Content uploaded by Daniel Peppe
Author content
All content in this area was uploaded by Daniel Peppe on Nov 21, 2018
Content may be subject to copyright.
Content uploaded by Daniel Peppe
Author content
All content in this area was uploaded by Daniel Peppe on Nov 21, 2018
Content may be subject to copyright.
Chapter 13
Reconstructing Paleoclimate and Paleoecology Using Fossil
Leaves
Daniel J. Peppe, Aly Baumgartner, Andrew Flynn, and Benjamin Blonder
Abstract Plants are strongly influenced by their surround-
ing environment, which makes them reliable indicators of
climate and ecology. The relationship between climate,
ecology, plant traits and the geographic distribution of plants
based on their climatic tolerances have been used to develop
plant-based proxies for reconstructing paleoclimate and
paleoecology. These proxies are some of the most accurate
and precise methods for reconstructing the climate and
ecology of ancient terrestrial ecosystems and have been
applied from the Cretaceous to the Quaternary. Despite their
utility, the relationships between plant traits and climate that
underlie these methods are confounded by other factors such
as leaf life-span and phylogenetic history. Work focused on
better understanding these confounding factors, incorporat-
ing the influence of phylogeny and leaf economic spectrum
traits into proxies, expanding modern leaf trait-climate and
ecology calibration datasets to additional biogeographic
areas and climate regimes, and developing automated
computer algorithms for measuring leaf traits are important
growing research areas that will help considerably improve
plant-based paleoclimate and paleoecological proxies.
Keywords Paleobotany Leaf physiognomy Leaf margin
analysis Leaf area analysis CLAMP Digital leaf
physiognomy Nearest living relative Leaf economic
spectrum Leaf vein density Leaf mass per area
Introduction
Plants are stationary, primary producers whose productivity
is based on their ability to photosynthesize and use water
efficiently. Terrestrial plants use roots and rhizoids to take in
water and minerals from the soil, which are essential for
photosynthesis and for maintaining the water content of the
plant. Leaves, a plant’s primary photosynthetic organ, are
used for gas and water exchange between the plant and the
atmosphere, and generally have high surface to volume
ratios. Plants must therefore balance water loss and uptake,
light and carbon dioxide fixation, and heat gain and loss
from their environment as these affect their water use effi-
ciency and photosynthetic capacity. Thus for plants to be
successful, they must have biochemical and physiological
adaptions that are well matched to the conditions they
experience in their environment. Given these relationships
between plants and their surrounding environment, a plant’s
traits, such as the size and shape of its leaves, are intrinsi-
cally linked to the climate and environment in which it is
growing.
A variety of methods have been developed over the last
100 years using the relationships between plant traits, cli-
mate, and ecology to develop paleoclimate and paleoeco-
logical proxies that can be applied to fossil leaves. In this
review, we discuss univariate, multivariate, and taxonomic
methods for reconstructing paleoclimate and paleoecology
using fossil plants and their advantages and disadvantages.
We also discuss new developments and future research
directions for leaf-based methods of reconstructing paleo-
climate and paleoecology.
D. J. Peppe (&)A. Baumgartner A. Flynn
Terrestrial Paleoclimatology Research Group, Department
of Geosciences, Baylor University, Waco, TX 76706, USA
e-mail: Daniel_Peppe@baylor.edu
A. Baumgartner
e-mail: Aly_Baumgartner@baylor.edu
A. Flynn
e-mail: Andrew_Flynn@baylor.edu
B. Blonder
School of Life Sciences, Arizona State University,
427 E Tyler Mall, Tempe, AZ 85287, USA
e-mail: bblonder@gmail.com
©Springer International Publishing AG, part of Springer Nature 2018
D. A. Croft et al. (eds.), Methods in Paleoecology: Reconstructing Cenozoic Terrestrial Environments
and Ecological Communities, Vertebrate Paleobiology and Paleoanthropology, https://doi.org/10.1007/978-3-319-94265-0_13
289
Leaf Traits and Climate
The relationships between plant traits, and more specifically
leaf traits, and the climate and environment have been noted
for over 100 years (e.g., Billings 1905; Bailey and Sinnott
1915,1916). In particular, the sizes and shapes (physiog-
nomy) of leaves correlate strongly with temperature and
moisture, and these relationships have been used to develop
proxies for reconstructing paleoclimate (e.g., Bailey and
Sinnott 1915,1916; Webb 1959,1968; Wolfe 1971,1978,
1979,1993,1995; Dilcher 1973; Dolph and Dilcher 1980a,
b; Greenwood 1991,1992; Wing and Greenwood 1993;
Wilf 1997; Wilf et al. 1998; Jacobs 1999; Gregory-Wodzicki
2000; Jacobs 2002; Kowalski and Dilcher 2003; Greenwood
et al. 2004; Traiser et al. 2005; Spicer 2007; Adams et al.
2008; Peppe et al. 2011; Teodoridis et al. 2011; Yang et al.
2011,2015; Spicer 2016). Additionally, the climatic toler-
ances of plants very strongly influence their geographic
distributions (e.g., Von Humboldt and Bonpland 1807;
Merriam 1894;Küchler 1964; Holdridge 1967; Whittaker
1975; Larcher and Winter 1981). Thus, a plant community,
which is comprised of species with overlapping geographic
ranges and environmental tolerances, should broadly reflect
the climate in which it is growing, assuming the plant
community is in equilibrium with climate (e.g., Svenning
and Sandel 2013). The assumption that a plant community is
in equilibrium with climate is probably broadly true; however,
there are notable situations when plant communities can be
out of equilibrium with climate, often referred to as dise-
quilibrium or community disequilibrium, particularly when
climate is rapidly changing (e.g., Davis 1986; Svenning and
Sandel 2013; Blonder et al. 2015a). The relationship between
plant geographic distributions and climate has been used to
develop a variety of qualitative and quantitative “nearest
living relative”approaches, in which fossil taxa are sys-
tematically assigned to a closely related modern plant taxon
(i.e., their “nearest living relative”) and then the fossil spe-
cies assemblage is matched to the climatic conditions of its
“nearest living relatives”(e.g., Heer 1870,1878a,b,1882;
Chaney 1938; Axelrod 1948,1952; MacGinitie 1953;
Axelrod and Bailey 1969; Hickey 1977; Kershaw and Nix
1988; Mosbrugger and Utescher 1997).
Univariate and multivariate leaf physiognomic methods
and nearest living relative approaches have been applied
hundreds, if not thousands, of times to Cretaceous and
Cenozoic angiosperm-dominated fossils floras to reconstruct
paleoclimate (as an example, see supplement in Little et al.
(2010) for a relatively complete list of papers from 1902 to
2010 that have utilized univariate and multivariate leaf
physiognomic methods). These climate reconstructions are
probably what fossil leaves are best known for among the
broader geological and paleoanthropological communities,
and there are a number of other excellent detailed reviews on
using angiosperm leaves to reconstruct climate (e.g., Chal-
oner and Creber 1990; Greenwood 2007; Wilf 2008; Jordan
2011; Royer 2012a).
The basic observations underpinning leaf physiognomic
paleoclimate methods are the relationships between leaf teeth
and mean annual temperature (MAT) and leaf size and mean
annual precipitation (MAP). In particular, the average leaf
size of all species at a site positively correlate with MAP and
the percentage of woody, dicotyledonous (dicot) angiosperm
species at a site with toothed leaves and variables related to
tooth count and size all negatively correlate with MAT
(Figs. 13.1,13.2; e.g., Bailey and Sinnott 1916; Webb 1968;
Dilcher 1973; Wolfe 1979,1993; Dolph and Dilcher 1980a,
b; Givnish 1984; Greenwood 1992; Wilf 1997; Wilf et al.
1998; Jacobs 1999,2002; Huff et al. 2003; Greenwood et al.
2004; Royer et al. 2005; Peppe et al. 2011; Royer 2012b;
Yang et al. 2015). A number of studies have demonstrated
that leaf shape in at least some species of angiosperms
changes in response to climate both within the lifetime of a
single plant (e.g., Royer et al. 2009b; Chitwood et al. 2016a,
b) and over evolutionary time scales (e.g., Schmerler et al.
