Content uploaded by Irene Esteban
Author content
All content in this area was uploaded by Irene Esteban on Oct 03, 2017
Content may be subject to copyright.
Modern soil phytolith assemblages used as proxies for Paleoscape
reconstruction on the south coast of South Africa
Irene Esteban
a
, Jan C. De Vynck
b
, Elzanne Singels
c
, Jan Vlok
b
, Curtis W. Marean
d
,
e
,
Richard M. Cowling
b
, Erich C. Fisher
d
, Dan Cabanes
a
,
f
, Rosa M. Albert
a
,
g
,
*
a
ERAAUB, Dept. de Prehist
oria, Hist
oria Antiga i Arqueologia, Universitat de Barcelona, Montalegre, 6, 08001, Barcelona, Spain
b
Centre for Coastal Paleosciences, Nelson Mandela Metropolitan University, P.O. Box 77000, Port Elizabeth, 6031, South Africa
c
Dept. of Archaeology, Beattie Building, University of Cape Town, University Avenue, Cape Town, South Africa
d
Institute of Human Origins, School of Human Evolution and Social Change, PO Box 872402, Arizona State University, Tempe, AZ, 85287-2402, USA
e
Faculty of Science, Nelson Mandela Metropolitan University, Port Elizabeth, Eastern Cape, 6031, South Africa
f
Plant Foods in Hominin Dietary Ecology Research Group, Max Planck Institute for Evolutionary Anthropology, Deutscher Platz 6, 04103, Leipzig, Germany
g
Catalan Institution for Research and Advanced Studies (ICREA), Spain
article info
Article history:
Available online 15 March 2016
Keywords:
Paleoenvironment
Landscape reconstruction
Human origins
Phytoliths
Modern soils
South African coast
abstract
South Africa continues to receive substantial attention from scholars researching modern human origins.
The importance of this region lies in the many caves and rock shelters containing well preserved evi-
dence of human activity, cultural material complexity and a growing number of early modern human
fossils dating to the Middle Stone Age (MSA). South Africa also hosts the world's smallest floral kingdom,
now called the Greater Cape Floristic Region (GCFR), with high species richness and endemism. In
paleoanthropological research, improving our capacity to reconstruct past climatic and environmental
conditions can help us to shed light on survival strategies of hunteregatherers. To do this, one must use
actualistic studies of modern assemblages from extant habitats to develop analogies for the past and
improve paleoenvironmental reconstructions. Here, we present a phytolith study of modern surface soil
samples from different GCFR vegetation types of the south coast of South Africa. In this study, the
phytolith concentration and morphological distribution are related to the physicochemical properties of
soils, the environmental conditions and the characterization of the vegetation for the different study
areas. Our results show that phytolith concentration relates mostly to vegetation types and the dominant
vegetation rather than to the type of soils. More abundant phytoliths from Restionaceae and woody/
shrubby vegetation are also noted from fynbos vegetation and grass phytoliths are a recurrent compo-
nent in all the vegetation types in spite of being a minor component in the modern vegetation. The grass
silica short cells from these plants, however, suggest a mix of C
3
and C
4
grasses in most of the vegetation
types with a major presence of the rondels ascribed to C
3
grasses. The exceptions are riparian, coastal
thicket and coastal forest vegetation, which are characterized by the dominance of C
4
grass phytoliths.
©2016 Elsevier Ltd and INQUA. All rights reserved.
1. Introduction
Recent studies on hominin evolution have emphasized the use
and effectiveness of phytoliths for paleoenvironmental re-
constructions (e.g. Barboni et al., 1999, 2007, 2010; Albert et al.,
2006, 2009, 2015; Bamford et al., 2006, 2008; Mercader et al.,
2009; Neumann et al., 2009; Rossouw, 2009; Cordova and Scott,
2010; Albert and Bamford, 2012; Cordova, 2013; Novello et al.,
2015). These studies are often paired with actualistic studies of
extant habitats, which can be used to develop analogies to help
reconstruct past environments. Phytolithsdthe silica micro-
remains formed in living plant tissuesdreproduce via the intra-
and extra-cellular structures of certain plants, and their composi-
tion makes them extremely durable and able to be preserved in
ancient soils and sediments for up to millions of years (Str€
omberg,
2002, 2004; Str€
omberg et al., 2007; Str€
omberg and McInerney,
2011; Dunn et al., 2015). Phytolith studies of palms, sedges, and
woody eudicots are commonplace in Africa, but most phytolith
studies of African paleoenvironments have focused on the grass
*Corresponding author. ERAAUB, Dept. de Prehist
oria, Hist
oria Antiga i
Arqueologia, Universitat deBarcelona, Montalegre, 6, 08001, Barcelona, Spain.
E-mail address: rmalbert@ub.edu (R.M. Albert).
Contents lists available at ScienceDirect
Quaternary International
journal homepage: www.elsevier.com/locate/quaint
http://dx.doi.org/10.1016/j.quaint.2016.01.037
1040-6182/©2016 Elsevier Ltd and INQUA. All rights reserved.
Quaternary International 434 (2017) 160e179
family since it exhibits a variety of different shapes that have been
linked to differences in paleovegetation. Commonalities between
phytolith studies allows them to be subdivided into three broad
types of research: 1) studies of modern plants that are aimed at
identifying diagnostic morphotypes, which can later be used for the
identification of botanical remains in fossil records (e.g. Runge,
1996, 1999; Runge and Runge, 1997; Albert and Weiner, 2001;
Bamford et al., 2006; Fahmy, 2008; Albert et al., 2009; Mercader
et al., 2009; Rossouw, 2009; Cordova and Scott, 2010; Eichhorn
et al., 2010; Mercader et al., 2010; Radomski and Neumann, 2011;
Cordova, 2013; Novello and Barboni, 2015); 2) studies of modern
soils that investigate the limitations of phytoliths for reconstructing
vegetation and climatic conditions (e.g. Bremond et al., 2005a,
2005b, 2008; Albert et al., 2006, 2015; Barboni et al., 2007;
Barboni and Bremond, 2009; Mercader et al., 2011; Novello et al.,
2012; Cordova, 2013; Garnier et al., 2013); and 3) combined
studies of modern plants and modern soils for understanding
deposition processes of phytoliths and postdepositional effects that
may affect their preservation under different mineralogical and
climatic conditions (Albert et al., 2006).
Fossil and genetic records indicate that the modern human
lineage evolved in sub-Saharan Africa during the Middle Pleisto-
cene, approximately 200,000 years ago (Clark et al., 2003; White
et al., 2003; McDougall et al., 2005; Fagundes et al., 2007; Gonder
et al., 2007; Smith et al., 2007; Behar et al., 2008). Based on
recent genetic studies, the progenitor populations of modern
humans may have been located in southern Africa (e.g. Schlebusch
et al., 2012, 2013), and detailed records from numerous South Af-
rican archaeological sites shows the subsequent cultural and
behavioral evolution characteristic of modern humans, under-
scoring the importance of this region in our understanding of
modern human origins (e.g. Watts, 1999, 20 02; Henshilwood et al.,
2001, 2002, 2004; d'Errico et al., 2005; d'Errico and Henshilwood,
2007; Brown et al., 2009, 2012; Jerardino and Marean, 2010;
Texier et al., 2010; Marean et al., 2014). The south coast of South
Africa is also located in the hyper-diverse Greater Cape Floristic
Region (GCFR), which contains a high variety of edible plants
(fruits, geophytes, etc.) (De Vynck et al., 2015) and a rich marine
ecosystem (Menge and Branch, 2001; De Vynck et al., 2015) that
may have supported human populations throughout the Pleisto-
cene (Marean, 2010, 2011). Phytolith studies have already contrib-
uted to our understanding of modern human evolution in the GCFR
(Albert and Marean, 2012), but much remains to be resolved about
the plant resources that Pleistocene hunteregatherers used for
food, fuel, and other activities and the environmental and climatic
conditions present then.
2. Background to the research region
The GCFR comprises the world's most diverse extratropical flora,
both in terms of richness and endemism (Colville et al., 2014), and it
is comprised of five floristically and structurally-defined biomes
that are distributed based on rainfall amount and type, soil con-
ditions, and the underlying geological substrate. Listed in terms of
the area of spatial extent, these biomes are: Fynbos, Renosterveld,
Succulent Karoo, Thicket, and Forest. Our study area is located along
the southern coast of the GCFR (Fig. 1) in a centralized area that
receives rain from both the winter-driven circumpolar westerly
systems and post-frontal events that are caused by moist air, which
is advected across the warm Indian Ocean to produce rain at any
time of the year (Deacon et al., 1992; Engelbrecht et al., 2014). The
annual rainfall for the study area varies from 398 to 510 mm
[supplied by South African Weather Service (Linnow, 2012 n.p)]
with the driest and warmest periods occurring in late summer. The
area generally has a moderate climate with an annual mean
temperature of around 18
C, minimum average of 6
C and
maximum average temperature not higher than 30
C. Frost occurs
rarely.
The study area is underlain by the Palaeozoic deposits of the
Cape Supergroup that are characterized by Table Mountain Group
sandstones (associated with mountains and ridges and coastal cliffs
of the Cape Folded Belt) and Bokkeveld shales (associated with
coastal forelands). In the eastern part of the study area the Palae-
ozoic sediments dip in an east-west fault and are filled by the
Cretaceous Enon Formation comprising conglomerates and mud-
stones (Rogers, 1984; Malan, 1987). Much of the southern coastal
plain is underlain by Pliocene limestone of the Bredasdorp For-
mation, which has been covered by alkaline Pliocene-Pleistocene
sands of marine origin near the coastal margin. Inland of these
sands are found patches of older aeolian sands that are leached and
acidic (Rebelo et al., 1991). In comparison, shale and mudstone-
derived soils are moderately fertile, yet those associated with
leached sands are infertile. The calcareous sands associated with
limestone, calcrete and coastal dunes are also relatively infertile
due to their high alkalinity and subsequent low levels of plant-
available phosphorous (Thwaites and Cowling, 1988).
The vegetation of the region has been mapped at 1:30,000 scale
by Vlok and De Villiers (2007) and we used this assessment for
identifying vegetation units within our study area. Four biomes are
represented within the study area: Fynbos, Renosterveld, Thicket,
and Forest (Bergh et al., 2014). Fynbos is characterized as scle-
rophyllous, fire-prone shrubland, which is associated with acidic,
infertile sands, and having a graminoid component that is domi-
nated by Restionaceae. Renosterveld is a fire-prone shrubland
found on shales and mudstones where the graminoid component is
dominated by a mixture of C
4
and C
3
Poaceae. Thicket is a dense
shrubland dominated by large and leathery-leaved shrubs and low
trees that is associated with areas that are fire-protected (deep
valleys and the coastal margin) and located on clay-rich soils and
alkaline sands along the coast. Graminoids only comprise a few
shade-tolerant grasses within the thicket. Lastly, forests are char-
acterized by a dense cover of evergreen trees that are associated
with fire-free enclaves located in the foothills of the Folded Belt.
Graminoids within these forests comprise shade tolerant Poaceae
and Cyperaceae.
There is also a long and detailed record of human occupation in
the study area. Blombos cave, located near Still Bay (Western Cape
Province), preserves a detailed record of novel modern human
behaviors dating during the Late Pleistocene, approximately
70e100 ka, which includes early evidence for bone tools, marine
shell beads, pigment processing, and geometric artwork
(Henshilwood and van Niekerk, 2014 and references therein). Near
Mossel Bay, the Pinnacle Point Archaeological Complex (PPAC)
consists of a series of caves and rock shelters that formed in excess
of 1 million years ago (Marean et al., 2007). Preserved archaeo-
logical deposits in several of these sites shows distinctive behav-
ioral complexity that includes early evidence for marine resource
exploitation (Marean et al., 2007; Jerardino and Marean, 2010); use
of pigment (Marean et al., 2007; Watts, 2010); heat-treatment of
lithic raw materials (Brown et al., 2009); and the production of
microlith-tipped projectile weapons (Brown et al., 2012). Within
the immediate area of Pinnacle Point are several vegetation types
with different taxonomic and functional assemblages that would
have been accessible within the daily foraging range for a hunter-
egatherer (~10 km). This highly diverse flora would extend the
foraging opportunities for modern humans occupying the Pinnacle
Point area (Marean et al., 2014).
Studies of
d
18
O and
d
13
C isotope records in speleothems, have
provided high resolution records of climatic and environmental
changes at PP dating from the late Middle Pleistocene to the Late
I. Esteban et al. / Quaternary International 434 (2017) 160e179 161
Holocene. The speleothem record from Crevice Cave documents a
strongly C
3
plant flora from ~90 to 74 ka, and then an increase in C
4
plants during the cooler MIS4, which is coincident with a shift in
d
18
O that is interpreted to suggest increasing amounts of summer
rain (Bar-Matthews et al., 2010).
Diverse and detailed paleoenvironmental records have also
been derived from the analysis of large mammal and micro-
mammal fauna, charcoal, phytoliths and pollen from archaeological
and paleoanthropological sites at Pinnacle Point. Changes in
micromammal taxa from deposits dating to MIS 6 and MIS5e-a at
cave PP9C indicated climatic shifts, from colder conditions and
more open grassy environment to warmer and wetter conditions
and dense vegetation (Matthews et al., 2011). The phytolith study
from PP13B cave showed a low grass representation during the
earlier occupations, from ~160 to ~120 ka, which was inferred to
indicate the presence of Fynbos that naturally has a low grass
component (Albert and Marean, 2012). The presence of Grass Silica
Short Cell phytoliths (GSSCsdcharacteristic of C
4
grasses) during
MIS5c (Shelly Brown Sand) was interpreted as being indicative of at
least some summer rain (Albert and Marean, 2012).
This study aims to assess the potential for using phytolith
studies from modern surface soil samples to characterize the
vegetation of the study area of the GCFR so as to better identify past
environments and climate changes on the south coast of South
Africa. The ultimate goal of this study is the identification of phy-
tolith morphotypes and phytolith assemblages with the potential
to be used as proxies for the paleoenvironment, paleoclimate and
paleovegetation reconstructions at Pinnacle Point and other south
coast sites. The present study builds on, and differs from, previous
studies in South Africa that have been carried out by Cordova (2013)
and Cordova and Scott (2010) based on three aspects: i) our study
focuses specifically on the south coast; ii) our study incorporates
the whole phytolith assemblage that is preserved in modern soils,
including eudicotyledoneous plants (here after ‘eudicots’) and
graminoids; and iii) our study accounts for the taphonomic pro-
cesses that might have affected phytolith preservation and thus
their representation in the phytolith record.
