Content uploaded by Nigel P Barker
Author content
All content in this area was uploaded by Nigel P Barker on Apr 14, 2016
Content may be subject to copyright.
A
NN
.M
ISSOURI
B
OT
.G
ARD
. 92: 445–462. P
UBLISHED ON
28 D
ECEMBER
2005.
Volume 92
Number 4
2005
Annals
of the
Missouri
Botanical
Garden
A REVIEW AND SURVEY OF
BASICARPY, GEOCARPY,
AND AMPHICARPY IN THE
AFRICAN AND
MADAGASCAN FLORA
1
Nigel P. Barker
2
A
BSTRACT
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 traits, and a survey detailing basicarpy, 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.
Key words: African flora, amphicarpy, basicarpy, geocarpy, seed dispersal.
Geocarpy was defined by van der Pijl (1982: 94)
as ‘‘the burying near the mother plant of all dia-
spores . . . ,’’ a definition taken in turn from the
work of Zohary (1962). As such, this mechanism
results in seeds being dispersed short distances
1
I am deeply indebted to Henry Connor for extensive and constructive review of this manuscript, and for drawing
my attention to additional examples of plants with unusual reproductive strategies, as well as additional terminology.
Further advice and suggestions from the editor and an anonymous reviewer are also acknowledged. I would also like
to thank the following for their discussions and suggestions on amphicarpic and geocarpic African taxa: Sue Edwards
(National Herbarium, Ethiopia), John Manning (Compton Herbarium, Kirstenbosch), Graham Rowe (University of Cape
Town), Brian Schrire (Royal Botanic Gardens, Kew), Yashica Singh (Natal Herbarium, Durban), Dee Snijman (Compton
Herbarium, Kirstenbosch), Marc Sosef (Wageningen University, The Netherlands), Simon van Noort (Iziko Museums,
Cape Town). Anjanette Haller-Barker and Robert McKenzie are thanked for their careful proofreading of the manuscript.
I would like to acknowledge the Rhodes University Joint Research Council for providing funding for me to attend the
XIIIth AETFAT meeting in Addis Ababa, Ethiopia, and to participate in the post-congress tour to the Bale Mountain
region, where I first observed the geocarpic Haplocarpha schimperi, which stimulated this investigation.
2
Molecular Ecology and Systematics Group, Department of Botany, Rhodes University, Grahamstown, 6140, South
Africa. n.barker@ru.ac.za.
from the parent plant. While this may be construed
as a method to ensure that the next generation have
a suitable habitat in which to germinate, it also
means that gene flow via the seed is extremely lim-
ited. This was considered by van der Pijl to be a
446 Annals of the
Missouri Botanical Garden
defensive strategy, and geocarpy was thus viewed
as one of a number of mechanisms that result in
atelechory, defined by van der Pijl (1982: 91) as
‘‘. . . the avoidance of too much of any dispersal
and the inhibitory mechanisms for obtaining this.’’
There appears to be some confusion in the lit-
erature differentiating what van der Pijl (1982) re-
ferred to as full geocarpy (where flowers, and sub-
sequently fruits, are formed underground) and a
second type of geocarpy (where the flowers are
borne above ground, but all the fruits subsequently
develop below the soil surface). Note, however, that
the definition of geocarpy given in the opening sen-
tence relies on the fact that all fruits must develop
underground, and thus the situations where plants
bear both aerial and subterranean fruits are not
covered by this definition. Instead, this latter strat-
egy is generally referred to as amphicarpy (Che-
plick, 1987, and references therein). Underground
flowering is perhaps better referred to as geoflory
and is rarely known, especially in the absence of
the production of aerial flowers as well. Somewhat
confusingly, Kaul et al. (2000: 40) used the term
amphicarpic to refer to plants that bear fruits from
flowers ‘‘which are ab initio underground’’ (with no
mention of the co-existence of aboveground flow-
ers), a different use of the term from van der Pijl
(1982).
Amphicarpy was defined by Cheplick (1987) as
plants in which at least some fruits are produced
below the soil surface on specialized structures, in
addition to aerial fruits on the same plant. This
definition, if strictly applied, emphasizes where the
seed is produced. As noted by Cheplick (1994:
119), the term amphicarpic really means ‘‘having
two types of seed.’’ Amphicarpic species sensu
Cheplick thus produce aboveground and below-
ground (geoflorous) flowers. However, some species
that have been considered amphicarpic bury some
of their developing fruits following aerial flowering,
a completely different process, but with the same
result: subterranean and aerial seed set. Geocarpy
may thus occur as a component of amphicarpy. Am-
phicarpic plants have only some of their seed un-
derground, while other seeds develop and disperse
in the normal fashion above ground. Amphicarpy
appears to be usually found in plants from arid re-
gions that occupy a niche that is ephemeral or sus-
ceptible to stochastic disturbances (Manda´k, 1997,
and references therein).
Amphicarpy is also often associated with cleis-
togamy (self-fertilization in an enclosed flower), es-
pecially of the basal or underground flowers, and
heterodiaspory. The latter occurs as either or both
heterocarpy (multiple forms of fruit) and heteros-
permy (multiple forms of seed, also termed seed
heteromorphism by Imbert, 2002). To further com-
plicate matters, seeds (either homospermic or het-
erospermic) from the same plant may be a mix of
tachysporous (fast germinating) or bradysporous
(slow germinating).
One of the more complex combinations of these
strategies is found in the amphicarpic daisy Catan-
anche lutea L. This species has both aerial and
basal capitula. The aerial capitula are the ‘‘normal’’
daisy type, while the basal capitula are produced
and flower above the soil surface, set seed, and are
subsequently retracted 5–20 mm below the soil sur-
face by contractile roots. Aerial capitula produce
three different types of cypselae, and the subter-
ranean capitula two kinds. The two different sub-
terranean cypselae possess differing germination
rates (De Clavijo, 1995). All these strategies pro-
vide a range of options for seeds derived from a
single plant to survive and to reproduce in sto-
chastic environments (van der Pijl, 1982). Manda´k
(1997) provided a detailed review of seed hetero-
morphism.
As might be apparent from the above, one prob-
lem encountered in researching this subject is that
the terminology available appears to be used in-
consistently. Some examples fail to draw the dis-
tinction between species that show both geoflory
and subsequent geocarpy (full geocarpy sensu van
der Pijl, which is almost always found in amphi-
carpic species) and species that have only above-
ground flowers, but which, through specialization
and growth of certain structures, bury the devel-
oping seed. Added to this distinction, the term bas-
icarpy has been used for those plants that flower
and fruit at ground level (van der Pijl, 1982). This
term could, however, apply to all plants that are
prostrate or trail along the ground, because it fol-
lows that all flowers will be produced at or close to
ground level, rendering this definition somewhat ir-
relevant. While basicarpy is a meaningful term, it
needs to be carefully defined to obviate the auto-
matic inclusion of plants such as those that are
prostrate. In an effort to clarify the confusion sur-
rounding the use of various terms, a summary of
these existing and newly created terms is presented
in Table 1, and is explained (and used in subse-
quent examples) below.
I use the term basicarpy exclusively for those
species that have flowers (including the ovary) at
or slightly above ground level at all times during
the flowering and fruiting cycle. I also use this term
to further subdivide the variably applied term am-
phicarpy. As noted above, if the definition of am-
phicarpy is to be strictly applied, then geoflory
Volume 92, Number 4
2005 447Barker
Basicarpy, Geocarpy & Amphicarpy
Table 1. Terminology to distinguish the different subtypes of amphicarpy and geocarpy.
Type Subtype Terminology
Subterranean fruiting from subterranean flowers full geocarpy sensu van
der Pijl (1982)
Subterranean fruiting from aboveground flowers aboveground flowering, subse-
quent burial by plant
active geocarpy
aboveground flowering, subse-
quent burial by environ-
ment
passive geocarpy
aboveground flowering, with ova-
ry buried below ground, of-
ten associated with under-
ground storage organ
geophytic geocarpy
Fruiting at ground level basicarpy
Two different floral and fruiting structures, one
chasmogamous and aerial, the other often cleis-
togamous, usually at or below ground level
flowers and fruit produced
above ground level
aerial amphicarpy
flowers and fruit produced above
and below ground level
amphi-geocarpy
flowers and fruit produced above
and at ground level
amphi-basicarpy
must occur. Amphicarpic species, at least as inter-
preted by some workers such as Stopp (1958), com-
prise plants with two kinds of flowers: aerial and
basal (as opposed to subterranean). Often, the basal
flower is cleistogamous. Clearly, this syndrome of
aerial and basal flowers does not fit the definition
of amphicarpy sensu Cheplick (1987). I thus refer
to this kind of amphicarpy as amphi-basicarpy, in
order to distinguish it from amphicarpy as defined
by Cheplick (1987), which requires aerial and un-
derground flowers and fruits. There is an additional
variant of this, where the hidden (cleistogamous)
flowers are not basal, but produced elsewhere on
the plant. This is known to occur in some grasses,
where cleistogamous flowers (cleistogenes) are hid-
den in the leaf sheaths associated with the aerial
culms or aerial nodes (as opposed to a basal posi-
tion). This form of seed production is known as
cleistocarpy. This has been well studied in the aptly
named grass Amphicarpum purshii Kunth (see Che-
plick, 1994, for a summary). Other examples of
grasses with this reproductive strategy include Mi-
crolaena R. Br. (notably M. polynoda Hook. f. and
M. stipoides R. Br.; Connor & Matthews, 1977; Wat-
son & Dallwitz, 1992), Achnatherum Beauv. (in-
cluding the weedy A. caudatum (Trin.) S. W. L. Ja-
cobs & J. Everett (Caro & Sanchez, 1971; Watson
& Dallwitz, 1992), and some species of Nassella E.
Desv. and Stipa L. (Watson & Dallwitz, 1992). I use
the term aerial amphicarpy for this variant. Finally,
the category of amphicarpic plants where aerial and
subterranean fruits are produced, but where the
subterranean fruits originate from aerial flowers, is
termed here amphi-geocarpy.