2012). Further, it appears that distinct molecular pathways
have evolved in plants to mediate changes in leaf shape in
response to changes in photosynthesis, temperature, and light
quality (e.g., Chitwood and Sinha 2016), suggesting that
plants have evolved mechanisms for changing leaf physiog-
nomy in response to climate and environment.
Leaf size typically scales with local water availability
(Fig. 13.2B; e.g., Webb 1968; Givnish 1984; Wilf et al. 1998;
Jacobs 1999; Malhado et al. 2009; Wright et al. 2017). Larger
leaves tend to be warmer because they have a thicker
boundary layer (e.g., Vogel 1970,2009; Parkhurst and
Loucks 1972; Gates 1980; Givnish 1984). Photosynthetic and
transpiration rates both increase with increasing leaf tem-
perature up to a maximum of *35–40°C, at which point
plants begin to experience heat stress and photosynthetic and
transpiration rates rapidly decrease (e.g., Mooney et al. 1978;
Salvuci and Crafts-Bandner 2004; Wahid et al. 2007;
Michaletz et al. 2015). As a result, plants tend to have smaller
leaves in drier climates to reduce evaporative cooling and
larger leaves in wet climates where water loss related to
increased transpiration is less important, as a means of bal-
ancing heat loss via evaporative cooling and water loss via
transpiration (i.e., leaf energy balance) (e.g., Givnish 1984;
Montheith and Unsworth 2007). Smaller leaves usually have
higher densities of major veins, which makes them less vul-
nerable to embolisms caused by hydraulic disruptions that
occur during drought conditions (Scoffoni et al. 2011), which
may also help explain why plants in dry climates have smaller
leaves than in wetter climates.
290 D. J. Peppe et al.
There are many hypotheses about the functional rela-
tionship between leaf teeth in angiosperms and climate. Leaf
teeth are sites of increased transpiration (Canny 1990). Thus
in cold climates, teeth could be an adaptation for increased
carbon uptake via enhanced sap flow early in the growing
season, which would allow rapid early season growth (“early
season gas-exchange hypothesis”; e.g., Billings 1905; Bailey
and Sinnott 1916; Baker-Brosh and Peet 1997; Wing et al.
2000; Royer and Wilf 2006). Leaf teeth also commonly have
hydathodes, which are fixed pores in teeth that probably
allow for water loss via guttation (Feild et al. 2005). The
hydathodes likely also help the plant enhance sap flow,
which allows it to release excess root pressure, preventing
flooding of intercellular spaces in the leaf and, in colder
climates, freeze-thaw embolisms (“root-pressure hypothe-
sis”; Feild et al. 2005). Leaf shape, and in particular dis-
sections of a leaf, are also probably related to bud packing
(“bud-packing hypothesis”; Kobayashi et al. 1998; Couturier
et al. 2011,2012; Edwards et al. 2016). Additionally, at least
some plants exhibit seasonal heteroblasty, in which there are
systematic differences in leaf form associated with position
along a branch (“seasonal heteroblasty hypothesis”). More
specifically, studies have shown that “preformed”leaves,
which are leaves that undergo a period of arrested devel-
opment within a bud, differ in shape from “neoformed”
leaves, which are leaves that develop continuously within a
bud (Critchfield 1960,1971; Edwards et al. 2016).
The early season gas-exchange hypothesis provides a
functional explanation for the relationship between leaf teeth
and MAT because, based on this hypothesis, species with
teeth are able to grow and photosynthesize more rapidly
early in the growing season than untoothed species, which is
a considerable advantage in colder climates with short
growing seasons. The root-pressure hypothesis (e.g., Feild
et al. 2005) is not as clearly related to MAT because it is
related to water movement and freeze-thaw cycles through
the plants, which has at best a weak link to climate. How-
ever, this hypothesis might help provide a mechanism to
explain how plants maintain enhanced sap flow during rapid
leaf expansion early in the growing season. The relationship
between leaf teeth, enhanced sap flow and increased water
loss also helps explain why toothed species are often more
abundant in locally wet environments where the potential
negative implications of water loss through leaf teeth are
mitigated (e.g., Wolfe 1993; Burnham et al. 2001; Kowalski
and Dilcher 2003; Greenwood 2005; Royer et al. 2009a;
Peppe et al. 2011). The seasonal heteroblasty hypothesis is
also not directly linked to MAT. However, it provides a
functional explanation for how temperature may influence
leaf development and may help at least partially explain why
leaf trait variability appears to be more strongly related to
annual climate instead of growing season climate. For
example, in Viburnum, preformed leaves tend to be rounder,
more lobed and toothier than neoformed leaves, which were
more elliptical with greatly reduced teeth (Edwards et al.
2016). This seasonal variation in leaf shape could be due to
two co-occurring forces: first, the different scenarios of early
development (e.g., winter dormancy vs. continual develop-
ment from bud primordia to leaf flush) and second, the
plastic response to different light and temperature regimes
experienced in the early versus late season (Edwards et al.
2016). The bud-packing hypothesis is based on the premise
that there are constraints on leaf shape that are driven by
how the folds and boundaries of leaf primordia are arranged
Fig. 13.1 Examples of modern and fossil leaves of woody dicotyledonous angiosperms with entire (A, C) and toothed (B, D) margins. A, Cleared
and stained leaf of Calophyllujm calabea with an entire margin (NCLC4372). B, Cleared and stained leaf of Tilia heterophylla with a toothed
margin (NCLC5403). C, Unidentified Paleocene fossil leaf from the Fort Union Formation, Williston Basin, North Dakota, USA with an entire
margin (YPM52359). Black arrow indicate location of petiolar attachment. D, Davidia antiqua, Paleocene fossil leaf from the Fort Union
Formation, Powder River Basin, Montana, USA with a toothed margin (DMNH28307). Large black arrow indicates the leaf’s petiole. Small black
arrow in upper left indicates one of the leaf’s teeth
13 Paleoclimate and Paleoecology Using Fossil Leaves 291
Fig. 13.2 Univariate leaf physiognomic models for estimating mean annual temperature (MAT) and mean annual precipitation (MAP). A,
Relationship between the proportion of untoothed species in a flora and MAT for 1488 globally distributed sites. Linear regression fit shown (r
2
=
0.43, F=1113, P< 0.001). Data from Wolfe (1979,1993), Midgley et al. (1995), Wilf (1997), Kennedy, (1998), Jacobs, (1999,2002), Burnham
et al. (2001), Greenwood et al. (2004), Royer et al. (2005), Hinojosa et al. (2006), Aizen and Ezcurra (2008), Su et al. (2010), Peppe et al. (2011),
Chen et al. (2014), Yang et al. (2015). B, Relationship between mean leaf area and MAP for 184 globally distributed sites. Linear regression fit
shown (r
2
= 0.41, F=126, P< 0.001). Data from Wilf et al. (1998), Jacobs (1999,2002), Gregory-Wodzicki (2000), and Peppe et al. (2011)
292 D. J. Peppe et al.
in a bud, which is unlikely to be driven by climate. However,
the bud-packing and seasonal heterblasty hypotheses are
very similar in that both argue that climate influences leaf
development within the bud, which in turn influences leaf
shape, suggesting that climate plays an important role in
influencing leaf shape from the beginning of bud develop-
ment. Although each of these hypotheses help explain some
aspects of the relationship between leaf physiognomy and
climate, we still lack a physiological or evolutionary mecha-
nism that fully explains how and why the empirical rela-
tionships that have been observed between leaf physiognomy
and climate in modern floras exists. Work to better under-
stand the mechanism linking climate and leaf physiognomy
is ongoing (see for example review in Chitwood and Sinha
2016) and help offer insights that will likely significantly
improve leaf physiognomy-climate models in the future.
The occurrence of leaf teeth are likely also influenced by
a variety of factors not directly related to climate, such as
phylogeny, leaf life-span, and leaf thickness. Phylogenetic
history influences the occurrence of leaf teeth (e.g., Little
et al. 2010; Burnham and Tonkovich 2011; Hinojosa et al.