3. Materials and methods
3.1. Materials
Fieldwork was undertaken over four field seasons (2006e2011)
on the south coast of South Africa (Fig. 1). A total of seventy-two
modern soil samples were collected during June (2006 and 2009),
November (2010) and from August to October (2011) from twelve
vegetation types corresponding to the general 5 biomes, namely
limestone fynbos, sand fynbos, mountain fynbos, grassy fynbos
(Fynbos biome), renosterveld (Renosterveld biome), strandveld,
dune cordon, subtropical thicket, coastal thicket (Thicket biome),
coastal forest (Forest Biome), riparian and wetlands (both azonal)
(Table 1). Soils were collected beneath intact (not impacted by
human-managed activities) vegetation at 5e10 cm depth in the
different vegetation types in proportion to their relative extents in
the study area. The study area has not experienced any tectonic
activity or significant climate shifts over the past few thousand
years (Marean et al., 2014) thus the phytolith assemblages in the
topsoil should be representative of the modern vegetation. Since
renosterveld in the southern Cape has been extensively trans-
formed for agriculture most samples were from slopes too steep for
cultivation (Kemper et al., 2000). Several related aspects that might
have influenced the configuration of phytolith assemblages and
their concentration and preservation in modern soils have also
been taken into account, including: vegetation type, dominant
plant species, soil texture, and soil pH (Table 1).
Fig. 1. Map showing the location of the modern surface soil samples by vegetation types and the major GCFR biomes after Mucina and Rutherford (2006).
I. Esteban et al. / Quaternary International 434 (2017) 160e179162
Table 1
Description of samples provenance and main phytolith results: soil type and soil pH, estimated number of phytoliths per gram of sediment, total number of phytoliths
identified, percentage of weathered morphotypes and D/Pº, Fy, Iph and Ic indices.
Sample
number
Vegetation
type
Coordinates
(Long, Lat)
Dominant taxa Soil
type
Soil
pH
Estimated#
of phytoliths
g/sed
# Phytoliths
identified
% Weathered
morphotypes
D/Pº
index
Fy
index
Iph
index
Ic
index
Fynbos biome
LF11-19 Limestone
fynbos
34,42828333,
21,3311
Restionaceae; Asteraceae;
Carpobrotus acinaciformus
Sand 7.6 7700 22 15 eeee
LF11-23 Limestone
fynbos
34,34333333,
21,2214
Themeda triandra;Olea europea
subs. Africana; Restionaceae
Sand 7.3 60,000 97 0 0.8 1.0 0.1 0.8
LF11-25 Limestone
fynbos
34,29508333,
21,27883333
Restionaceae; Protea repens;P.
obtusifolium;Leucadendron
platyspermum
Sand 7.2 74,000 30 4 eeee
LF11-26 Limestone
fynbos
34,31408333,
21,33561667
Restionaceae; Erica sp;
Asteraceae
Sand 7.5 61,000 21 0 eeee
LF11-68 Limestone
fynbos
34,31581667,
21,03331667
Leucadendron coniferum;
Restionaceae
Sand 7 160,000 62 5.6 0.6 1.3 0.0 0.6
LF11-74 Limestone
fynbos
34,35955,
20,89986667
Leucadendron coniferum;
Restionaceae
Sand 7.6 69,000 49 8.9 eeee
LF11-85 Limestone
fynbos
34,29261667,
21,7305
Restionaceae Sand 7.3 71,000 68 5.2 0.6 1.0 0.4 0.6
SF10-05 Sand fynbos 34,34465,
21,86868333
Freesia sp. Sand 5.8 99,000 83 3.5 0.4 0.4 0.4 0.7
SF10-06 Sand fynbos 34,34513333,
21,86923333
Restionaceae Sand 5.9 690,000 74 7.5 0.2 0.4 0.3 0.4
SF11-37 Sand fynbos 34,35335,
21,69413333
Leucadendron salignum;
Leucospermum praecox;
Restionaceae
Sand 5.3 300,000 89 2.5 0.7 0.8 0.3 0.7
SF11-42 Sand fynbos 34,20858333,
21,70006667
Poaceae; Aloe ferox;Osyris
compressa
Sand 3.5 2,900,000 211 1.4 0.0 0.0 0.3 0.8
SF11-43 Sand fynbos 34,20115,
21,69343333
Elytropappus rhinocerotis;Erica;
Poaceae; Schotia afra
Loam 4.4 1,290,000 210 0.5 0.0 0.1 0.4 0.9
SF11-45 Sand fynbos 34,22953333,
21,6038
Leucadendron salignum;Protea
repens;Leucospermum praecox;
Erica sp.
Sand 4.8 1,000,000 163 3.9 0.5 1.3 0.1 0.8
SF11-47 Sand fynbos 34,2928,
21,5772
Leucospermum praecox;Erica spp;
Leucadendron spp.
Sand 5.8 290,000 109 1 1.0 1.6 0.4 0.7
SF11-62 Sand fynbos 34,21151667,
21,5496
Restionaceae; Leucospermum
muirii;Brunia
Sand 4.2 225,000 109 7.8 0.3 0.5 0.7 0.6
SF11-82 Sand fynbos 34,26395,
21,59831667
Thamnochortus insignis;
Leucadendron galpinii;
Leucospermum praecox
Sand 5.3 154,000 93 5.7 0.3 0.4 0.0 0.7
GF10-10 Grassy fynbos 34,08936667,
21,91323333
Themeda triandra;Eragostris
capensis;E. curvula
Sand 5.5 2,330,000 182 7.6 0.0 0.0 0.3 0.8
GF10-11 Grassy fynbos 34,08915,
21,91358333
Pentaschistis colorata Sand 5.9 2,340,000 175 6.4 0.0 0.1 0.2 0.7
GF10-14 Grassy fynbos 34,0533,
22,99021667
Poaceae Sand 5.9 1,310,000 153 4.4 0.0 0.1 0.2 0.4
MF12-01 Mountain
fynbos
33,880997,
22,043091
Restionaceae; Cyperaceae Sand 5 900,000 127 0 0.2 0.2 0.2 0.9
F/RV10-07 Fynbos/
Renosterveld
34,03696667,
22,18781667
Tritoniopsis spp. Restionaceae Sand 5.5 1,890,000 115 2.8 0.3 0.3 0.7 0.4
Renosterveld biome
RV09-01 Renosterveld 34,15241389,
22,0015
Themeda tiandra;Brachiaria
serrate;Eragrostis sp.
Loam 5.4 2,410,000 155 27.7 0.1 0.1 0.2 0.7
RV09-02 Renosterveld 34,15241389,
22,0015
Restionaceae; Poaceae Loam 5.2 1,980,000 172 11.4 0.1 0.1 0.7 0.7
RV09-03 Renosterveld 34,15229444,
22,00163056
Elytropappus rhinocerotis;Searsia
lucida
Loam 5 3,400,000 130 11.8 0.0 0.0 0.1 0.8
RV09-07 Renosterveld 34,08574444,
21,25053333
Elytrotrappus rhinocerotis Sandy 5.4 2,040,000 104 19.7 0.1 0.1 0.5 0.6
RV09-08 Renosterveld 34,08585833,
21,25053056
Eragrostis curvula;Elytrotrappus
rhinocerotis
Sandy 5.8 2,250,000 193 19.4 0.0 0.0 0.3 0.7
RV10-01 Renosterveld 34,16553056,
22,005425
Romulea flava Loam 6.2 3,000,000 201 4 0.0 0.0 0.2 0.7
RV10-02 Renosterveld 34,03324722,
22,30835
Moraea sp. Loam 5.9 2,500,000 179 2.4 0.0 0.0 0.4 0.7
RV10-03 Renosterveld 34,01595,
22,00685
Hypoxis villosa Loam 5.8 2,740,000 196 9.6 0.1 0.2 0.1 0.6
RV10-04 Renosterveld 34,10858611,
22,04568333
Ehrharta bulbosa Loam 5.6 2,990,000 171 4.8 0.1 0.2 0.4 0.9
RV10-09 Renosterveld 34,28396111,
21,76596111
Elytropappus rhinocerotis Sand 5.8 1,380,000 177 5.8 0.1 0.1 0.8 0.5
RV11-04 Renosterveld 34,28876667,
21,89245
Eriocephalus africanus;
Elytopappus rhinocerotis;Ruschia
sp.
Sand 5 1,480,000 230 0.4 0.1 0.1 0.6 0.8
RV11-11 Renosterveld 34,1148,
21,25161667
Elytropappus rhinocerotis;
Eriocephalus africanus;Themeda
triandra
Loam 5.3 4,850,000 220 2.4 0.1 0.2 0.2 0.9
(continued on next page)
I. Esteban et al. / Quaternary International 434 (2017) 160e179 163
Table 1 (continued )
Sample
number
Vegetation
type
Coordinates
(Long, Lat)
Dominant taxa Soil
type
Soil
pH
Estimated#
of phytoliths
g/sed
# Phytoliths
identified
% Weathered
morphotypes
D/Pº
index
Fy
index
Iph
index
Ic
index
RV11-15 Renosterveld 34,23876667,
21,79173333
Elytropappus rhinocerotis;
Restionaceae; Asteraceae
Loam 6.8 130,000 64 0 0.4 0.5 0.6 0.7
RV11-22 Renosterveld 34,10561667,
21,30705
Elytropappus rhinocerotis;Searsia
graveolens; Poaceae
Loam 4.9 2,550,000 186 0.5 0.0 0.0 0.2 0.8
RV11-32 Renosterveld 34,11173333,
21,01186667
Elytropappus rhinocerotis;Aloe
ferox;Ehrharta villosa;Themeda
triandra
Sand 4.8 2,260,000 200 1.5 0.0 0.0 0.4 0.9
RV11-58 Renosterveld 34,29145,
21,01003333
Roepera morgsana;Euphorbia
mauritanica;Aloe ferox
Loam 6.3 620,000 187 3.3 0.1 0.1 0.5 0.7
RV11-67 Renosterveld 34,25323333,
20,9884
Searsia glauca;Aloe ferox;
Euphorbia mauritanica
Loam 5.2 1,200,000 186 3.3 0.0 0.0 0.5 0.8
Thicket biome
StT09-04 Subtropical
thicket
34,235825,
21,91684722
Sideroxylon inerme Sandy/
loam
6.1 450,000 94 13.8 0.1 0.1 0.3 0.7
StT09-05 Subtropical
thicket
34,235825,
21,91685
Eriocephalus africanus Sand 7 260,000 79 16.8 0.2 0.2 0.8 0.8
StT09-06 Subtropical
thicket
34,25073056,
21,90236111
Searsia pterota Sand 5.3 1,190,000 127 15.9 0.1 0.1 0.5 0.6
CT06-01 Coastal
thicket
34,20847222,
22,08953889
Dense shrubby vegetation Sand 8 150,000 61 39 0.1 0.1 0.5 0.7
CT06-02 Coastal
thicket
34,20847222,
22,08953889
Dense shrubby vegetation Sand 8 230,000 68 26.9 0.2 0.2 0.3 0.6
CT06-03 Coastal
thicket
34,20847222,
22,08953889
Dense shrubby vegetation Sand 7.9 78,000 39 46.6 eeee
CT06-04 Coastal
thicket
34,20847222,
22,08953889
Dense shrubby vegetation Sand 7.5 23,000 12 29.4 eeee
CT06-05 Coastal
thicket
34,20847222,
22,08953889
Open grassy area of Eragrostis
curvula
Sand 8 76,000 71 5.3 0.0 0.0 1.0 0.1
CT06-06 Coastal
thicket
34,20847222,
22,08953889
Acacia karroo Sand 6.6 100,000 54 35.7 0.1 0.1 0.7 0.6
CT06-07 Coastal
thicket
34,20847222,
22,08953889
Dense shrubby vegetation Loam 8.3 920,000 20 56.5 eeee
CT10-12 Coastal
thicket
34,05521667,
22,38678333
Eragrostis rehmannii;Asparagus
sp.; Chasmanthe aethiopica;
Rhoicissus sp.
Loam 6 28,000 45 4.3 eeee
STV11-01 Strandveld 34,37883333,
21,41118333
Stoebe cinorera;Erharta villosa.
Restionaceae
Sand 5.3 185,000 103 4.8 0.3 0.7 0.2 0.8
STV11-16 Strandveld 34,4132,
21,37246667
Poaceae; Restionaceae;
Sideroxylon inerme
Sand 5.9 513,000 122 4.5 0.1 0.2 0.8 0.7
STV11-20 Strandveld 34,36381667,
21,47653333
Agathosma sp.; Asteraceae Sand 7.6 20,000 24 14.3 eeee
STV11-21 Strandveld 34,3731,
21,64463333
Asteraceae; Poaceae Sand 7.4 83,000 60 5 0.1 0.1 0.0 0.8
STV11-51 Strandveld 34,37893333,
20,85816667
Restionaceae; Olea europaea
subsp. Africana;Metalasia sp.;
Agathosma sp.
