Content uploaded by Fernando Ojeda
Author content
All content in this area was uploaded by Fernando Ojeda on Sep 09, 2020
Content may be subject to copyright.
ORIGINAL
ARTICLE
Variation in plant diversity in
mediterranean-climate ecosystems: the
role of climatic and topographical
stability
Richard M. Cowling
1
*, Alastair J. Potts
1
, Peter L. Bradshaw
1,2
, Jonathan
Colville
3
, Margarita Arianoutsou
4
, Simon Ferrier
5
, Felix Forest
6
, Nikolaos
M. Fyllas
4
, Stephen D. Hopper
6,7
, Fernando Ojeda
8
,S
ßerban Proches
ß
9
, Rhian
J. Smith
6
, Philip W. Rundel
10
, Emmanuel Vassilakis
11
and Brian R. Zutta
12
1
Department of Botany, Nelson Mandela
Metropolitan University, Port Elizabeth 6032,
South Africa,
2
South African National Parks,
Nelson Mandela Metropolitan University,
Port Elizabeth 6031, South Africa,
3
Applied
Biodiversity Research Division, South African
National Biodiversity Institute, Claremont
7735, South Africa,
4
Department of Ecology
and Systematics, Faculty of Biology,
University of Athens, 15784 Athens, Greece,
5
CSIRO Ecosystem Sciences, Canberra, ACT
2601, Australia,
6
Royal Botanic Gardens,
Kew, Richmond, Surrey TW9 3DS, UK,
7
Centre of Excellence in Natural Resource
Management and School of Plant Biology,
University of Western Australia, Albany 6330,
Australia,
8
Departamento de Biolog
ıa,
Universidad de C
adiz, Campus R
ıo San
Pedro, 11510 Puerto Real, Spain,
9
School of
Biological and Conservation Sciences,
University of KwaZulu-Natal, Scottsville 3209,
South Africa,
10
Department of Ecology and
Evolutionary Biology, University of California
(UCLA), Los Angeles, CA 90095, USA,
11
Department of Dynamic, Tectonic, Applied
Geology, Faculty of Geology and Geo-
Environment, University of Athens, Athens
15784, Greece,
12
Jet Propulsion Laboratory,
California Institute of Technology, Pasadena,
CA 91109, USA
*Correspondence: Richard Cowling,
Department of Botany, Nelson Mandela
Metropolitan University, P.O. Box 77000, Port
Elizabeth 6031, South Africa.
E-mail: richard.cowling@nmmu.ac.za
ABSTRACT
Aim Although all five of the major mediterranean-climate ecosystems (MCEs)
of the world are recognized as loci of high plant species diversity and ende-
mism, they show considerable variation in regional-scale richness. Here, we
assess the role of stable Pleistocene climate and Cenozoic topography in
explaining variation in regional richness of the globe’s MCEs. We hypothesize
that older, more climatically stable MCEs would support more species, because
they have had more time for species to accumulate than MCEs that were his-
torically subject to greater topographic upheavals and fluctuating climates.
Location South-western Africa (Cape), south-western Australia, California, cen-
tral Chile and the eastern (Greece) and western (Spain) Mediterranean Basin.
Methods We estimated plant diversity for each MCE as the intercepts of
species–area curves that are homogeneous in slope across all regions. We used
two down-scaled global circulation models of the Last Glacial Maximum
(LGM) to quantify climate stability by comparing the change in the location of
MCEs between the LGM and present. We quantified the Cenozoic topographic
stability of each MCE by comparing contemporary topographic profiles with
those present in the late Oligocene and the early Pliocene.
Results The most diverse MCEs –Cape and Australia –had the highest Ceno-
zoic environmental stability, and the least diverse –Chile and California –had
the lowest stability.
Main conclusions Variation in plant diversity in MCEs is likely to be a con-
sequence not of differences in diversification rates, but rather the persistence of
numerous pre-Pliocene clades in the more stable MCEs. The extraordinary
plant diversity of the Cape is a consequence of the combined effects of both
mature and recent radiations, the latter associated with increased habitat heter-
ogeneity produced by mild tectonic uplift in the Neogene.
Keywords
California, Cape Floristic Region, central Chile, diversification rate, mature
radiation, Mediterranean Basin, OCBIL, recent radiation, south-western
Australia, YODFEL.
INTRODUCTION
The world’s five mediterranean-climate ecosystems (MCEs)
have attracted interest as loci for studying ecosystem and
evolutionary convergence for almost 150 years (Grisebach,
1872; Cody & Mooney, 1978; Specht & Moll, 1983; Keeley
et al., 2012; Cardillo & Pratt, 2013). More recently, MCEs
have been the focus of numerous molecular-level studies on
552 http://wileyonlinelibrary.com/journal/jbi ª2014 John Wiley & Sons Ltd
doi:10.1111/jbi.12429
Journal of Biogeography (J. Biogeogr.) (2015) 42, 552–564
the diversification of plants (e.g. Hopper et al., 2009; Verboom
et al., 2009; Esp
ındola et al., 2010; Valente et al., 2011; Buerki
et al., 2012; Lancaster & Kay, 2013). Interest in plant evolution
in MCEs stems from their harbouring the world’s richest
extratropical floras (Cowling et al., 1996; Kreft & Jetz, 2007)
and having high levels of endemism across all spatial scales
(Cowling & Holmes, 1992; Hopper & Gioia, 2004; Georghiou
& Delipetrou, 2010; Kraft et al., 2010). All five MCEs have
been identified as biodiversity hotspots: areas which share
large numbers of endemic taxa that are being increasingly
threatened by human impacts (Myers et al., 2000).
Although MCEs exhibit some of the best-documented
examples of ecological convergence (Mooney, 1977; Cowling
& Witkowski, 1994), there are equally striking examples of
divergence (Cody & Mooney, 1978; Cowling & Campbell,
1980). Researchers have resorted to categorizing MCEs in
terms of their selective regimes in order to better explain
convergences and divergences among them. Thus, MCEs
have been differentiated in terms of climate (e.g. amount of
summer rain, reliability of winter rainfall) (Cowling et al.,
2005), soil nutrient status (Specht & Moll, 1983), fire regime
(Keeley et al., 2012), topography (Carmel & Flather, 2004)
and the interactions between climate, fire and soil nutrient
status (Keeley et al., 2012).
Contemporary ecological factors do not explain all the
variation in regional-scale plant diversity of MCEs (Cowling
et al., 1996; Valente & Vargas, 2013). Energy regimes, as
measured by potential evapotranspiration, are highest in the
South West Australian Floristic Region and the California
Floristic Province (Bradshaw & Cowling, 2014), which show
contrasting diversity patterns. Moreover, topographical heter-
ogeneity –a surrogate for habitat diversity –is highest in the
(relatively species-poor) eastern Mediterranean Basin and
central Chile and lowest in the (relatively species-rich) South
West Australian Floristic Region, whereas the Cape Floristic
Region, which is the most species-rich area by far, has mod-
erate heterogeneity (Bradshaw & Cowling, 2014). It appears
that variation in contemporary plant diversity patterns in
MCEs is profoundly influenced by their respective environ-
mental histories and their impacts on patterns of diversifica-
tion (speciation minus extinction) (Cowling et al., 1996,
2009; Jansson & Dynesius, 2002; Linder, 2008; Hopper, 2009;
Valente et al., 2011; Valente & Vargas, 2013).
In this regard, Hopper’s (2009) categorization of MCEs in
terms of environmental stability is informative: he differenti-
ated those associated with old, climatically buffered, infertile
landscapes (OCBILs), namely the South West Australian Flo-
ristic Region (hereafter Australia) and the Cape Floristic
Region of south-western South Africa (hereafter, ‘the Cape’),
and those dominated by young, often-disturbed, fertile land-
scapes (YODFELs), namely the Mediterranean Basin, the Cal-
ifornia Floristic Province (hereafter, ‘California’) and central
Chile (hereafter, ‘Chile’). Hopper (2009) predicted higher
plant richness and endemism in OCBIL than in YODFEL
MCEs, owing to higher rates of lineage persistence in
OCBILs. In other words, old, stable landscapes have had
more time for species to accumulate than landscapes that
have been subject to greater topographic upheavals and fluc-
tuating climates. Hopper (2009) did not define his categories
quantitatively, however, which we hope to redress here.