2011). This suggests that the leaf teeth-climate relationship
observed in modern floras may reflect a combination of
plants’responses on both short term (i.e., seasonal) and long
term (i.e., evolutionary) time scales and processes of plant
community assembly (Schmerler et al. 2012). Plant leaf
life-span and leaf thickness also influence the occurrence of
leaf teeth as both deciduous species and species with thinner
leaves more commonly have teeth than evergreen species
and species with thicker leaves (Royer et al. 2012). How-
ever, neither plant life-span (i.e., deciduous vs. evergreen)
nor leaf thickness correlate strongly with MAT (e.g., Reich
et al. 1999; Wright et al. 2004,2005; Royer et al. 2012),
indicating that non-climatic factors also play a role in the
occurrence of leaf teeth. The influence of phylogeny, leaf
life-span, leaf thickness and other non-climatic factors on
leaf physiognomic methods for reconstructing climate is
unclear, but may inherently bias paleoclimate estimates (see
Royer 2012a for more details). Leaf physiognomic methods
that incorporate phylogenetic and leaf ecological information
are in development and aim to address the potential influ-
ence of phylogeny and other non-climatic drivers of leaf
teeth on paleoclimate estimates.
Methods of Paleoclimate
Reconstruction
Methods that reconstruct paleoclimate using fossil leaves
can be broadly divided into two main types: leaf physiog-
nomic methods and nearest living relative (NLR) methods.
Leaf physiognomic methods use the relationship between
leaf traits and climate, while NLR approaches use the cli-
matic, and less commonly, the ecological requirements of
the taxon that is the nearest living relative to the fossil taxon
being studied.
Leaf physiognomic methods –univariate approaches:The
relationship between leaf physiognomy and climate was first
noted a century ago by Bailey and Sinnott (1915,1916)and
this relationship has become the cornerstone of leaf physiog-
nomic paleoclimate methods (e.g., Wolfe 1971,1979; Dilcher
1973; Wing and Greenwood 1993;Wolfe1993;Wilf1997;
Wilf et al. 1998;Jacobs1999; Gregory-Wodzicki 2000;
Greenwood et al. 2004; Spicer 2007,2016; Peppe et al. 2011).
In particular, Bailey and Sinnott (1915,1916) noted that the
proportion of a flora with entire-margined leaves and with
large leaves increased with decreasing latitude and suggested
that these relationships could be used to estimate ancient cli-
mate. Following this pioneering work, paleobotanists used
similarities between the relative abundance of species with
entire margins and the different size classes of leaves in
modern vegetation, as well as Paleogene and Neogene floras to
qualitatively infer paleoclimate (e.g., MacGinitie 1953;Wolfe
and Hopkins 1967;Hickey1977; Christophel and Blackburn
1978). Despite the acknowledgement of the relationship
between leaf physiognomy and climate and its application to
fossil floras and ancient climates, it took more than a half a
century before the qualitative relationships noted by Bailey
and Sinnott (1915,1916) were demonstrated quantitatively.
Using floral lists from East Asia, Wolfe (1979) demon-
strated a remarkably strong relationship between the per-
centage of woody dicot species at a site with entire margins,
defined as species lacking either marginal teeth or lobes, and
MAT. However, although Wolfe (1979) presented a linear
regression in his figure that showed the relationship between
the percent of species with entire margins and MAT, the
statistics and the equation for the regression were not pre-
sented. This relationship was then quantified by Wing and
Greenwood (1993) using linear regression to develop a pre-
dictive equation to estimate MAT based on the percentage of
woody dicot species at a site with entire margined species:
LMAT ¼30:6Pþ1:14 ð1Þ
where LMAT is mean annual temperature estimated using
leaf margin analysis and Pis defined as:
P¼r
nð2Þ
where nis the total number of species sampled, ris the
number of species at a site with entire margins, and (n–r)is
the number of species with toothed margins. In cases where
a species has both entire and non-entire margins, the rscore
for the species is 0.5.
13 Paleoclimate and Paleoecology Using Fossil Leaves 293
Table 13.1 Leaf physiognomic models for climate from published literature
Method Reconstructed
climate variable
Equation Region r
2
nSE Authors
LMA MAT (1) MAT = 0.306E + 1.141 East Asia 0.98 34 0.80 Wolfe (1979), Wing and
Greenwood (1993)
LMA MAT (2) MAT = 0.276E + 1.038 China 0.79 50 1.90 Su et al. (2010)
LMA MAT (3) MAT = 0.223E + 6.68 China 0.53 732 3.10 Chen et al. (2014)
LMA MAT (4) MAT = 0.2338E + 4.6321 Japan 0.7396 35 1.87 Kennedy et al. (2014)
LMA MAT (5) MAT = 0.314E + 0.512 Europe 0.6 1835 1.70 Traiser et al. (2005)
LMA MAT (6) MAT = 0.291E −0.266 North, Central America and
Japan
0.76 106 3.40 Wilf (1997)
LMA MAT (7) MAT = 0.286E + 2.240 North, Central and South
America
0.94 9 2.00 Wilf (1997)
LMA MAT (8) MAT = 0.363E + 2.233 North America 0.8 10 3.60 Kowalski and Dilcher (2003)
LMA MAT (9) MAT = 0.290E + 1.320 North, Central America 0.91 84 –Miller et al. (2006)
LMA MAT (10) MAT = 0.2471E + 2.3305 North America 0.7919 104 2.63 Kennedy et al. (2014)
LMA MAT (11) MAT = 0.250E + 3.42 Temperate Northern Hemisphere 0.87 144 2.20 Wolfe (1993)
LMA MAT (12) MAT = 0.275E + 1.36 Temperate Northern Hemisphere 0.768 173 3.30 Gregory-Wodzicki (2000)
LMA MAT (13) MAT = 0.249E + 3.53 Temperate Northern Hemisphere
(alpine excluded)
0.869 144 2.20 Gregory-Wodzicki (2000)
LMA MAT (14) MAT = 0.2475E + 2.9066 Northern Hemisphere 0.81236 149 2.55 Kennedy et al. (2014)
LMA MAT (15) MAT = 0.759E −0.013 New Zealand, the Andes, the
Alps
0.571 19 –Halloy and Mark (1996)
LMA MAT (16) MAT = 0.316E −0.059 South America 0.89 14 1.60 Gregory-Wodzicki (2000)
LMA MAT (17) MAT = 0.385E −10.24 Tropical South America 0.47 30 3.40 Kowalski, (2002)
LMA MAT (18) MAT = 0.3825E −10.9 South America (isotherm) 0.733 16 –Aizen and Ezcurra (2008)
LMA MAT (19) MAT = 0.4236E −11.37 South America (cell) 0.535 16 –Aizen and Ezcurra (2008)
LMA MAT (20) MAT = 0.2603E + 1.31 South America 0.82 51 2.80 Hinojosa et al. (2011)
LMA MAT (21) MAT = 0.4396E −8.3637 South America 0.9256 14 1.73 Kennedy et al. (2014)
LMA MAT (22) MAT = 0.2399E + 3.6916 Southern Africa 0.8567 25 –Steart et al. (2011)
LMA MAT (23) MAT = 0.1094E + 9.5173 Southern Africa 0.3708 14 2.30 Kennedy et al. (2014)
LMA MAT (24) MAT = 0.270E −2.120 Australia 0.63 74 –Greenwood et al. (2004)
LMA MAT (25) MAT = 0.0862E + 17.6547 Pacific Islands 0.3491 6 1.33 Kennedy et al. (2014)
LMA MAT (26) MAT = 0.0571E + 16.4785 Australia 0.0559 21 2.34 Kennedy et al. (2014)
LMA MAT (27) MAT = 0.0239E + 10.561 New Zealand 0.0174 35 1.69 Kennedy et al. (2014)
LMA MAT (28) MAT = 0.257E −1.8412 Southern Hemisphere 0.40406 90 4.83 Kennedy et al. (2014)
LMA MAT (29) MAT = 0.204E + 4.600 North and South America, Japan,
and Oceania
0.58 92 4.80 Peppe et al. (2011)
LMA MAT (30) MAT = 0.2039E + 3.6562 Oceania, Japan, North America,
South America, Southern Africa
0.5733 239 3.97 Kennedy et al. (2014)
LMA MAT (31) MAT = 0.194E + 5.884 Global 0.42 1488 4.54 This study
(continued)