Sand 7.5 75,000 48 8.3 eeee
STV11-54 Strandveld 34,40275,
21,21333333
Metalasia sp.; Searsia graviolens;
Chrysanthemoides monilifera;
Thamnochortus insignis
Sand 7.9 7500 11 10 eeee
STV11-55 Strandveld 34,38378333,
21,22213333
Sideroxylon inerme;Chasmanthe
aethiopica;Searsia glauca;
Thamnochortus insignis
Sand 7.6 7500 11 22.2 eeee
DC11-18 Dune cordon 34,41866667,
21,35148333
Restionaceae; Chrysanthemoides
monilifera;Sideroxylon inerme
Sand 5.6 83,000 65 3.2 0.3 0.3 0.9 0.7
DC11-38 Dune cordon 34,39383333,
21,41353333
Roepera morgsana;Carpobrotus
accinaciformus;Euphorbia
mauritanica
Sand 6.7 7000 36 8.8 eeee
DC11-39 Dune cordon 34,43165,
21,32283333
Osteospernum moniliferum;
Carpobrotus edulis;Roepera
morgsana
Sand 7.4 27,000 19 5.3 eeee
DC11-71 Dune cordon 34,3667,
20,92925
Searsia glauca;Phylica ericoides;
Asteraceae
Sand 7.9 22,000 9 11.1 eeee
DC11-72 Dune cordon 34,36693333,
20,92425
Sideroxylon inerme Sand 7.3 37,000 21 0 eeee
DC11-73 Dune cordon 34,36998333,
20,91175
Roepera morgsana; Asteraceae;
Carpobrotus edulis
Sand 7.8 28,000 27 0 eeee
DC11-75 Dune cordon 34,39611667,
20,8443
Pterocarpus tricuspidatus;Phylica
ericoides
Sand 7.8 38,000 13 0 eeee
Forest biome
CF10-13 Coastal Forest 34,06155,
23,03056667
Celtis sp.; Vepris sp. Loam 6 345,000 108 2.7 0.0 0.0 0.2 0.3
Azonala
RP11-13 Riparian 34,25971667,
21,78031667
Acacia karroo; Poaceae;
Asteraceae
Sand 5.5 2,440,000 166 0.6 0.0 0.0 0.7 0.5
RP11-14 Riparian 34,26373333,
21,77966667
Acacia karroo; Poaceae;
Asteraceae
Sand 6.2 1,240,000 203 3 0.0 0.0 0.8 0.4
I. Esteban et al. / Quaternary International 434 (2017) 160e179164
The preliminary phytolith results of modern soils from the area
showed the presence of some morphotypes, similar to those
described by Cordova and Scott (2010) and Cordova (2013) as
papillated and non-papillated rondels and which these authors
describe as being characteristic of Restionaceae. In order to clarify
whether our morphotypes corresponded to this family we present
here the first results of the analyses of four Restionaceae plant
specimens collected in the study area (Elegia juncea,Thamnochortus
insignis and T.rigidus and Restio triticeus) (Table A1).
3.2. Methods
3.2.1. Physico-chemical analyses
The physico-chemical analyses (soil texture and pH) of the 2011
samples werecarried out at Bemlab laboratoriesin Cape Town, South
Africa. The pH analysisof the samples from 2006, 2009, and 2010 was
carried out using a CRISON BASIC 20 þpH-meter and mineralogy
was analyzed using Fourier Transformed Infrared Spectroscopy
(FTIR), all at the Laboratory of Archaeology of the University of Bar-
celona. Infrared spectra were obtained using KBr pellets (potassium
bromide) at 4 cm
1
resolution with a Nicolet iS5 spectrometer.
3.2.2. Phytolith analysis
The phytolith extraction from modern soils and Restionaceae
plant specimens was carried out at the Laboratory of Archaeology of
the University of Barcelona. Methods for the modern soils followed
Katz et al. (2010). An initial sediment weight of between 30 and
50 mg was required. Carbonateminerals were dissolvedadding 50
m
L
of hydrochloric acid (6 N HCl). After the bubbling ceased, 450 ml of
2.4 g/ml sodium polytungstate solution Na
6
(H
2
W
12
O
40
)vH
2
O] was
added. The tube was vortexed, sonicated and centrifuged for 5 min at
5000 rpm (MiniSpin plus, Eppendorf). The supernatant was subse-
quently removed to a new 0.5 ml centrifuge tube and vortexed. For
examination underthe optical microscope, an aliquot of 50 ml of the
supernatant was placed on a microscope slide and covered with a
24 mm 24 mm cover-slip. Quantificationof the total phytolithswas
based on 20 fields at 200magnification whereas morphological
identification of phytoliths took place at 400magnification. Mini-
mally, 200 phytoliths were counted for the morphological analysis
and when this was notpossible we analyzed only those samples with
a minimum number of 50 phytoliths in order to obtain as much in-
formation as possible (Albert and Weiner, 2001).
The phytolith extraction process from Restionaceae modern
plants is a modification of the methods described in Albert and
Weiner (2001) and Parr et al. (2001). Washed and air dried aliquots
were weighed and burned in a muffle furnace at 500
C for 2.5 h. The
ash was treated with an equivolume solution of 1 N HCl for 30 min at
100
C, to leave only the siliceous minerals where phytoliths are
found. The inorganicacid insoluble fraction (AIF)was centrifuged, re-
suspended in deionized water, and centrifuged again. The superna-
tant was discarded and the washing was repeated three times. The
pellet was transferred to a glass Petri dish and about 10 ml of 30%
hydrogenperoxide (H
2
O
2
) was added. The samplewas evaporated on
a hot plate at 70
C. More hydrogen peroxide was added as needed
until all bubbling ceased. The remaining residue was carefully
removed from the Petri dish, weighed, and transferred into an
Eppendorf tube for storage. Microscope slides were prepared using
around 1 mg of the final fraction with Entellan mounting media
(Merck). Phytolith counting and identification followed the pro-
cedures described above for the soil samples.
3.2.3. Phytolith morphological identification and classification
Morphological identification of phytoliths was based on our
modern plant reference collection that hasbeen compiled from plant
species that are found in the study area (Esteban, unpublished data).
The study also includes the four Restionaceae specimens described
here. Reference collections from other study areas have also been
consulted (Albert and Weiner, 2001; Bamford et al., 2006; Albert
et al., in press). Additionally, the results were also compared to the
modern reference collection of graminoids from South Africa
developed by Rossouw (2009), Cordova and Scott (2010), and
Cordova (2013). Standard literature (Twiss et al.,1969; Piperno,1988,
2006; Mulholland and Rapp, 1992 and references therein) was also
accessed when necessary. The terminology for describing phytolith
morphotypes was based on the anatomical and/or taxonomic origin
of the phytoliths.When this was not possible, geometricaltraits were
followed. The International Code for Phytolith Nomenclature (ICPN)
was also followed where possible (Madella et al., 2005).
3.2.4. Statistical analysis
Statistical analysis one-way ANOVA and a post hoc Tukey
Honest Significant Differences (HSD) tests were performed on the
dataset in order to identify those morphotypes that are statistically
representative of specific vegetation types or biomes. These tests
account for the variation in total numbers of phytolith morpho-
types counted for each vegetation type and biome. The null hy-
pothesis in each test assumes that there is no significant difference
in the distribution of phytolith morphotypes between the different
vegetation types and biomes from the GCFR.
The non-parametric Kruskal Wallis test was also used to identify
significant differences in the mean values of phytolith indices be-
tween vegetation types since data were non-normally distributed
(ShapiroeWilk test for normality).
All statistical procedures were performed with the JMP-
SAS12.1.0 software. Samples from mountain fynbos, fynbos/renos-
terveld, dune cordon, coastal forest and wetlands were excluded
Table 1 (continued )
Sample
number
Vegetation
type
Coordinates
(Long, Lat)
Dominant taxa Soil
type
Soil
pH
Estimated#
of phytoliths
g/sed
# Phytoliths
identified
% Weathered
morphotypes
D/Pº
index
Fy
index
Iph
index
Ic
index
RP11-48 Riparian 34,33896667,
21,02705
Euphorbia mauritanica;Roepera
morgsana
Sand 7.3 97,000 55 3.8 0.1 0.1 0.8 0.5
RP11-49 Riparian 34,32151667,
20,77593333
Acacia karroo; Poaceae Loam 6.9 4,300,000 220 2.7 0.0 0.0 0.3 0.6
RP11-59 Riparian 34,30161667,
21,01116667
Acacia karroo; Poaceae Sand 7.3 850,000 156 1.9 0.0 0.0 0.8 0.7
RP11-79 Riparian 34,04986667,
21,75665
Acacia karroo;Albuca maxima;
Poaceae; Aloe ferox
Sand 4.7 1,170,000 193 3.1 0.0 0.0 0.7 0.5
RP11-80 Riparian 34,08223333,
21,77531667
Acacia karroo;Albuca maxima Sand 5.4 340,000 64 0 0.0 0.0 0.8 0.5
W10-08 Wetland 34,0608,
22,10786667
Cyperus textilis;Stenotaphrum
secundatum;Arundo sp.
Loam 6 1,590,000 183 7.6 0.1 0.1 0.2 0.7
I. Esteban et al. / Quaternary International 434 (2017) 160e179 165
from the statistical analysis due to the absence of replicated sam-
ples from the same vegetation type or because phytoliths were
absent or identified in few concentrations as occurred for dune
cordon vegetation. Only samples from riparian vegetation repre-
sent Azonal vegetation.
4. Results
Table 1 lists the results of the analysis of the sediment samples,
which includes information about the samples, vegetation type,
geographical coordinates, the dominant taxa, the type of soils and
the pH of soils as well as the estimated phytolith concentration in
sediments, the total number of phytoliths that were identified, the
percentage of weathered morphotypes and the D/Pº, Fy, Iph and Ic
indices values. Table 2 lists the distribution of phytolith morpho-
types in the samples and their plant attribution (plant types and
plant parts). These morphotypes were later related to the vegeta-
tion type, the dominant plants, and the type of soil. All of the
samples showed an acidic to a moderate alkaline pH. FTIR results
further indicated that clay and quartz are the main mineral com-
ponents in all the samples but these minerals occurred in different
proportions depending on the type of soils.
Phytoliths were identified in different concentrations in all of
the analyzed samples. Renosterveld, riparian vegetation, and grassy
fynbos vegetation provided the highest amount of phytoliths per
gram of sediment (g/sed) (Table 1). In contrast, the samples that
were collected from limestone fynbos, dune cordon, coastal thicket
and strandveld vegetation had the lowest amount of phytoliths.
Weathered morphotypes showing irregular shapes and pitted
surfaces were identified in low numbers in all the samples and
almost never above 20%, with the exception of samples from coastal
thicket. Together with phytoliths, diatoms and sponge spicules
were also recovered from most of the samples, albeit in varying
amounts. Samples from riparian, subtropical and coastal thicket
vegetation contained the highest number of diatoms.
Grass phytoliths were common to all vegetation types analyzed.
Samples from limestone fynbos showed a low grass phytolith
component (Table 2). Among grasses, GSSCs were the most repre-
sentative morphotypes, together with the presence of prickles,
bulliforms fan-shaped as well as long cell with decorated margins
(mainly echinates) (Fig. 2aec, respectively). GSSCs have been used
widely as climatic indicators due to their characteristic morphol-
ogies that generally relate to different Poaceae subfamilies. For
example, rondels, which include trapezoids are common in C
3
Pooideae, bilobates and crosses in C
4
Panicoideae and saddles in C
4
Chloridoideae (Twiss et al., 1969; Twiss, 1987, 1992; Fredlund and
Tieszen, 1994, 1997). In addition to Pooideae, GSSC rondels have
also been identified in South Africa in the C
3
Ehrhartoideae and
Danthonioideae subfamilies (Cordova and Scott, 2010; Cordova,
2013). Finally, oblongs are common in C
3
Pooideae grasses
(Rossouw, 2009). Accordingly, GSSCs have been classified in four
categories: rondels (rondels, trapezoids, towers), lobates (bilobates,
crosses and polylobates), saddles and oblongs. Rondels were the
most common GSSCs recognized in our samples (Fig. 2def), while
lobates (Fig. 2g and h) and saddles (Fig. 2i) were also represented,
albeit in lower numbers. The distribution of grasses in the phytolith
record was traced using a ternary diagram highlighting composi-
tion differences among vegetation types and GSSCs (Fig. 3). Here
the category rondels include also the oblong morphotypes. The
latters dominate most of the vegetation types while samples from
riparian vegetation present higher proportion of saddles (Fig. 3).
Among eudicot plants the most representative phytoliths were
parallelepiped blockys, epidermal ground mass polyhedral (Fig. 2j),
eudicot hair cells (Fig. 2k) and sclerenchyma (Fig. 2l). Psilate and
rugulate spheroids were also identified in high frequencies mainly
in samples from limestone fynbos vegetation.
Other common morphotypes that were identified included large
spheroids, sometimes showing protuberances, with granulate or
verrucate decoration (~25
m
m) (Fig. 4aec) and spheroids showing
spiraling decoration, sometimes showing a double ring on the edges
(10e15
m
m) (Fig. 4def). These morphotypes resemble those
described by Cordova and Scott (2010) and Cordova (2013) as being
representative of Restionaceae plants. They were mainly found in
fynbos, strandveld, and dune cordon vegetation (both thicket-
fynbos mosaics), and marginally renosterveld vegetation. Our
comparison with the modern Restionaceae plants (Elegia juncea,T.
insignis,T.rigidus and Restio triticeus) confirmed that these mor-
photypes correspond to Restionaceae plants (Fig. 4gel). In this work
we have chosen not to follow Cordova's (Cordova and Scott, 2010;
Cordova, 2013) distinction between papillated and non-papillated
rondels since they have, in fact, spheroidal shapes that can be
recognized when rotating phytoliths under the microscope. Elon-
gate phytolith morphotypes were also observed in Restionaceae
plants. Because these are also produced by other plants they were
not used here as representative of Restionaceae. Furthermore, psi-
late and rugulate spheroids (~10
m
m) were also identified in our
modern plants (Fig. 5aed), each having similar shapes, decorations,
and sizes to the ones that were identified in the wood and bark of
trees and shrubs (Fig. 5eeh) from our modern plant reference
collection in the study area (Esteban, unpublished data). In order to
circumvent any confusion with psilate or rugulate spheroid phyto-
liths, which are both produced in Restionaceae (Fig. 5aed) and
eudicot wood (Fig. 5eeh) (Table A1), the characteristic large gran-
ulate spheroids and verrucate spheroids showing spiraling decora-
tion from Restionaceae are both referred to here as restio phytoliths.