The term geocarpy needs to be further refined,
and I have subdivided this category, deliberately
distinguishing between active, passive, and geo-
phytic geocarpy, as some of the records of geocarpy
given in this paper merit this distinction. In active
geocarpy the plant physically responds in an active
manner (by means of some form of growth or tro-
pism, usually of the peduncle) to bury the seed.
The classic example of this form of geocarpy is the
peanut, Arachis hypogaea L., but Kaul et al. (2000)
also cited Trifolium subterraneum L., Vigna subter-
ranea (L.) Verdc. (
5
Voandzeia subterranea (L.)
Thouars) and Macrotyloma geocarpum (Harms) Ma-
re´ chal & Baudet (
5
Kerstingiella geocarpa Harms)
as additional examples. Interestingly, all these taxa
are of the family Leguminosae. In the Compositae,
the amphicarpic Catananche lutea (discussed
above) also shows active geocarpy in that the basal
capitula are retracted underground after flowering.
In contrast, I suggest the term passive geocarpy to
include those species in which seeds are not dis-
persed from the parent but that are nonetheless
buried while still attached to the parent plant. This
burial is, however, not a result of the plant’s actions,
but of the actions of the environment: deposition of
soil over the plant (or the infructescence) by the
action of wind or water.
Another type of geocarpy is that in which the
flower is produced above the ground, but the ovary
and subsequent fruit remain below ground level at
448 Annals of the
Missouri Botanical Garden
all times. I propose the term geophytic geocarpy for
plants that fit into this category, as many of the
species that adopt this strategy are geophytes. This
type of subterranean fruit retention is linked to the
retention of the ovary underground, as described in
several monocotyledons such as Colchicum L., Cro-
cus L., and Crinum L. (Burtt, 1970). However, in
the majority of the examples cited by Burtt, the
fruits are subsequently elevated above the soil sur-
face when ripe so as to ensure seed dispersal. The
retention of the ovary underground might be a strat-
egy to protect the developing seeds for as long as
possible prior to their dispersal (Burtt, 1970). The
geophytic geocarpy category as employed here is,
however, not restricted to geophytes.
T
HE
I
NCIDENCE OF
A
MPHICARPY AND
G
EOCARPY
This phenomenon is not widely reported, and the
numbers of known amphicarpic and geocarpic spe-
cies are low. Cheplick (1987) noted that amphicar-
py was found in 26 species from nine plant fami-
lies, and most frequently in the Leguminosae and
Poaceae. Prior to this, Zohary (1937) reported ap-
proximately 30 species of amphicarpic plants, and
this work highlighted the abundance of these spe-
cies from Palestine, and their association with de-
sert environments. Van der Pijl (1982) made ref-
erence to a number of taxa, but no clear number of
amphicarpic or geocarpic taxa was offered. Kaul et
al. (2000) cited 36 plant species that bear under-
ground, cleistogamous flowers. Lev-Yadun (2000:
289) expanded this number to ‘‘. . . about 50 am-
phicarpic species and 20 geocarpic species . . .’’
The term amphicarpic as used in this latter paper
refers to plants with both aerial and subterranean
flowers, which is what I term amphi-geocarpic. In
Israel, Lev-Yadun (2000) recorded eight amphicar-
pic species and eight geocarpic species out of a
total flora of approximately 2500 species. The flora
of Israel thus has a disproportionately high number
of amphi- and geocarpic species, and this is attri-
buted to the fact that these features are found more
frequently in annual species (Kaul et al., 2000),
and that the Israeli flora is rich in annuals (Shmida,
1981). Additional literature and internet searches
have unearthed reports and observations of several
additional taxa that are noted as either amphicarpic
or geocarpic.
Active geocarpy (as defined here) has been re-
ported in the Australian Tribulopis R. Br. (Zygo-
phyllaceae) by Keighery (1982), the Nepalese and
Himalayan Lignariella hinkuensis Kats. Arai, H.
Ohba & Al-Shehbaz (Brassicaceae) by Al-Shehbaz
et al. (2000), the Central Asian Taphrospermum
himalaicum (Hook. f. & Thomson) Al-Shehbaz,
Kats. Arai & H. Ohba (Brassicaceae) by Al-Sheh-
baz (2000) and Euploca hypogaea (Urb. & Ekman)
Diane & Hilger (Heliotropaceae) by Diane et al. (in
press). Kalin Arroyo (1981) listed Astragalus L.,
Trigonella L., Trifolium L., Arachis L. (mentioned
above), Tephrosia Pers., and Dolichos L. as having
species that are geocarpic.
In addition, Meney et al. (1990) reported an un-
usual form of amphicarpy in three species of the
dioecious Alexgeorgea Carlquist (Restionaceae),
where the female flowers are geoflorous and the
male flowers aerial. Halodule uninervis (Forssk.)
Asch. is a dioecious marine angiosperm seagrass
that has geoflorous (and submerged) female flowers,
and the fruits are released below the surface of ma-
rine sediments (Inglis, 2000). The family placement
of this genus is confusing, as IPNI, TROPICOS,
and the online Flora of North America variously
place it in the Cymodoceaceae, Zosteraceae, Na-
jadaceae, Hydrocharitaceae, and Potamogetona-
ceae. The first-mentioned family is the placement
used for the Flora of North America.
Bruhl (1994) reported on amphicarpy in the Cy-
peraceae, citing examples of amphicarpic species
in the genera Schoenoplectus (Rchb.) Palla and
Eleocharis R. Br. The South African Trianoptiles
Fenzl (comprising three species) is entirely amphi-
carpic (Bruhl, 1994; Haines & Lye, 1977). Note,
however, that all these Cyperaceae taxa show what
I termed here as amphi-basicarpy. Calvin˜o and Gal-
etto (2003) reported amphicarpy in Cryptantha cap-
ituliflora (Clos) Reiche, a high-altitude species of
Boraginaceae from the Andes. The legumes have a
number of amphicarpic genera; Kalin Arroyo
(1981), citing Rivals (1953), reported amphicarpy
from species of Vicia L., Amphicarpa S. Elliott ex
Nuttall, Vigna Savi, and Trifolium L., and Tindale
and Craven (1989) reported on amphicarpy in some
species of Glycine Willd. In addition, web searches
have revealed a number of other unpublished stud-
ies on amphicarpic species: Centrosema rotundifol-
ium Mart. ex Benth., a South American legume
(Mu¨ ller et al., 2001), Trifolium polymorphum Poir.
(Speroni & Izaguirre, 2003), and Schoenoplectus
hallii (A. Gray) S. G. Sm. (Cyperaceae; Beatty et
al., 2004). However, given the number of African
species that my research has revealed as being am-
phicarpic, basicarpic, or geocarpic, there are prob-
ably many more taxa that can be added to this list,
if they are adequately investigated.
B
ASICARPY
,A
MPHICARPY
,
AND
G
EOCARPY IN THE
A
FRICAN AND
M
ADAGASCAN
F
LORA
Literature surveys, web searches, and personal
communications with a number of botanists result-
Volume 92, Number 4
2005 449Barker
Basicarpy, Geocarpy & Amphicarpy
ed in a list of over 50 African or Madagascan taxa
that are either amphicarpic, basicarpic, or geocar-
pic in some form. I have relied in particular on the
works by Stopp (1958) and Agnew and Hedberg
(1969) as starting points. It is quite possible that
many more species have been noted as being am-
phicarpic or geocarpic, but the likely sources of
such observations, notably flora treatments, are
generally not available in electronic formats, and
thus not searchable using modern computer-based
methods. The recent exception to this is Flora
Zambeziaca, launched in 2004,
^
http://www.kew.
org/efloras/search.do
&
, but this is not searchable for
anything other than taxon names, a limitation that
I hope will be addressed as technologies develop.
Agnew and Hedberg (1969: 215, 216) reported
seven afroalpine species as being ‘‘strongly geocar-
pic,’’ and listed a further twelve taxa that are ‘‘de-
positers’’ sensu Hylander (1929). Geocarpy in the
afroalpine environment is considered by Agnew
and Hedberg (1969) to be an adaptation to constant
solifluction (the disturbance of the soil by the often
daily cycle of freezing and thawing of soils at high
altitudes). The considerable number of geocarpic
taxa in both afroalpine and arid areas such as Israel
(Lev-Yadun, 2000) is intriguing. Arid and afroal-
pine regions are two very different habitats, but
both can comprise soils that are mobile. This makes
seedling establishment difficult, and the burial of
seeds may ensure reproductive success in such
habitats.
I have only considered flowering plants in this
list, but it must be noted that Stopp (1958) included
some species of the aquatic fern genus Marsilea L.,
as well as a species of the lycophyte Isoetes stellen-
bossiensis A. V. Duthie. Table 2 lists angiosperm
species that have either or both been observed or
reported as geocarpic. The so-called ‘‘depositers’’
(sensu Agnew & Hedbeg, 1969), which simply re-
lease their seed at the base of the plant, have not
been included in this list, as I feel that deposition
is a general description that cannot readily be re-
searched, and that deposition could be a prelude
to further dispersal by means of specialist syn-
dromes. Furthermore, some species that could be
described as depositors may fit this description be-
cause they happen to be basicarpic. The distinction
is thus difficult to make.
Each of the families represented in Table 2 are
briefly discussed below, in alphabetical order. As
noted by Cheplick (1987), confirming or verifying
some of these reports is difficult, and I welcome
communications from botanists who may have ad-
ditional information or conflicting opinions.
APIACEAE
Agnew and Hedberg (1969) cited Haplosciadium
abyssinicum Hochst., now known as Trachydium
abyssinicum (Hochst.) Hiern, as an alpine geocar-
pic species. However, I have been unable to obtain
any additional information on this species, and it
is included here on the basis of this initial report.