There are many examples in the literature of climatic and
topographic stability influencing processes that determine the
number of species supported by particular regions (e.g. Qian
& Ricklefs, 2000; Graham et al., 2006; Ara
ujo et al., 2007;
Mittelbach et al., 2007). No attempt has yet been made to
comprehensively assess the role of measures of historical cli-
matic and topographical heterogeneity in explaining regional
patterns of plant diversity in MCEs. We attempt to do this
here, by quantifying the climatic and topographic stability of
MCEs during the Pleistocene (climate) and the Neogene
(topography) in order to assess the OCBIL–YODFEL catego-
rization of MCEs. We then test whether OCBIL MCEs have
more diverse floras than YODFEL ones, as Hopper (2009)
predicted.
MATERIALS AND METHODS
Study area
We examined the five MCE biodiversity hotspots, namely Cal-
ifornia, the Cape, Chile, Australia and the Mediterranean
Basin. Because the Mediterranean Basin MCE occupies a huge
area (2.3 million km
2
) and includes a wide array of subcli-
mates (Blondel et al., 2010), we chose two subregions for fur-
ther analysis, one in the western basin (Spain) and one in the
eastern basin (Greece). These regions differ substantially in cli-
mate and biogeography. The western basin, west of the Sicily–
Cap Bon line, has less intense summer aridity than the eastern
basin and a higher ratio of autumn and spring rainfall to win-
ter rainfall (Blondel et al., 2010). Furthermore, climate-driven
isolation and vicariant differentiation of many Mediterranean
taxa seems to have occurred in the western and eastern ends
of the Mediterranean Basin during the Messinian Salinity
Crisis (Migliore et al., 2012) and, particularly, the Pleistocene
glaciations (Rodr
ıguez-S
anchez & Arroyo, 2008). This,
together with the west–east climatic differences, translates into
different biogeographical regions (Blondel et al., 2010). Thus,
we identified a total of six MCE regions for analysis.
Delimitation of mediterranean-climate ecosystems
There is no consensus regarding a global classification of
MCEs, although the general definition of cool, wet winters
and warm, dry summers applies to all. To avoid using an
MCE delimitation that may bias our results, we used three
models of MCE classification. The first is the long-standing,
expert-opinion classification proposed by K€
oppen (Geiger,
1961), and the remainder use algorithms widely applied in
the field of distribution modelling (Franklin, 2010): maxi-
mum entropy (Maxent; Phillips et al., 2006) and generalized
additive models (GAM; Hastie & Tibshirani, 1990). Both
expert opinion and statistical models were used to establish
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
553
Plant diversity and stability in mediterranean-climate ecosystems
the distribution of, and then changes in distribution of, med-
iterranean climate under current and past climate states.
The K€
oppen climate classification system defines MCE
areas (Csa and Csb classes) as warm temperate, with mini-
mum monthly temperatures between 3°C and 18 °C, and
with at least four months above 10 °C. K€
oppen (Geiger,
1961) applied three criteria to ensure that winter precipita-
tion predominates: (1) that the minimum summer monthly
rainfall is less than the minimum winter monthly rainfall;
(2) that maximum winter monthly rainfall is three times the
minimum summer monthly rainfall; and (3) that the
minimum summer monthly rainfall is less than 40 mm. To
differentiate between mediterranean and more arid winter-
rainfall regions, K€
oppen (Geiger, 1961) applied an aridity
index, so that annual precipitation (mm) exceeds 20 times
mean annual temperature (°C) (where two-thirds of precipi-
tation occurs in winter).
The two statistical models were trained using the distribu-
tion of mediterranean-type vegetation and five bioclimatic
variables (annual mean temperature, maximum temperature
of the warmest month, annual temperature range, precipita-
tion of the warmest quarter and precipitation of the wettest
quarter; Hijmans et al., 2005). These climate variables cap-
ture aspects of temperature, precipitation, seasonality and
continentality. There are, however, two potential problems.
First, there is no standardized global floristic or vegetation
map of mediterranean-type vegetation. Second, mediterra-
nean climate and mediterranean-type vegetation are not
always strongly associated (Keeley et al., 2012); the fynbos of
the Cape Floristic Region, for example, extends far into a
non-seasonal rainfall regime. We identified the distribution
of mediterranean-type vegetation from available floristic or
vegetation maps, and where these maps were misleading (e.g.
the eastern Cape Floristic Region) (Keeley et al., 2012), we
employed expert opinion to refine the area further (see
Appendix S1 in Supporting Information for further details).
We selected 500 localities at random from within each regio-
nal vegetation map. We used these in conjunction with
10,000 background points sampled across the Earth’s terres-
trial surface to train and test the distribution models. To
avoid any bias that may have been introduced by the non-
standardized vegetation maps and any expert-opinion alter-
ation thereof, we trained the statistical models for any given
MCE region using all localities except those from the region
in question. This also avoids any potential circularity of
including mapped vegetation for a given MCE region that
may, in part, have included climate as a delimiter. These
excluded locality points were then used to test the model
(i.e. k-folding with each fold representing an MCE) and to
calculate the threshold to convert the continuous probability
of occurrence maps into binary presence/absence. The con-
version to the binary maps utilized the ‘equal sensitivity plus
specificity’ threshold criterion, although the relative ranking
of the results were robust to the threshold criterion selected
(e.g. maximum kappa or maximum sensitivity plus specific-
ity; results not shown).
We constructed the expert-opinion climate model and sta-
tistical models in R 3.0.1 (R Core Team, 2013) and using cli-
mate layers from the WorldClim database (Hijmans et al.,
2005). The distribution models were constructed in R using
Maxent 3.3.3e called from the dismo library 0.9-3 (Hijmans
et al., 2013), while GAMs were conducted using the mgcv
library 1.7-29 (Wood, 2014).
Plant diversity
We compiled regional richness data (native species only) for
MCEs from Cowling et al. (1996, 1997). Additional sources
were: Australia, NatureMap (http://naturemap.dec.wa.gov.au/
default.aspx); Cape, Cowling & Lombard (2002); and Spain,
Ojeda et al. (2000). Because there are no suitable data from
mainland Greece, we had to use data from the larger islands of
the Aegean Sea (see Appendix S2 for data sources). With the
exception of Thira (Santorini), which is an active volcanic cal-
dera, and Kos, the south-western part of which is characterized
by contemporary volcanic activity, the islands are all continen-
tal islands. We would nonetheless expect that their species–area
relationship would have a steeper curve and lower intercept
than mainland samples (Rosenzweig, 1995). All sites were
located within the strictly MCE area of each region. We fitted
the species–area data for each of the six regions to a double-
logarithmic regression model and tested for homogeneity of
slopes (z) and intercepts (c), following Rosenzweig (1995).
We assessed the relationships between an index of diversity
(quantified as the intercept of the respective species–area
model) and measures of climatic and topographic stability
described below.
Climate stability
We assessed the climate stability of MCEs as the difference in
climate between the Last Glacial Maximum (LGM) and the
present day, representing one of the largest fluctuations in cli-
mate conditions experienced during the Quaternary. We used
two down-scaled global circulation models (GCMs) of LGM
climate: the Community Climate System Model (CCSM; Col-
lins et al., 2004) and the Model for Interdisciplinary Research
on Climate (MIROC; Hasumi & Emori, 2004). These LGM
climate estimates have been statistically down-scaled using
the WorldClim data set (Hijmans et al., 2005) and GCM data
from the Paleoclimate Modelling Intercomparison Project II
(PMIP2) and are available at http://www.worldclim.org/. We
performed analyses using both GCMs for each of the MCE
delimitations. We calculated climate stability by comparing
the change in the location of MCE between the LGM and the
present, i.e. the percentage of the current MCE area that also
existed under LGM conditions.