4.1. Phytolith indices
Phytolith indices have been used widely as indicators of aridity
[Iph (%) index] (Diester-Haass et al., 1973; Alexandre et al., 1997;
Novello et al., 2012); climate conditions based on the C
3
eC
4
grass
distribution [Ic index] (Twiss, 1992; Barboni et al., 2007; Bremond
et al., 2008); tree cover density [D/Pºindex] (Alexandre et al.,
1997eD/P index; Barboni et al., 1999; Bremond et al., 2005a,
2008); evapotranspiration and water stress [Fs index] (Bremond
et al., 2005b; Novello et al., 2012; Fisher et al., 2013); and aquatic
and xerofitic grass dominance [Iaq and Ixe indices] (Novello et al.,
2012, 2015). Here, we used the D/Pºindex (Bremond et al., 2008)
to identify shrubby vegetation/tree cover density, which is defined
as follows:
psilate spheroids þrugulate spheroids
PGSSCs
The Ic and Iph indices were also used to further characterize
grass distribution among vegetation types. We modified the for-
mula by eliminating the percentage, so they are defined as follows:
Ic ¼GSSC Rondelsðrondels;towers;trapezoidsandoblongsÞ
PGSSCs
Iph ¼GSSC Saddles
PGSSCs Saddles and Lobates
Furthermore, we have defined a new phytolith index that is
specific to the identification of fynbos vegetation, which we term
the Fy index. Since fynbos is shrub-dominated vegetation with a
high Restionaceae and a low grass component (Bergh et al., 2014),
I. Esteban et al. / Quaternary International 434 (2017) 160e179166
Table 2
List of morphotypes identified, their plant type and plant part attribution and the average of their presence in the vegetation types.
Total PPlant type Phytolith
morphotypes
Plant attribution Fynbos biome Fynbos/
Renosterveld
transition
Renosterveld
biome
Thicket biome Forest biome Azonal vegetation
Limestone
Fynbos
Sand
Fynbos
Grassy
Fynbos
Mountain
Fynbos
Fynbos/
renosterveld
Renosterveld Subtropical
Thicket
Coastal
Thicket
Strandveld Dune
Cordon
Coastal
Forest
Riparian Wetland
Bulliform Grass leaves 1.7 3.6 7.0 4.7 13.9 4.5 1.6 1.5 7.3 6.7 1.9 4.8 6.0
Papillae Grass leaves 0.0 0.0 0.0 0.0 0.0 0.2 3.9 1.4 0.0 0.0 0.0 0.0 0.0
Prickle Grass leaves 6.0 4.8 6.7 3.9 7.0 5.7 8.5 2.4 4.6 3.3 0.0 4.2 7.1
Long cell with
decorated margins
Grass
inflorescences
2.9 2.5 2.0 1.6 0.0 2.7 1.1 0.8 2.8 3.3 0.0 2.8 3.3
GSSC Rondel Grasses 20.9 26.0 24.6 33.1 10.4 37.5 22.5 24.5 30.1 31.7 21.3 28.3 26.2
GSSC Lobate Grasses 8.5 8.0 12.7 5.5 5.2 8.2 7.3 7.3 5.9 1.7 49.1 8.7 9.8
GSSC Saddle Grasses 2.2 3.8 3.4 1.6 12.2 6.3 6.3 26.4 5.0 13.3 10.2 22.4 2.7
GSSC Oblong Grasses 1.7 4.2 5.2 10.2 2.6 5.0 7.1 3.0 4.9 3.3 3.7 4.5 7.1
Total grasses 43.9 52.7 61.7 60.6 51.3 70.0 58.2 67.3 60.6 63.3 86.1 75.8 62.3
Restio morphotype Restionaceae 13.0 7.1 2.2 3.9 0.9 0.9 0.0 0.0 6.6 1.7 0.0 0.4 0.0
Total restios 13.0 7.1 2.2 3.9 0.9 0.9 0.0 0.0 6.6 1.7 0.0 0.4 0.0
Hat-shape Cyperaceae 0.7 1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0
Total sedges 0.7 1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0
Spheroid echinate Arecaceae 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.1 0.6 0.0 0.0 0.0 0.5
Total palms 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.1 0.6 0.0 0.0 0.0 0.5
Sclerenchyma Eudicot leaves 0.4 0.3 0.0 0.0 0.0 0.4 0.3 0.0 0.6 0.0 0.0 0.5 0.5
Ellipsoid Eudicot wood
and bark
1.0 0.6 0.8 2.4 0.0 0.1 1.5 0.0 0.0 0.0 0.0 0.1 0.0
Hair cell Eudicot Eudicot leaves 0.0 1.0 1.4 0.8 0.0 0.5 0.4 0.0 0.3 0.0 0.0 0.6 0.0
Hair Base Eudicot Eudicot leaves 0.0 0.0 0.0 0.0 0.0 0.0 0.4 0.0 0.0 0.0 0.0 0.0 0.0
Stomata Eudicot leaves 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0
Parallelepiped blocky Eudicot wood
and bark
3.6 2.2 2.2 0.8 2.6 1.6 0.7 0.4 1.2 0.0 0.0 0.2 4.4
Spheroid Eudicot wood
and bark
22.0 11.6 0.9 7.9 9.6 4.4 5.5 4.3 6.0 13.3 2.8 1.9 5.5
Vascular elements Eudicot leaves 0.4 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4 0.0
Fibers Eudicots 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0
Epidermal ground-
mass polyhedral
Eudicot leaves 0.0 0.8 0.2 0.0 0.0 0.2 0.0 0.0 2.0 1.7 0.9 0.6 0.0
Fruit phyotliths Eudicot fruits 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0
Total eudicots 27.4 16.7 5.4 11.8 12.2 7.3 8.8 4.7 10.2 15.0 3.7 4.5 10.4
Cylindroid Non-diagnostic 5.8 6.6 8.5 5.5 6.1 5.3 6.3 2.3 3.0 0.0 1.9 4.5 10.9
Parallelepiped
elongate
Non-diagnostic 3.4 10.8 18.1 15.7 17.4 11.2 11.8 6.2 13.6 11.7 5.6 9.2 12.6
Parallelepiped
thin
Non-diagnostic 2.7 1.7 0.6 1.6 6.1 1.2 3.1 2.4 2.9 3.3 0.0 2.1 1.6
Total elongates 11.9 19.0 27.2 22.8 29.6 17.6 21.3 11.0 19.5 15.0 7.4 15.8 25.1
Irregular/
Indeterminate
Non-diagnostic 3.3 3.2 3.5 0.8 6.1 4.0 11.7 16.0 2.5 5.0 2.8 3.3 1.6
I. Esteban et al. / Quaternary International 434 (2017) 160e179 167
Fig. 2. Microphotographs of phytoliths identified in samples analyzed. Pictures taken at 400. a) prickle (RV10-04), b) bulliform cell (GF10-10), c) elongate with echinate margin
(RV11-58), def) rondel GSSCs (SF11-42, SF11-43 and RV11-15), geh) bilobates GSSCs (GF10-14, CF10-13), i) saddle GSSCs (CT06-05), j) epidermal ground mass polyhedral (STV11-
01), k) eudicot hair cell (SF11-37), l) sclerenchyma (RP11-59). Scale bar represents 10
m
m.
Fig. 3. Ternary plot showing the grass silica short cells, grouped by the three fundamental categories: rondels, lobates and saddles, distribution among vegetation types. Rondel
category includes also the oblong morphotypes. Legend: CT, coastal thicket; GF, grassy fynbos; LF, limestone fynbos; RP, riparian; RV, renosterveld; SF, sand fynbos; StT, subtropical
thicket; STV, strandveld.
I. Esteban et al. / Quaternary International 434 (2017) 160e179168
Fig. 4. Microphotographs of four restio phytoliths identified in modern soils and Restionaceae modern plants. Phytoliths are shown from three different points of view by its
rotation on the slide. Pictures taken at 400.aec) large spheroid showing protuberances, with granulate decoration (LF11-68); def) spheroids showing spiraling decoration (LF11-
85); gei) spheroid showing protuberances, with granulate/verrucate decoration (Thamnochortus insignis); jel) spheroids showing spiraling decoration with a double ring on the
edges (Elegia juncea). Scale bar represents 10
m
m.
Fig. 5. Microphotographs of phytoliths spheroids psilate and rugulate identi fied in Restionaceae plants (aed) and the wood and bark of eudicot plants (eeh). Pictures taken at 400.
aeb) Restio triticeus,ced) Thamnochortus insignis,e)Protea lanceolata,f)Leucospermum praecox,g)Elytropappus rhinocerotis and h) Protea repens. Scale bar represents 10
m
m.
I. Esteban et al. / Quaternary International 434 (2017) 160e179 169
we defined the Fy index as the ratio of restio phytoliths and
spheroids psilate and rugulate phytoliths to the sum of GSSCs,
which is defined as follows:
restio þpsilate spheroids þrugulate spheroids phytoliths
PGSSCs
As we pointed out above, spheroids psilateand rugulate phytoliths
were identified in the wood of trees and shrubs and in Restionaceae
from our plant reference collection making it impossible to distin-
guish each phytolithmorphotype from one-another. Considering that
fynbos is a shrub-dominated vegetation and Restionaceae are the
diagnostic family of the Fynbos biome, we have therefore assumed
that the identification of spheroids psilateand rugulate, whetherthey
are produced by woody eudicots or Restionaceae, are representative
of fynbos vegetation. Thus the Fy index provides a signal for the
presence of vegetation characteristic of the Fynbos biome. Further-
more, we assume that in samples from vegetation types devoid of
Restionaceae plants, and also having no restio phytoliths, spheroid
morphotypes might be representative of woody eudicots.
The KruskaleWallis test showed a statistically significant dif-
ference between the D/Pºand Fy index values by vegetation type
(x
2
¼24.92, p¼0.0008 and x
2
¼24.84, p¼0.008, D/Pºand Fy
indices respectively) with the highest mean ranks for limestone
fynbos (46.33 for both indices) and the lowest mean ranks for
grassy fynbos (8.5 and 19.5, D/Pºand Fy indices respectively) and
riparian vegetation (12.86 and 11.07, D/Pºand Fy indices respec-
tively) (Fig. 6a and b). Iph values also detected statistically signifi-
cant differences between vegetation types (x
2
¼16.84, p¼0.0184)
with the highest mean rank of 39.93 for riparian, 33.75 for coastal
thicket and 32.5 subtropical thicket (Fig. 6c). There were no sig-
nificant differences in the Ic index between vegetation types
(x
2
¼12.81, p¼0.0769). Nonetheless, samples from renosterveld
had the largest numbers of GSSC rondels, while riparian and coastal
thicket the lowest (Fig. 6d).
4.2. Fynbos biome
4.2.1. Limestone fynbos
The dominant vegetation of the sampled areas consisted mainly
of restios, Asteraceae shrubs and other eudicot small trees and
shrubs. Grasses were either absent or barely present (Table 1). The
few grasses that were observed corresponded to either Ehrharta or
Themeda.
Phytoliths from limestone fynbos vegetation were identified in
low concentration ranging from 7000 to 160,000 phytoliths g/sed.
This low phytolith concentration might be related to the restio and
Asteraceae dominance, which produce low amounts of phytoliths,
combined with the low presence of grasses (Table 1). Interestingly,
the only sample (LF11-23) where Themeda triandra (C
4
Panicoideae)
was part of the dominant vegetation also showed low phytolith
concentration.
Out of the seven samples analyzed from this vegetation type,
three showed enough phytoliths for a reliable morphological
interpretation (LF11-23, LF11-68 and LF11-85). Although grasses
were not common in the vegetation type, grass phytoliths were
well represented in the phytolith record. Rondel phytoliths were
most common among GSSCs (mean: 51.3%) whereas lobates were
also less frequent (mean: 38.7%). Interestingly, we observed that
Fig. 6. Box plot showing the phytolith indices values among vegetation types. a) D/Pºindex [P(psilate and rugulate spheroids)/grass silica short cells] and the standard deviation
among the different vegetation types; b) Fy index [P(psilate and rugulate spheroids and restio phytoliths)/grass silica short cells] and the standard deviation among the different
vegetation types; c) Iph index [GSSC saddles/P(GSSC saddles and lobates)]; d) Ic index [GSSC rondels and oblongs/P(GSSC rondels, lobates and saddles)]. The mean values (mid-
line), standard error ±(box) and standard deviation (whiskers) are given for the four indices. Legend: LF, limestone fynbos; SF, sand fynbos; GF, grassy fynbos; RV, renosterveld; StT,
subtropical thicket; CT, coastal thicket; STV, strandveld; RP, riparian. Different letters indicate means that are significantly different based on a KruskaleWallis test.
I. Esteban et al. / Quaternary International 434 (2017) 160e179170
despite Themeda trianda was the dominant grass species from
sample LF11-23 area, which produces GSSC bilobates in high
numbers (Rossouw, 2009), it had the lowest counts of GSSC lobates.
Consistent with the vegetation, restio phytoliths were identified in
high frequencies in all the samples (Table 2). Hateshape phytoliths
from the Cyperaceae family were identified in samples LF11-68
although sedges were not identified in the field. Eudicot phyto-
liths, mainly from wood and bark, were identified in high quantities
in all the samples analyzed, where spheroid psilates were the most
representative morphotypes (Table 2).
4.2.2. Sand fynbos
The vegetation present during sampling varied greatly in the
different sampled areas. However, eudicots such as Leucadendron
salignum,Leucospermum praecox and L. muirii,Protea repens, and
some Erica spp. were highly represented. Restios were also well
represented and grasses dominated in some of the areas (Table 1).
Out of the nine samples that we analyzed from sand fynbos,
those from grass-dominated areas exhibited the highest phytolith
concentration (SF11-42 and SF11-43). The samples having the
lowest phytolith concentration corresponded to areas where
eudicot plants dominated (SF10-05, SF10-06, SF11-37, SF11-47,
SF11-62, SF11-82) (Table 1). The exception was sample SF11-45,
which showed a high phytolith concentration and little grass-cover.
Overall, grass phytoliths dominated sand fynbos samples and
among those grass phytoliths that were identified, GSSC rondels
were the most representative morphotypes, accounting for 57%
(mean) of the different GSSCs (Table 2 and Fig. 3). Restio phytoliths
were also identified in high frequencies. Interestingly, samples
SF11-45 and SF11-47, which were collected from areas where res-
tios did not dominate, showed the highest concentration of these
morphotypes (Tables 1 and 2). Cyperaceae hateshape phytoliths
were also identified in some sand fynbos samples, and these
morphotypes were also observed in samples from limestone fynbos
vegetation, even though they were not dominant in the vegetation
(SF10-05, SF11-37, SF11-47 and SF11-62). Eudicot phytoliths
(mainly spheroids from the wood/bark) were identified in all the
samples in high numbers with the exception of samples SF11-42
and SF11-43.