ASTERACEAE
As noted above, fascinating examples of the di-
versity of amphicarpy may be found in members of
this family, but examples of amphicarpy or geocar-
py in African representatives of this family have
not been well documented. Agnew and Hedberg
(1969) reported geocarpy in Haplocarpha rueppellii
(Sch. Bip.) Beauverd, and I have independently ob-
served this in the field in the Bale Mountains, Ethi-
opia. While this species is obviously geocarpic, re-
flexing the short peduncle to force the capitulum
onto the ground, another species in the genus, Hap-
locarpha schimperi (Benth. & Hook.) Beauverd, is
even more strongly and obviously geocarpic. This
species, known only from the Ethiopian and Eri-
trean highlands, has capitula borne on short pe-
duncles. After flowering, these peduncles invert,
forcing the capitulum with the developing cypselae
into the soil (pers. obs.). Both species of Haplocar-
pha Less. prefer disturbed sites along streams and
drainage lines, and both occur in the afroalpine
zone, although H. schimperi is also quite wide-
spread at lower altitudes, even being found in damp
places in urban areas such as Addis Ababa (pers.
obs.). Thus, although Agnew and Hedberg (1969)
considered geocarpy to be an adaptation to soli-
fluction of afroalpine soils, it appears that H. schim-
peri can successfully reproduce in areas that are
not subject to solifluction, but that are, nonetheless,
disturbed.
A third species of Haplocarpha, H. nervosa
(Thunb.) P. Beauv., has also been observed both in
the field and greenhouse as being actively geocar-
pic (see Fig. 1), and this can also be noted from
herbarium specimens (e.g., S. Schonland 3014, R.
A. Dyer, 786, both at GRA). This morphologically
variable species is found in moist boggy habitats or
along stream banks in montane regions of the East-
ern Cape and Drakensberg of Kwa-Zulu Natal.
Haplocarpha nervosa in the south, and H. schimperi
and H. rueppellii in the north, may be examples of
afromontane disjunct sister taxa.
Arctotheca populifolia (Bergius) Norl. is also list-
ed in Table 2 as passively geocarpic. It inhabits
sandy beaches of the southern and eastern coastline
of Africa, and is one of the three main dune pio-
450 Annals of the
Missouri Botanical Garden
Table 2. African and Madagascan plants recorded as being amphicarpic or geocarpic. Where information is available, the nature of amphicarpy (as defined in Table 1) is described.
Family Taxon Form of reproduction Distribution Source
Apiaceae Trachydium abyssinicum (Hochst.) Hiern geocarpy (subtype un-
known)
East Africa Agnew & Hedberg (1969), as Haploscia-
dium abyssinicum Hochst.
Asteraceae Arctotheca populifolia (Bergius) Norl. passive geocarpy Southern Africa Barker (pers. obs.)
Haplocarpha schimperi (Sch. Bip.) Beauverd active geocarpy African Mountain Chain, Ethio-
pia to Tanzania
Barker (pers. obs.); Barker 1899 (ETH,
GRA)
Haplocarpha rueppellii Sch. Bip.) Beauverd active geocarpy African Mountain Chain, Ethio-
pia to Tanzania
Agnew & Hedberg (1969); Barker 1906
(ETH)
Haplocarpha nervosa (Thunb.) P. Beauv. active geocarpy South Africa, Zimbabwe Barker (pers. obs.); McKenzie 991
(GRA)
Balanophoraceae Mystropetalon polemannii Harv. basicarpy Western Cape, South Africa Kuijt (1969); Visser (1981)
Mystropetalon thomii Harv. basicarpy Western Cape, South Africa Kuijt (1969); Visser (1981)
Sarcophyte sanguinea Sparrm. basicarpy South Africa Kuijt (1969); Visser (1981)
Begoniaceae Begonia laporteifolia Warb. depositor (active geocarpy
reported by van der Pijl,
1982)
Tropical Africa van der Pijl (1982), as B. hypogaea
Winkler; M. Sosef, (pers. comm.)
Commelinaceae Commelina benghalensis L. amphi-geocarpy Central & East Africa, Mada-
gascar
Stopp (1958); Cheplick (1987); Kaul et
al. (2002)
Commelina forsskalaei Vahl amphi-geocarpy Central & East Africa Stopp (1958); Cheplick (1987)
Convolvulaceae Falkia repens L.f. active geocarpy S and SW Cape, South Africa van der Pijl (1982); Stopp (1958)
Nephrophyllum abyssinicum A. Rich. active geocarpy Ethiopia and Eritrea S. Edwards (pers. comm.)
Cucurbitaceae Cucumis humofructus Stent
Kedrostis psammophila Bruyns
active geocarpy
geophytic geocarpy
Southern Africa
Southern Africa
Stopp (1958)
Bruyns (1993)
Cyperaceae Bulbostylis humilis Kunth
Bulbostylis glaberrima Ku¨k.
Bulbostylis sp.
Scirpus lateriflorus J. F. Gmel.
Scirpus articulatus L.
amphi-basicarpy
amphi-basicarpy
amphi-basicarpy
amphi-basicarpy
amphi-basicarpy
East Africa
East Africa
East Africa
East Africa
East Africa
Haines (1971)
Haines (1971)
Haines (1971)
Haines (1971)
Haines (1971)
Scirpus supinus L.
Scirpus praelongatus Poir.
Scirpus muricinux C. B. Clarke
Scirpus aberrans Cherm.
Scirpus perrieri Cherm.
Scirpus reductus Cherm.
amphi-basicarpy
amphi-basicarpy
amphi-basicarpy
amphi-basicarpy
amphi-basicarpy
amphi-basicarpy
Southern Africa
East Africa
East Africa
Madagascar
Madagascar
Madagascar
Haines (1971)
Haines (1971)
Haines (1971)
Haines (1971)
Haines (1971)
Haines (1971)
Trianoptiles solitaria (C. B. Clarke) Levyns amphi-basicarpy Western Cape, South Africa Levyns (1943); Haines & Lye (1977)
Trianoptiles capensis Harv. amphi-basicarpy Western Cape, South Africa Levyns (1943); Haines & Lye (1977)
Trianoptiles stipitata Levyns amphi-basicarpy Western Cape, South Africa Levyns (1943)
Volume 92, Number 4
2005 451Barker
Basicarpy, Geocarpy & Amphicarpy
Table 2. Continued.
Family Taxon Form of reproduction Distribution Source
Cytinaceae Cytinus capensis Marloth
Cytinus sanguineus (Thunb.) Fourc.
Cytinus sp.
basicarpy
basicarpy
basicarpy
Western Cape, South Africa
Western Cape
Western Cape
Visser (1989)
Visser (1989)
Visser (1989)
Hydnoraceae Hydnora africana Thunb.
Hydnora esculenta Jumelle & H. Perrier
geophytic geocarpy
geophytic geocarpy
Cape to Swaziland
Madagascar
Pole Evans (1926)
Jumelle & Perrier (1912)
Hydnora johannis Becc. geophytic geocarpy South Africa to Arabia Musselman (1997), as H. abyssinica;
Musselman & Visser (1987)
Hydnora triceps Drege & E. Mey. geophytic geocarpy Namaqualand, South Africa Visser (1989)
Hypoxidaceae Empodium Salisb. (approx. 10 species) basicarpy Western, Central and Southern
Africa
D. Snijman (pers. comm.); Hilliard &
Burtt (1973)
Pauridia longituba M. E. Thompson geophytic geocarpy Cape region, South Africa G. Rowe (pers. comm.)
Saniella occidentalis (Nel) B. L. Burtt basicarpy SW Cape, South Africa J. Manning & D. Snijman (pers. comm.);
Burtt (2000)
Iridaceae Babiana bainesii Baker geophytic geocarpy (possi-
bly basicarpy)
Central South Africa J. Manning (pers. comm.)
Babiana hypogaea Burch. geophytic geocarpy (possi-
bly basicarpy)
Central South Africa J. Manning (pers. comm.)
Duthiastrum linifolium (Phill.) M. P. de Vos geophytic geocarpy Central South Africa De Vos (1999a); J. Manning (pers.
comm.)
Ixia acaulis Goldblatt & J. C. Manning geophytic geocarpy Cape region, South Africa De Vos (1999b); J. Manning (pers.
comm.)
Romulea stellata M. P. de Vos geophytic geocarpy Cape region, South Africa J. Manning (pers. comm.)
Syringodea bifurcata M. P. de Vos geophytic geocarpy Cape region, South Africa Stopp (1958); De Vos (1983)
Syringodea concolor (Baker) M. P. de Vos geophytic geocarpy Cape region, South Africa Stopp (1958); De Vos (1983)
Syringodea derustensis M. P. de Vos geophytic geocarpy Cape region, South Africa Stopp (1958); De Vos (1983)
Syringodea flanaganii Baker geophytic geocarpy Cape region, South Africa Stopp (1958); De Vos (1983)
Syringodea longituba (Klatt) Kuntze geophytic geocarpy Cape region, South Africa Stopp (1958), as S. leipoldtii;DeVos
(1983)
Syringodea pulchella Hook. f. geophytic geocarpy Cape region, South Africa Stopp (1958); De Vos (1983)
Syringodea saxatilis M. P. de Vos geophytic geocarpy Cape region, South Africa Stopp (1958); De Vos (1983)
Syringodea unifolia Goldblatt geophytic geocarpy Cape region, South Africa Stopp (1958); De Vos (1983)
Leguminosae Aeschynomene nematopoda Harms active geocarpy East Africa Stopp (1958); Gillett et al. (1971)
Amphicarpaea africana (Hook.f.) Harms amphi-geocarpy Tropical Africa B. Schrire (pers. comm.)
Macrotyloma geocarpum (Harms) Mare´chal &
Baudet
active geocarpy West and Central Africa Hepper (1963), as Kerstingiella geocar-
pa Harms
Tephrosia lupinifolia DC. amphi-geocarpy Tropical Africa Cheplick (1987), Gillett et al. (1971)
Trifolium subterraneum L. active geocarpy Cosmopolitan Stopp (1958); B. Schrire (pers. comm.)