Topographic stability
Within each MCE, we located c. 350-km transects that are
representative of each region’s topography. Using published
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
554
R. M. Cowling et al.
sources (Table 1), we manually reconstructed topographic
profiles for each region during the late Oligocene (c. 20 Ma)
and the early Pliocene (c. 5 Ma) (Fig. 1). We chose the late
Oligocene as a starting point because it preceded a period of
prolonged relative aridity which, in conjunction with fire,
would have promoted the expansion of sclerophyllous vegeta-
tion allied to that in present-day MCEs (Cowling et al., 2009;
Keeley et al., 2012). The Oligocene also witnessed radiation of
many MCE lineages, especially in Australia and the Cape
(Linder, 2005; Hopper et al., 2009). In order to quantify the
topographic stability for the two geological periods (late Oli-
gocene to late Miocene and early Pliocene to present), we
developed a stability index based on the information on topo-
graphical change generated from the geological descriptions
(Table 1, Fig. 1). This stability index captures the elevational
change relative to the horizontal axis of the MCE region. It
was calculate as 1
DElev
Dist , where DElev is the change in eleva-
tion over the horizontal distance. A value of 1 indicates com-
plete stability, whereas a value of 0 or less indicates equal or
greater vertical movement than the horizontal distance.
RESULTS
Diversity patterns
Although the intercepts (c) of the species–area regressions for
the six MCEs (Fig. 2) were significantly different among the
MCEs (F=34.967, d.f.
1
=5, d.f.
2
=67, P<0.0001), the
slopes were homogeneous (F=0.484, d.f.
1
=5, d.f.
2
=72,
P=0.787). It was thus acceptable to compute the c-ratio for
any two curves (Gould, 1979). This value –the ratio of the val-
ues of the intercepts in arithmetical space –provides a measure
of relative species densities (Rosenzweig, 1995). We computed
c-ratios comparing the Cape, the region with the highest value,
with each of the other MCEs. This showed that for similar-
sized areas, the Cape was 1.23 times richer than Australia, 1.62
times richer than Spain, 1.91 times richer than California, 2.08
times richer than Greece and 2.63 times richer than Chile.
Climate stability
Estimates of the area of MCE predicted to have also been
MCE at the LGM varied depending on the model (K€
oppen,
Maxent and GAM) and global climate model (MIROC and
CCSM) used, resulting in high intra- and inter-region vari-
ance of stability estimates (Fig. 3, Appendix S3). Some gen-
eral patterns were nonetheless evident. The Cape and
Australia had consistently high stability, with one exception
for Australia (Appendix S3). Chile had the lowest climatic
stability, followed by California. Spain and Greece had inter-
mediate levels of stability.
Topographic stability
Chile, California and the Mediterranean Basin represent
zones of tectonic plate convergence, whereas the Cape and
Australia are located away from plate boundaries (Table 1).
Hence, the former three regions have experienced consider-
able, albeit varied topographic instability during the Neo-
gene. Within the Mediterranean Basin, Cenozoic topographic
stability has been almost twice as high in the east (Greece) as
in the west (Spain), the west having experienced the least sta-
ble conditions of all the MCEs. In California and Chile, over-
all instability was driven largely by the very high values
between the late Pliocene and the present. Palaeogene land-
scapes in the two regions were much more stable than Neo-
gene ones. On the other hand, the Cape and Australia have
remained largely unchanged during the Cenozoic (Fig. 1).
Even so, the Cape experienced some Neogene uplift and this
has exposed clay-rich substrata on the coastal forelands,
areas that had previously been mantled with duricrusts and
sandplains.
Relationships between diversity and stability
The highest plant diversities –quantified as the intercept
(c-value) of the respective MCE –were generally recorded
for the most climatically and topographically stable MCEs,
namely Cape and Australia (Fig. 3). Spain, an MCE having
the lowest topographic stability but relatively high climatic
stability, showed intermediate diversity. California and
Chile, the MCEs with the lowest stability, including very
low measures of topographic stability since the late Pliocene,
had the lowest diversity values. Greece was an anomaly,
having lower diversity that expected on the basis of its
stability.
DISCUSSION
Our results are consistent with research for many different
biomes and taxa: regions of high environmental stability dur-
ing the Cenozoic are associated with high species (and
genetic) diversity (e.g. Graham et al., 2006; Ara
ujo et al.,
2007; Carnaval et al., 2009; Werneck et al., 2011, 2012). To
our knowledge, no studies have attempted to quantify the
degree of both climatic and topographic stability within all
examples of a particular biome on Earth, and to link this to
contemporary diversity of plant species. In this sense, our
results are novel and interesting. However, we acknowledge
the limitations of this study. First, our statistical inference is
constrained by the few degrees of freedom associated with
our study system, although the global scale of the system is
in our favour. Second, there is an unavoidable mismatch
between the temporal scales we used for climate and topo-
graphic stability. Unfortunately, it is not yet feasible to hind-
cast climates much beyond the mid-Pleistocene at the scale
required for this study. Climatic instability during the Pleis-
tocene was nonetheless more intense than any other time
during the Cenozoic (Zachos et al., 2001) and this is likely
to have had a profound effect on plant extinctions (Jansson
& Dynesius, 2002). Furthermore, it is not yet feasible to
hindcast topographic dynamics to the point that accurate
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
555
Plant diversity and stability in mediterranean-climate ecosystems
Table 1 Major geomorphic processes affecting topographic dynamics since the late Cenozoic in six mediterranean-climate ecosystems
(MCEs).
MCE Late Oligocene to late Miocene Early Pliocene to Pleistocene References
California Multiple orogenies, dating back to the middle
Mesozoic and resulting from subduction
associated with the Farallon and Pacific plates,
lifted up the proto-Sierra Nevada, but to a much
lesser extent than the contemporary mountains.
The current area of the Coast Ranges remained
beneath the Pacific Ocean. The Central Valley,
which began as a trough associated with tectonic
forces of subduction in the Mesozoic, formed a
large marine embayment for most of the
Cenozoic.
Although it was thought until recently that
fault-block tilting during the past 10–5 Myr
was largely responsible for the uplift of the
high elevation of the modern Sierra Nevada,
new data from analyses of tectonics suggest
that the range achieved heights >3000 m in
the Palaeogene and remained high through
subsequent millennia. The form, topography
and elevation of the modern Sierra Nevada
were, however, strongly influenced by
extensional and faulting processes over the
past 3 Myr, which added major uplift to the
southern Sierra Nevada. This orogeny left a
legacy of volcanic geomorphology in the
northern portion of the range. All but the
lowest peaks were severely glaciated during
the Pleistocene glacials. Significant uplift of
the Coast Ranges began in the Pliocene,
about 3.5 Ma, and is still ongoing. The uplift
is a response to compressional deformation
associated with strike-slip movement of the
Pacific and North American plates along the
San Andreas Fault. These ranges have a
complex geological history with granitic
basement rock lying west of the fault and
deep-ocean metamorphic and sedimentary
rocks, including serpentines, east of the fault.
Scattered volcanic activity exists all along the
ranges. Rotational movements of microplates
along a bend in the fault produce the east–
west orientation of the Transverse Ranges of
contemporary southern California. The onset
of uplift of the Coast Ranges blocked the
connection of the Central Valley to the
Pacific Ocean, and the basin is filled with
continental sediments derived from the
growing Sierra Nevada. As late as the late
Pleistocene, however, large areas of the valley
formed a freshwater lake of glacial meltwater.
Montgomery (1993); Small
& Anderson (1995); Wolfe
et al. (1998); Kellogg &
Minor (2005); Mix et al.
(2011)
Spain The Baetic Cordillera at the southern end of the
Iberian Peninsula achieved moderate uplift
resulting from a collision between the Alboran
Terrane and the Iberian microplate, but only the
proto-Sierra Nevada, the westernmost mountain
range of the European Alpine belt, is emergent
from the sea at the time.
The ongoing orogeny resulted in the uplift of
mountains in the western Baetic Cordillera by
the middle Miocene. By the late Messinian (c.
5 Ma), the Baetic Cordillera approximated its
present topography.
Michard et al. (2002); Braga
et al. (2003); Iribarren
et al. (2009); Mart
ın et al.
(2009)
Greece Except for the easternmost range, including
present-day Mount Olympus, the region was
submerged. Considerable topography, trending
along an east–west axis, existed in this zone and
molassic sediments were being deposited on its
western margin (Meteora), building a thick
conglomerate cover derived from continental
material. This topography was a consequence of
the subduction of the African Plate beneath
Eurasia.