4.2.3. Grassy fynbos
Grassy fynbos vegetation was dominated by grasses such as
Themeda trianda (C
4
Panicoideae), Eragrostis capensis and E. curvula
(C
4
Chloridoideae) and Pentachistis colorata (C
3
Danthonioideae)
(Table 1). Phytolith results showed a high phytolith concentration
with a dominance of grass characteristic phytoliths (mainly GSSCs
and bulliform cells) (Table 2). Samples SF11-42 and 43 presented
the highest grass content of all the samples. Among GSSCs, rondels
are represented in high frequencies in the two samples (GF10-10
and GF10-11), where Pentachistis colorata,Themeda trianda and
Eragrostis curvula and E. capensis were dominant (Fig. 3). Even
though chloridoids dominated in sample GF10-10, GSSC saddles
were identified in low frequencies. Conversely, sample GF10-14,
which was collected from a hilltop in the Brenton Lake area, con-
tained high frequencies of GSSC lobates and saddles. Restios and
eudicot phytoliths (leaves and wood and bark) were also identified
although in very low frequencies (Table 2). These results are
consistent with the low presence of these plants in this vegetation
type.
4.2.4. Mountain fynbos
Only one sample was collected from mountain fynbos vegeta-
tion (MF12-01). The sample was collected from an area where
restios and sedges dominated the plant cover, but there were some
associated shrubs (Table 1). Phytoliths were identified in relatively
high concentration and were dominated by rondel GSSCs (Table 2;
Fig. 3). Restio phytoliths were not recorded in high numbers,
although these plants dominated the vegetation at the collection
area. Spheroids phytoliths were also identified in relatively high
concentrations (Table 2).
4.3. Renosterveld biome
4.3.1. Renosterveld
Renosterveld vegetation on the Cape south coast (south of
Langeberg and Riviersonderend Mountains) is considered to be a
distinct vegetation type on account of the high abundance of
largely C4 grasses and 50e70% plant cover (Mucina and Rutherford,
2006). The species composition of renosterveld differs significantly
from any fynbos vegetation types due the lack of Proteaceae and
Ericaceae species. The dominant vegetation in the 17 samples
analyzed from renosterveld was composed of shrubs such as Ely-
tropappus rhinocerotis (commonly known as renosterbos), other
Asteraceae shrubs and various family grasses (C
4
Panicoideae and
Chloridoideae, and C
3
Ehrhartoideae and Danthonioideae)
(Table 1).
Quantitatively most of the samples from renosterveld showed
the highest phytolith concentration in comparison to other vege-
tation types, ranging from 4.8 to 1.2 million phytoliths g/sed. The
only exceptions were samples RV11-15 and RV11-58, with much
lower counts (Table 1). As with other vegetation types, samples
containing the highest phytolith concentration were collected from
areas where grasses were dominant.
All the samples showed similar phytolith assemblages with
grasses dominating and restios and eudicot plants represented in
lesser amounts (Table 2). GSSC rondels were the most abundant
morphotypes identified among grasses. The exceptions were
samples RV09-01, RV11-15 and RV11-58, which contained a higher
presence of GSSC saddles and lobates, that each made up around
30%. Sample RV09-01 was collected from an area where C
4
grasses
dominated, such as Themeda trianda and Brachiaria serrata
(C
4
ePanicoideae) and Eragrostis curvula and E.capensis
(C
4
eChloridoideae) along with Cymbopogon marginatus
(C
4
eAndropogonae). Restio phytoliths were also identified but in
very low frequencies (Table 2). Though restios were recognized in
most of the sampled areas, they did not represent a dominant
component. Among eudicots, spheroid phytoliths were in slightly
higher frequencies with the exception of samples RV09-03 and
RV11-67 despite vegetation growing in these plots being similar to
the other ones. Nonetheless, we need to bear in mind that because
grass phytoliths were identified in very high numbers both restio
and woody plants, which are low phytolith producers, might
appear underrepresented in the phytolith record.
4.4. Transition fynbos/renosterveld
One sample was collected from this area where vegetation
showed characteristic elements from both fynbos and renosterveld
vegetation that had high grass presence, restios, and shrubs (F/
RV10-07). This sample was collected below Tritoniopsis sp. and
Restionaceae. The phytolith concentration was similar to both
grassy fynbos and renosterveld samples (Table 1). The phytolith
assemblage was dominated by grasses although restios and eudicot
phytoliths were also well represented. GSSC saddles dominated
among other different GSSCs making up 50% of the assemblage
(Table 2). No further characterization of transition vegetation could
be inferred through the phytolith assemblage.
I. Esteban et al. / Quaternary International 434 (2017) 160e179 171
4.5. Thicket biome
4.5.1. Subtropical thicket
Subtropical thicket vegetation was dominated by eudicot plants
characterized mainly by Sideroxylon inerme,Eriocephalus africanus
and Searsia pterota. Of the three samples analyzed from this vege-
tation type, only sample StT09-06 contained a high phytolith con-
centration (1.2 million phytoliths g/sed) (Table 1). The phytolith
assemblage from this sample showed low morphological variety
with a dominance of grasses even though wood/bark phytoliths
were also well represented. Among the grasses, GSSCs rondelswere
the dominant morphotype and lobates and saddles were identified
in much lower frequencies (Table 2).
4.5.2. Coastal thicket
Six samples were collected from the densely vegetated cliff area
around Pinnacle Point where Euclea racemosa,Euphorbia spp.,
Sideroxylon inerme,Grewia occidentalis,Lycium sp., Zygophyllum
morgsana and Cassine peragua dominated. Grasses were also pre-
sent in some of the areas, and Eragrostis spp. dominated in sample
CT06-05 (Table 1). Our results showed very low phytolith concen-
tration overall, with the only exception being sample CT06-07
(900,000 phytoliths g/sed), which was collected from a loamy
soil. Out of eight samples collected only four (CT06-01, CT06-02,
CT06-05 and CT06-06) contained enough phytoliths for a reliable
interpretation of their data. Furthermore, despite the diversity of
eudicot plants in coastal thicket vegetation, our samples showed
low morphological variety with characteristic grass phytoliths be-
ing the most abundant. Among these grasses, GSSC saddles domi-
nated along with lobates morphotypes (Table 2). For example, in
sample CT06-05, where Eragrostis spp. was the dominant plant
component, GSSC saddles represented the 92%. Whereas no palm
species occur in the GCFR today, spheroid echinates, characteristics
of the Arecaceae family, were identified in samples CT06-02 and
CT06-05 although in low frequencies. Spheroids and characteristic
eudicot leaf phytolith morphotypes were identified in low fre-
quencies in spite of the denser woody vegetation occurring in the
area (Table 2).
4.5.3. Strandveld
The dominant vegetation in our sampled areas was grasses
(mainly C
3
Ehrhartoideae), restios, asteraceous plants and other
trees or shrubs such as Stoebe cinorera,Sideroxylon inerme,Aga-
thosma spp., Searsia spp. and Metalasia sp. (Table 1).
Out of the seven samples that we analyzed, only three samples
(STV11-1, STV11-16 and STV11-21) contained enough phytoliths for
reliable interpretations. As with other vegetation types, the highest
phytolith concentration corresponded to areas where grasses were
dominant (Table 1). The phytolith assemblage recovered from these
samples is most characteristic of fynbos vegetation, whereas grass
phytoliths dominated in all the samples. Among GSSCs, rondels
were dominant and associated with lobates and saddles in most of
the samples (Table 2). This is consistent with Ehrhartoideae being
the most representative grass family in the area. Restio phytoliths
and spheroids were present in high frequencies in sample STV11-1.
4.5.4. Dune cordon
The vegetation present in dune cordon areas was characterized
by asteraceous plants and other shrubby vegetation with few grassy
elements (only Ehrharta villosa was identified). Dune cordon sam-
ples also had the lowest phytolith concentrations. No sample had
phytolith concentration exceeding 100,000 phytoliths g/sed
(Table 1) and only sample DC11-18 contained enough phytoliths for
a reliable interpretation of the data. Furthermore, grass phytoliths
dominated these samples even though grasses were rarely
observed in the understorey of dune cordon vegetation (Tables 1
and 2). Among grasses, GSSC rondels were the dominant mor-
photypes while saddles were also well-represented. Restio phyto-
liths were noted although in low numbers. Eudicot phytoliths,
mainly spheroids from the wood and bark, were identified in
relatively high frequencies in all the samples (Table 2).
4.6. Forest biome
4.6.1. Coastal forest
Coastal forest area was densely vegetated and dominated by
eudicot trees such as Celtis sp. and Vepris sp., as well as other
eudicot herbs and shrubs. The understorey was composed of
grasses, although they were not part of the dominant vegetation.
Our quantitative results showed relatively high phytolith concen-
tration (345,000 phytolith g/sed) (Table 1). In relation to
morphology, grasses dominated the phytolith assemblage making
up 86%. Contrary to other samples, GSSC lobates dominated here,
making up 50% of the total phytolith morphotypes identified. Few
phytoliths characteristics of eudicot leaves as well as few parallel-
epiped blocky morphotypes probably from the wood and bark of
eudicot were identified (Table 2).
4.7. Azonal vegetation
4.7.1. Riparian
The winter-deciduous Acacia karroo and other eudicot trees or
tall shrubs, as well as several asteraceous plants, dominated in most
of the sampled areas. Grasses were present as part of the dominant
vegetation for most of the areas in association to other herbaceous
plants (Table 1).
Together with grassy fynbos and renosterveld, samples from
riparian vegetation contained the highest phytolith concentration,
ranging from 4.3 million to 97,000 phytoliths g/sed. Samples with
the highest phytolith concentration corresponded to areas domi-
nated by grasses (Table 1). Morphologically, grass phytoliths were
the major component with a minor presence of phytoliths from
other plants (Table 2). GSSC saddles were the most representative
morphotypes. Although restios were not present in the sampled
areas, Restio phytoliths were identified in samples RP11-48, RP11-
59 and RP11-79. Hateshape phytoliths, characteristic of the
Cyperaceae family, were only identified in sample RP11-79, and in
very small abundance (~1%). Eudicot phytoliths, both from wood/
bark and leaves, were noted in low frequencies in all the samples
(Table 2).
4.7.2. Wetlands
The wetlands in our study area are seasonally inundated sites
that are adjacent to rivers and floodplains. Water levels may vary
throughout the year, but the sites are inundated for at least several
weeks per annum. Our sample was collected from the riverside of
the Little Brak River where Cyperaceae and Arundo grasses domi-
nated. Some eudicot trees and shrubs were also present. Phytoliths
were identified in high numbers (1,590,000 phytolith g/sed)
(Table 1). The phytolith assemblage was dominated by grass and
wood/bark phytoliths. Despite C
4
Panicoideae grasses dominated
the vegetation, GSSC lobates were identified in low frequencies
(Table 2). Characteristic sedge phytoliths (hat-shape) were absent
from the phytolith assemblage.
4.8. Statistical analysis
The results from the ANOVA test showed statistically significant
differences in nine out of the twenty-two phytolith morphotypes
between vegetation types and in eight out of the twenty-two
I. Esteban et al. / Quaternary International 434 (2017) 160e179172
phytolith morphotypes between biomes (p <0.05) (Tables A2 and
A3, respectively). Subsequently, we ran those significant phytolith
morphotypes through a post hoc Tukey's HSD test, which is rep-
resented in Fig. 7 (phytolith morphotypes in regards to vegetation
types) and Fig. 8 (phytolith morphotypes in regards to vegetation
biomes). These results indicate that the null hypothesis can be
rejected, which stated that no significant difference can be found
between the phytolith morphological distribution from each of the
vegetation types or biomes on the south coast of South Africa.
The ANOVA analysis also showed that parallelepiped blocky
(ANOVA, p ¼0.0378), spheroids (psilate and rugulate) (ANOVA,
p¼<0.0001) and restio phytoliths (ANOVA, p¼<0.0001) are the
defining phytolith morphotypes of limestone fynbos vegetation
(Table A2 and Fig. 7). No other vegetation type showed this pattern
(Fig. 7). These results are in agreement with those observed
through the D/Pºand Fy indices (Fig. 6a and b). The post hoc Tukey's
HSD test confirmed a clear separation among limestone fynbos
vegetation and the rest of the vegetation types (Fig. 7). Contrari-
wise, the statistical analysis conducted did not show any defining
phytolith morphotype in samples from sand and grassy fynbos
(Table A2 and Fig. 7). Nevertheless grassy fynbos vegetation is
characterized by the low presence of woody and restio elements
(Fig. 7).
Renosterveld vegetation is characterized mainly by a very low
presence of GSSC saddles (ANOVA, p ¼0.0024), spheroids (psilate
and rugulate) (ANOVA, p ¼0.0001) and restio phytoliths (ANOVA,
p¼0.0001) (Table A2 and Fig. 7). The post hoc Tukey's HSD test
showed that Renosterveld mainly differ from those vegetation
types belonging to the Thicket biome regarding the presence of
ellipsoids, GSSC saddles and irregular and indeterminate morpho-
types, and with limestone and sand fynbos vegetation types
regarding the presence of spheroids and restio phytoliths (Fig. 7).
ANOVA and post hoc Tukey's HSD test results identified hair
base phytoliths (ANOVA, p ¼0.0195) and ellipsoids (ANOVA,
p¼0.026) as the defining morphotypes for subtropical thicket
vegetation (Fig. 7). Coastal thicket vegetation is correlated with an
abundance of GSSC saddles (ANOVA, p ¼0.0024), irregular and
indeterminate morphotypes (ANOVA, p¼<0.0001), and spheroid
echinates from the Arecaceae subfamily (Table A2 and Fig. 7). The
statistical analysis did not show any defining phytolith distribution
in samples from strandveld vegetation (Table A2 and Fig. 7). Finally,
ANOVA and post hoc Turkey's HSD method demonstrated a clear
separation of riparian vegetation among the other vegetation types
by the abundance of GSSC saddles (ANOVA, p ¼0.0024) (Table A2
and Fig. 7).