294 D. J. Peppe et al.
Table 13.1 (continued)
Method Reconstructed
climate variable
Equation Region r
2
nSE Authors
LAA MAT MAT = 0.115A + 1.192 New Zealand, the Andes, the
Alps
0.872 19 –Halloy and Mark (1996)
DiLP MAT MAT = 0.210E + 42.296D −2.609I −16.004 North and South America, Japan,
and Oceania
0.7 92 4.00 Peppe et al. (2011)
MLR MAP (1) MAP = 167.948(apex attenuate) + 377.735
(mesophyll2) + 11.489
Temperate Northern Hemisphere 0.497 74 57.96 Wing and Greenwood (1993)
modified
LAA
MAP (2) MAP = 6.18(mesophyll1 + mesophyll2) + 47.5 Temperate Northern Hemisphere 0.439 74 –Wilf et al. (1998)
LAA MAP (3) lnMAP = 0.484(MlnA) + 1.78 Temperate Northern Hemisphere 0.612 144 0.47 Gregory-Wodzicki (2000)
LAA MAP (4) lnMAP = 0.548(MlnA) + 0.786 North, Central and South
America, Africa
50 50 0.36 Wilf et al. (1998)
LAA MAP (5) lnMAP = 0.298(MlnA) + 2.64 Bolivia 0.52 12 0.22 Gregory-Wodzicki (2000)
LAA MAP (6) lnMAP = 1.369(mesophyll1 + mesophyll2) +
3.982
Africa and Bolivia 0.826 42 0.18 Jacobs (2002)
LAA MAP (7) lnMAP = 0.309(MlnA) + 2.566 Tropical Africa and Bolivia 0.734 42 –Jacobs and Heredeen (2004)
MLR MAP (8) lnMAP = 0.6903(mesophyll) + 0.7059(apex
acuminate) + 0.884(apex acute) + 0.7542E +
4.9415
Africa 0.794 30 0.19 Jacobs (1999)
MLR MAP (9) lnMAP = 1.274(mesophyll1 + mesophyll2) −
1.013(shape elliptic) + 5.198
Africa 0.806 30 0.18 Jacobs (2002)
modified
LAA
MAP (10) lnMAP = 1.321(mesophyll1 + mesophyll2) +
4.018
Africa 0.764 30 0.20 Jacobs (2002)
LAA MAP (11) lnMAP = 0.321(MlnA) + 2.476 Africa 0.661 30 0.24 Jacobs (2002)
LAA MAP (12) lnMAP (<260 cm) = 0.354(MlnA) + 2.167 Tropical Africa, Bolivia, tropical
to subtropical Western
Hemisphere
0.709 79 –Jacobs (2002)
LAA MAP (13) lnMAP = 0.429(MlnA) + 1.705 Western Hemisphere, Bolivia,
Africa
0.713 92 0.34 Jacobs (2002)
LAA MAP (14) MAP = 0.001A + 0.551 New Zealand, the Andes, the
Alps
0.713 19 –Halloy and Mark (1996)
LAA MAP (15) lnMAP = 0.283(MlnA) + 2.92 North and South America, Japan,
and Oceania
0.23 92 0.61 Peppe et al. (2011)
LAA MAP (16) lnMAP = 0.346(MlnA) + 2.404 Global 0.41 184 0.93 This study
DiLP MAP lnMAP = 0.298A(MlnA)−2.717(lnP)+
0.279(lnI)−3.033
North and South America, Japan,
and Oceania
0.27 92 0.60 Peppe et al. (2011)
MLR GSP (1) GSP (<222 cm) = 141.368(mesophyll2) −
136.340(L:W <1) + 93.936(shape elliptic) −
79.774(base round) −52.386(teeth acute) + 48.050
Temperate Northern Hemisphere 0.804 74 16.00 Gregory and McIntosh (1996)
(continued)
13 Paleoclimate and Paleoecology Using Fossil Leaves 295
Table 13.1 (continued)
Method Reconstructed
climate variable
Equation Region r
2
nSE Authors
MLR GSP (2) GSP = −3.393(leptophyll2) +2.40(apex
attenuate) −2.671(base cordate) + 2.360 (L:W 2-3)
+ 3.122 (L:W 3-4:1) + 31.6
Temperate Northern Hemisphere
(alpine excluded)
0.796 144 –Wiemann et al. (1998)
MLR GSP (3) GSP (<222 cm) = 1.60(apex attenuate) + 2.80
(L:W 2-3) −45.2
Temperature Northern
Hemisphere (alpine excluded)
0.63 144 27.00 Gregory-Wodzicki (2000)
MLR Wet months
precipitation
(1) ln Wet Ppn = 0.8368(mesophyll) + 0.819 (apex
acuminate) + 1.1718(apex acute) + 0.8246 (no
teeth) + 4.4993
Africa 0.795 30 0.22 Jacobs (1999)
modified
LAA
Wet months
precipitation
(2) ln Wet Ppn = 1.546(mesophyll1 + mesophyll2)
+ 6.112
Africa 0.779 30 0.22 Jacobs (2002)
modified
LAA
Wet months
precipitation
(3) ln Wet Ppn = 1.601(mesophyll1 + mesophyll2)
+3.777
Africa and Bolivia 0.833 42 0.21 Jacobs (2002)
modified
LAA
Wet months
precipitation
(4) ln Wet Ppn = 0.367(MlnA) + 2.07 Tropical Africa and Bolivia 0.748 42 –Jacobs and Heredeen (2004)
MLR Wet months
precipitation
3-WET = 110.841(apex attenuate) + 320.457 (L:W
2-3:1) + 179.775(mesophyll2) −172.859
Temperate Northern Hemisphere 0.583 74 47.23 Wing and Greenwood (1993)
MLR Dry months
precipitation
3-DRY = 45.54(apex attenuate) + 38.186(L:W
2-3:1) −24.489
Temperature Northern
Hemisphere
0.55 74 8.91 Wing and Greenwood (1993)
LMA = leaf margin analysis, LAA = leaf area analysis, MLR = multiple linear regression model, DiLP = digital leaf physiognomy, MAT = mean annual temperature, MAP = mean annual
precipitation, GSP = growing season precipitation, wet months precipitation = precipitation during the three wettest months of the year, dry months precipitation = precipitation during the three
driest months of the year, E = percent of woody dicots with entire margins, A = leaf area, D = Feret’s diameter ratio, I = number of teeth: internal perimeter, P = perimeter ratio, MlnA= average
leaf area of all dicot species
296 D. J. Peppe et al.
Since Wing and Greenwood’s(1993) recalculation of
Wolfe’s(1979) East Asian leaf margin analysis (LMA)
regression, the relationship between leaf margin and MAT
has been tested on floras from most major geographic
locations and climate biomes on Earth and many global,
regional, and site specific LMA regressions have been
developed (Fig. 13.1; Table 13.1). Significantly, since Wilf
(1997), leaf margin type has been classified as toothed or
entire (i.e., non-toothed). Although many factors confound
the relationship between margin percentage and MAT, as
discussed above and reviewed by others (e.g., Greenwood
2007; Jordan 2011; Peppe et al. 2011; Royer 2012a; Royer
et al. 2012), the relationship between leaf margin and MAT
is typically statistically significant and similar across geo-
graphic regions (Fig. 13.2A). Based on these LMA regres-
sions, LMA has been applied to numerous of Cretaceous and
Cenozoic fossil floras to estimate MAT using fossil leaves
(see for example studies cited in supplementary materials in
Little et al. 2010).
The accuracy and precision of MAT estimates using LMA
has been shown to improve when more species are sampled. In
his original LMA work, Wolfe (1971) suggested 30 species
were necessary to calculate a reliable MAT estimate, though
he suggested that the minimum bar was 20 species, which
has been the practice of most paleobotanists. While the total
number of species sampled is important, the proportional
recovery of the total flora is more significant. For example,
Burnham et al. (2001,2005) found that in very diverse, extant
floras with 55 to >400 species, 25 of the most common
species had to be sampled to make an accurate MAT estimate
using LMA. This suggests that LMA should only be applied to
well-sampled floras and that MAT estimates made using
poorly sampled floras and/or floras with low species richness
are unlikely to be accurate.