Shortly after the late Miocene, mainland
Greece –a more-or-less rigid slab –was
rotated clockwise, a consequence of the
westward movement of the Anatolian Plate
towards the Aegean Sea. This produced the
contemporary NNW-trending direction of the
Hellenides and resulted in the uplift of the
Pindos mountain chain in the western
mainland as well as the emergence of
numerous basins and ranges (the Hellenides)
between this chain and the Olympus massif,
which remains active to this day.
Kahle et al. (1998); Royden
& Husson (2006);
Papanikolaou & Royden
(2007); Reilinger et al.
(2010); Vassilakis et al.
(2011); Pearce et al. (2012)
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
556
R. M. Cowling et al.
assessments of changes in landscape structure can be made
or to identify thresholds of topostability relevant to evolu-
tionary processes. Nevertheless, the results presented here
provide new information on the quantification of MCEs’ cli-
matic and topographic stability in relation to OCBIL and
YODFEL categorization, and address some fundamental
issues about the evolution of plant diversity.
Generally, more stable MCEs –both climatically and
topographically (Cape and Australia) –have the highest con-
temporary plant diversity, whereas the least stable (Califor-
nia and Chile) have the lowest diversity. Spain, an area of
low topographic stability and high climatic stability, occu-
pied an intermediate position in terms of diversity. Greece
had a lower diversity than expected from its stability (see
Table 1 Continued
MCE Late Oligocene to late Miocene Early Pliocene to Pleistocene References
Chile Proto-Andean orogeny, dating from the late
Proterozoic to the breakup of Pangaea at the end
of the Permian, produced considerable relief,
albeit much less than at present. The region was
still exposed to the incursion of moist air from
the Atlantic Ocean to the east.
The uplift of the Andean Cordillera was
initiated in the early Miocene, a consequence
of the subduction of the Nazca Plate beneath
the continental South American Plate. The
compressional forces along the western
margin of the South American Plate, resulted
in the uplift, faulting and folding of ancient
sedimentary and metamorphic cratons to the
east. By 15 Ma, the Andes had been lifted
enough to develop the hyperarid conditions
of the Atacama Desert. South of 33°S, the
dip angle of the subducting plate was
relatively steep, producing active volcanism.
North of this, at 28–33°S, the dip angle was
gentle and volcanism over this latitude ended
9–10 Ma. The Cordillera de la Costa (coastal
range) in central Chile separated from the
Andes in the Neogene as the result of the
subsidence that formed the Central Valley.
This range is dominated by granites of
Carboniferous to Permian age, which
represent part of a proto-Andean orogeny.
Alpers & Brimhall (1988);
Gregory-Wodzicki (2000);
Charrier et al. (2002);
Hartley (2003)
Cape The quartzitic sandstone core of the Cape Folded
Belt, exhumed by the break-up of Gondwana in
the Cretaceous (140–65 Ma), was subject to
erosion that decreased markedly in the Cenozoic.
Essential features of Cretaceous erosion persist in
the complex topography of the mountains, with
topography similar to present. Lowlands and
intermontane valleys were capped with silcretes
and ferricretes that were deposited in the early
Palaeocene.
Mild tectonic uplift during the Miocene and
early Pliocene, and the consequent
denudation of the lowlands –underlain by
shales –resulted in the erosion of sil-
ferricretes and the exposure of clay-rich
substrata. Owing to slow denudation rates
associated with the dominant quartzitic
sandstones, mountain topography remained
largely unchanged, although some incision of
softer (shale) intermontane valleys probably
occurred. The entire region was tectonically
stable during the mid-Pliocene, although
regression during Pleistocene glacials exposed
a large area (up to 200 km offshore from the
present day) of the Agulhas Bank in the
south east and 60–80 km offshore along the
west coast.
Partridge & Maud (1987);
Tinker et al. (2008);
Cowling et al. (2009);
Fisher et al. (2010);
Erlanger et al. (2012);
Scharf et al. (2013)
Australia Tectonically stable since the mid-Proterozoic, the
landscape was a gently-dissected and slowly
eroding palaeosurface on basement granitoid
rock. The Darling Scarp and its south coast
equivalent (the Ravensthorpe Ramp) are elevated
–starting in the mid Cenozoic –by minor
marginal up-warping. On the Darling Plateau, a
phase of haematite formation occurred 10 Ma.
The Stirling Range is an old intrusion of
quartzite, offering the highest peaks (to 1100 m)
on the otherwise subdued palaeoland surface.
Very slow erosion and weathering was
associated with accentuated aridity and the
onset of mediterranean climate. Some of the
oldest persistent landscapes on Earth survived
this period with relatively little change.
Shorelines oscillated with sea-level change
during the Pleistocene, up to 100 km offshore
from present day at the Last Glacial
Maximum.
Finkl & Fairbridge (1979);
Kendrick et al. (1991);
Anand & Paine (2002);
Anand (2005); Pillans
(2007); Jakica et al. (2011)
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
557
Plant diversity and stability in mediterranean-climate ecosystems
below). Differences in Neogene plant diversification rates are
unlikely to have produced these patterns: these rates,
although variable, are highest in the Mediterranean Basin,
intermediate in California and the Cape, and lowest in Aus-
tralia (Linder, 2008; Madri~
n
an et al., 2013). Studies on lin-
eages shared between MCEs indicate that contrasting
diversification rates do not explain differences in extant spe-
cies richness (Hopper, 2009; Sauquet et al., 2009; Valente
et al., 2011; Buerki et al., 2012; Valente & Vargas, 2013).
Instead, we suggest that the persistence of older clades in
the more stable MCEs of the Cape and Australia is a more
likely explanation for contemporary diversity patterns. These
regions include many extant species-rich clades that began
to diversify as early as the Eocene but mostly diversified
from the early Oligocene to the mid-Miocene (Linder, 2005;
Hopper et al., 2009; Sauquet et al., 2009; Verboom et al.,
2009; Schnitzler et al., 2011; Cardillo & Pratt, 2013; Valente
& Vargas, 2013), a feature largely absent from other MCEs,
where ancient lineages are rare and, when extant, show little
evidence of diversification, i.e. they are relicts (e.g.
Rodr
ıguez-S
anchez & Arroyo, 2008; Sauquet et al., 2009;
Lancaster & Kay, 2013).
Andes
Transverse
Valleys
California (36.3° N)
Spain (37.0° N)
Chile (32.4° S)
Cape (34.0° S)
Australia (34.2° S)
present-day coastline
Coast
Range
Table
Mt
HoƩentots Holland Mts Langeberg
SƟrling Range
Pampas
0 50 100 150 200 250 300 350 km
6000 m
5000 m
4000 m
3000 m
2000 m
1000 m
sea level
Sierra Nevada
Sierra Nevada
Central Valley
Coast Range
Guadalquivir Delta
Greece (40.0° N)
Corfu
Mt Olympus
Ioannina
Pindos Mts Meteora
Figure 1 Topographic profiles of the six mediterranean-climate ecosystems (MCEs) in the late Oligocene (dotted), early Pliocene
(dashed) and present day (solid lines). Vertical arrows mark the historical coastlines at the corresponding times. Shaded horizontal bars
delimit the extent of the MCE along the profiles. Because of the low topographic change in Australia, differences in the profiles for each
period have been exaggerated.
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
558
R. M. Cowling et al.
There are, of course, factors other than environmental
stability that could determine differences in patterns of regio-
nal plant diversity among MCEs, notably differences in distur-
bance regimes (history of human impacts; fire regimes)
(Blondel et al., 2010; Keeley et al., 2012) and soil fertility
(Wisheu et al., 2000; Lambers et al., 2010). It may be argued
that the association between landscape stability and low soil
fertility (Hopper, 2009) makes it impossible to untangle the
independent effects of either factor. Young landscapes can,
however, have infertile soils (e.g. the pine barrens of eastern
North America and the mountain heathlands of south-western
Spain) and old landscapes can have fertile soils (e.g. exposures
of mudstones in the Cape and dolerite outcrops in Australia).
The patterns of plant diversity in these regions are more
consistent with OCBIL theory than a theory based on soil
fertility per se (Ojeda et al., 2001; R.M.C. & S.D.H., pers. obs.).