Vicia amphicarpa Dorthes amphi-geocarpy Mediterranean North Africa Plitmann (1973), as Vicia sativa subsp.
amphicarpa (Dorth.) Ashers. & Graebn.
452 Annals of the
Missouri Botanical Garden
Table 2. Continued.
Family Taxon Form of reproduction Distribution Source
Vigna subterranea (L.) Verdc. active geocarpy West and Central Africa Hepper (1963), as Voandzeia subterranea
(L.) Thouars
Moraceae Ficus sur Forssk. geocarpy (occasional) Central-southern Africa Van Noort (pers. comm.); http://www.
figweb.org/ficus/subgenus
p
Sycomorus/
section
p
Sycomorus/subsection
p
Sycomorus/Ficus
p
sur.htm
Ficus trichoclada Baker geocarpy Madagascar Van Noort (pers. comm.); http://www.
figweb.org/ficus/subgenus
p
Sycomorus/
section
p
Sycomorus/subsection
p
Sycomorus/Ficus
p
trichoclada.htm
Ficus vogeliana (Miq.) Miq. geocarpy Tropical Africa Van Noort (pers. comm.); http://www.
figweb.org/ficus/subgenus
p
Sycomorus/
section
p
Sycomorus/subsection
p
Sycomorus/Ficus
p
vogeliana.htm
Poaceae Enneapogon desvauxii P. Beauv. amphi-basicarpy Namibia, Northern Cape of
South Africa
Stopp (1958)
Libyella cyrenaica Pamp. amphi-geocarpy Libya & Morocco Watson & Dallwitz (1992)
Polygonaceae Emex australis Steinh. amphi-geocarpy Cosmopolitan Stopp (1958)
Primulaceae Anagallis kingaensis Engl. active geocarpy Tropical Africa Taylor (1958); Stopp (1958)
Anagallis oligantha P. Taylor active geocarpy Tropical Africa Taylor (1958); Stopp (1958)
Ranunculaceae Ranunculus cryptanthus Milne-Redh. & Turrill possible active geocarpy East Africa Milne-Redhead & Turrill (1952); Agnew
& Hedberg (1969)
Ranunculus oreophytus Delile active geocarpy East Africa Milne-Redhead & Turrill (1952); Agnew
& Hedberg (1969)
Ranunculus stagnalis Hochst. ex A. Rich. possible active geocarpy East Africa Milne-Redhead & Turrill (1952); Agnew
Hedberg (1969)
Rubiaceae Galium ankaratrense Puff active geocarpy Madagascar Kiehn & Puff (1987)
Scrophulariaceae Limosella africana Gluck geocarpy (subtype un-
known)
East Africa Agnew & Hedberg (1969)
Limosella capensis Thunb. geocarpy (subtype un-
known)
South Africa Stopp (1958)
Limosella macrantha R. E. Fr. geocarpy (subtype un-
known)
East Africa Agnew & Hedberg (1969)
Urticaceae Laportea ovalifolia (Schumach.) Chew amphi-geocarpy Tropical Africa Engler (1895); Rendle (1917); Cheplick
(1987), all as Fleurya podocarpa
Wedd. var. amphicarpa Engl.; Friis
(1991)
Volume 92, Number 4
2005 453Barker
Basicarpy, Geocarpy & Amphicarpy
neers in this habitat. It has a rapid growth rate and
forms small nabkha dunes (hummocks of wind-
blown sand trapped by the leaves; Hesp & Mc-
Lachlan, 2000). The capitula are held more or less
erect and are obviously insect-pollinated. However,
following anthesis, the capitulum closes and the pe-
duncle relaxes such that it becomes procumbent.
Subsequent burial by the accumulation of wind-
blown sand occurs, and the seeds are either not
released from the infructescence, or are only re-
leased if it is damaged by subterranean herbivores.
The seeds are quite large, lack a well-developed
pappus, and are woolly (McKenzie et al., 2005, this
issue). This latter feature is possibly a water-reten-
tion mechanism, and these cypselae will germinate
in situ, close to the parent plant (Knevel, 2002;
Knevel et al., 2002).
Interestingly, these three species of Haplocarpha
and A. populifolia are from the predominantly Af-
rican subtribe Arctotidinae (tribe Arctotideae), and
it is tempting to infer a close phylogenetic relation-
ship of these species on the basis of this reproduc-
tive strategy. Results from preliminary molecular
systematic studies (Funk et al., 2004; McKenzie et
al., unpublished) suggest that Haplocarpha is high-
ly polyphyletic, but relationships within this tribe
are not yet fully resolved so this hypothesis awaits
testing.
BALANOPHORACEAE
In southern Africa, Sarcophyte sanguinea
Sparrm. parasitizes the roots of Acacia L. species.
This species is dioecious and relies on flies for pol-
lination, as the small flowers, produced at ground
level, smell of carrion. This species, which I con-
sider to be basicarpic, has compound fruits made
up of the fused ovaries, which ripen into a fleshy
mass. It is thought that these are animal-dispersed
(Kuijt, 1969). Mystropetalon Harv. is a monoecious
genus of two species from the southwestern Cape,
South Africa, and produces spike-like inflorescenc-
es slightly above ground level. Male flowers are ter-
minal to the female flowers on these spikes, and
the inflorescences are protogynous. The seeds pos-
sess an elaiosome and are ant-dispersed (Kuijt,
1969; Visser, 1981). As these spikes are produced
at ground level, these species are considered bas-
icarpic.
BEGONIACEAE
Van der Pijl (1982) reported geocarpy in Begonia
hypogaea Winkler (now known as B. laporteifolia
Warb.), an occupant of African rain forests. This
species grows near streams and produces a berry,
but that is all that appears to be known about this
species. However, M. Sosef (pers. comm.) considers
this species to merely hold its fruits close to the
ground, where they disintegrate, releasing the seeds
in the vicinity of the parent. This species should
thus be considered as a depositor sensu Agnew and
Hedberg (1969), but is listed here owing to its pre-
vious categorization as geocarpic.
COMMELINACEAE
Kaul et al. (2000) cited five species in this family
as having both cleistogamous underground flowers
and aboveground flowers (amphi-geocarpy). As with
many such species, the fruits from these two types
of flowers are dimorphic. Commelina benghalensis
L. is well studied in this regard (cf. Kaul et al.,
2002).
CONVOLVULACEAE
Nephrophyllum abyssinicum A. Rich. is a mem-
ber of the small tribe Dichondreae, the basal tribe
in the family (Austin, 1997). It is an actively geo-
carpic species that is found in the highlands of
Ethiopia and Eritrea, where it is recorded as oc-
curring in heavily grazed pastures, open ground,
and between crevices of rocks (Demissew & Austin,
1995). It is a prostrate herb, rooting at the nodes,
with axillary flowers. The fruit is a utricle that de-
velops underground from an elongating pedicel that
pushes it below the soil surface. There are also re-
ports of geocarpy within the closely related genus
Falkia L. f., another member of the tribe Dichon-
dreae, with species found in both Ethiopia and
southern Africa. Van der Pijl (1982) included the
southern African Falkia repens L. f. in his discus-
sion of geocarpy, but omitted further details. This
species was described by Stopp (1958, who cited it
as F. dichondroides Baker) as a prostrate plant, in
which the developing fruit is driven into the ground
(also noted by Meeuse & Welman, 2000). In this
species it was noted that the success of fruit burial
was dependent on the substrate conditions. Stopp
(1958) also considered Dichondra J. R. Forst. & G.
Forst. to be similar in this regard, but provided no
further discussion. Only the cosmopolitan Dichon-
dra micrantha Urb. is known from Southern and
East Africa (Verdourt, 1969; Austin, 1998; Meeuse
& Welman, 2000), but none of these works makes
any reference to fruit burial in this species. The
seed development and dispersal of all members of
the tribe Dichondreae are worthy of further inves-
tigation.
454 Annals of the
Missouri Botanical Garden
Volume 92, Number 4
2005 455Barker
Basicarpy, Geocarpy & Amphicarpy
←
Figure 1. —A. Haplocarpha nervosa, showing capitulum on short peduncle, rosette growth form, and two reflexed
peduncles of developing geocarpic infructescences at 9 and 7 o’clock. —B. Hirsute form of H. nervosa in the field,
showing peduncle reflexing from the center of the rosette of leaves. —C. Details of reflexed peduncle and developing
geocarpic infructescence. (A & C photographed in cultivation at Rhodes University from McKenzie 991, GRA.)
CUCURBITACEAE
Two geocarpic species are known from this fam-
ily. Stopp (1958) included Cucumis humofructus
Stent in his list of geocarpic taxa from southern
Africa, as the plant actively buries the fruit of this
species. However, as noted by van der Pijl (1982),
the burial of fruit in this instance may not be a
mechanism to ensure atelochory (lack of dispersal,
also termed antitelochory by workers such as Ellner
& Shmida, 1981) as the fruits are dispersed by
aardvarks, which seek out the fruit as a source of
water, and incidentally disperse the seeds far and
wide following passage through the animal’s gut
(Meeuse, 1958). Thus this species is actively geo-
carpic, but the animal dispersal agent ensures that
the seeds are widely dispersed, a situation that is
probably unusual in geocarpic taxa.
Kedrostis psammophila Bruyns is found in Na-
maqualand, South Africa, and is usually prostrate,
with separate male and female flowers. The female
flowers develop from underground stems, and the
ovary (and developing fruit) is retained below
ground at all times, while the perianth is produced
above ground (Bruyns, 1993). This species thus
falls into the geophytic geocarpy category. The
fruits are spherical, approximately 20 mm in di-
ameter, and have 8 to 10 seeds. The fruits are pos-
sibly dispersed by moles, but this has not been
observed or tested (Bruyns, 1993).
