):
97–109.
Thunberg, C.P. (1798) Nova genera plantarum. J. Edman, Upsaliæ, 194 pp.
Thunberg, C.P. (1800) Prodromus plantarum capensium: quas in promontorio bonæ spei Africes, annis 1772–1775, collegit Carol. Pet.
Thunberg 2. J. Edman, Upsaliae, 191 pp.
Thunberg, C.P. (1823) Flora capensis, sistens plantas promontorii bonæ spei Africes, secundum systema sexuale emendatum, redactas ad
classes, ordines, genera et species, cum differentiis specificis, synonymis et descriptionibus. J.G. Cottae, Stuttgardtiae, 803 pp.
Vahl, M. (1791) Om Perdicium og dens arter, og om en ny slaegt Rohria, henhørende til Compositas. Skrivter af Naturhistorie-Selskabet
1 (2): 1–17.
Vahl, M. (1793) Tillæg til de tvende plante-slægter, Perdicium og Rohria. Skrivter af Naturhistorie-Selskabet 2 (2): 31–41.
Vincent, P.L.D. & Wilson, S.L. (1997) The systematic value of the surface micromorphology and anatomy of cypselae of some members
of the Senecioneae, Liabeae and Vernonieae (Asteraceae). South African Journal of Botany 63: 382–399.
http://dx.doi.org/10.1016/S0254-6299(15)30790-0
Walpers, G.G. (1847) Repertorium botanices systematicae 6. Sumtibus Friderici Hofmeister, Lipsiae, 834 pp.
Willdenow, C.L. (1803) Caroli a Linné species plantarum exhibentes plantas rite cognitas ad genera relatas cum differentiis specificis,
nominibus trivialibus, synonymis selectis, locis natalibus secundum systema sexuale digestas (ed. 4) 3. G.C. Nauk, Berolini, 1474
pp.
Zahlbruckner, A. (1910) Plantae Pentherianae: Aufzählung der von Dr. A. Penther und in seinem Auftrage von P. Krook in Südafrika
gesammelten Pflanzen, pars IV. Annales des K.K. Naturhistorischen Museums in Wien 24: 293–326.
Zhang, J.W., Boufford, D.E. & Sun, H. (2013) Systematic significance of achene morphology in Soroseris, Syncalathium and
Parasyncalathium (Asteraceae: Cichorieae). Botanical Journal of the Linnean Society 173: 476–486.
http://dx.doi.org/10.1111/boj.12046
TAXONOMIC RELATIONSHIPS IN THE BERKHEYA CLADE Phytotaxa 246 (1) © 2016 Magnolia Press 21
Appendix I. Collection and voucher information for the specimens examined.
All vouchers are deposited in the Selmar Schonland Herbarium, Grahamstown, unless indicated otherwise. NBG,
Compton Herbarium, Cape Town; PRE, National Herbarium, Pretoria.
* Achene immature.
Berkheya acanthopoda (DC.) Roessler: N.P. Barker s.n., 2005; R.G. Strey 10807*.
B. angustifolia (Houtt.) Merr.: H.H. Burrows 3955; E.C. Feyt s.n., 10 Apr 1991; R.J. McKenzie 3047; Taylor &
Edwards 8846*.
B. annectens Harv.: J.P.H. Acocks 13193* (PRE); R.O. Peter 01 (PRE).
B. armata (Vahl) Druce: R.J. McKenzie 2568*, 3145*.
B. barbata (L.f.) Hutch.: R.J. McKenzie 2591*, 3121*; sine coll., 1899.
B. bipinnatifida subsp. bipinnatifida: R.J. McKenzie 1839*.
B. buphthalmoides (DC.) Schltr.: A. Dieterlen 1280* (PRE); H.G. Flanagan 1604* (PRE); N. Phaliso 89*; J.E. Victor
1445* (PRE).
B. cf. caffra MacOwan: L. Mucina 200207/1.
B. canescens DC.: C. Evrard 8856 (PRE); L. Mucina 290806/32*; R. Schlechter 11335*.
B. cardopatifolia (DC.) Roessler: Clark, Barker, Ramdhani & Kelly 41*; Clark & Pienaar 460*; Clark, Pienaar &
Daniels 672*; Clark & Te Water Naude 365*.