Within the overall pattern we document, there are some
interesting anomalies. Why does Australia have lower diver-
sity than the Cape, given their similar environmental histo-
ries? We can think of at least two reasons. The lower
topographic heterogeneity of Australia both now (Bradshaw
& Cowling, 2014) and during the Tertiary, may have con-
strained opportunities for radiation owing to the shallower
environmental gradients there than in the Cape, with its
impressive and ancient topography (Tinker et al., 2008). Sec-
ond, mild Neogene uplift in the Cape eroded duricrusts and
sandplains to create large tracts of novel, moderately fertile
habitats associated with shale bedrock (Cowling et al., 2009),
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Cape
2.5 3.0 3.5
r² = 0.94
LogS = 0.25logA + 2.53
●
●●
●●
●
●
●
●
●
●
●
Australia
r² = 0.91
LogS = 0.23logA + 2.45
●
●●
●
●
●
●
Spain
2.5 3.0 3.5
r² = 0.78
LogS = 0.24logA + 2.33
●
●
●
●
●
●
●
●
●
●
●
California
r² = 0.94
LogS = 0.23logA + 2.26
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Greece
2.5 3.0 3.5
01234
r² = 0.75
LogS = 0.28logA + 2.22
●
●
●
●
●
●
●
●
●
●
●
●
Chile
01234
r² = 0.92
LogS = 0.23logA + 2.12
Log(area [km²])
Log(species richness)
Figure 2 Plant species–area relationships
for six mediterranean-climate ecosystems
(MCEs) with regression models.
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
559
Plant diversity and stability in mediterranean-climate ecosystems
which initiated a flurry of recent diversification (Verboom
et al., 2009). Moreover, the now-submerged Agulhas Bank
was a relatively fertile landscape, covering an area more than
half the size of the present-day Cape, which was exposed for
much of the Pleistocene (Fisher et al., 2010); this YODFEL
supported a largely extinct fauna of grazing mammals (Mare-
an et al., 2014) associated in the context of the present-day
Cape with novel habitats that are likely to have been loci of
Pleistocene plant radiations. Products of these radiations are
likely to persist on the present-day Cape littoral and adjacent
lime-rich hinterland (Cowling & Holmes, 1992). In Australia,
more recent YODFEL-like landscapes are mainly restricted to
the Swan Coastal Plain and south coast and associated off-
shore shelf, a much smaller area than in the Cape. Like the
Cape, however, this region is associated with recent diversifi-
cation, albeit on a much smaller scale than in the Cape, and
with evident persistence of old lineages as well (Coates et al.,
2003; Nevill et al., 2014). Overall, as shown by Linder
(2008), the poorer Australian flora may be dominated by
mature (pre-Pliocene) radiations, whereas the richer Cape
flora includes an abundance of both recent (Plio-Pleistocene)
radiations associated with the younger, lowland landscapes
and mature radiations associated with the ancient Cape
mountains (Verboom et al., 2009; Slingsby et al., 2014).
Why does Greece have lower plant diversity than expected
on the basis of its environmental history? An obvious reason
is our use of island data to assess regional richness patterns
in Greece. In particular, as predicted by island biogeographi-
cal theory, the depauperate floras of the smaller islands ele-
vated the slope, albeit not significantly, and depressed the
c-value for the Greek data; the larger islands had comparable
diversity to similar-sized regions in nearby Spain. Moreover,
the MCE part of mainland Greece, which comprises an area
similar to the Cape Floristic Region, supports half the num-
ber of species (Valente & Vargas, 2013); this is consistent
with our reported Cape–Greece c-ratio of 2.08. Thus, there
may be a longitudinal (west–east) gradient of declining plant
diversity in the Mediterranean Basin, as has been demon-
strated for the Cape (Cowling & Lombard, 2002) and
Australia (Sniderman et al., 2013).
Our results are consistent with the notion that given suffi-
cient stability, plant hyperdiversity can develop outside the
humid tropics (Cowling et al., 1996, 2009; Hopper et al.,
2009; Sniderman et al., 2013) implying that water and energy
variables are not consistent predictors of high regional-scale
plant richness (Davies et al., 2004, 2005; Kreft & Jetz, 2007;
Cowling et al., 2009; Hopper, 2009). The concentration of
plant species in the humid tropics of the world is likely to be
a consequence of Cenozoic environmental stability at these
latitudes (Ricklefs, 2004). Moreover, rich floras can be the
product of mature radiations (Australia and the Cape moun-
tains) or recent radiations, such as the succulent karoo and
the Cape lowlands (Linder, 2008; Verboom et al., 2009) and
alpine habitats in the tropics (Hughes & Eastwood, 2006;
Madri~
n
an et al., 2013). The extraordinary plant diversity of
the Cape is a consequence of the combined effects of both
mature and recent radiations (Linder, 2008). This raises the
issue of the importance of relative stability, or how much
stability is necessary for the evolution of hyperdiversity. Too
much stability, especially topographic stability, leads to a
drop in diversification rates, as may be the case in Australia
(Linder, 2008) and on the Cape mountains (Slingsby et al.,
2014). Too much instability, as occurred in California and
Chile during the Pleistocene, results in high extinction rates
and a reduction in species numbers. The Cape appears to
have had –at least for many of its component lineages –just
the right amount of environmental heterogeneity for the
preservation of old clades and the radiation of younger ones.
The hypothesis presented here –that environmentally sta-
ble MCEs have higher diversity owing to a greater persistence
of lineages over time –yields predictions that can be tested
using dated molecular phylogenies. For example, lineages-
through-time plots would differ systematically among the
five MCEs. The environmentally stable MCEs –Cape and
Australia –would show a greater spread of lineages across
time-slices of the Cenozoic (i.e. a relatively constant rate of
diversification), whereas the less stable MCEs would show
patterns skewed in favour of younger lineages (i.e. increased
diversification rates towards the present). The limited avail-
able data are largely consistent with this hypothesis (Hopper
0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.4 0.6 0.8 1.0
Climate stability
Topographic stability
●
●
●
●
●
●
●
●
●
●
●
●
●
California
Chile
Greece
Spain
Cape
Austr.
Diversity index
2.12 2.22 2.26 2.33 2.45 2.53
Figure 3 Relationships between diversity indices (c-value of
species–area regression; see Fig. 2) and mean values of
topographic and climate stability in six mediterranean-climate
ecosystems (MCEs); a topographic stability index of 1 indicates
complete stability. The downward-pointing triangles are the
index for the period from the late Oligocene to the late
Miocene; upward-pointing triangles are for the period from the
early Pliocene to the present-day. Mean climate stability values
for mediterranean climate models (K€
oppen, Maxent, GAM) and
down-scaled global climate simulations (MIROC and CCSM)
are shown as black dots, and individual combinations are shown
in grey. (See Appendix S2 for values).
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
560
R. M. Cowling et al.
et al., 2009; Sauquet et al., 2009; Valente et al., 2011; Buerki
et al., 2012; Valente & Vargas, 2013).
In conclusion, the patterns we have shown and the pro-
cesses invoked to explain them are largely consistent with the
predictions of Hopper’s (2009) OCBIL theory: old, climati-
cally buffered landscapes are associated with high contempo-
rary diversity, owing to the persistence of old lineages.
However, the hyperdiversity of the Cape may well be a con-
sequence of the juxtaposition of an ancient and topographi-
cally heterogeneous landscape (the Cape Fold Belt) and a
relatively young lowland landscape. The former provided a
pool of lineages for colonizing and diversifying recently and
rapidly on these lowlands.
ACKNOWLEDGEMENTS
We thank the South African National Biodiversity Institute
for hosting the workshop on which this contribution is
based. R.M.C. acknowledges the National Research Founda-
tion and Nelson Mandela Metropolitan University for fund-
ing; A.J.P. was supported by funding from the Claude Leon
Foundation; P.W.R. was supported by funding from the
Stunt Ranch Santa Monica Mountains Reserve; M.A. thanks
M. Panitsa and I. Bazos for providing references for the flora
of the Aegean islands.
REFERENCES
Alpers, C.N. & Brimhall, G.H. (1988) Middle Miocene cli-
matic change in the Atacama Desert, northern Chile: evi-
dence from supergene mineralization at La Escondida.