CYPERACEAE
There have been several reports of amphicarpy
in the form of amphi-basicarpy in this family, which
appears to have the greatest number of amphicarpic
species in the African flora. Haines’s (1971) report
of amphicarpic taxa in East Africa probably only
scratches the surface of the actual number of spe-
cies in this family with this trait. A thorough in-
vestigation by cyperologists will probably find many
more such taxa. In the cases noted in Table 2, the
species possess aerial flowers as well as basal flow-
ers protected by the leaf sheaths. The basal flowers
usually produce larger seeds that may have a sim-
ilar or different surface ornamentation (Haines,
1971). The unusual Cape genus Trianoptiles Fenzl
has been studied quite extensively (Levyns, 1943;
Haines & Lye, 1977; Bruhl, 1994). The compara-
tively large number of reports of amphicarpic spe-
cies in this family might indicate that this strategy
is an adaptation to ephemerally moist environ-
ments.
CYTINACEAE
This small family comprising the holoparasitic
genera Bdallophyton Eichler and Cytinus L. was
previously considered part of the Rafflesiaceae, but
Nickrent et al. (2004) demonstrated that these two
genera (as Cytinaceae) are related to the Malvales.
There are three African and at least one Madagas-
can species of Cytinus, as well as species from the
Mediterranean region (
^
http://www.science.siu.edu/
parasitic-plants/Cytinaceae
&
). These species are ei-
ther monoecious (the European C. hypocistis L.) or
dioecious (C. capensis Marloth and C. sanguineus
(Thunb.) Fourc.), and like the Hydnoraceae, they
produce flowers from underground shoots and the
fruits are borne at ground level. They are thus con-
sidered to be basicarpic. The fruits contain ‘‘several
thousand’’ seeds (Visser, 1981: 63), which are
thought to be dispersed when the fruit dries and
cracks open, although Kuijt (1969) reported the
presence of an elaiosome, suggesting ant dispersal.
HYDNORACEAE
This family of holoparasites is thought to be a
Gondwanan relict. There are three species of Hyd-
nora Thunb. in Africa and one in Madagascar, all
of which live almost entirely below ground. The
fruit has been described by Musselman (1997: 17)
as ‘‘. . . a subterranean fleshy berry with a woody
pericarp; seeds numerous in a flesh-coloured pulp,
minute.’’ The African species produce hermaphro-
dite flowers, while the Madagascan H. esculenta Ju-
melle & H. Perrier is monoecious. Flowers are pre-
dominantly subterranean, opening at the surface. In
H. triceps Drege & E. Mey., the flowers open below
ground, and pollinating insects (blowflies) enter
through cracks in the soil (Visser, 1989). In H. af-
ricana Thunb. the flowers open above ground, and
attract dermestid beetles. Following pollination the
fruits mature below ground and are dispersed by
jackals, porcupines, and baboons (Pole Evans,
1926). They are also eaten by humans (Musselman,
1997; Musselman & Visser, 1989). Hydnora johan-
456 Annals of the
Missouri Botanical Garden
nis Becc. is the most widespread species, and its
fleshy fruits are large (10–15 cm) and completely
subterranean (Musselman & Venter, 1987). The
fruits of the Madagascan species have been docu-
mented as having a juicy, white pulp, with a sour
taste, being eaten by humans and lemurs (Jumelle
& Perrier, 1912;
^
http://bibliophile.mc.duke.edu/
gww/Berenty/Plants/Hydnora-esculenta/index.html
&
).
These species thus best fit the category of geophytic
geocarpy, as the entire plant spends its life under-
ground. For more on this intriguing group see
^
http://www.odu.edu/webroot/instr/sci/plant.nsf/
pages/hydnoraplant
&
and
^
http://www.science.siu.
edu/parasitic-plants/Hydnoraceae
&
.
HYPOXIDACEAE
Considerable confusion arose during my inves-
tigations into this family. The Hypoxidaceae (and
the Iridaceae) contain a number of species that
have subterranean ovaries (Burtt, 1970), suggesting
that these taxa would also be geocarpic. However,
consultation with experts in these families has re-
sulted in reports (published and unpublished ob-
servations) that suggest that the subterranean ovary
may often become aerial as the fruit ripens and
approaches dehiscence. Saniella occidentalis (Nel)
B. L. Burtt possesses basal or subterranean ovaries,
which are retained below ground when mature
(Burtt, 2000). In contrast, the closely related Em-
podium Salisb. (a genus of 8 to 10 species) has a
similar morphology (Hilliard & Burtt, 1973), but it
is not clear if all species are basicarpic or geocar-
pic. D. Snijman (pers. comm.) reports that in San-
iella occidentalis, Empodium plicatum Salisb., E.
namaquensis (Baker) M. F. Thompson, and E. flexile
M. F. Thompson ex Snijman, the mature fruit is
pushed from the soil by an extension of the pedicel.
This report for Saniella conflicts with Burtt’s (2000)
observations. Until clarity on the phenology of these
species has been obtained, these taxa may be best
considered as basicarpic, possibly even fitting with
Agnew and Hedberg’s (1969) depositor category.
Similarly, Rhodohypoxis rubella Nel bears its fruit
at or below ground level (basicarpic), while another
(undescribed) species from marshy habitats is de-
scribed as having fruit on an erect flower stalk that
bends over to deposit the fruit on the ground (Hil-
lard & Burtt, 1973).
Pauridia longituba M. E. Thompson is a small
geophyte that produces its flowers above ground
while retaining the ovary underground throughout
its life. These mature fruit capsules are thin-walled
and indehiscent, fragmenting irregularly to release
the seeds. A second species, P. minuta T. Durand
& Schinz, produces the flower entirely above
ground, but post-flowering the pedicel recurves and
releases the seeds at the base of the parent plant
(G. Rowe, pers. comm.). This species thus fits the
category of a depositor sensu Agnew and Hedberg
(1969).
IRIDACEAE
There are several taxa in this predominantly
southern African geophyte family that have evolved
flowers that retain the inferior ovary underground,
and as noted above, Burtt (1970) reported this mor-
phology for the European Crocus as well. In these
taxa, the long, tubular perianth is exserted above
ground, but developing fruits remain underground.
The taxa in which this occurs are all from the sub-
family Ixioideae, tribe Ixieae, suggesting that there
may be a phylogenetic predisposition for this pat-
tern of reproduction.
The genus Syringodea Hook. f. was first reported
by Stopp (1958) as fitting what he considered to be
the definition of basicarpy. The eight species in this
genus all flower above ground, with the ovary re-
tained underground, where it develops and is re-
tained, and thus falls in my subtype of geophytic
geocarpy. The fruit capsule in this genus is hygro-
chastic, and seed dispersal occurs when the cap-
sule opens in damp or wet conditions, or when the
capsule disintegrates with age. Burtt (1970) reit-
erated Stopp’s (1958) reports that in S. leipoldtii
(now known as S. longituba (Klatt) Kuntze) the up-
per part of the capsule emerges from the soil, where
the hygrochastic valves split when wetted and some
of the seeds are dispersed. The remaining seeds are
retained and will germinate in situ.
Like species of Syringodea, Duthiastrum linifol-
ium (Phill.) M. P. de Vos also possesses flowers with
subterranean bases, and according to J. Manning
(pers. comm.) the ovary and fruit remain buried.
Similarly, Ixia acaulis Goldblatt & J. C. Manning
has a buried ovary and is the only species in the
genus with this feature. J. Manning (pers. comm.)
also indicated that certain species of Tritonia Ker
Gawl. and Hesperantha Ker Gawl. may also have
this morphology, but details of these taxa have not
been included here as no clear indication of the
floral structure and the possible type of geocarpy
could be found in the literature.
Two species of Babiana Ker Gawl. exhibit either
geophytic geocarpy or basicarpy, the exact nature
of the position of the mature fruit being unclear.
This genus has several species that produce flowers
at or near ground level, and in the instance of the
two species listed in Table 2, J. Manning (pers.
Volume 92, Number 4
2005 457Barker
Basicarpy, Geocarpy & Amphicarpy
comm.) observed the presence of an underground
ovary (and developing fruit), which may be a mech-
anism to avoid seasonal temperature extremes, as
it takes several months for the fruit to mature.
LEGUMINOSAE
Vigna subterranea (the bambara groundnut) and
Macrotyloma geocarpum are actively geocarpic le-
gumes from dry savanna areas of northern Came-
roon and central Africa (Hepper, 1963; Kalin Ar-
royo, 1981; van der Pijl, 1982). Both these species
produce aerial flowers but bury the developing
fruit. When first discovered, these species were un-
der limited cultivation as a food source for humans
(Hepper, 1963). Subsequently, V. subterranea has
become an important crop in many countries of
tropical Africa (Linnemann, 1993; Pasquet et al.,
1999), but M. geocarpum is disappearing from tra-
ditional food production (Pasquet et al., 2002).
From southern Africa, Stopp (1958) reported Tri-
folium subterraneum L. as being geocarpic. This
species originates from the Mediterranean region,
and Stopp (1958) considered it to be a neophyte in
the Cape region. Indeed, internet searches suggest
that this species is of agricultural value, and is cos-
mopolitan, but it is also listed as a weed by the
Global Compendium of Weeds (
^
http://www.hear.
org/gcw/html/index.html
&
). Few additional details
are provided by Stopp (1958) who noted that the
fruits are shallowly buried, but internet resources
(such as
^
http://www.fao.org/ag/AGP/AGPC/doc/
Gbase/DATA/pf000351.HTM
&
) imply that the fruits
are actively buried.
Aeschynomene nematopoda Harms is a scandent
or trailing herb and is considered here to be ac-
tively geocarpic. This was first recorded by Stopp
(1958), who provided a quotation from Harms, com-
paring this plant to the peanut, Arachis hypogaea.
Gillett et al. (1971) noted in the Flora of Tropical
East Africa that the pods develop below the ground,
and suggested that herbarium specimens lack fruits
because they break off when the plant is removed
from the soil.
Vicia amphicarpa Dorthes from the Mediterra-
nean region of North Africa is amphicarpic (ac-
cording to Plitmann, 1973, as Vicia sativa subsp.
amphicarpa (Dorthes) Ashers. & Graebn.), and this
species is considered here to be amphi-geocarpic.