When analyzing the phytolith assemblage in a broader scale
(vegetation biomes) it was observed that ANOVA and post hoc
Tukey's HSD test clearly differentiate the Fynbos biome from others
local biomes. Like our results of the limestone fynbos vegetation,
the Fynbos biome itself can be defined based on parallelepiped
blockys (ANOVA, p ¼0.005), spheroids psilate and rugulate
(ANOVA, p ¼0.0021), and restio phytoliths (ANOVA, p ¼0.0001)
(Table A3 and Fig. 8). ANOVA showed that Renosterveld biome is
mainly characterized by the abundance of GSSC rondels (ANOVA,
p¼0.0128) and a low presence of GSSC saddles (Fig. 8). The post
hoc Tukey's HSD test showed that the highest differences in the
presence of GSSC rondels are observed between Renosterveld and
Fynbos biome (Fig. 8). Thicket biome is mainly characterized by the
abundance of irregular and indeterminate morphotypes (ANOVA,
p¼0.0015) as well as the abundance of spheroid echinate phyto-
liths (ANOVA, p ¼0.0103). The post hoc Tukey's HSD test showed
that Thicket vegetation can be clearly differentiate among biomes
by the abundance in the presence of irregular phytoliths (ANOVA,
p¼0.0015) (Fig. 8). Finally, ANOVA showed that riparian vegetation
is correlated with the abundance of GSSC saddles (ANOVA,
p¼0.0027). The post hoc Tukey's HSD test showed that riparian
vegetation differs mainly from Fynbos and Renosterveld biomes
regarding GSSC saddle concentration and placed closely to the
Thicket biome (Table A3 and Fig. 8).
5. Discussion
5.1. Phytolith deposition and preservation in modern soils
Pre- and postdepositional processes may affect phytolith
deposition and preservation in soils (Piperno, 1988, 2006). Pre-
depositional processes are the factors that may influence the
plant accumulation in soils and their subsequent release of phy-
toliths after organic material decay. External factors like the degree
of vegetation cover, the differential phytolith production in
different plants, and the life cycles of specific plants have all been
shown to influence the pre-deposition of phytoliths in modern soils
(Piperno, 1988, 2006 and references therein). Other external factors
may relate to wildlife and consumption of certain plant parts such
as fruits or seeds, which when consumed will not contribute to the
soil component. Aeolian or fire impact may also influence phytolith
deposition in soils (see below).
South African grasses, for example, are high phytolith producers
while other plants such as trees, shrubs, restios and most geophytes
from the Iridaceae family are low phytolith producers (Esteban,
unpublished data). Furthermore, phytoliths from the wood and
bark of trees are often less well represented because these parts of
the plants live longer and their tissues are not often deposited in
the soils. In contrast, grasses and other herbaceous plants often
contribute a higher phytolith input in soils because these plants
have shorter life cycles. In areas with a dense understory it is
therefore more likely to find phytoliths from graminoids and herbs
rather than from the trees themselves.
Our analysis of the sediment samples from our study area
showed a similar phytolith pattern. The samples that were
collected from shrubby vegetation, such as limestone fynbos, dune
cordon and strandveld, or samples that were collected from thicket
(coastal thicket) did not contain high phytolith concentration.
Conversely, samples from dense grassy vegetation, such as renos-
terveld and riparian vegetation, contained the highest phytolith
concentration. The differences in these cases are less likely to be the
product of actual vegetation changes rather than the dominance of
grassy phytoliths at the expense of plants that produce much fewer
phytoliths and the implication is that when grasses are present
then it makes accurate vegetation identification more difficult.
After their deposition, preservation of phytoliths in modern soils
is associated with the silicon cycle in which phytoliths are recycled
by plant roots immediately upon their deposition in the soil A
horizon (Alexandre et al., 1997; Derry et al., 2005). Soils with slow
development and under constant biomass activity might have low
preservation of the phytolith assemblage. Soils in the Southern
Cape are from slightly fertile to infertile (Thwaites and Cowling,
1988), with a slow development and under a low biomass activ-
ity. We assume that the rate of deposition of phytoliths in fynbos
vegetation (where perennial plants dominate) is lower than that of
other types of vegetation with a large representation of grasses
(annual plants). Consequently, the differences between the depo-
sition rate and the rate of recycling of phytoliths in the fynbos
vegetation account for their low concentration in soils.
Moreover, phytolith dissolution cannot be considered as cause
for their low concentration in samples since all of the samples
showed an acidic to a moderate alkaline pH.
Low phytolith concentration in very coarse soils or in very active
bioturbation soils has also been shown to be the result of phytolith
translocation (e.g. Fiskish et al., 2010). These authors showed size
I. Esteban et al. / Quaternary International 434 (2017) 160e179 173
dependence on phytolith percolation by water, with phytoliths
with a size diameter of 5
m
m being the most susceptible to move
downward. In our study, if phytolith translocation had occurred in
our samples with low number of phytoliths (limestone fynbos,
coastal thicket, strandveld vegetation types) then we would expect
to not see small morphotypes, such as GSSC rondels and saddles or
spheroids psilates and rugulates, what is not the case. Nonetheless,
this should be further tested by analyzing the bottom of soil
profiles.
The presence of restio phytoliths in some riparian plots
where Restionaceae plants were not present might be related to
the alluvial character of these soils, and thus, restio phytoliths
might represent the runoff soil from up stream. Another likely
explanation of its presence can be explained due to their
transport in ash clouds after fire events in fynbos vegetation as
it has been observed in other, although not-homologous, envi-
ronments (Aleman et al., 2014). Thus, either water flow or
aeolian transport or both could account for the small quantities
Fig. 7. Box-plots of the nine phytolith morphotypes pointed out by ANOVA and a post hoc Tukey Honest Significant Differences (HSD) as statistically significant different among
vegetation types. The mean values (mid-line), standard error ±(box) and standard deviation (whiskers) are given for the nine phytolith morphotypes. Legend: CT, coastal thicket;
GF, grassy fynbos; LF, limestone fynbos; RP, riparian; RV, renosterveld; SF, sand fynbos; StT, subtropical thicket; STV, strandveld. Different letters indicate means that are significantly
different based on the post hoc Tukey (HSD) test.
I. Esteban et al. / Quaternary International 434 (2017) 160e179174
of restio phytoliths in some riparian plots, despite the lack of
restio plants.
5.2. Recognition of GCFR vegetation types through modern
phytolith assemblages
Fynbos constitutes one of the most characteristic vegetation
types of the GCFR and thus the identification of key plants like
restios and shrubby vegetation in fossil phytolith assemblages is
critical to undertake accurate paleovegetation reconstructions. Our
results showed that limestone fynbos can be characterized through
the analyses of phytoliths from modern soils as well as by the
identification of certain characteristic phytolith morphotypes
(Fig. 7), such as restio phytoliths and spheroids. Our modern plant
reference study, however, has shown that both restios (Table A1)
and the wood of eudicot plants (e.g. Albert and Weiner, 2001;
Tsartsidou et al., 2007; Esteban unpublished data) produce very
few phytoliths per gram of plant material. The identification of
these morphotypes in our samples may be related to their good
preservation conditions. Previous studies have shown that the
stability of phytolith morphotypes do differ from each other with
spheroids showing the highest stability (Cabanes and Shahack-
Gross, 2015). Whereas no dissolution experiments have been con-
ducted on restio plants, their heavily silicified body suggests that
they might be relatively stable in soils and sediments.
Fire regime is another important factor in fynbos ecology, but its
role in phytolith deposition in soils has been discussed rarely. Fire
regime is responsible for the decay of shrubby and herbaceous
plants and the subsequent regeneration of shrubs and herbaceous
plants. Because modern soil phytolith assemblages represent an
amalgamation of the vegetation mosaic over a period of years to
several decades, the effects of fire may favor a higher phytolith
deposition rate in soils thereby increasing the representation of
fynbos elements (shrubs and restios) in the phytolith record.
Therefore, an increased presence of spheroids and restio phytoliths
in fynbos soilsddespite low phytolith production of eudicots and
Restionaceaedmight, in fact, be related to the effects ofa persistent
fire regime.
Furthermore, grass phytoliths were identified in all of our fyn-
bos samples even though mature fynbos (i.e. unburnt for ca.
Fig. 8. Box-plots of the eight phytolith morphotypes pointed out by ANOVA and a post hoc Tukey Honest Significant Differences (HSD) as statistically significant different among
vegetation biomes. The mean values (mid-line), standard error ±(box) and standard deviation (whiskers) are given for the eight phytolith morphotypes. Different letters indicate
means that are significantly different based on the post hoc Tukey (HSD) test.
I. Esteban et al. / Quaternary International 434 (2017) 160e179 175
10e20 yr) is characterized by a restio understorey with few grasses.
Grasses produce large amounts of phytoliths per gram of sediment
(Albert and Weiner, 2001; Bamford et al., 2006; Tsartsidou et al.,
2007; Esteban, unpublished data) and they are also most abun-
dant immediately following burning episodes (Linder and Ellis,
1990). Thus, the effect of fire might be also responsible of leaving
grass phytoliths as component of the soils.
In addition to the characterization of limestone vegetation, our
study also showed how D/Pºand Fy (high values) indices can be
used for the identification of vegetation types with a high grass
component, like grassy fynbos, renosterveld, and riparian vegeta-
tion (low values) (Fig. 6a and b). This observation is especially
important because it provides insights into the true character of the
paleo-vegetation regardless of any overbearing plant components.
Sand fynbos samples, for example, showed a mean Fy index value of
0.6 with a high standard deviation of 0.5. This high deviation
among samples is caused by the presence of grasses in samples
SF11-42 and SF11-43, which showed a very low Fy values (Fig. 6b).
These results show that when grasses are present, its occurrence
might mask the correct identification of vegetation types what
might constitute a problem facing paleoenvironmetal studies
across South Africa.
Forest and Thicket biomes are characterized by dense vegetation
and dominated by trees and shrubs that are often succulent. D/P
(Alexandre et al., 1997; Barboni et al., 1999; Bremond et al., 2005a)
and D/Pº(Bremond et al., 2008) indices have been used to estimate
the density of woody elements in different habitats in West and East
Africa, but they have not been applied to South African studies. In
our study, we found that low D/Pºvalues (<0.20) are not represen-
tative of dense tree and shrub vegetation from the Thicket and the
Forest Biomes rather than being from Renosterveld and Strandveld
(Fig. 6a). These findings imply that the D/Pºindex may not be a useful
tool to identify thicket or forest vegetation through fossil phytolith
assemblages when working on the south coast of South Africa.
However, ANOVA and post host Tukey's HSD test showed that
Thicket biome might be characterized by the abundance of irreg-
ular and other unknown phytolith morphotypes (probably pro-
duced by eudicot plants) as well as GSSC saddles as a secondary
term (Table A3 and Fig. 8). In regards to vegetation types, the sta-
tistical analysis showed that epidermal hair base (from eudicot
plants) and ellipsoid phytoliths are the defining features of sub-
tropical thicket vegetation. Nevertheless, because few phytoliths
were identified in these samples, results must be taken with
caution. Similarly, the statistical analysis showed that coastal
thicket vegetation might be characterized through fossil phytolith
assemblages by the presence of GSSC saddles, spheroid echinates
and irregular phytoliths (Table A2 and Fig. 7). Palms are not part of
the vegetation from the GCFR, whereas spheroids echinate from the
Arecaceae family were identified in samples from coastal thicket
and strandveld vegetation, although in very low percentage. Are-
caceae family produce large amount of phytoliths, namely the
spheroid echinate morphotypes (Bamford et al., 2006; Albert et al.,
2009). Thus the low amount recovered does not support the
presence of these plants but seems to be more related to contam-
ination by different factors, including human introduction of palms
for horticulture in the area.
As GSSCs have been linked to different subfamilies, their mor-
photypes can be used to differentiate between C
3
and C
4
grasses
and thus offer information on paleoecological conditions. We
observed that even though all the samples from the different
vegetation types show a mix of C
3
and C
4
GSSCs, there is a higher
presence of GSSC rondels, ascribed to the C
3
grass subfamilies such
as Ehrhartoideae, Danthonoideae and Pooideae (Fig. 3). The sta-
tistical analysis showed a significant dominance of GSSC rondels in
samples from Renosterveld biome.
GSSC lobates (mainly bilobates) have also been recovered in
most of the samples from the different vegetation types while
dominating in the sample collected from coastal forest in Brenton
lake area (Table 2). GSSCs lobates are common in C
4
Panicoideae
(Twiss et al., 1969), but have also been identified in the C
3
Ehr-
hartoideae and Danthonioideae grasses (Cordova and Scott, 2010).
Therefore, the grass attribution of bilobates should be made with
caution. Despite our results indicating a clear dominance of GSSC
lobates in coastal forest vegetation, more samples from this vege-
tation are needed to corroborate the statistically significance of
these results.
The C
4
Chloridoideae subfamily is a short tropical subfamily
adapted to dry climates and/or low available soil moisture (xeric
environments) (Vogel et al., 1978; Tieszen et al., 1979). GSSC sad-
dles, characteristic of C
4
chloridoids, were identified in all the
vegetation types but dominating in samples from riparian vegeta-
tion (ANOVA, p ¼0.0024) and coastal thicket (ANOVA, p ¼0.0024)
(Fig. 7). These results correlate well with those from the Iph index,
which showed a statistically significance correlation between high
Iph index values and riparian vegetation (Fig. 6c). Cynodon dactylon
would be a dominant grass in riparian (floodplain) habitats of the
Gouritz valley and are generally very common between thicket
clumps in coastal thicket areas. Thus the phytolith assemblage from
riparian and coastal thicket vegetation seems to record the pres-
ence of this grass species. Furthermore, this species occurs on
almost all soil typesdbut especially in fertile (loamy) soilsdand it
is common in disturbed areas like riparian environments. Conse-
quently, our results demonstrate that the distribution of GSSC
saddles in phytolith assemblages, as well as the Iph index, can be
used for identifying riparian and coastal thicket habitats through
the fossil phytolith record on the south coast of South Africa. Finally,
distinguishing between both riparian and coastal thicket vegeta-
tion (Thicket Biome) might be assessed by the higher presence of
irregular and unknown phytoliths probably from the eudicot group.