B. carduoides (Less.) Hutch.: N.P. Barker 1924*; M. Koekemoer 4144 (PRE); R.J. McKenzie s.n., 5 Mar 2009.
B. carlinifolia (DC.) Roessler: Clark and Midgley 302; Clark & O’Connor 358; Clark, Pienaar and Daniels 653; R.J.
McKenzie 2762*, 3043, 3139*, 3140, 3142, 3143*, 3153; L. Mucina 211006/29.
B. carlinopsis subsp. magalismontana (Bolus) Roessler: R. Leendertz 3714*; N. Phaliso 64*, 66*.
B. chamaepeuce (S.Moore) Roessler: J.P.H. Acocks 18011 (PRE); B. de Winter 3282 (PRE); H. Merxmüller and W.
Giess 28130 (PRE).
B. cirsiifolia (DC.) Roessler: J. Fearon 09-005*; Jacot-Guillarmod, Getliffe & Mzamane 245; R.A. Lubke 6220*; R.J.
McKenzie 973; N. Phaliso 88.
B. coddii Roessler: L.E. Codd 10338 (PRE).
B. coriacea Harv.: J. Muir 1462; R.J. McKenzie 3121.
B. cruciata (Houtt.) Willd. subsp. cruciata: J.P.H. Acocks 18291 (PRE); R.J. McKenzie 2600.
B. cruciata subsp. subintegra Roessler: H.C. Taylor 4714 (PRE).
B. cuneata (Thunb.) Willd.: J.J. Meyer 1759 (PRE).
B. decurrens (Thunb.) Willd.: E.E.A. Archibald 5042; P. Brown 00/05/10/PB/219; E. Brink 219; G.E. Gibbs Russell
3674*; R.J. McKenzie 1189, 1838; A.A. Merti 1045*; P.T. Robertshaw s.n., 20 Feb 1979*.
B. densifolia Bohnen ex Roessler: R.J. McKenzie 2051*.
B. discolor (DC.) O.Hoffm. & Muschl.: R.J. McKenzie 998; N. Phaliso 22, 28.
B. echinacea (Harv.) O.Hoffm. ex Burtt Davy subsp. echinacea: N. Phaliso 63*; H.J. Schlieben 9495* (PRE).
B. eriobasis (DC.) Roessler: J.P.H. Acocks 17388* (PRE); E.E. Galpin 11160* (PRE); R.J. McKenzie 2772*, 3016*, 3019*.
B. ferox O.Hoffm. var. ferox: L.E. Codd 5898 (PRE).
B. francisci Bolus: R.J. McKenzie 2599; E. Pienaar T707.
B. fruticosa (L.) Ehrh.: F.W., S.K. & R.W. Gess 95/96/220; S.K., F.W. & D.W. Gess 89/90/14, 89/90/68; Kolbe s.n.,
Sep 1917*; R.J. McKenzie 1301*, 2779*.
B. glabrata (Thunb.) Fourcade: M. Koekemoer 2046 (PRE).
B. griquana: N. Phaliso 27.
B. herbacea (L.f.) Druce: R.J. McKenzie 1176.
B. heterophylla (Thunb.) O.Hoffm. var. heterophylla: K.Å. Dalstrand 2513; T. Dold s.n., 19 Oct 2011*; P. Philipson
5572; S.C. Troughton 209.
B. heterophylla var. radiata (DC.) Roessler: E. Brink 220, 316; L. Britten s.n., 10 Nov 1919; F.R. Long 26; R.J.
McKenzie 1578, 2559, 2895, 3001, 3002, 3028*, 3031, 3033, 3040, 3041, 3112*, 3136; S. Schonland 840, 3469, 3652;
W. Tyson 56.
B. jardineana J.C.Manning & Goldblatt: R.J. McKenzie 3042.
B. latifolia J.M.Wood & M.S.Evans: K.E. Codd 7880 (PRE); G.B. Deall 1980 (PRE); M. Koekemoer 2223* (PRE);
S. Kurzweg K1103 (PRE).
B. sp. cf. latifolia J.M.Wood & M.S.Evans: N. Phaliso 56, 57.
B. mackenii (Harv.) Roessler: J. Medley Wood 5310 (PRE); R.J. McKenzie 2043 & N. Phaliso*.
PHALISO ET AL.