Tephrosia lupinifolia DC. has been reported by
Stopp (1958) as amphicarpic and is annual and in-
habits sandy regions. Gillett et al. (1971: 167–169)
illustrated and described this species as having
normal inflorescences as well as occasionally pro-
ducing axillary pseudoracemes near the base of the
stem. These push underground when growing in
sandy soil and bear small cleistogamous flowers.
The subterranean pods are indehiscent and usually
contain a single seed or sometimes two. This spe-
cies is thus amphi-geocarpic.
Amphicarpaea africana (Hook. f.) Harms is the
only African species of this small genus. The Amer-
ican species A. bracteata (L.) Fern is known as the
‘‘Hog Peanut.’’ It appears that this genus is also
amphi-geocarpic, but interestingly, Gillett et al.
(1971) made no mention of amphi-geocarpy in their
description of A. africana in the Flora of Tropical
East Africa.
MORACEAE
The genus Ficus L. is large, and van der Pijl
(1982) noted that some (unlisted) species are geo-
carpic. Some species of Ficus sect. Sycomorus bear
syconia at or just below ground level on branchlets
produced from the trunk and lower branches. In
some species (listed in Table 2) these can become
geocarpic. Other species such as F. mauritiana
Lam., F. tiliifolia Baker, F. polyphlebia Baker, and
F. botryoides Baker may also be potentially geocar-
pic in this manner, but this is not the usual form
of fruiting (S. van Noort, pers. comm.), and the in-
clusion of this genus in a list of amphicarpic or
geocarpic taxa is thus tentative. Given the attractive
nature of the fruit, and the long life span of the
species, it is unlikely that this is a strategy to retain
seeds in the same location as the parent, but rather
to make them accessible to dispersal agents.
POACEAE
Cleistogamy is well known in the grasses (see
Campbell et al., 1983; Connor, 1987, for reviews)
and usually occurs as a variant of amphicarpy (I
list several examples of aerial amphicarpic grasses
above). Campbell et al. (1983) indicated that eight
species of cleistogamous grasses are known from
Africa, but these are not named. I have not been
able to find records of that many taxa. It is impor-
tant to note that cleistogamy does not equate to
geocarpy and that only two genera of grasses show
underground cleistogamy (i.e., are geoflorous).
Campbell et al. (1983) used Dobrenz and Beetle’s
(1966) term rhizanthogene for the underground
flowers borne on specialized rhizomes, and indi-
cated that there are eight species from four genera
that possess these.
Connor (1979) reported several grass species
that have cleistogenes—flowers at ground level en-
closed by leaf sheaths. This is the same syndrome
as reported in the Cyperaceae (above) and may be
458 Annals of the
Missouri Botanical Garden
considered as a borderline case of geocarpy: that
termed basicarpy by van der Pijl (1982) and amphi-
basicarpy here. This syndrome appears predomi-
nantly in New World grasses, and the majority of
the examples of aerial amphicarpy cited above are
taxa of the subfamily Pooideae. However, it is also
known from tribe Pappophoreae (subfamily Chlor-
idoideae), an Old World group, a southern African
example being Enneapogon desvauxii P. Beauv., an
amphi-basicarpic annual species found in the arid
western parts of southern Africa (Stopp, 1958, as
the synonym E. brachystachyus Stapf). In this spe-
cies, as is the general trend in amphicarpic grasses,
the basal flowers produce a seed that is consider-
ably larger than those from the aerial flowers. This
larger seed is not readily released, as the disinte-
gration of the leaf sheath is required to liberate it.
Amphi-geocarpy is also found in Libyella cyrenaica
Pamp., a monotypic genus of subfamily Pooideae.
This plant, described by Clayton and Renvoize
(1986) as a minute annual, grows in coastal sands
and possesses hidden cleistogenes in the leaf
sheaths as well as underground (Watson & Dallwitz,
1992).
POLYGONACEAE
Stopp (1958) cited Emex australis Steinh., a
weed of South African origin, as possessing aerial
and subterranean fruit. Surprisingly, given its de-
clared weed status in several countries, I could find
no mention of this seed heteromorphism on any
weed-related web site I visited. Its sister species,
Emex spinosa (L.) Campd., has also been docu-
mented as amphi-geocarpic, having both aerial and
subterranean seeds (Evenari et al., 1977). In this
species, the geocarpic seeds are approximately four
times the size of the aerial seed and have different
germination requirements (Weiss, 1980). It is thus
possible that E. australis has a similar amphi-geo-
carpic nature.
PRIMULACEAE
In a postscript, Stopp (1958) cited Taylor’s
(1958) work on tropical African Primulaceae in
which the actively geocarpic nature of Anagallis
kingaensis Engl. is first described. Stopp (1958)
stated that this is the first record of geocarpy in the
family, but that there are some basicarpic species
in Anagallis L. as well as Cyclamen L. and Lysi-
machia L. Taylor (1958) noted that in A. kingaensis,
the pedicel elongates, reflexes, and becomes buried
as the fruits ripen. Taylor (1958) noted that A. oli-
gantha P. Taylor is closely related to A. kingaensis
and made no mention of geocarpy in this species,
but Stopp (1958) considered it to be geocarpic.
RANUNCULACEAE
Agnew and Hedberg (1969) listed three geocar-
pic species of Ranunculus L. (Table 2). However,
of these three species, only R. oreophytus Delile is
clearly described by Milne-Redhead and Turrill
(1952) as being actively geocarpic. In this species,
curvature of the pedicels buries the spherical heads
in the soil. The remaining two species listed in Ta-
ble 2 are described merely as having achenes in
spherical heads on reflexed pedicels, a behavior
they have in common with R. oreophytus, so it is
possible that these two species are also actively
geocarpic.
RUBIACEAE
Geocarpy has been reported in the Madagascan
Galium ankaratrense Puff by Kiehn and Puff
(1987). This species is found at high altitudes in
the Ankaratra mountains. It is cushion-forming and
grows in moist, sheltered places. After flowering,
the pedicels elongate up to 40 mm in length, push-
ing the developing fruits into the ground. The seeds
germinate under the mother plants, further adding
to the dense cushion growth form (Kiehn & Puff,
1987).
SCROPHULARIACEAE
Two species of Limosella L. noted in Table 2 are
afroalpine species recorded by Agnew and Hedberg
(1969) as being geocarpic. The third species, L.
capensis, is listed as geocarpic by Stopp (1958).
Limosella comprises small aquatic plants, common-
ly known as mudworts. Some species of the genus
have widespread distributions, thought to be me-
diated by bird dispersal of the small seeds that
stick together along with mud to birds’ legs (Korn-
hall, 2004). To the best of my knowledge, this dis-
persal explanation has not been tested, but I have
observed the fruit of one of the southern African
taxa being produced both on the surface and be-
neath the mud of shallow temporary water bodies
(voucher Barker 1922, GRA). An accurate identi-
fication of this specimen to species level is not pos-
sible, and Hilliard and Burtt (1986) noted that a
taxonomic revision of the African species is need-
ed. The presence of geocarpy in other species of
Limosella needs to be investigated.
URTICACEAE
Little is known about Fleurya podocarpa var. am-
phicarpa, the amphicarpic status of which was first
Volume 92, Number 4
2005 459Barker
Basicarpy, Geocarpy & Amphicarpy
noted by Engler (1895) and again cited by Cheplick
(1987). This taxon is now known as Laportea oval-
ifolia (Schumach.) Chew, and Friis (1991: 89) in
Flora Zambesiaca merely noted that ‘‘geocarpic
fruits may occur,’’ while earlier, Rendle (1917) not-
ed that one specimen comprised fruits that seemed
to have been buried. These reports suggest that this
species may be amphi-geocarpic.
C
AN
A
MPHICARPIC AND
G
EOCARPIC
S
PECIES
B
E
C
HARACTERIZED
?
Cheplick (1994) listed four adaptive advantages
of amphi-geocarpy. These are the retention of off-
spring in suitable microhabitats (the mother site
theory sensu Zohary, 1937, 1962), and the protec-
tion of seeds from microenvironmental extremes,
fire, and herbivores or predators. Underground
fruits may thus be important in maintaining the
population in situ, unless they have specific adap-
tations for animal dispersal as noted, for example,
in Hydnora and Cucumis discussed above. Assum-
ing a lack of dispersal, the consequences of anti-
telochoric subterranean seed production include
the lack of genetic exchange between the cleistog-
amous flowers leading to inbreeding depression, the
high energy costs associated with the production of
larger underground reproductive structures, limited
dispersal, sibling competition among offspring, and
exposure to underground herbivores and seed pred-
ators (Cheplick, 1994). For species that are exclu-
sively geocarpic, these consequences (notably the
genetic ones) would be further exacerbated by the
lack of xenogamy. In light of these trade-offs, it
might be expected that an amphicarpic or geocarpic
strategy would only evolve under certain environ-
mental conditions: conditions that, if known, could
be used to predict the presence of amphicarpy, bas-
icarpy, or geocarpy.
While the known amphicarpic and geocarpic
taxa listed here are restricted to one continent (Af-
rica, including Madagascar), this list has added to
our knowledge of this unusual form of reproductive
biology. The limited numbers of species make it
difficult to draw any meaningful correlations with
habit or habitat that would support any predictions.
However, as noted above, species that bury their
seeds usually have flowers or inflorescences at or
near ground level. Thus species that are prostrate,
scandent, cushion-forming, or rosette-forming may
be candidates for geocarpy, and should be carefully
investigated for it. From my investigations, it is ap-
parent that some observations of geocarpy have
been accidental and depend on good field obser-
vations and thorough plant-collecting techniques.
Similarly, observations of amphi-basicarpy in caes-
pitose taxa such as sedges and grasses require care-
ful dissection of basal leaf and sheathing structures
to locate these well-hidden fruits and seeds. Even
herbaceous taxa that bear inflorescences from basal
or ground-level branches should be treated with
suspicion, and carefully examined and collected.