6. Conclusions
This study shows the potential and limitations for using phy-
tolith assemblages from modern soils to reconstruct GCFR vegeta-
tion. We found that each of the vegetation types in our study are
consistent with a mix of grass subfamilies from the C
3
and the C
4
photosynthetic pathway, which is consistent with the distribution
of grasses in the Southern Cape. We also showed that the co-
occurrence of restio phytoliths, high presence of spheroid mor-
photypes, and a high presence of rondel GSSCs from C
3
grasses are
the characteristic feature for the identification of fynbos vegetation,
mainly limestone and sand fynbos. Our demonstration of the Fy
index also showed that it can be used to identify limestone and
sand fynbos vegetation through phytolith assemblages. However,
as it was observed in samples SF11-42 and SF11-43, if grass plants
are present in the vegetation, it is not possible to identify fynbos
vegetation accurately. The D/Pºindex was useful in identifying
limestone fynbos vegetation, but it was not able to characterize
other vegetation types, such as forest and shrubland/thicket.
Moreover, our results also provided key details on identifying
specific GCFR vegetation types through phytolith analysis. In
particular, coastal forest vegetation appears to be related to the
dominance of GSSC lobates though a larger number of samples are
needed to corroborate this association. Similarly, saddles from the
C
4
chloridoid grass subfamily may be used to identify riparian and
coastal thicket vegetation when studying fossil phytolith assem-
blages. However, we contend that further studies focused on the
vegetation types occurring in Western (C
3
dominance) and Eastern
Cape (C
4
dominance) are needed to fully understand the phytolith
morphological distribution along different South African habitats
I. Esteban et al. / Quaternary International 434 (2017) 160e179176
with different environmental and climatic conditions. This, along
with the results provided in this study, might contribute to our
understanding of past environments and climate changes on the
south coast of South Africa.
Acknowledgements
Field and laboratory work was supported by financial support
from the Spanish Ministry of Science and Innovation and Economy
and Competitivity (HAR2010-15967 and HAR2013-42054-P) and
from the Catalan Government (2014-SGR0845). CWM acknowl-
edges the support of the National Science Foundation (BCS-
0524087, and BCS-1138073), the Hyde Family Foundation, and the
Institute of Human Origins at Arizona State University. This
research was also funded by the Oppenheimer Memorial Trust. We
also thank the two anonymous reviewers for their constructive
comments on the paper.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.quaint.2016.01.037.
References
Albert, R.M., Weiner, S., 2001. Study of phytoliths in prehistoric ash layers using a
quantitative approach. In: Meunier, J.D., Coline, F. (Eds.), Phytoliths: Applica-
tions in Earth Sciences and Human History. A.A. Balkema Publishers, Lisse,
pp. 251e266.
Albert, R.M., Bamford, M., 2012. Vegetation during UMBI and deposition of Tuff IF at
Olduvai Gorge, Tanzania (ca. 1.8 Ma) based on phytoliths and plant remains.
Journal of Human Evolution 63, 342e350.
Albert, R.M., Marean, C.W., 2012. The exploitation of plant resources by early Homo
sapiens: the phytolith record from pinnacle point 13B Cave, South Africa.
Geoarchaeology 27 (4), 363e384.
Albert, R.M., Bamford, M.K., Cabanes, D., 2006. Taphonomy of phytoliths and mac-
roplants in different soils from Olduvai Gorge (Tanzania) and the application to
Plio-Pleistocene palaeoanthropological samples. Quaternary International 148,
78e94.
Albert, R.M., Bamford, M.K., Cabanes, D., 2009. Palaeoecological significance of
palms at Olduvai Gorge, Tanzania, based on phytolith remains. Quaternary In-
ternational 193, 41e48.
Albert, R.M., Bamford, M.K., Esteban, I., 2015. Reconstruction of ancient palm
vegetation landscapes using a phytolith approach. Quaternary International
369, 51e66.
Albert, R.M., Ruíz, J.A., Sans, A., 2016. PhytCore ODB: a new tool to improve effi-
ciency in the management and exchange of information on phytoliths. Journal
of Archaeological Science. Available online 12 December 2015, ISSN 0305-4403.
http://dx.doi.org/10.1016/j.jas.2015.10.014 (in press).
Aleman, J.C., Canal-Subitani, S., Favier, C., Bremond, L., 2014. Influence of the local
environment on lacustrine sedimentary phytolith records. Palaeogeography,
Palaeoclimatology, Palaeoecology 414, 273e283.
Alexandre, A., Meunier, J.-D., L
ezine, A.-M., Vincens, A., Schwartz, D., 1997. Phyto-
liths indicators of grasslands dynamics during the late Holocene in intertropical
Africa. Palaeogeography, Palaeoclimatology, Palaeoecology 136, 213e219.
Bamford, M.K., Albert, R.M., Cabanes, D., 2006. Assessment of the lowermost Bed II
Plio-Pleistocene vegetation in the eastern palaeolake margin of Olduvai Gorge
(Tanzania) and preliminary results from fossil macroplant and phytolith re-
mains. Quaternary International 148, 95e112.
Bamford, M.K., Stanistreet, I.G., Stollhofen, H., Albert, R.M., 2008. Late Pliocene
grassland from Olduvai Gorge, Tanzania. Palaeogeography, Palaeoclimatology,
Palaeoecology 257 (3), 280e293.
Barboni, D., Bremond, L., 2009. Phytoliths of East African grasses: an assessment of
their environmental and taxonomic significance based on floristic data. Review
of Palaeobotany and Palynology 158 (1e2), 29e41.
Barboni, D., Bonnefille, R., Alexandre, A., Meunier, J., 1999. Phytoliths as paleo-
environmental indicators, west side Middle Awash Valley, Ethiopia. Palae-
ogeography, Palaeoclimatology, Palaeoecology 152, 87e100 .
Barboni, D., Bremond, L., Bonnefille, R., 2007. Comparative study of modern
phytolith assemblages from inter-tropical Africa. Palaeogeography Palae-
oclimatology Palaeoecology 246, 454e470.
Barboni, D., Ashley, G.M., Dominguez-Rodrigo, M., Bunn, H.T., Mabulla, A.Z.P.,
Baquedano, E., 2010. Phytoliths infer locally dense and heterogeneous paleo-
vegetation at FLK North and surrounding localities during upper Bed I time,
Olduvai Gorge, Tanzania. Quaternary Research 74, 344e354.
Bar-Matthews, M., Marean, C.W., Jacobs, Z., Karkanas, P., Fisher, E.C., Herries, A.I.R.,
Brown, K., Williams, H.M., Bernatchez, J., Ayalon, A., Nilssen, P.J., 2010. A high
resolution and continuous isotopic speleothem record of paleoclimate and
paleoenvironment from 90 to 53 ka from Pinnacle point on the south coast of
South Africa. Quaternary Science Review 29, 2131e2145.
Behar, D.M., Villems, R., Soodyall, H., Blue-Smith, J., Pereira, L., Metspalu, E.,
Scozzari, R., Makkan, H., Tzur, S., Comas, D., 2008. The Dawn of human Matri-
lineal diversity. The American Journal of Human Genetics 82, 1130e1140 .
Bergh, N.G., Verboom, G.A., Rouget, M., Cowling, R.M., 2014. Vegetation types of the
greater cape floristic region. In: Allsopp, N., Colville, J.F., Verboom, T. (Eds.),
Fynbos: Ecology, Evolution, and Conservation of a Megadiverse Region. Oxford
University Press, Oxford, pp. 1e25.
Bremond, L., Alexandre, A., H
ely, C., Guiot, J., 2005a. A phytolith index as a proxy of
tree cover density in tropical areas: calibration with leaf area index along a
forestesavanna transect in southeastern Cameroon. Global and Planetary
Change 45 (4), 277e293.
Bremond, L., Alexandre, A., H
ely, C., Peyron, O., Guiot, J., 2005b. Grass water stress
estimated from phytoliths in West Africa. Journal of Biogeography 32 (2),
311e327.
Bremond, L., Alexandre, A., Wooller, M.J., H
ely, C., Williamson, D., Sch€
afer, P.A.,
Majule, A., Guiot, J., 2008. Phytolith indices as proxies of grass subfamilies on
East African tropical mountains. Global and Planetary Change 61 (3),
209e224.
Brown, K.S., Marean, C.W., Herries, A.I., Jacobs, Z., Tribolo, C., Braun, D., Roberts, D.L.,
Meyer, M.C., Bernatchez, J., 2009. Fire as an engineering tool of early modern
humans. Science 325, 859e862.
Brown, K.S., Marean, C.W., Jacobs, Z., Schoville, B.J., Oestmo, S., Fisher, E.C.,
Bernatchez, J., Karkanas, P., Matthews, T., 2012. An early and enduring advanced
technology originating 71000 years ago in South Africa. Nature 491, 590e593.
Cabanes, D., Shahack-Gross, R., 2015. Understanding Fossil phytolith preservation:
the role of partial dissolution in paleoecology and archaeology. PLoS One 10 (5)
e0125532.
Clark, J.D., Beyene, Y., WoldeGabriel, G., Hart, W.K., Renne, P.R., Gilbert, H.,
Defleur, A., Suwa, G., Katoh, S., Ludwig, K.R., Boisserie, J.R., Asfaw, B., White, T.D.,
2003. Stratigraphic, chrono- logical and behavioural contexts of Pleistocene
Homo sapiens from Middle Awash, Ethiopia. Nature 423, 747e752.
Colville, J.F., Potts, A.J., Bradshaw, P.L., Measey, G.J., Snijman, D., Picker, M.D.,
Proches, S¸ ., Bowie, R.C.K., Manning, J.C., 2014. Floristic and faunal Cape bio-
choria: do they exist?. In: Fynbos: Ecology, Evolution, and Conservation of a
Megadiverse Region. Oxford University Press United Kingdom, pp. 73e92.
Cordova, C.E., 2013. C3 Poaceae and Restionaceae phytoliths as potential proxies for
reconstructing winter rainfall in South Africa. Quaternary International 287,
121e140.
Cordova, C.E., Scott, L., 2010. The Potential of Poaceae, Cyperaceae, and Restionaceae
Phytoliths to Reflect Past Environmental Conditions in South Africa. Palae-
oecology of Africa. CRC Press Taylor and Francis Group, Boca Raton, Florida,
pp. 107e133.
d’Errico, F., Henshilwood, C.S., 2007. Additional evidence for bone technology in the
southern African Middle Stone Age. Journal of Human Evolution 52, 142e163.
d’Errico, F., Henshilwood, C.S., Vanhaeren, M., van Niekerk, K., 2005. Nassarius
kraussianus shell beads from Blombos Cave: evidence for symbolic behaviour in
the Middle Stone Age. Journal of Human Evolution 48, 3e24.
De Vynck, J.C., Van Wyk, B.E., Cowling, R.M., 2015. Indigenous edible plant use by
contemporary Khoe-San descendants of South Africa's Cape South Coast. South
African Journal of Botany 102, 60e69.
Deacon, H.J., Jury, M.R., Ellis, F., 1992. Selective regime and time. In: Cowling, R.M.
(Ed.), The Ecology of Fynbos: Nutrients, Fire and Diversity. Oxford University
Press, Cape Town, South Africa, pp. 6e22.
Derry, L.A., Kurtz, A.C., Ziegler, K., Chadwick, O.A., 2005. Biological control of
terrestrial silica cycling and export fluxes to watersheds. Nature 433, 728e731.
Diester-Haass, L., Schrader, H.J., Thiede, J., 1973. Sedimentological and Paleoclimato-
logical investigations of two Pelagic Ooze cores off cape Barbas, North-West
Africa. Meteor Forsh-Ergebnisse C 16, 19e66.
Dunn, R.E., Str€
omberg, C.A.E., Madden, R.H., Kohn, M.J., Carlini, A.A., 2015. Linked
canopy, climate, and faunal change in the Cenozoic of Patagonia. Science 347
(6219), 258e261.
Eichhorn, B., Neumann, K., Garnier, A., 2010. Seed phytoliths in West African
Commelinaceae and their potential for palaeoecological studies. Palae-
ogeography, Palaeoclimatology, Palaeoecology 298, 300e310.
Engelbrecht, C.J., Landman, W.A., Engelbrecht, F.A., Malherbe, J., 2014. A synoptic
decomposition of rainfall over the Cape south coast of South Africa. Climate
Dynamics 44 (9e10), 2589e2607.
Fagundes, N.J.R., Ray, N., Beaumont, M., Neuenschwander, S., Salzano, F.M.,
Bonatto, S.L., Excoffier, L., 2007. Statistical evaluation of alternative models of
human evolution. Proceedings of the National Academy of Sciences 104, 17614.
Fahmy, A.G., 2008. Diversity of lobate phytoliths in grass leaves from the Sahel
region, west tropical Africa: tribe Paniceae. Plant Systematics and Evolution
270, 1e23.
Fisher, E.C., Albert, R.M., Botha, G., Cawthra, H.C., Esteban, I., Harris, J., Jacobs, Z.,
Jerardino, A., Marean, C.W., Neumann, F.H., Pargeter, J., Poupart, M., Venter, J.,
2013. Archaeological reconnaissance for Middle Stone Age sites along the
Pondoland coast, South Africa. PaleoAnthropology 104e137.
Fiskish, O., Ingwersena, J., Lamers, M., Denysenko, D., Streck, T., 2010. Phytolith
transport in soil: a laboratory study on intact soil cores. European Journal of Soil
Science 61, 445e455.
Fredlund, G.G., Tieszen, L.T., 1994. Modern phytolith assemblages from the North
American Great Plains. Journal of Biogeography 21, 321e335.
I. Esteban et al. / Quaternary International 434 (2017) 160e179 177
Fredlund, G.G., Tieszen, L.L., 1997. Calibrating grass phytolith assemblages in cli-
matic terms: application to late Pleistocene assemblages from Kansas and
Nebraska. Palaeogeography, Palaeoclimatology, Palaeoecology 136, 199e211.
Garnier, A., Neumann, K., Eichhorn, B., Lespez, L., 2013. Phytolith taphonomy in the
middle- to late-Holocene fluvial sediments of Ounjougou (Mali, West Africa).
The Holocene 23 (3), 416e431.
Gonder, M.K., Mortensen, H.M., Reed, F.A., de Sousa, A., Tishkoff, S.A., 2007. Whole-
mtDNA Genome Sequence analysis of ancient African lineages. Molecular
Biology and Evolution 24, 757e768.