Geological Society of America Bulletin,100, 1640–1656.
Anand, R.R. (2005) Weathering history, landscape evolution
and implications for exploration. Regolith landscape evolu-
tion across Australia: a compilation of regolith- landscape
case studies and landscape evolution models (ed. by R.R.
Anand and P. de Broekert), pp. 15–45. Cooperative
Research Centre for Landscape Environments and Mineral
Exploration, Perth, Australia.
Anand, R.R. & Paine, M. (2002) Regolith geology of the
Yilgarn Craton, Western Australia: implications for
exploration. Australian Journal of Earth Sciences,49,
3–162.
Ara
ujo, M.B., Nogu
es-Bravo, D., Diniz-Filho, J.A.F., Hay-
wood, A.M., Valdes, P.J. & Rahbek, C. (2007) Quaternary
climate changes explain diversity among reptiles and
amphibians. Ecography,31,8–15.
Blondel, J., Aronson, J., Boudiou, J.Y. & Boeuf, G. (2010)
The Mediterranean region: biological diversity in space and
time. Oxford University Press, Oxford.
Bradshaw, P.L. & Cowling, R.M. (2014) Landscapes, rock
types and climate of the Greater Cape Floristic Region.
Fynbos: ecology, evolution and conservation of a megadiverse
region (ed. by N. Allsopp, J.F. Colville and G.A. Verboom).
Oxford University Press, Oxford (in press).
Braga, J.C., Mart
ın, J.M. & Quesada, C. (2003) Patterns and
average rates of late Neogene–Recent uplift of the Betic
Cordillera, SE Spain. Geomorphology,50,3–26.
Buerki, S., Jose, S., Yadav, S.R., Goldblatt, P., Manning, J.C.
& Forest, F. (2012) Contrasting biogeographic and diversi-
fication patterns in two Mediterranean-type ecosystems.
PLoS ONE,7, e39377.
Cardillo, M. & Pratt, R. (2013) Evolution of a hotspot genus:
geographic variation in speciation and extinction rates in
Banksia (Proteaceae). BMC Evolutionary Biology,13, 155.
Carmel, Y. & Flather, C.H. (2004) Comparing landscape
scale vegetation dynamics following recent disturbance in
climatically similar sites in California and the Mediterra-
nean basin. Landscape Ecology,19, 573–590.
Carnaval, A.C., Hickerson, M.J., Haddad, C.F.B., Rodrigues,
M.T. & Moritz, C. (2009) Stability predicts genetic diver-
sity in the Brazilian Atlantic forest hotspot. Science,323,
785–789.
Charrier, R., Baeza, O., Elgueta, S., Flynn, J.J., Gans, P., Kay,
S.M., Mu~
noz, N., Wyss, A.R. & Zurita, E. (2002) Evidence
for Cenozoic extensional basin development and tectonic
inversion south of the flat-slab segment, southern Central
Andes, Chile (33°–36°S.L.). Journal of South American
Earth Sciences,15, 117–139.
Coates, D.J., Carstairs, S. & Hamley, V.L. (2003) Evolution-
ary patterns and genetic structure in localized and wide-
spread species in the Stylidium caricifolium complex
(Stylidiaceae). American Journal of Biology,90, 997–1008.
Cody, M.L. & Mooney, H.A. (1978) Convergence versus
nonconvergence in mediterranean-climate ecosystems.
Annual Review of Ecology and Systematics,9, 265–321.
Collins, W.D., Bitz, C.M., Blackmon, M.L., Bonan, G.B.,
Bretherton, C.S., Carton, J.A., Chang, P., Doney, S.C.,
Hack, J.J., Henderson, T.B., Kiehl, J.T., Large, W.G., McK-
enna, D.S., Santer, B.D. & Smith, R.D. (2004) The Com-
munity Climate System Model Version 3 (CCSM3).
Journal of Climate,19, 2122–2143.
Cowling, R.M. & Campbell, B.M. (1980) Convergence in
vegetation structure in the mediterranean communities of
California, Chile and South Africa. Vegetatio,43, 191–197.
Cowling, R.M. & Holmes, P.M. (1992) Endemism and speci-
ation in a lowland flora from the Cape Floristic Region.
Biological Journal of the Linnean Society,47, 367–383.
Cowling, R.M. & Lombard, A.T. (2002) Heterogeneity, speci-
ation/extinction history and climate: explaining regional
plant diversity patterns in the Cape Floristic Region.
Diversity and Distributions,8, 163–179.
Cowling, R.M. & Witkowski, E.T.F. (1994) Convergence and
non-convergence of plant traits in climatically and edaphi-
cally matched sites in Mediterranean Australia and South
Africa. Australian Journal of Ecology,19, 220–232.
Cowling, R.M., Rundel, P.W., Lamont, B.B., Arroyo, M.K. &
Arianoutsou, M. (1996) Plant diversity in mediterranean-
climate regions. Trends in Ecology and Evolution,11, 362–
366.
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
561
Plant diversity and stability in mediterranean-climate ecosystems
Cowling, R.M., Richardson, D.M., Schulze, R.J., Hoffman,
M.T., Midgley, J.J. & Hilton-Taylor, C. (1997) Species
diversity at the regional scale. Vegetation of southern Africa
(ed. by R.M. Cowling, D.M. Richardson and S.M. Pierce),
pp. 447–473. Cambridge University Press, Cambridge.
Cowling, R.M., Ojeda, F., Lamont, B.B., Rundel, P.W. &
Lechmere-Oertel, R. (2005) Rainfall reliability, a neglected
factor in explaining convergence and divergence of plant
traits in fire-prone mediterranaean-climate ecosystems.
Global Ecology and Biogeography,14, 509–519.
Cowling, R.M., Proches
ß,S
ß. & Partridge, T.C. (2009) Explain-
ing the uniqueness of the Cape flora: incorporating geomor-
phic evolution as a factor for explaining its diversification.
Molecular Phylogenetics Evolution,51,64–74.
Davies, T.J., Savolainen, V., Chase, M.W., Moat, J. & Barrac-
lough, T.G. (2004) Environmental energy and evolutionary
rates in flowering plants. Proceedings of the Royal Society B:
Biological Sciences,271, 2195–2200.
Davies, T.J., Savolainen, V., Chase, M.W., Goldblatt, P. &
Barraclough, T.G. (2005) Environment, area, and diversifi-
cation in the species-rich flowering plant family Iridaceae.
The American Naturalist,166, 418–425.
Erlanger, E.D., Granger, D.E. & Gibbon, R.J. (2012) Rock
uplift rates in South Africa from isochron burial of fluvial
and marine terraces. Geology,40, 1019–1022.
Esp
ındola, A., Buerki, S., Bedalov, M., K€
upfer, P. & Alvarez,
N. (2010) New insights into the phylogenetics and bioge-
ography of Arum (Araceae): unravelling its evolutionary
history. Botanical Journal of the Linnean Society,163,
14–32.
Finkl, C.W. & Fairbridge, R.W. (1979) Paleogeographic
evolution of a rifted cratonic margin: S.W. Australia. Pal-
aeogeography, Palaeoclimatology, Palaeoecology,26, 221–
252.
Fisher, E.C., Bar-Matthews, M., Jerardino, A. & Marean,
C.W. (2010) Middle and Late Pleistocene paleoscape mod-
eling along the southern coast of South Africa. Quaternary
Science Reviews,29, 1382–1398.
Franklin, J. (2010) Mapping species distributions: spatial infer-
ence and prediction. Cambridge University Press, Cam-
bridge, UK.
Geiger, R. (1961) K€
oppen–Geiger Klima der Erde. 1:16 million
scale wall map. Klett-Perthes, Gotha, Germany.
Georghiou, K. & Delipetrou, P. (2010) Patterns and traits of
the endemic plants of Greece. Botanical Journal of the Lin-
nean Society,162, 130–422.
Gould, S.J. (1979) An allometric interpretation of species-
area curves: the meaning of the coefficient. The American
Naturalist,114, 335–343.
Graham, C.H., Moritz, C. & Williams, S.E. (2006) Habitat
history improves prediction of biodiversity in rainforest
fauna. Proceedings of the National Academy of Sciences
USA,103, 632–636.