Those geophytic taxa with ovaries held at or below
ground level are obviously candidates for the geo-
phytic geocarpy category, but careful observation of
flower and fruiting phenology is required in order
to rule out active geocarpy.
The species in Table 2 are often found in either
arid areas where the substrate can be unstable, or
high-altitude afroalpine regions in damp to wet
habitats, also with potentially unstable soils. Thus
the hypothesis of a link between geocarpy and life
on unstable substrates, in either arid environments
(as noted by Zohary, 1962) or afroalpine soils sub-
ject to erosion or solifluction (noted by Agnew &
Hedberg, 1969) may have some merit. Ellner and
Shmida (1981) noted that basicarpy is common in
desert plants, and is often found in annual species
that have hygrochastic diaspores. Several taxa list-
ed here are found in arid regions, and some (no-
tably the legumes) appear to occupy regions of low
and strongly seasonal rainfall. Some of the Cyper-
aceae (e.g., Trianoptiles) occupy ephemerally wet
habitats, as do the examples of Rhodohypoxis and
Limosella noted here. The species of Haplocarpha
(Compositae) occupy damp to wet disturbed sites,
often along stream banks, which would be prone to
flooding and associated erosion. Burial of seeds to
avoid dispersal by runoff and associated erosion in
this environment would be advantageous, an aspect
also discussed in the context of flash floods in arid
regions by Ellner and Shmida (1981).
As intimated by Cheplick (1994), basicarpy and
geocarpy could limit seed-mediated gene flow, lead-
ing to increased isolation of populations, possibly
resulting in increased speciation rates. Conversely,
reduced gene flow between populations may result
in genetic drift, fixation of alleles, and inbreeding
depression, which could ultimately lead to local
population (and possibly species) extinctions. Giv-
en this unusual reproductive strategy, it is surpris-
ing that no population genetic studies on geocarpic
species (or comparing geocarpic and non-geocarpic
congeners) have been carried out.
An additional reproductive phenomenon appears
to be associated with some of the taxa in Table 2:
that of floral dimorphism and various forms of mon-
oecy. Species of Ficus (Moraceae) usually have
male and female flowers within a single syconium,
and in Haplocarpha (Compositae) the ray florets are
460 Annals of the
Missouri Botanical Garden
female, while disk florets are hermaphrodite, but
this system is not uncommon in the family. Several
of the holoparasitic species in Kedrostris (Cucur-
bitaceae), Laportea (Urticaceae), and Emex (Poly-
gonaceae) are also monoecious. However, the spe-
cies listed here have not been researched
exhaustively, so the co-occurrence of amphicarpy,
basicarpy, and geocarpy with monoecy could be co-
incidental.
In summary, amphicarpy, basicarpy, and geocar-
py in all of their forms occur across a diverse range
of flowering plants, edaphic and climatic condi-
tions, and continents. In the context of the size of
the flora of Africa and Madagascar, correlating the
small number of amphicarpic, basicarpic, and geo-
carpic species listed here to environmental param-
eters provides very limited predictive power, but an
association with unstable substrates appears to be
a fairly common theme, and this biological enigma
deserves further study.
Literature Cited
Agnew, A. D. Q. & O. Hedberg. 1969. Geocarpy as an
adaptation to afroalpine solifluction soils. J. E. Africa
Nat. Hist. Soc. Natl. Mus. 27: 216–217.
Al-Shehbaz, I. A. 2000. A revision of the Himalayan and
Central Asian genus Taphrospermum (Brassicaceae).
Harvard Pap. Bot. 5: 99–108.
, K. Arai & H. Ohba. 2000. A revision of the
genus Lignariella (Brassicaceae). Harvard Pap. Bot. 5:
113–121.
Austin, D. F. 1997. Parallel and convergent evolution in
the Convolvulaceae. Pp. 201–234 in P. Matthews & M.
Sivadasan (editors), Diversity and Taxonomy of Tropical
Flowering Plants. Mentor Books, Calicut. [
^
http://www.
fau.edu/divdept/biology/people/evolution.htm
&
(ac-
cessed 23 Oct. 2003)].
. 1998. The indiscriminate vector: Human distri-
bution of Dichondra micrantha (Convolvulaceae). Econ.
Bot. 52: 88–106.
Beatty, B. L., W. F. Jennings & R. C. Rawlinson. 2004.
Schoenoplectus hallii (Gray) S. G. Sm. (Hall’s bulrush):
A Technical Assessment. USDA Forest Service, Rocky
Mountain Region, Species Conservation Project.
^
http://www.fs.fed.us/r2/projects/scp/assessments/
schoenoplectushallii.pdf
&
.
Bruhl, J. J. 1994. Amphicarpy in the Cyperaceae, with
novel variation in the wetland sedge Eleocharis caespi-
tosissima Baker. Austral. J. Bot. 42: 441–448.
Bruyns, P. 1993. A new species of Kedrostis from the
Western Cape. Bothalia 23: 233–235.
Burtt, B. L. 1970. The evolution and taxonomic signifi-
cance of a subterranean ovary in certain monocotyle-
dons. Israel J. Bot. 19: 77–90.
. 2000. Saniella and its relation to other South
African genera of Hypoxidaceae. Edinburgh J. Bot. 57:
63–70.
Calvin˜o, A. & L. Galetto. 2003. Cleistogamy in the rare
high Andean perennial herb Cryptantha capituliflora
(Boraginaceae). Pl. Syst. Evol. 237: 41–50.
Campbell, C. S., J. A. Quinn, G. P. Cheplick & T. J. Bell.
1983. Cleistogamy in grasses. Annual Rev. Ecol. Syst.
14: 411–441.
Caro, J. A. & E. Sanchez. 1971. La identidad de Stipa
brchychaete Gordon, S. caudata Trinius y S. bertrandii
Philippi. Darwinia 16: 637–653.
Cheplick, G. P. 1987. The ecology of amphicarpic plants.
Trends Ecol. Evol. 2: 97–101.
. 1994. Life history evolution in amphicarpic
plants. Pl. Spec. Biol. 9: 119–131.
Clayton, W. D. & S. A. Renvoize. 1986. Genera Gramin-
um. Grasses of the World. Royal Botanic Gardens, Kew.
Connor, H. E. 1979. Breeding systems in the grasses: A
survey. New Zealand J. Bot. 17: 547–574.
. 1987. Reproductive biology in grasses. Pp. 117–
132 in T. R. Soderstrom, K. W. Hilu, C. S. Campbell &
M. E. Barkworth (editors), Grass Systematics and Evo-
lution. Smithsonian Institution Press, Washington, DC.
& B. A. Matthews. 1977. Breeding systems in
New Zealand grasses VII. Cleistogamy in Microlaena.
New Zealand J. Bot. 15: 531–534.
De Clavijo, E. R. 1995. The ecological significance of
fruit heteromorphism in the amphicarpic species Ca-
tananche lutea (Asteraceae). Int. J. Pl. Sci. 156: 824–
833.
Demissew, S. & D. F. Austin. 1995. The genus Nephro-
phyllum (Convolvulaceae, tribe Dichondreae) in Ethio-
pia. Kew Bull. 50: 103–108.
De Vos, M. P. 1983. Syringodea. Pp. 1–9 in O. A. Leist-
ner (editor), Flora of Southern Africa, Vol. 7, Part 2,
Fascicle 2. Government Printer, South Africa.
. 1999a. Duthiastrum. Pp. 139–141 in G. Ger-
mishuizen (editor), Flora of Southern Africa, Vol. 7, Part
2, Fascicle 1. National Botanical Institute, Pretoria.
. 1999b. Ixia. Pp. 3–87 in G. Germishuizen (ed-
itor), Flora of Southern Africa, Vol. 7, Part 2, Fascicle
1. National Botanical Institute, Pretoria.
Diane, N., H. Fo¨rther, H. H. Hilger & M. Weigend. (in
press). Heliotropaceae Schrad. In K. Kubitzki (editor),
Families and Genera of the Flowering Plants. Springer,
Berlin.
Dobrenz, A. K. & A. A. Beetle. 1966. Cleistogenes in
Danthonia. J. Range Managem. 19: 292–296.
Ellner, S. & A. Shmida. 1981. Why are adaptations for
long-range seed dispersal rare in desert plants? Oec-
ologia 51: 133–144.
Engler, A. H. G. 1895. Uber Amphicarpie bei Fleurya
podocarpa Wedd., nebst einigen allgemeinen Bermer-
kungen uber die Erscheinung der Amphicarpie und
Geocarpie. Sitzungsber. Ko¨nigl. Preuss. Akad. Wiss.
Berlin 1: 57–66.
Evenari, M., A. Kadouri & Y. Gutterman. 1977. Eco-
physiological investigations on the amphicarpy of Emex
spinosa (L.) Campd. Flora 166: 223–238.
Friis, I. 1991. Urticaceae. Pp. 79–116 in E. Launert &
G. V. Pope (editors), Flora Zambesiaca, Vol. 9, Part 6.
Whitstable Litho Printers, Whitstable, Kent.
Funk, V. A., R. Chan & S. C. Keeley. 2004. Insights into
the evolution of the tribe Arctoteae (Compositae: sub-
family Chichorioideae) using trnL-F, ndhF and ITS.
Taxon 53: 637–655.
Gillett, J. B., R. M. Polhill & B. Verdcourt. 1971. Le-
guminosae Part 3, subfamily Papillionoideae Part 1. Pp.
1–501 in E. Milne-Redhead & R. M. Polhill (editors),
Flora of Tropical East Africa. Royal Botanic Gardens,
Kew.
Haines, R. W. 1971. Amphicarpy in East African Cyper-
aceae. Mitt. Bot. Staatssamml. Mu¨ nchen 10: 534–538.
Volume 92, Number 4
2005 461Barker
Basicarpy, Geocarpy & Amphicarpy
& K. A. Lye. 1977. Studies in African Cypera-
ceae XV: Amphicarpy and spikelet structure in Tri-
anoptiles solitaria. Bot. Not. 130: 235–240.