22 Phytotaxa 246 (1) © 2016 Magnolia Press
B. macrocephala J.M.Wood: O. West 1387* (PRE).
B. maritima J.M.Wood & M.S.Evans: W. Tyson 2740.
B. montana J.M.Wood & M.S.Evans: J.Medley Wood 6958*; N. Phaliso 34*, 41.
B. onobromoides (DC.) O.Hoffm. & Muschl. var. onobromoides: J.P.H. Acocks 17358* (PRE); M.M. Page s.n., Oct
1921* (PRE);
B. onopordifolia var. glabra Bohnen ex Roessler: N. Phaliso 40.
B. onopordifolia (DC.) O.Hoffm. ex Burtt Davy var. onopordifolia: N. Phaliso 83.
B. pauciflora Roessler: D. McMurtry 12633* (PRE); N. Phaliso 65.
B. pinnatifida (Thunb.) Thell. subsp. pinnatifida: J.P.H. Acocks 1502* (PRE); Baaschers and Scheepers 2* (PRE);
C.A. Smith 908* (PRE).
B. pinnatifida subsp. ingrata (Bolus) Roessler: N. Phaliso 72.
B. pinnatifida subsp. stobaeoides (Harv.) Roessler: L. Henderson 787 (PRE); N. Phaliso 23.
B. purpurea (DC.) Masters: J. Fearon s.n., 4 Apr 2009; D.J.B. Killick & J. Vahrmeijer 3793* (PRE); M. Koekemoer
3448 (PRE); N. Phaliso 24, 90; J.P. Roux 1385 (PRE); R. Story 3735 (PRE).
B. radula (Harv.) De Wild.: O.J. Hansen 3410 (PRE); M. Koekemoer 3090*, 4169 (PRE); R.J. McKenzie 1151*; J.
Woollard 256 (PRE).
B. rhapontica subsp. aristosa (DC.) Roessler var. aristosa: R.J. McKenzie 2084*, N. Phaliso 31*.
B. rhapontica (DC.) Hutch. & Burtt Davy subsp. rhapontica: V.A. Funk 12422 & M. Koekemoer (PRE).
B. rigida (Thunb.) Bolus & Wolley-Dod ex Ewart, Jean White & B.Rees: R.H. Compton 164, 15399, 16659, 19025,
21852 (NBG); S.M. Johnson 9 (NBG); R.J. McKenzie 1805*, 2501*, 2517*; L. Mucina 281203/6.
B. robusta Bohnen ex Roessler: N. Phaliso 62; F.A. Rogers 5195 (PRE).
B. rosulata Roessler: M. Schmitz 6901 (PRE); Clark & Keevey 10.
B. schinzii O.Hoffm. ex Hutch.: F.W. & S.K. Gess 03/04/78, 05/06/75.
B. setifera DC.: N. Phaliso 37; F.A. Rogers s.n., 29 May 1905.
B. speciosa subsp. lanceolata Roessler: J.P.H. Acocks 11245* (PRE); P.M. Burgoyne 966 (PRE); G. Germishuizen
3883 (PRE); R.G.M. Young 26596 (PRE).
B. speciosa (DC.) O.Hoffm. subsp. speciosa: J.P.H. Acocks 13079*; C.L. Bredenkamp 1350*; E. Cloete 1059*; K.Å.
Dahlstrand 1817*; E.E. Galpin 14711* (PRE); L. Mucina 010303/3*; N. Phaliso 32; H.G. Schweickert 826 (PRE); R.
Story 4136* (PRE); R.G. Strey 5946* (PRE); T. Martin, L. De Wett and R.A. Lubke 17 (PRE); Watt and Brandwyk
1257 (PRE).
B. sphaerocephala (DC.) Roessler: N. Phaliso 26.
B. spinosa (L.f.) Druce: J.P.H. Acocks 14567* (PRE); P. Bohnen 8512*; H.H. Burrows 2933; M. Koekemoer 1457
(PRE); R. Marloth 13140* (PRE); L.W. Powrie 889*; B. Sachse 605* (PRE).