Henshilwood, C.S., van Niekerk, K.L., 2014. Blombos cave: the Middle Stone Age
levels. In: Smith, C. (Ed.), Encyclopedia of Global Archaeology. Springer, New
York, pp. 915e922.
Henshilwood, C.S., d’Errico, F., Marean, C.W., Milo, R.G., Yates, R.J., 2001. An early
bone tool industry from the Middle Stone Age, Blombos Cave, South Africa:
implications for the origins of modern human behaviour, symbolism and lan-
guage. Journal of Human Evolution 41, 631e678.
Henshilwood, C.S., d'Errico, F., Yates, R., Jacobs, Z., Tribolo, C., Duller, G.A.T.,
Mercier, N., Sealy, J.C., Valladas, H., Watts, I., Wintle, A.G., 2002. Emergence of
modern human behavior: Middle Stone Age engravings from South Africa.
Science 295, 1278e1280.
Henshilwood, C., d'Errico, F., Vanhaeren, M., van Niekerk, K., Jacobs, Z., 2004. Middle
Stone Age shell beads from South Africa. Science 304, 404.
Jerardino, A., Marean, C.W., 2010. Shellfish gathering, marine palaeoecology and
modern human behavior: perspectives from Cave PP13B, Pinnacle point, South
Africa. Journal of Human Evolution 59, 412e429.
Katz, O., Cabanes, D., Weiner, S., Maeir, A., Boaretto, E., Shahack-Gross, R., 2010.
Rapid phytolith extraction for analysis of phytolith concentrations and assem-
blages during an excavation: an application at Tell es-Safi/Gath, Israel. Journal of
Archaeological Science 37 (7), 1557e1563.
Kemper, J., Cowling, R.M., Richardson, D.M., Forsyth, G.G., McKelly, D.H., 2000.
Landscape fragmentation in South Coast Renosterveld, South Africa, in relation
to rainfall and topography. Austral Ecology 25 (2), 179e186.
Linder, H.P., Ellis, R.P., 1990. A revision of Pentaschistis (Arundineae: poaceae).
Contributions from the Bolus Herbarium 12, 1e124.
Linnow, G. 2012. Client Liaison Officer. Regional Office_Western Cape. South African
Weather Service. Unpublished.
Madella, M., Alexandre, A., Ball, T., 2005. International code for phytolith nomen-
clature 1.0. Annals of Botany 96, 253e260.
Malan, J.A., 1987. The bredasdorp group in the area between Gans Bay and Mossel
Bay. South African Journal of Science 83, 506e507.
Marean, C.W., 2010. Pinnacle point Cave 13B (Western Cape Province, South Africa)
in context: the Cape Floral kingdom, shellfish, and modern human origins.
Journal of Human Evolution 59 (3e4), 425e443.
Marean, C.W., 2011. Coastal South Africa and the co-evolution of the modern human
lineage and coastal adaptations. Settlement. In: Bicho, N., Haws, J.A., Davis, L.G.
(Eds.), Trekking the Shore: Changing Coastlines and the Antiquity of Coastal.
Springer, New York, pp. 421e440.
Marean, C.W., Bar-Matthews, M., Bernatchez, J., Fisher, E.C., Goldberg, P.,
Herries, A.I., Jacobs, Z., Jerardino, A., Karkanas, P., Minichillo, T., 2007. Early
human use of marine resources and pigment in South Africa during the Middle
Pleistocene. Nature 449, 905e908.
Marean, C.W., Cawthra, H.C., Cowling, R.M., Esler, K.L., Fisher, E., Milewski, A.,
Potts, A.J., Singels, E., De Vynck, J., 2014. Stone Age people in a changing South
African Greater Cape floristic region. In: Allsopp, N., Colville, J.F., Verboom, G.A.
(Eds.), Fynbos: Ecology, Evolution, and Conservation of a Megadiverse Region.
Oxford University Press.
Matthews, T., Rector, A., Jacobs, Z., Herries, A.I.R., Marean, C.W., 2011. Environmental
implications of micromammals accumulated close to the MIS 6 to MIS 5 tran-
sition at Pinnacle point cave 9 (Mossel Bay, western Cape province, South Af-
rica). Palaeogeography, Palaeoclimatology, Palaeoecology 302, 213e229.
McDougall, I., Brown, F.H., Fleagle, J.G., 2005. Stratigraphic placement and age of
modern humans from Kibish, Ethiopia. Nature 433, 733e736.
Menge, B.A., Branch, G.M., 2001. Rocky intertidal communities. In: Bertness, M.D.,
Gaines, S.D., Hay, M.E. (Eds.), Marine Community Ecology. Sinauer Associates,
Sunderland, pp. 221e251.
Mercader, J., Bennett, T., Esselmont, C., Simpson, S., Walde, D., 2009. Phytoliths in
woody plants from the miombo woodlands of Mozambique. Annals of Botany
104 , 91 e113.
Mercader, J., Astudillo, F., Barkworth, M., Bennett, T., Esselmont, C., Kinyanjui, R.,
Grossman, D.L., Simpson, S., Walde, D., 2010. Poaceae phytoliths from the Niassa
rift, Mozambique. Journal of Archaeological Science 37 (8), 1953e1967.
Mercader, J., Bennett, T., Esselmont, C., Simpson, S., Walde, D., 2011. Soil phytoliths
from miombo woodlands in Mozambique. Quaternary Research 75 (1),
138 e150 .
Mulholland, S.C., Rapp Jr., C., 1992. A morphological classification of grass
silica-bodies. In: Rapp Jr., G., Mulholland, C.S. (Eds.), Phytolith Systematics.
Emerging Issues, Advances in Archaeological and Museum Science. Plenum
Press, New York, pp. 65e89.
Mucina, L., Rutherford, M.C., 2006. The Vegetation Map of South Africa, Lesotho and
Swaziland. SANBI, Pretoria.
Neumann, K., Fahmy, A., Lespez, L., Ballouche, A., Huysecom, E., 2009. The Early
Holocene palaeoenvironment of Ounjougou (Mali): phytoliths in a multiproxy
context. Palaeogeography, Palaeoclimatology, Palaeoecology 276 (1), 87e106.
Novello, A., Barboni, D., 2015. Grass inflorescence phytoliths of useful species and
wild cereals from sub-Saharan Africa. Journal of Archaeological Science 59,
10e22.
Novello, A., Barboni, D., Berti-Equille, L., Mazur, J.C., Poilecot, P., Vignaud, P., 2012.
Phytolith signal of aquatic plants and soils in Chad, Central Africa. Review of
Palaeobotany and Palynology 178, 43e58.
Novello, A., Lebatard, A.E., Moussa, A., Barboni, D., Sylvestre, F., Bourl
es, D.L.,
Paill
es, C., Buchet, G., Decarreau, A., Duringer, P., Ghienne, J.F., Maley, J.,
Mazur, J.C., Roquin, C., Schuster, M., Vignaud, P., 2015. Diatom, phytolith, and
pollen records from a
10
Be/
9
Be dated lacustrine succession in the Chad basin:
insight on the MioceneePliocene paleoenvironmental changes in Central Africa.
Palaeogeography, Palaeoclimatology, Palaeoecology 430, 85e103.
Parr, J.F., Dolic, V., Lancaster, G., Boyd, W.E., 2001. A microwave digestion method for
the extraction of phytoliths from herbarium specimens. Reviewof Palaeobotany
and Palynology 116 (3), 203e212.
Piperno, D.R., 1988. Phytolith Analysis: an Archaeological and Geological Perspec-
tive. Academic Press, San Diego.
Piperno, D.R., 2006. Phytoliths: a Comprehensive Guide for Archaeologists and
Paleoecologists. AltaMira Press, Lanham, MD.
Radomski, K.U., Neumann, K., 2011. Grasses and grinding stones: inflorescence
phytoliths from modern West African Poaceae and archaeological stone arte-
facts. In: Fahmy, A.G., Kahlheber, S., D'Andrea, A.C. (Eds.), Windows on the Af-
rican Past. Current Approaches to African Archaeobotany. Africa Magna Verlag,
Frankfurt am Main, pp. 153e166.
Rebelo, A.G., Cowling, R.M., Campbell, B.M., Meadows, M.E., 1991. Plant commu-
nities of the Riversdale Plain. South African Journal of Botany 57, 10e18.
Rogers, J.C., 1984. The association between the North Atlantic Oscillation and the
Southern Oscillation in the northern hemisphere. Monthly Weather Review 112
(10), 1999e2015.
Rossouw, L., 2009. The Application of Fossil Grass-phytolith Analysis in the
Reconstruction of Late Cainozoic Environments in the South African Interior.
Doctoral dissertation. University of the Free State.
Runge, F., 1996. Leaf Phytoliths and Silica Skeletons from East African Plants. CD-
Rom Database. Germany, Department of Physical Geography. University of
Paderborn.
Runge, F., 1999. The opal phytolith inventory of soils in central Africadquantities,
shapes, classification, and spectra. Review of Palaeobotany and Palynology 107
(1), 23e53.
Runge, F., Runge, J., 1997. Opal Phytoliths in East African Plants and Soils. The State-
of-the-art of Phytoliths in Soils and Plants. Centro de Ciencias Medi-
oambientales, Madrid, pp. 71e81.
Schlebusch, C.M., Skoglund, P., Sj€
odin, P., Gattepaille, L.M., Hernandez, D., Jay, F.,
Li, S., de Jongh, M., Singleton, A., Blum, M.G.B., Soodyall, H., Jakobsson, M., 2012.
Genomic variation in seven Khoe-San groups reveals Adaptation and complex
African history. Science 338 (6105), 374e379.
Schlebusch, C.M., Lombard, M., Soodyall, H., 2013. MtDNA control region variation
affirms diversity and deep sub-structure in populations from southern Africa.
BMC Evolutionary Biology 13 (1), 56.
Smith, T.M., Tafforeau, P., Reid, D.J., Grün, R., Eggins, S., Boutakiout, M., Hublin, J.J.,
2007. Earliest evidence of modern human life history in North African early
Homo sapiens. Proceedings of the National Academy of Sciences 104,
6128e6133.
Str€
omberg, C.A.E., 2002. The origin and spread of grass-dominated ecosystems in
the Late Tertiary of North America: preliminary results concerning the evolu-
tion of hypsodonty. Palaeogeography, Palaeoclimatology, Palaeoecology 177,
59e75.
Str€
omberg, C.A.E., 2004. Using phytolith assemblages to reconstruct the origin and
spread of grass-dominated habitats in the great plains of North America during
the late Eocene to early Miocene. Palaeogeography, Palaeoclimatology, Palae-
oecology 207, 239e275.
Str€
omberg, C.A.E., McInerney, F.A., 2011. The Neogene transition from C3 to C4
grasslands in North America: assemblage analysis of fossil phytoliths. Paleobi-
ology 37 (1), 50e71.
Str€
omberg, C.A.E., Werdelin, L., Friis, E.M., Saraç, G., 2007. The spread of grass-
dominated habitats in Turkey and surrounding areas during the Cenozoic:
phytolith evidence. Palaeogeography, Palaeoclimatology, Palaeoecology 250
(1e4), 18e49.
Texier, P.J., Porraz, G., Parkington, J., Rigaud, J.P., Poggenpoel, C., Miller, C., Tribolo, C.,
Cartwright, C., Coudenneau, A., Klein, R., Steele, T., Verna, C., 2010. A Howiesons
Poort tradition of engraving ostrich eggshell containers dated to 60,00 0 years
ago at Diepkloof Rock Shelter, South Africa. Proceedings of the National Acad-
emy of Sciences 107, 6180e6185.
Thwaites, R.N., Cowling, R.M., 1988. Soil-vegetation relationships on the Agulhas
plain, South Africa. Catena 15 (3), 333e345.
Tieszen, L.L., Senyimba, M.M., Imbamba, S.K., Troughton, J.H., 1979. The distribution
of C3 and C4 grasses and carbon isotope discrimination along an altitudinal and
moisture gradient in Kenya. Oecologia 37 (3), 337e350.
Tsartsidou, G., Lev-Yadun, S., Albert, R.M., Miller-Rosen, A., Efstratiou, N., Weiner, S.,
2007. The phytolith archaeological record: strengths and weaknesses evaluated
based on a quantitative modern reference collection from Greece. Journal of
Archaeological Science 34 (8), 1262e1275.
Twiss, P.C., 1987. Grass-opal phytoliths as climatic indicators of the Great Plains
Pleistocene. Quaternary Environments of Kansas 5, 179e188.
I. Esteban et al. / Quaternary International 434 (2017) 160e179178
Twiss, P.C., 1992. Predicted world distribution of C3 and C4 grass phytoliths. In:
Rapp Jr., G., Mulholland, S.C. (Eds.), Phytolith Systematics: Emerging Issues,
Advances in Archaeological and Museum Science. Plenum Press, New York,
pp. 113e128.
Twiss, P.C., Suess, E., Smith, R.M., 1969. Morphological classification of grass
phytoliths. Soil Science Society of America Journal 33, 109e115.
Vlok, J.H.J., De Villiers, M.E., 2007. Vegetation Map for the Riversdale Domain. Un-
published 1:50 000 maps and report supported by the CAPE Fine Scale Planning
Task Team and CapeNature.
Vogel, J.C., Fuls, A., Ellis, R.P., 1978. Geographical distribution of Kranz grasses in
South Africa. South African Journal of Science 74, 209e215.
Watts, I., 1999. The origin of symbolic culture. In: Dunbar, R., Knight, C., Power, C.
(Eds.), The Evolution of Culture. Edinburgh University Press, Edinburgh,
pp. 113e146.
Watts, I., 2002. Ochre in the Middle Stone Age of southern Africa: ritualized display
or hide preservative? South African Archaeological Bulletin 57, 1e14 .
Watts, I., 2010. The pigments from Pinnacle point cave 13B, western Cape, South
Africa. Journal of Human Evolution 59 (3), 392e411.
White, T.D., Asfaw, B., DeGusta, D., Gilbert, H., Richard, G.D., Suwa, G., Howell, F.C.,
2003. Pleistocene Homo sapiens from Middle Awash, Ethiopia. Nature 423,
742 e747.
I. Esteban et al. / Quaternary International 434 (2017) 160e179 179