Sample size also affects the uncertainty of LMA esti-
mates. Based on their reanalysis of the Wolfe (1979) East
Asian LMA, Wing and Greenwood (1993) calculated the
standard error of LMA estimates of MAT to be ±0.8°C.
However, the standard errors of the East Asian LMA
regression alone considerably underestimate the uncertainty
of leaf margin analysis mean annual temperature (LMAT)
estimates. To account for additional uncertainty in LMAT
estimates, Wilf (1997) proposed using the binomial sam-
pling error for LMA estimates, which accounts for the
standard error of the LMA regression and uncertainty related
to sampling the species richness of a fossil flora:
rLMAT½¼cpP1PðÞ
nð3Þ
where cis the slope of the MAT vs. leaf-margin regression
in the dataset used, Pis the proportion of species with entire
margins, and nis the number of species sampled.
Wilf (1997) suggested the minimum uncertainty of a
LMAT estimate was at least ±2.0°C and that the larger of
the two errors given by the binomial sampling error and the
standard error of the regression if it were >2.0°C was the
appropriate uncertainty for LMAT estimates. In Wilf ’s
(1997) study, he noted that the binomial sampling error only
accounted for part of the total uncertainty in LMAT esti-
mates. Miller et al. (2006) incorporated the uncertainties
associated with the estimation of the proportion of species at
a site with teeth (P) into the estimation of uncertainty of
LMAT estimates:
rLMAT½¼cpf1þun1ðÞP1PðÞðÞ
P1PðÞ
n
ð4Þ
where cis the slope of the MAT vs. leaf-margin regression
in the dataset used, Pis the proportion of species with entire
margins, nis the number of species sampled and uis the
overdispersion factor for P, which was calculated by Miller
et al. (2006) to be 0.052.
The Miller et al. (2006) estimation of uncertainty pro-
vides the best measurement of the precision of LMAT;
however, they noted that there may be additional systematic
uncertainties in LMAT based on sampling differing forested
environments. Greenwood et al. (2004) documented this
effect in a series of sites from Australia that showed a sig-
nificant correlation between margin percentage and MAT (r2
= 0.57, P< 0.001), but a larger uncertainty than datasets
from East Asia and North and Central America (r=±3°C).
Based on a large global compilation of sites, Peppe et al.
(2011) showed that the uncertainty associated with LMAT
estimates using a globally derived LMA is at least ±4°C.
Conservatively, given uncertainties associated with sampling
(accounted for in Eq. 4) and all of the potential confounding
factors that affect the relationship between climate and the
proportion of species with teeth uncertainties, which are
currently not accounted for in any uncertainty calculation, a
minimum uncertainties for LMAT estimates is probably ±5°C
(see also Jordan 1997,2011; Royer 2012a). Despite these
uncertainties, LMAT estimates are typically in agreement
with other proxies and are generally accurate, and reason-
ably precise, making the LMA method a useful tool for
estimating MAT in the past.
Building on another of Bailey and Sinnott’s(1915,1916)
observations that large leaves were more common in warm,
mesic environments, several researchers have noted a strong
positive relationship between leaf size and mean annual
precipitation (MAP) (Webb 1968; Dilcher 1973; Dolph and
Dilcher 1980a,b; Hall and Swaine 1981; Givnish 1984). The
correlation between leaf size and MAP was first quantified
by Givnish (1984) for floras from South America, Central
America and Australia. Wilf et al. (1998) expanded on
Givnish’s(1984) work and used the correlation between the
13 Paleoclimate and Paleoecology Using Fossil Leaves 297
mean of the natural log of species’leaf area (MlnA) and the
natural log of MAP (lnMAP) at 50 sites from North
America, Central America, South America, and Africa to
develop a predictive equation for MAP:
ln MAPðÞ¼0:548MlnA þ0:768 ð5Þ
MlnA was calculated for the entire flora by first deter-
mining the Raunkier-Webb (Webb 1959) size category for
each species and then determining the proportion of species
in each of the seven size categories:
MlnA ¼Xaipið6Þ
where a
i
is the seven means of the natural log areas of the
size categories and p
i
is the proportion of species in each size
category. This method of estimating MAP using leaf area is
known as leaf area analysis (LAA; Table 13.1). The standard
error for ln(MAP) for LAA is 0.359 (Wilf et al. 1998).
Following the work of Wilf et al. (1998), the relationship
between leaf area and MAP has been tested on a series of
floras from Bolivia and predominately Northern Hemisphere
temperate regions (Gregory-Wodzicki 2000), floras from
equatorial Africa (Jacobs 1999,2002), and floras from North
America, Central America, South America, Asia, and
Oceania (Peppe et al. 2011) (Fig. 13.2). These correlations
have been used to develop several additional LAA equations
to estimate MAP (Table 13.1, Fig. 13.2; Jacobs 1999,2002;
Gregory-Wodzicki 2000; Peppe et al. 2011). Using a dataset
of sites from equatorial Africa and Bolivia, Jacobs and
Herendeen (2004) also documented a significant relationship
between leaf area and precipitation during the three wettest
months, which they used to develop a proxy for wet month
precipitation (Table 13.1).
Although the relationship between leaf area and mean
annual precipitation is significant, LAA estimates of MAP
are not accurate and the uncertainties are relatively large.
Peppe et al. (2011) used a jackknife-type approach for a
dataset of 92 globally distributed sites to gauge the accuracy
of LAA, and found that LAA tended to underestimate MAP
at sites with relatively high precipitation and overestimate
MAP at sites with relatively low precipitation. These results
and the original work of Wilf et al. (1998) suggest that LAA
is not a precise estimator of MAP and instead is primarily a
useful tool for roughly estimating MAP.
Leaf size is also influenced by a variety of other factors
including temperature, elevation, soil fertility, and leaf
life-span (e.g., Webb 1968; Dolph and Dilcher 1980a,b;
Givnish 1984; Wolfe 1995; Halloy and Mark 1996;
McDonald et al. 2003; Peppe et al. 2011) confounding the
relationship between MAP and leaf size leading to the low
accuracy and precision of MAP estimates made using LAA.
Additionally, available water appears to be the main control
on leaf size (Givnish 1984). In addition to precipitation,
plant available water is influenced by many factors such as
rates of evapotranspiration, temperature, seasonality, the
position of the water table, and hydrologic conductance
within the soil. As a result, MAP may not be a good proxy
for available water in many situations, which perhaps
explains the relatively poor accuracy and precision of LAA.
Although LMA, and to a lesser extent LAA, have been
widely adopted by the paleobotanical community and the
fossil leaf based MAT and MAP estimates made using the
methods are often considered some of the most robust esti-
mates for terrestrial paleoclimates, both methods rely on
single characters. Leaf size and shape is remarkablely
diverse and many aspects of leaf physiognomy vary with
climate (e.g., Wolfe 1993,1995; Royer et al. 2008,2009b;
Schmerler et al. 2012). Additionally, in at least some cases,
multiple leaf physiognomic traits appear to co-vary with
climate variables (Dolph and Dilcher 1980a,b; Peppe et al.
2011). Further, although LMA and LAA commonly agree
with other climate proxy reconstructions, in many cases,
these proxies provide cooler and drier estimates of MAT and
MAP than other proxies (e.g., Liang et al. 2003; Fricke and
Wing 2004; Wing et al. 2009; Adams et al. 2011).
Because LMA and LAA are univariate approaches, the
incorporation of additional leaf characters may lead to
improvements in paleoclimate estimates.
Leaf physiognomic methods –multivariate approaches:
In an attempt to improve on the shortcomings of univariate
leaf physiognomic methods, Wolfe (1990,1993,1995)
developed a multivariate method called Climate-Leaf Anal-
ysis Multivariate Program (CLAMP), which currently uses
31 categorical leaf physiognomic traits (see Spicer 2016).