Gregory-Wodzicki, K.M. (2000) Uplift history of the Central
and Northern Andes: a review. Geological Society of Amer-
ica Bulletin,112, 1091–1105.
Grisebach, A. (1872) Die Vegetation der Erde nach ihrer kli-
matischen Anordnung [A climatic classification of the Earth’s
vegetation]. W. Engelmann, Leipzig.
Hartley, A.J. (2003) Andean uplift and climate change.
Journal of the Geological Society,160,7–10.
Hastie, T.J. & Tibshirani, R.J. (1990) Generalized additive
models. Chapman and Hall, London.
Hasumi, H. & Emori, S. (2004) K-1 coupled GCM (MIROC)
description. Center for Climate System Research, University
of Tokyo, Tokyo.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G. & Jarvis,
A. (2005) Very high resolution interpolated climate sur-
faces for global land areas. International Journal of Clima-
tology,25, 1965–1978.
Hijmans, R.J., Phillips, S., Leathwick, J. & Elith, J. (2013)
dismo: species distribution modeling. R package version 0.9-
3. Available at: http://cran.r-project.org/package=dismo.
Hopper, S.D. (2009) OCBIL theory: towards an integrated
understanding of the evolution, ecology and conservation
of biodiversity on old, climatically buffered, infertile land-
scapes. Plant and Soil,322,49–86.
Hopper, S.D. & Gioia, P. (2004) The Southwest Australian
Floristic Region: evolution and conservation of a global
hotspot of biodiversity. Annual Review of Ecology, Evolu-
tion, and Systematics,35, 623–650.
Hopper, S.D., Smith, R.J., Fay, M.F., Manning, J.C. & Chase,
M.W. (2009) Molecular phylogenetics of Haemodoraceae
in the Greater Cape and Southwest Australian Floristic
Regions. Molecular Phylogenetics and Evolution,51,19–30.
Hughes, C. & Eastwood, R. (2006) Island radiation on a con-
tinental scale: exceptional rates of plant diversification
after uplift of the Andes. Proceedings of the National Acad-
emy of Sciences USA,103, 10334–10339.
Iribarren, L., Verg
es, J. & Fern
andez, M. (2009) Sediment
supply from the Betic-Rif orogen to basins through Neo-
gene. Tectonophysics,475,68–84.
Jakica, S., Quigley, M.C., Sandiford, M., Clark, D., Fifield,
L.K. & Alimanovic, A. (2011) Geomorphic and cosmo-
genic nuclide constraints on escarpment evolution in an
intraplate setting, Darling Escarpment, Western Australia.
Earth Surface Processes and Landforms,36, 449–459.
Jansson, R. & Dynesius, M. (2002) The fate of clades in a
world of recurrent climatic change: Milankovitch oscilla-
tions and evolution. Annual Review of Ecology and System-
atics,33, 741–777.
Kahle, H.-G., Straub, C., Reilinger, R., McClusky, S., King,
R., Hurst, T., Veis, G., Kastens, K. & Cross, P. (1998) The
strain rate field in the eastern Mediterranean region, esti-
mated by repeated GPS measurements. Tectonophysics,
294, 237–252.
Keeley, J.E., Bond, W.J., Bradstock, R.A., Pausas, J.G. & Run-
del, P.W. (2012) Fire in Mediterranean ecosystems: ecology,
evolution and management. Cambridge University Press,
Cambridge, UK.
Kellogg, K.S. & Minor, S.A. (2005) Pliocene transpressional
modification of depositional basins by convergent
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
562
R. M. Cowling et al.
thrusting adjacent to the “Big Bend” of the San Andreas
fault: an example from Lockwood Valley, southern Califor-
nia. Tectonics,24, TC1004.
Kendrick, G.W., Wyrwoll, K.-H. & Szabo, B.J. (1991) Plio-
cene-Pleistocene coastal events and history along the
western margin of Australia. Quaternary Science Reviews,
10, 419–439.
Kraft, N.J.B., Baldwin, B.G. & Ackerly, D.D. (2010) Range
size, taxon age and hotspots of neoendemism in the Cali-
fornia flora. Diversity and Distributions,16, 403–413.
Kreft, H. & Jetz, W. (2007) Global patterns and determinants
of vascular plant diversity. Proceedings of the National
Academy of Sciences USA,104, 5925–5930.
Lambers, H., Brundrett, M.C., Raven, J.A. & Hopper, S.D.
(2010) Plant mineral nutrition in ancient landscapes: high
plant species diversity on infertile soils is linked to func-
tional diversity for nutritional strategies. Plant and Soil,
334,11–31.
Lancaster, L.T. & Kay, K.M. (2013) Origin and diversification
of the California flora: re-examining classic hypotheses
with molecular phylogenies. Evolution,67, 1041–1054.
Linder, H.P. (2005) Evolution of diversity: the Cape flora.
Trends in Plant Sciences,10, 536–541.
Linder, H.P. (2008) Plant species radiations: where, when,
why? Philosophical Transactions of the Royal Society B: Bio-
logical Sciences,363, 3097–3105.
Madri~
n
an, S., Cort
es, A.J. & Richardson, J.E. (2013) P
aramo
is the world’s fastest evolving and coolest biodiversity hot-
spot. Frontiers in Genetics,4, 192.
Marean, C.W., Cawthra, H.C., Cowling, R.M., Esler, K.J.,
Fisher, E., Milewski, A., Potts, A.J., Singels, E. & deVynck,
J. (2014) Stone age people in a changing South African
Greater Cape Floristic Region. Fynbos: ecology, evolution
and conservation of a megadiverse region (ed. by N. All-
sopp, J.F. Colville and G.A. Verboom). Oxford University
Press, Oxford (in press).
Mart
ın, J.M., Braga, J.C., Aguirre, A. & Puga-Bernab
eu,
A.
(2009) History and evolution of the North-Betic Strait
(Prebetic Zone, Betic Cordillera): a narrow, early Torto-
nian, tidal-dominated, Atlantic–Mediterranean marine pas-
sage. Sedimentary Geology,216,80–90.
Michard, A., Chalouan, A., Feinberg, H., Goff
e, B. & Monti-
gny, R. (2002) How does the Alpine belt end between
Spain and Morocco. Bulletin de la Soci
et
eG
eologique de
France,173,3–15.
Migliore, J., Baumel, A., Juin, M. & M
edail, F. (2012) From
Mediterranean shores to central Saharan mountains: key
phylogeographical insights from the genus Myrtus.Journal
of Biogeography,39, 942–956.
Mittelbach, G.G., Schemske, D.W., Cornell, H.V. et al.
(2007) Evolution and the latitudinal diversity gradient:
speciation, extinction and biogeography. Ecology Letters,
10, 315–331.
Mix, H.T., Mulch, A., Kent-Corson, M.L. & Chamberlain,
C.P. (2011) Cenozoic migration of topography in the
North American cordillera. Geology,39,87–90.
Montgomery, D.R. (1993) Compressional uplift in the cen-
tral California Coast Ranges. Geology,21, 543–546.
Mooney, H.A. (1977) Convergent evolution in Chile and Cali-
fornia. Mediterranean climate ecosystems. Dowden, Hutch-
inson & Ross, Stroudsburg, PA.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca,
G.A.B. & Kent, J. (2000) Biodiversity hotspots for conser-
vation priorities. Nature,403, 853–858.
Nevill, P.G., Bradbury, D., Williams, A., Tomlinson, S. &
Krauss, S.L. (2014) Genetic and palaeo-climatic evidence
for widespread persistence of the coastal tree species Euca-
lyptus gomphocephala (Myrtaceae) during the Last Glacial
Maximum. Annals of Botany,113,55–67.
Ojeda, F., Mara~
n
on, T. & Arroyo, J. (2000) Plant diversity
patterns in the Aljibe Mountains (S. Spain): a comprehen-
sive account. Biodiversity and Conservation,9, 1323–1343.
Ojeda, F., Simmons, M.T., Arroyo, J., Mara~
n
on, T. & Cowl-
ing, R.M. (2001) Biodiversity in South African fynbos and
Mediterranean heathland. Journal of Vegetation Science,12,
867–874.