B. spinosissima (Thunb.) Willd. subsp. spinosissima: F.W. & S.K. Gess 05/06/67.
B. subulata var. wilmsiana Roessler: R.J. McKenzie 2046 & N. Phaliso.
B. umbellata DC.: H.G. Flanagan 1134*; L. Mucina 081103/2.
B. zeyheri subsp. rehmannii (Thell.) Roessler var. rehmannii: N. Phaliso 70.
B. zeyheri (Sond. & Harv.) Oliv. & Hiern subsp. zeyheri: N. Phaliso 67.
Cullumia aculeata (Houtt.) Roessler var. aculeata: R.J. McKenzie 2574*.
C. bisulca (Thunb.) Less.: J.P.H. Acocks 23703 (PRE); M. Koekemoer 2604 (PRE).
C. decurrens Less.: L. Hugo 2054 (PRE).
C. patula (Thunb.) Less. subsp. patula: H.G. Fourcade 2491* (PRE).
C. patula subsp. uncinata Roessler: J.P.H. Acocks 20544 (PRE).
C. reticulata (L.) Greuter, M.V.Agab. & Wagenitz subsp. reticulata: L. Mucina 021106/05.
C. sulcata (Thunb.) Less. var. sulcata: P. Goldblatt 5909 (PRE).
C. sp.: R.J. McKenzie 3038 (Gouritzmond).
Cuspidia cernua subsp. annua (Less.) Roessler: E.E.A. Gledhill s.n., 18 Nov 1983; R.J. McKenzie 2606*; L. Mucina
261105/3*.
Didelta carnosa (L.f.) Aiton var. carnosa: B. de Winter 3564* (PRE); M. Koekemoer 2479* (PRE); L. Mucina
150704/11*; I. Örtendahl 167* (PRE); L.M. Raitt 241* (PRE); E. van Hoepen 1901* (PRE).
D. carnosa var. tomentosa (Less.) Roessler: R. Marloth 4729*; H. Stummer 1869.
D. spinosa (L.f.) Aiton: H.H. Burrows 2936*; E.E. Galpin 9638 (PRE); G. Germishuizen 4536* (PRE); M. Koekemoer
1936 and V.A. Funk* (PRE); R.J. McKenzie s.n., 14.ix.2013; M.A. Pocock 705 (PRE); H.J.T. Venter 8146* (PRE);
I.A. Walters 43 (PRE).
Heterorhachis aculeata (Burm.f.) Roessler: R.J. McKenzie 2950*, 3014; E.G.H. Oliver 3867* (PRE).
... Indeed, morphological characters are one of the practical taxonomic sources to discriminate groups [14], trace taxonomic affinity of the species [15] and the approach to using morphology in taxonomic affinity is very important because it can cluster species in their taxon correctly [16]. A total of 67 species of Berkheya (family Asteraceae) were grouped into 14 groups based on the morphology of the achene [17]. ...
Article
Full-text available
Morphological characters are important characters in determining plant relationships. This study aims to analyze the phenetic relationship of the 7 member of Asteraceae family based on morphological characters. The study was conducted in May – July 2022. We used 7 species of Asteraceae including Ageratum conyzoides L., Ageratum houstonianum Mill, Tithonia diversifolia (Hemsl.) A.Gray., Sonchus arvensis L., Erechtites valerianifolia (Link ex Spreng.) DC., Eupatorium odoratum L., and Tagetes erecta L. The equipment used were cameras, plant description books, and NTSYS software for taxonomic affinity analysis were used in this study. The results of the Asteraceae dendogram analysis showed that 2 clusters were formed, the first cluster containing the species Ageratum conyzoides L., Ageratum houstonianum Mill, Tagetes erecta L., Sonchus arvensis L., and Erechtites valerianifolia (Link ex Spreng.) DC., while the second cluster consisted of Tithonia diversifolia (Hemsl.) and Eupatorium odoratum L. The closest similarity coefficient of 0.8 was in the genus Ageratus in the first cluster and Tithonia diversifolia (Hemsl.) and Eupatorium odoratum L., in the second cluster. Based on this, it can be concluded that the morphological characters that influence the clustering of 7 species Asteraceae in this study are flower type, flower color, leaf base, leaf venation, leaf margin, and leaf color.