Hepper, F. N. 1963. The bambara groundnut (Voandzeia
subterranea) and Kersting’s groundnut (Kerstingiella
geocarpa) wild in west Africa. Kew Bull. 16: 395–407.
Hesp, P. & A. McLachlan. 2000. Morphology, dynamics,
ecology and fauna of Arctotheca populifolia and Gazania
rigens nabkha dunes. J. Arid Environm. 44: 155–172.
Hilliard, O. M. & B. L. Burtt. 1973. Notes on some plants
from southern Africa, chiefly from Natal: III. Notes Roy.
Bot. Gard. Edinburgh 32: 308–314.
& . 1986. Notes on some plants from
southern Africa, chiefly from Natal: XII. Notes Roy. Bot.
Gard. Edinburgh 43: 189–228.
Hylander, N. 1929. Diasporenabtrennung und Diasporen-
Transport. Svensk Bot. Tidskr. 23: 184–218.
Imbert, E. 2002. Ecological consequences and ontogeny
of seed heteromorphism. Perspect. Pl. Ecol. Evol. Syst.
5: 13–36.
Inglis, G. J. 2000. Disturbance-related heterogeneity in
the seed banks of a marine angiosperm. J. Ecol. 88:
88–99.
Jumelle, H. & H. Perrier de la Baˆthie. 1912. Quelques
phane´rogames parasites de Madagascar. Rev. Ge´ n. Bot.
24: 321–328.
Kalin Arroyo, M. T. 1981. Breeding systems and polli-
nation biology in Leguminosae. Pp. 723–769 in R. M.
Polhill & P. H. Raven (editors), Advances in Legume
Systematics Part 2. Royal Botanic Gardens, Kew.
Kaul, V., A. K. Koul & M. C. Sharma. 2000. The under-
ground flower. Curr. Sci. 78: 39–44.
, M. C. Sharma & A. K. Koul. 2002. Reproductive
effort and sex allocation strategy in Commelina bengh-
alensis L., a common monsoon weed. Bot. J. Linn. Soc.
140: 403–413.
Keighery, G. J. 1982. Geocarpy in Tribulopsis R. Br. (Zyg-
ophyllaceae). Flora 172: 329–333.
Kiehn, M. & C. Puff. 1987. Karyological and dispersal
biological notes on Galium ankaratrense Puff Rubi-
aceae. Bull. Mus. Natl. Hist. Nat., B, Adansonia 9:
403–406.
Knevel, I. C. 2002. The Life History of Selected Coastal
Foredune Species of South Africa. Ph.D. Thesis,
Rhodes University
^
http://www.ru.ac.za/library/theses/
2002/knevel
&
.
, H. G. Venema & R. A. Lubke. 2002. The search
for indigenous dune stabilizers: Germination require-
ments of selected South African species. J. Coastal Con-
servation 8: 169–178.
Kornhall, P. 2004. Phylogenetic Studies in the Lamiales
with Special Focus on Scrophulariaceae and Stilbaceae.
Ph.D. Thesis, Uppsala Universitet.
Kuijt, J. 1969. The Biology of Parasitic Flowering Plants.
Univ. California Press, Berkeley and Los Angeles.
Lev-Yadun, S. 2000. Why are underground flowering and
fruiting more common in Israel than anywhere else in
the world? Curr. Sci. 79: 289.
Levyns, M. R. 1943. A revision of Trianoptiles Fenzl. J.
S. African Bot. 9: 21–26.
Linnemann, A. R. 1993. Phenological development in
Bambara groundnut (Vigna subterranea) at constant ex-
posure to photoperiods of 10 and 16h. Ann. Bot. 71:
445–452.
Manda´k, B. 1997. Seed heteromorphism and the life cy-
cle of plants: A literature review. Preslia 69: 129–159.
McKenzie, R. J., J. Samuel, E. M. Muller, A. K. W. Skin-
ner & N. P. Barker. 2005. Morphology of cypselae in
subtribe Arctotidinae (Compositae–Arctotideae) and its
taxonomic implications. Ann. Missouri Bot. Gard. 92:
569–594.
Meeuse, A. D. J. 1958. A possible case of interdepen-
dence between a mammal and a higher plant. Arch.
Ne´erl. Zool. 13: 314–318.
& W. G. Welman. 2000. Convolvulaceae. Pp. 1–
138 in G. Germishuizen (editor), Flora of Southern Af-
rica 28 Part 2. NBI, Pretoria.
Meney, K. A., J. S. Pate & K. W. Dixon. 1990. Compar-
ative morphology, anatomy, phenology and reproductive
biology of Alexgeorgia spp. (Restionaceae) from south-
western Western Australia. Austral. J. Bot. 38: 523–
541.
Milne-Redhead, E. & W. B. Turrill. 1952. Ranuncula-
ceae. Pp. 1–22 in W. B. Turrill & E. Milne-Redhead
(editors), Flora of Tropical East Africa. Whitefriars
Press, London.
Mu¨ller, S., R. Schultze-Kraft & I. Rodriguez. 2001. Am-
phicarpy in the tropical legume Centrosema rotundifol-
ium: A research project in eastern Venezuela. Pp. 559–
560 in Proceedings, XIX International Grassland Con-
gress, Sao Pedro, Sa˜o Paulo (
^
http://www.
uni-hohenheim.de/i3v/00068900/23019041.htm
&
).
Musselman, L. J. 1997. Hydnoraceae. Pp. 16–18 in G.
V. Pope (editor), Flora Zambesiaca. Royal Botanic Gar-
dens, Kew.
& J. H. Visser. 1987. Hydnora johannis in South-
ern Africa. Dinteria 19: 77–82.
& . 1989. Taxonomy and natural history
of Hydnora (Hydnoraceae). Aliso 12: 317–326.
Nickrent, D. L., A. Blarer, Y-L. Qiu, R. Vidal-Russell &
F. E. Anderson. 2004. Phylogenetic inference in Raf-
flesiales: The influence of rate heterogeneity and hori-
zontal gene transfer. BMC Evolutionary Biology 4: 40
[
^
http://www.biomedcentral.com/1471–2148/4/40
&
].
Pasquet, R. S., S. Schwedes & P. Gepts. 1999. Isozyme
diversity in bambara groundnut. Crop Sci. 39: 1228–
1236.
, G. Mergeai & J-P. Baudoin. 2002. Genetic di-
versity of the African geocarpic legume Kersting’s
groundnut, Macrotyloma geocarpum (Tribe Phaseoleae:
Fabaceae). Biochem. Syst. Ecol. 30: 943–952.
Plitmann, U. 1973. Biological flora of Israel. 4 Vicia sa-
tiva subsp. amphicarpa (Dorth.) Aschers. & Graebn. Is-
rael. J. Bot. 22: 178–194.
Pole Evans, I. B. 1926. Hydnora aficana. Plate 207 in I.
B. Pole Evans (editor), Flowering Plants of Southern
Africa, Vol. 6. The Specialty Press of South Africa,
Cape Town.
Rendle, A. B. 1917. Urticaceae. Pp. 240–306 in D. Prain
(editor), Flora of Tropical Africa, Vol. VI Sect. 2. L.
Reeve, Ashford, Kent.
Rivals, A. 1953. Sur quelques Leguminoseuses ge´ocar-
pes et amphicarpes. Rev. Int. Bot. Appl. Agric. Trop.
33: 244–249.
Shmida, A. 1981. Mediterranean vegetation in California
and Israel: Similarities and differences. Israel J. Bot.
30: 105–123.
Speroni, G. & P. Izaguirre. 2003. Caracterı´sticas biolo´-
gicas de la leguminosa nativa promisoria forrajera Tri-
folium polymorpha Poir. (Fabaceae, Faboideae). [Bio-
logical characteristics of the promissory forage native
legume Trifolium polymorphum Poir. (Fabaceae, Faboi-
deae)]. Agrociencia 7: 68–76;
^
http://www.fagro.edu.uy/
agrociencia/VOL7/1/p68–76.pdf
&
.
462 Annals of the
Missouri Botanical Garden
Stopp, K. 1958. Die verbreitungenshemmenden Einri-
chtungen in der Su¨dafrikanischen Flora. Bot. Stud. 8:
1–103.
Taylor, P. 1958. Tropical African Primulaceae. Kew Bull.
13: 133–149.
Tindale, M. D. & L. A. Craven. 1989. Three new species
of Glycine (Fabaceae: Phaseolae) from north-western
Australia, with notes on amphicarpy in the genus. Aus-
tral. Syst. Bot. 1: 399–410.
van der Pijl, L. 1982. Principles of Dispersal in Higher
Plants, 3rd ed. Springer Verlag, New York.
Verdcourt, B. 1969. Convolvulaceae. Pp. 1–161 in C. E.
Hubbard & E. Milne-Redhead (editors), Flora of Trop-
ical East Africa. Crown Agents for Oversea Govern-
ments and Administration, London.
Visser, J. H. 1981. Parasitic Flowering Plants of South
Africa. Juta, Cape Town.
. 1989. Hydnora triceps. Plate 1992 in D. J. B.
Killick & E. du Plessis (editors), Flowering Plants of
Southern Africa, Vol. 50. Creda Press, Cape Town.
Watson, L. & M. J. Dallwitz. 1992 onwards. Grass Genera
of the World: Descriptions, illustrations, identification,
and information retrieval; including synonyms, mor-
phology, anatomy, physiology, phytochemistry, cytology,
classification, pathogens, world and local distribution,
and references.
^
http://delta-intkey.com
&
.
Weiss, P. W. 1980. Germination, reproduction and inter-
ference in the amphicarpic annual Emex spinosa (L.)
Campd. Oecologia 45: 244–251.
Zohary, M. 1937. Die verbreitungsokologischen Verhalt-
nisse der Pflanzen Palaestinas. Beih. Bot. Centralbl.
A56: 1–155.
. 1962. Plant Life of Palestine, Israel and Jordan.
Ronald, New York.