Papanikolaou, D.J. & Royden, L.H. (2007) Disruption of the
Hellenic arc: Late Miocene extensional detachment faults
and steep Pliocene-Quaternary normal faults –or what
happened at Corinth? Tectonics,26, TC5003.
Partridge, T.C. & Maud, R.R. (1987) Geomorphic evolution
of southern Africa since the Mesozoic. South African Jour-
nal of Geology,90, 179–208.
Pearce, F.D., Rondenay, S., Sachpazi, M., Charalampakis, M.
& Royden, L.H. (2012) Seismic investigation of the transi-
tion from continental to oceanic subduction along the
western Hellenic Subduction Zone. Journal of Geophysical
Research: Solid Earth,117, B07306.
Phillips, S.J., Anderson, R.P. & Schapire, R.E. (2006) Maxi-
mum entropy modeling of species geographic distribu-
tions. Ecological Modelling,190, 231–259.
Pillans, B. (2007) Pre-Quaternary landscape inheritance in
Australia. Journal of Quaternary Science,22, 439–447.
Qian, H. & Ricklefs, R.E. (2000) Large-scale processes and
the Asian bias in species diversity of temperate plants.
Nature,407, 180–182.
R Core Team (2013) R: a language and environment for sta-
tistical computing. R Foundation for Statistical Computing,
Vienna, Austria.
Reilinger, R., McClusky, S., Paradissis, D., Ergintav, S. &
Vernant, P. (2010) Geodetic constraints on the tectonic
evolution of the Aegean region and strain accumulation
along the Hellenic subduction zone. Tectonphysics,448,
22–30.
Ricklefs, R.E. (2004) A comprehensive framework for global
patterns in biodiversity. Ecology Letters,7,1–15.
Rodr
ıguez-S
anchez, F. & Arroyo, J. (2008) Reconstructing
the demise of Tethyan plants: climate-driven range
dynamics of Laurus since the Pliocene. Global Ecology and
Biogeography,17, 685–695.
Rosenzweig, M.L. (1995) Species diversity in space and time.
Cambridge University Press, Cambridge, UK.
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
563
Plant diversity and stability in mediterranean-climate ecosystems
Royden, L.H. & Husson, L. (2006) Trench motion, slab
geometry and viscous stresses in subduction systems. Geo-
physical Journal International,167, 881–905.
Sauquet, H., Weston, P.H., Anderson, C.L., Barker, N.P.,
Cantrill, D.J., Mast, A.R. & Savolainen, V. (2009)
Contrasted patterns of hyperdiversification in Mediterra-
nean hotspots. Proceedings of the National Academy of Sci-
ences USA,106, 221–225.
Scharf, T.E., Codilean, A.T., de Wit, M., Jansen, J.D. &
Kubik, P.W. (2013) Strong rocks sustain ancient postoro-
genic topography in southern Africa. Geology,41, 331–334.
Schnitzler, J., Barraclough, T.G., Boatwright, J.S., Goldblatt,
P., Manning, J.C., Powell, M.P., Rebelo, T. & Savolainen,
V. (2011) Causes of plant diversification in the Cape bio-
diversity hotspot of South Africa. Systematic Biology,60,
343–357.
Slingsby, J.A., Britton, M.N. & Verboom, G.A. (2014) Ecol-
ogy limits the diversity of the Cape flora: phylogenetics
and diversification of the genus Tetraria.Molecular Phylog-
enetics and Evolution,72,61–70.
Small, E.E. & Anderson, R.S. (1995) Geomorphically driven
late Cenozoic rock uplift in the Sierra Nevada, California.
Science,270, 277–281.
Sniderman, J.M.K., Jordan, G.J. & Cowling, R.M. (2013) Fos-
sil evidence for a hyperdiverse sclerophyll flora under a
non-Mediterranean-type climate. Proceedings of the
National Academy of Sciences USA,110, 3423–3428.
Specht, R.L. & Moll, E.J. (1983) Mediterranean-type heath-
lands and sclerophyllous shrublands of the world: an over-
view. Mediterranean-type ecosystems: the role of nutrients
(ed. by F.J. Kruger, D.T. Mitchell and J.U.M. Jarvis), pp.
41–65. Springer, Berlin.
Tinker, J., de Wit, M. & Brown, R. (2008) Mesozoic exhu-
mation of the southern Cape, South Africa, quantified
using apatite fission track thermochronology. Tectonophys-
ics,455,77–93.
Valente, L.M. & Vargas, P. (2013) Contrasting evolutionary
hypotheses between two mediterranean-climate floristic
hotpots: the Cape of southern Africa and the Mediterra-
nean Basin. Journal of Biogeography,40, 2032–2046.
Valente, L.M., Savolainen, V., Manning, J.C., Goldblatt, P. &
Vargas, P. (2011) Explaining disparities in species richness
between Mediterranean floristic regions: a case study in
Gladiolus (Iridaceae). Global Ecology and Biogeography,20,
881–892.
Vassilakis, E., Royden, L. & Papanikolaou, D. (2011) Kine-
matic links between subduction along the Hellenic trench
and extension in the Gulf of Corinth, Greece: a multidisci-
plinary analysis. Earth and Planetary Science Letters,303,
108–120.
Verboom, G.A., Archibald, J.K., Bakker, F.T., Bellstedt, D.U.,
Conrad, F., Dreyer, L.L., Forest, F., Galley, C., Goldblatt,
P., Henning, J.F., Mummenhoff, K., Linder, H.P., Muasya,
A.M., Oberlander, K.C., Savolainen, V., Snijman, D.A., van
der Niet, T. & Nowell, T.L. (2009) Origin and diversifica-
tion of the Greater Cape flora: ancient species repository,
hot-bed of recent radiation, or both? Molecular Phylogenet-
ics and Evolution,51,44–53.
Werneck, F.P., Costa, G.C., Colli, G.R., Prado, D.E. & Sites,
J.W., Jr (2011) Revisiting the historical distribution of Sea-
sonally Dry Tropical Forests: new insights based on palae-
odistribution modelling and palynological evidence. Global
Ecology and Biogeography,20, 272–288.
Werneck, F.P., Nogueira, C., Colli, G.R., Sites, J.W., Jr &
Costa, G.C. (2012) Climatic stability in the Brazilian Cer-
rado: implications for biogeographical connections of
South American savannas, species richness and conserva-
tion in a biodiversity hotspot. Journal of Biogeography,39,
1695–1706.
Wisheu, I.C., Rosenzweig, M.L., Olsvig-Whittaker, L. &
Shmida, A. (2000) What makes nutrient-poor mediterra-
nean heathlands so rich in plant diversity? Evolutionary
Ecology Research,2, 935–955.
Wolfe, J.A., Forest, C.E. & Molnar, P. (1998) Paleobotanical
evidence of Eocene and Oligocene paleoaltitudes in midlati-
tude western North America. GSA Bulletin,110, 664–678.
Wood, S. (2014) mgcv: mixed GAM computation vehicle with
GCV/AIC/REML smoothness estimation. R package version
1.7-29. Available at: http://cran.r-project.org/pack-
age=mgcv.
Zachos, J., Pagani, M., Sloan, L., Thomas, E. & Billups, K.
(2001) Trends, rhythms, and aberrations in global climate
65 Ma to present. Science,292, 686–693.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Appendix S1 Vegetation layers used to build the Maxent
and GAM distribution models of mediterranean climate
ecosystems.
Appendix S2 Data sources for plant species richness of
Greek islands.
Appendix S3 The stability of the mediterranean climate
between current and Last Glacial Maximum conditions.
BIOSKETCH
Richard Cowling has a keen interest in the comparative
ecology and evolution of the world’s mediterranean-climate
ecosystems. He has collaborated on this and other topics
with many of the coauthors of this paper.
Author contributions: R.M.C., P.B., J.C., S.F., F.F., S.D.H.,
S.P. and P.W.R. conceived the project, contributed data and
edited the manuscript; R.M.C. led the writing; M.A., N.K.M.,
F.O., R.J.S. and E.V. and B.R.Z. contributed data and edited
the manuscript; R.M.C and A.J.P. performed the analyses.
Editor: Melodie McGeoch
Journal of Biogeography 42, 552–564
ª2014 John Wiley & Sons Ltd
564
R. M. Cowling et al.