Article
Full-text available
A review of amphicarpy, basicarpy and geocarpy is provided, and the definitions of these terms are clarified. Additional distinction is made between active and passive geocarpy. The literature on amphicarpy, basicarpy, and geocarpy is surveyed for reports of taxa displaying these trails and a survey detailing basically, amphicarpy, and geocarpy, in the African and Madagascan flora is presented. Amphicarpy, basicarpy, or geocarpy is reported for 21 families and 40 genera of angiosperms. Interestingly, a number of holoparasites are basicarpic or geocarpic, but other than a predominance of species that occur in habitats with unstable substrates, (arid or afroalpine), there appears to be little or no trend among the species listed that would enable the construction of a "profile" of features that would suggest that species or genera are predisposed toward being amphicarpic or geocarpic.
Article
Full-text available
Septate hairs are recorded and described for the first time from the capitula of various southern African asteraceous genera. Such hairs are present on the paleae of Eriocephalus L., the cypselas of Tarchonanthus L., Lasiospermum Lag., Arctotis L. and Arctotheca JC Wendl. and the receptacle of Lidbeckia PJ Bergius. The basal coma of hairs on the cypselas of Ursinia Gaertn. subgenus Sphenogyne (R.Br.) Prassler also consists of septate hairs. The hairs in Eriocephalus, Tarchonanthus, Lasiospermum, Arctotis and Arctotheca have oblique cell walls while in Lidbeckia the hairs have perpendicular cell walls. In Ursinia subgenus Sphenogyne the septate hairs also have oblique cell walls and, in addition, spiral wall thickenings. Multicellular, biseriate hairs mixed with some septate hairs were observed in Arctotheca. The basal coma of the cypselas of Arctotis and Haplocarpha Less. consists of twin hairs. Septate hairs can be considered a generic characteristic of the genera Tarchonanthus and Arctotheca and possibly of the genera Eriocephalus, Lasiospermom, Lidbeckia and Ursinia subgenus Sphenogyne.
Book
Trained as a gardener in his native Scotland, William Aiton (1731–93) had worked in the Chelsea Physic Garden prior to coming to Kew in 1759. He met Joseph Banks in 1764, and the pair worked together to develop the scientific and horticultural status of the gardens. Aiton had become superintendent of the entire Kew estate by 1783. This important three-volume work, first published in 1789, took as its starting point the plant catalogue begun in 1773. In its compilation, Aiton was greatly assisted with the identification and scientific description of species, according to the Linnaean system, by the botanists Daniel Solander and Jonas Dryander (the latter contributed most of the third volume). Aiton added dates of introduction and horticultural information. An important historical resource, it covers some 5,600 species and features a selection of engravings. Volume 2 continues to catalogue the plants, covering Octandria to Monadelphia.
Book
This seminal publication began life as a collaborative effort between the Irish botanist William Henry Harvey (1811–66) and his German counterpart Otto Wilhelm Sonder (1812–81). Relying on many contributors of specimens and descriptions from colonial South Africa - and building on the foundations laid by Carl Peter Thunberg, whose Flora Capensis (1823) is also reissued in this series - they published the first three volumes between 1860 and 1865. These were reprinted unchanged in 1894, and from 1896 the project was supervised by William Thiselton-Dyer (1843–1928), director of the Royal Botanic Gardens at Kew. A final supplement appeared in 1933. Reissued now in ten parts, this significant reference work catalogues more than 11,500 species of plant found in South Africa. Containing orders of the Calyciflorae with a monopetalous corolla and an inferior ovary, Volume 3 covers Rubiaceae to Campanulaceae.
Article
Twenty one characters of the surface micromorphology and 12 characters of the anatomy of cypselae of species in the tribes Liabeae (Liabum discolor, L. ovatum), Senecioneae (subtribe Senecioninae: Senecio inaequidens, S. madagascariensis, S. vulgaris, Brachyglottis repanda, subtribe Blennospermatinae: Abrotanella emarginata, A. scapigera, Blennosperma nanum) and Vernoneae (Vernonia oligocephala, V. pauciflora, V. poskeana) were investigated using light microscopy (LM) and scanning electron microscopy (SEM). The data were analysed phenetically with the assistance of cluster (SAHN) and principal component analyses. Both sources of data were biologically very informative. Taxonomically, the SEM data was most useful in recognising discontinuities between the taxa followed by the combined SEM and LM data sets. The placement of the Blennospermatinae in the Senecioneae is confirmed as being systematically discordant.