In CLAMP the leaf character scores and climate data from a
modern training set are analyzed using canonical corre-
spondence analysis (CCA; Wolfe 1995; Kovach and Spicer
1996). The resultant eigenvectors for the climate and leaf
physiognomic variables are projected onto the major axes of
variation, which roughly correlate with temperature (first
axis) and moisture (second axis) (Wolfe 1995). A sec-
ond-order polynomial fit for a climate variable to the
eigenvector is then used to develop a predictive equation for
that climate variable, which can then be applied to a fossil
flora (Wolfe 1995). In practical terms, this means that the
leaf character scores of a fossil flora are projected onto each
of the climate variable vectors in the CLAMP calibration
dataset and those scores are used to estimate each climate
variable. CLAMP estimates MAT, warm month mean tem-
perature, cold month mean temperatures, growing season
length, growing season precipitation, mean monthly growing
season precipitation, precipitation in the three wettest and
three driest months, relative humidity, specific humidity, and
enthalpy (Wolfe 1995; Spicer 2016). The standard errors for
CLAMP climate estimates are small (Wolfe 1993,1995;
298 D. J. Peppe et al.
Spicer et al. 2005); however, the errors are only based on the
analytical uncertainty of CCA and do not account for any
additional uncertainties related to the leaf character states,
character scoring, precision in the climate variables of the
calibration data set, or factors that potentially confound leaf
character-climate relationships. Thus, the standard errors
reported for CLAMP considerably underestimate uncertainty
and are artificially precise. For example, at minimum, the
uncertainties in MAT estimates made using CLAMP are
similar to, if not larger than those for leaf margin analysis
(see Fig. 13.1 in Royer 2012a).
CLAMP was originally calibrated with 106 sites with a
limited global coverage, and since has been expanded to two
primary calibration datasets primarily derived from floras
from the Northern Hemisphere: (1) PHYSG3BR, which
includes 144 temperature Northern Hemisphere sites, and
(2) PHYSG3AR, which includes 173 Northern Hemisphere
sites that include sites that experience very cold winter
temperatures (e.g., Wolfe 1993,1995; Wolfe and Spicer
1999; Spicer et al. 2004; Spicer 2016). Since then Spicer and
colleagues (see Spicer 2007,2016) have assessed regional
differences in the relationship between climate and CLAMP
characters in South America (Gregory-Wodzicki 2000;
Kowalski 2002), South Africa (Steart et al. 2011), New
Zealand and Australia (Kennedy et al. 2002,2014) and
expanded the calibration datasets to include floras in regions
that experience a monsoon climate (Jacques et al. 2011;
Khan et al. 2014), floras from the Southern Hemisphere
(Kennedy et al. 2014), and floras from temperate and high
latitudes in the Northern and Southern Hemispheres (Yang
et al. 2015). The CLAMP modern calibration originally used
climate data from local climate stations (Wolfe 1993,1995),
but more recent studies follow Spicer et al. (2009) and use
climate data extracted from the New et al. (2002) global,
gridded climate model. CLAMP analyses were originally
performed using a series of spreadsheets and conducting a
CCA using the software package CANACO 4.5 (Micro-
computer Power, Ithaca, NY, USA). CLAMP analyses can
now be performed online (Yang et al. 2011) through the
CLAMP website: http://clamp.ibcas.ac.cn/CLAMP_Run_
Analysis.html.
Although CLAMP has been widely applied (reviewed in
Spicer 2007,2016), the method has several shortcomings
that have been reviewed many times (e.g., Jordan 1997,
2011; Wilf 1997; Wiemann et al. 1998; Wilf et al. 1998;
Green 2006; Peppe et al. 2010). The most notable criticisms
are that (1) based on the character distributions, CCA may
not be the most appropriate technique for predicting climate
(Green 2006; Peppe et al. 2010), (2) at least some of the
leaves in the original calibration dataset were scored incor-
rectly (Peppe et al. 2010), (3) many of the CLAMP character
states are at least partially ambiguous and the original
scoring procedures were poorly defined (Wilf 1997;
Wiemann et al. 1998; Wilf et al. 1998,1999; Green 2006),
(4) the CLAMP characters are discrete and categorical
despite the fact that leaf traits are continuous (i.e., number of
teeth, leaf size) and many of the continuous traits correlate
strongly with temperature and/or precipitation (Huff et al.
2003; Royer et al. 2005,2008; Peppe et al. 2011), and
(5) some CLAMP character states only correlate weakly to
climate and/or have no physiological explanations for cli-
mate causality.
One of the underlying assumptions of CCA is that the
variables are unimodally distributed (after Braak 1986).
However, Green (2006) and Peppe et al. (2010) showed that
many of the CLAMP leaf character distributions were not
unimodally distributed. Following this work, Teodoridis
et al. (2011) developed transformations of the CLAMP
PHYSG3BR and PHYSG3AR calibration datasets that could
be used as “correction coefficients”to account for the
non-normality of the CLAMP leaf character state data.
Peppe et al. (2010) rescored a subset of Wolfe’s(1993,
1995) original calibration sites and demonstrated that the
site-mean leaf area of 38 sites in the CLAMP calibration and
the leaf area scores of the species in each site were consid-
erably underestimated relative to digitally measured leaf
area. When the influence of this underestimation of leaf area
was accounted for, it added a considerable amount of
uncertainty to all CLAMP climate estimates (Peppe et al.
2010). Spicer and Yang (2010) assessed the influence of leaf
size on MAT and enthalpy estimates made using CLAMP
and found that when the leaf size character state was
removed it increased uncertainty, but caused only relatively
small changes in MAT and enthalpy. This result indicates
that leaf size contributes relatively little information when
predicting at least some climate variables with CLAMP and
suggests that some climate predictions could be improved by
changing the leaf character states used in CLAMP. Spicer
and Yang (2010) only indirectly addressed the main criti-
cism of Peppe et al. (2010), that the leaf area scores in the
CLAMP calibration dataset are underestimated, by sug-
gesting that the measured leaf area bias was the result of
measuring leaf area directly rather than using the CLAMP
leaf area template. This is an important distinction because it
means that when performing a CLAMP analysis, leaf area
can only be scored using the CLAMP leaf area template and
if leaf area is measured directly, even if it is more accurate, it
will add considerable uncertainty to the CLAMP estimates.
Several studies have documented issues with the ambi-
guity of CLAMP character states and original scoring pro-
cedures (Wilf 1997; Wiemann et al. 1998; Wilf et al. 1998,
1999; Green 2006; Peppe et al. 2010). Recently, Spicer
(2016) updated the character state definitions and scoring
protocol in an attempt to eliminate the ambiguities. To test
these revisions, Spicer and Yang (2010) had students collect
and score leaves from a forest reserve near Beijing, China
13 Paleoclimate and Paleoecology Using Fossil Leaves 299
and found that the resulting CLAMP climate estimates were
different, though reasonably similar, suggesting the revised
CLAMP scoring protocol and character definitions have
helped eliminated some of the problems with the original
method.
Although there has been recent work to expand the
CLAMP calibration datasets to a wider range of biogeo-
graphic regions and climate regimes (e.g., Jacques et al.
2011; Kennedy et al. 2014; Yang et al. 2015), the inherent
issues with using categorical variables and leaf characters
that are not physiologically linked to climate have not been
addressed. Furthermore, many studies have demonstrated
that climate estimates made using CLAMP are similar to, or
less accurate than, estimates made using LMA and LAA
(e.g., Wilf 1997; Wiemann et al. 1998; Gregory-Wodzicki
2000; Kowalski 2002; Kowalski and Dilcher 2003; Hinojosa
et al. 2006a,b; Dilcher et al. 2009; Royer 2012a; Smith et al.
2012; West et al. 2015).
Rather than apply CCA, several researchers have used the
CLAMP calibration datasets or comparable datasets using
the CLAMP characters to develop multiple linear regression
models for paleoclimate (e.g., Wing and Greenwood 1993;
Gregory 1994; Gregory and McIntosh 1996; Jacobs and
Deino 1996; Stranks and England 1997; Wiemann et al.
1998; Gregory-Wodzicki 2000; Kowalski and Dilcher 2003;
Lielke et al. 2012). However, these methods have not been
widely adopted. Further, because they utilize the CLAMP
calibration dataset and leaf character states, they suffer from
many of the same problems as CLAMP.
Given the potential problems with CLAMP, and in par-
ticular the subjectivity of scoring leaf traits, Huff et al.
(2003) proposed digital leaf physiognomy (DiLP), which is a
method for measuring leaf traits digitally. The benefits of
DiLP are that it increases the reproducibility of leaf character
scores and that it measures continuous variables, such as
number of teeth and tooth size, that have functional or
physiological links to climate. In their preliminary study of
three floras, Huff et al. (2003) found that continuous vari-
ables related to tooth number and tooth shape varied
between floras and with temperature, demonstrating that the
incorporation of continuous variables had the potential to
improve fossil leaf based paleoclimate models. A subsequent
study by Royer et al. (2005) showed that leaves in cold
climates are more highly dissected and have more, larger
teeth. Using data from 17 sites from the east coast of the
United States and Panama, Royer et al. (2005) developed a
preliminary multiple linear regression model for predicting
MAT that was more accurate and precise than LMA and
CLAMP. However, the study was based on sites from a
relatively limited range in biogeography and precipitation.
Peppe et al. (2011) quantified DiLP leaf physiognomic
variables and climate using 92 globally distributed and cli-
matically diverse sites and noted the same patterns between
leaf traits and MAT as observed by Huff et al. (2003) and
Royer et al. (2005) and also that leaves in wet climates were
larger and had fewer, smaller teeth. Using these relation-
ships, Peppe et al. (2011) developed multiple linear regres-
sion models for MAT and MAP (Table 13.1):
MAT ¼0:210 EðÞþ42:296 FDRðÞþ2:609 T:IPðÞ
16:004 ð7Þ
ln MAPðÞ¼0:298 ln AðÞ2:71 ln PRðÞþ0:279 ln T:IPðÞþ3:033
ð8Þ
where Eis the proportion of species at the site with entire
margins, FDR is feret’s diameter ratio (diameter of a circle with
the same area as the blade area (cm)/longest measurable line
across blade area (cm)), T:IP is the number of teeth to internal
perimeter (number of teeth/perimeter of the leaf after teeth are
removed (cm)), Ais leaf area, and PR is perimeter ratio
(perimeter of the leaf blade/internal perimeter (cm)/perimeter of
the leaf after teeth are removed (cm)). The leaf character states
used in these models are defined in Peppe et al. (2011). The
standard error for the MAT model is ±4.0°C and for the MAP
model is 0.60.
DiLP estimates of MAT and MAP are significantly more
accurate than MAT estimates made using LMA and mar-
ginally better than MAP estimates using LAA (Peppe et al.
2011). DiLP MAT estimates made for several fossil floras
are also in better agreement with other proxy reconstructions
of MAT than LMAT estimates (Peppe et al. 2011). This
suggests that the DiLP method potentially offers improve-
ments over other existing leaf physiognomic methods. DiLP
also offers three additional advantages over CLAMP, LMA,
and LAA: (1) it minimizes ambiguity in leaf character
scoring because computer algorithms process most of the
measurements; (2) the DiLP models use mostly continuous
variables, such as feret’s diameter ratio and number of teeth:
internal perimeter; and (3) the leaf characters used in the
DiLP models have a functional connection to climate and
include traits that display phenotypic plasticity between
growing seasons (Royer et al. 2009b) indicating the traits
can respond quickly to climate change.
Despite its advantages, a major limitation of DiLP is that
some of the characters used in the DiLP MAT and MAP
models (feret’s diameter ratio and leaf area) require rela-
tively complete preservation of leaves. Unfortunately, fossil
leaves are typically fragmented (e.g., Greenwood 1992,
2005,2007; Spicer et al. 2005), which means that it may not
be possible to score all DiLP characters for all fossil taxa. In
a sensitivity experiment using modern leaves, Royer et al.
(2005) tested the effect of fragmentary leaves on LMA and
DiLP and found that in all cases fragmentary leaves led to
less precise climate estimates, but when leaves were 50%
intact, LMA estimates of MAT had larger errors than
300 D. J. Peppe et al.
estimates made with DiLP. Additionally, Royer et al. (2005)
found that while multiple specimens should be scored when
possible, only one specimen per species needed to be digi-
tally scored to provide an accurate MAT estimate using
DiLP. These results suggest that although there is some
information lost when leaves are fragmentary, DiLP may
still be able to provide reasonable estimates of paleoclimate
when fragmentary leaves are scored. In an attempt to address
the fact that fossil leaves are commonly incomplete, Peppe
et al. (2011) developed a detailed protocol for processing
incomplete leaves. In practice DiLP characters can be scored
in most fragmentary leaves, provided that at least 25% of the
leaf is preserved and the fragment preserves both the margin
and the mid-vein. Additionally, recent work documented a
strong relationship between leaf vein density and leaf size
(e.g., Sack et al. 2012), and was used to develop a “vein
scaling”transfer function for estimating leaf area from
fragmentary fossils (Merhofer et al. 2015). The incorpora-
tion of leaf area estimates made using vein scaling also has
the potential to increase the number of fragmentary fossil
specimens that can be used to reconstruct MAT and MAP.
DiLP has been successfully applied to several fossil floras
(Peppe et al. 2011; Flynn et al. 2016), which suggests that
the method has great potential to be more widely applied.
However, there are some examples in which the climate
estimates made using DiLP were unreasonable based on
other lines of evidence (Peppe et al. 2011). For example,
estimates for the Laguna del Hunco flora from Argentina,
were cooler and drier than expected (Peppe et al. 2011),
based on the habitat preference of some of the species pre-
sent in the flora (Wilf et al. 2009). This discrepancy may be
because the physiognomic-climate space of South America
is not fully characterized in the current DiLP calibration.
There are also some floras to which DiLP cannot be
applied. Peppe et al. (2011) recommend using CCA as an
initial check to confirm that a fossil floras physiognomic
space is captured by the calibration data. Peppe et al. (2011)
and Peppe (unpublished data) have analyzed two floras that
plotted outside the calibrated physiognomic space: Bonanza,
an Eocene flora from North America, and Rusinga R3, a
Miocene flora from Kenya (Michel et al. 2014; Fig. 13.3).
Peppe et al. (2011) argued that the Bonanza flora was likely
on outlier because the fossil flora was a mixture of two
habitats resulting in a unique site mean physiognomic space.
There are no modern floras from Africa in the current DiLP
calibration, and thus the Rusinga R3 flora from Kenya likely
samples a currently uncalibrated physiognomic space
(Fig. 13.3). Future work expanding the calibration dataset to
more sites from the Southern Hemisphere, and particularly
from Africa, may help improve DiLP so that the method can
be applied to a wider range of floras.
Shortcomings and limitations of physiognomic methods:
Although leaf physiognomic methods of reconstructing
paleoclimate have been widely applied and generate reason-
ably accurate and precise estimates, they are not without
their inherent disadvantages. There are a variety of factors
that affect and confound the relationships between leaf
physiognomy and climate. In this review, we will briefly
discuss three of the important factors that limit leaf phys-
iognomic methods: life history, phylogeny, and plant ecol-
ogy. For a review of some of the other confounding factors,
such as taphonomy, local water availability, position in the
canopy, soil fertility, and nutrient availability, see the recent
reviews by Greenwood (2007), Jordan (2011), and Royer
(2012a).
The original leaf margin analysis calibration data and
most subsequent calibrations, as well as most of the cali-
bration data that underlies CLAMP and DiLP, are primarily
derived from Northern Hemisphere temperate floras (e.g.,
Wolfe 1979,1993,1995; Wilf 1997; Kowalski and Dilcher
2003; Su et al. 2010; Peppe et al. 2011). However, the
relationship between leaf physiognomy and climate, and
particularly between the presence of leaf teeth and MAT, is
distinctly different in the Southern Hemisphere where floras
typically have a higher percentage of untoothed species than
their Northern Hemisphere counterparts (e.g., Wolfe 1979;
Fig. 13.3 Canonical correspondence analysis plot of the digital leaf
physiognomy (DiLP) calibration dataset and fossil sites based on leaf
physiognomy from Peppe et al. (2011) with modern biomes mapped
onto plot. Extant sites indicated by circles and fossil sites indicated by
squares. Site from the Northern Hemisphere are indicated by white
symbols and sites from the Southern Hemisphere by gray symbols. The
Rusinga R3 fossil site (Michel et al. 2014), which is indicated by a red
square, falls well outside the leaf physiognomic space of the extant
floras in the current DiLP calibration demonstrating a limitation of the
applicability of the method
13 Paleoclimate and Paleoecology Using Fossil Leaves 301
Upchurch and Wolfe 1987; Gregory-Wodzicki 2000;
Kowalski 2002; Greenwood et al. 2004; Hinojosa et al.
2006a; Aizen and Ezcurra 2008; Hinojosa et al. 2011; Steart
et al. 2011; Kennedy et al. 2014). Further, in Africa, Aus-
tralia, and New Zealand the correlation between leaf size and
MAT is as strong, if not stronger, than between leaf margin
and MAT (Webb 1968; Greenwood 1992,1994; Jacobs
1999,2002).
These hemispheric differences could be due to regional
differences in environment, such as thermal seasonality or
soil fertility, phylogenetic differences, leaf life-span (i.e.,
lack of temperate deciduous forests in the Southern Hemi-
sphere), or differences in plant ecology (e.g., Upchurch and
Wolfe 1987; Jordan 1997; Greenwood et al. 2004; Jordan
2011; Peppe et al. 2011). Researchers have also found that
there are considerable differences between physiognomy and
climate when comparing the CLAMP PHYSG3BR and
PHYSG3AR to regional floras in South America, Australia,
and New Zealand (e.g., Gregory-Wodzicki 2000; Kennedy
et al. 2002,2014; Kowalski 2002; Hinojosa et al. 2006a;
Steart et al. 2011). There are also notable differences in the
relationship between leaf physiognomy and climate in some
of the DiLP characters (e.g., number of teeth vs. temperature
and leaf area vs. temperature) between sites from the dif-
ferent hemisphere, and particularly from Australia and New
Zealand (Peppe et al. 2011). Finally, although much research
has focused on the regional differences in the relationships
between climate and physiognomy, most of the regional
datasets have not been added to the larger calibration dataset
or used to develop new models (e.g., Steart et al. 2011 from
South Africa). As a result, most LMA regressions and the
most commonly used CLAMP calibration datasets
PHYSG3BR and PHYSG3AR have few to no floras from
Africa, South America, and Oceania. The net effect is the
most leaf physiognomic models do a relatively poor job
reconstructing the paleoclimate of Southern Hemisphere
floras (Kowalsk 2002; Greenwood et al. 2004; Hinojosa
et al. 2006a; Peppe et al. 2011).
Two of the hypothesized explanations for the differences
between Northern and Southern Hemisphere floras, phy-
logeny and plant ecology, likely play a role in their differ-
ences and also strongly influence the relationships globally.
Although there is evidence that leaves can phenotypically
respond to climate change rapidly (Royer et al. 2009b;
Chitwood et al. 2016a,b), there is clearly a genetic control on
leaf shape and tooth characters (reviewed in Chitwood and
Sinha 2016). Additionally, there are many families of plants
that are overwhelmingly toothed (e.g., Betulaceae, Juglan-
daceae) or untoothed (Lauraceae, Myrtaceae). However, in
Viburnum there are evolutionary correlated shifts in leaf
shape and leaf tooth morphology that are coincident with
climate (Schmerler et al. 2012), suggesting that in at least
some taxa, leaf physiognomy responds to climate on evo-
lutionary timescales.
More generally, the phylogenetic composition of a flora
strongly can strongly influence leaf physiognomy (Little
et al. 2010). There appears to be a significant phylogenetic
signal in the relationship between leaf margin and MAT
(Jones et al. 2009; Little et al. 2010; Hinojosa et al. 2011)
and, using a restricted dataset of 17 sites from the eastern
United States and Central America, in the relationship
between DiLP characters and MAT as well (Little et al.
2010). Little et al. (2010) also found that phylogenetic
generalized least squares analysis that included phylogenetic
information improved the ability to predict margin state
(toothed vs. entire) from MAT, but with much larger
uncertainty and a flatter regression slope. Based on these
results, Little et al. (2010) argue that LMA is so strongly
affected by phylogenetic bias that it should not be used.
However, although there is phylogenetic signal in leaf traits,
there is a global relationship between leaf traits and climate
(Peppe et al. 2011; Yang et al. 2015) suggesting there is
likely geographic and evolutionary sorting of taxa and pre-
ferred plant morphology relative to climate. Further, recent
work by Lawing et al. (2016) suggests that the work of Little
et al. (2010) may be affected by phylogenetic autocorrelation
and that their conclusions may not be valid unless the geo-
graphic distribution of tree species also has strong phylo-
genetic inertia, which seems unlikely given the rapid
movement of plants in response to changing climate in the
Quaternary and today (e.g., Comes and Kadereit 1998;
Davis and Shaw 2001; Kelly and Goulden 2008). However,
if plant communities and/or species are in climate disequi-
librium (e.g., Svenning and Sandel 2013; Rokataorinivo
et al. 2013; Blonder et al. 2015a), there is the potential for
paleoclimate to have influenced modern plant distributions,
and thus modern leaf trait-climate relationships, which would
considerably confound leaf physiognomic based paleocli-
mate estimates. Regardless, phylogenetic history likely plays
an important role in influencing climate-leaf physiognomy
relationships, and future work should focus on incorporating
phylogeny into future leaf physiognomic methods.
Plant growth habit and leaf life-span also influences leaf
physiognomy. In particular, deciduous taxa are more likely
to be toothed than evergreen taxa (Peppe et al. 2011; Royer
et al. 2012) and leaf life-span (deciduous vs. evergreen) also
influences tooth size and number (Peppe et al. 2011). Thus,
the growth habits of species within a plant community likely
have strong influence on leaf physiognomy. As an example,
floras from the Southern Hemisphere tend to have more
species that are evergreen relative to Northern Hemisphere
floras; therefore, it is possible that leaf life-span plays a
significant role in driving the higher percentage of untoothed
species in the Southern Hemisphere (e.g., Upchurch and
302 D. J. Peppe et al.
Wolfe 1987; Greenwood et al. 2004; Peppe et al. 2011).
Using a data set of species, instead of sites for which all of
the leaf physiognomic calibrations are optimized, Royer
et al. (2012) found that the inclusion of leaf thickness and
deciduousness with other DiLP variables and a multiple
linear regression model of leaf thickness, deciduousness, and
leaf margin reduced the standard errors of both models
predictions of MAT. Leaf thickness and deciduousness are
not measureable in fossils, but leaf mass per area, which is a
proxy for leaf life-span (Royer et al. 2007), can be estimated.
The incorporation of leaf mass per area into a model with
other DiLP characters also reduces the standard error of the
models prediction of MAT (Royer et al. 2012). These pre-
liminary results suggest that although plant ecology clearly
has an effect on leaf physiognomy, its incorporation into
paleoclimate models might improve the accuracy and pre-
cision of the paleoclimate models.
Nearest living relative methods: The geographic distribu-
tions of plants are strongly influenced by their environmental
and climatic tolerances (e.g., Merriam 1894;Küchler 1964;
Holdridge 1967; Whittaker 1975). Under the assumption that
the climatic tolerances of fossil taxa and fossil plant
assemblages are similar to their modern “nearest living rel-
atives”(NLR), for more than 100 years paleobotanists have
used this information to estimate ancient climate based on
fossil plant assemblages (e.g., Heer 1870,1878a,b,1882;
Chaney 1938; Axelrod 1948,1952; MacGinitie 1953;
Axelrod and Bailey 1969; Hickey 1977; Kershaw and Nix
1988; Mosbrugger and Utescher 1997). Probably beginning
with the work of Heer (e.g., 1870,1878a,b,1882),
numerous NLR qualitative and quantitative methods have
been developed and applied primarily to Cenozoic floras.
Both qualitative and quantitative NLR methods use the cli-
matic and ecological requirements of extant plant taxa,
which means that these methods rely upon accurate taxo-
nomic identification of plant fossil material. Additionally,
NLR methods are not limited to woody dicot angiosperm
leaves alone, and any reliably identified plant organ can be
utilized.
Qualitative NLR methods rely on individual indicator
taxa to derive paleoclimate and paleoenvironment from a
fossil flora. For example, the occurrence of palms in a fossil
flora would be used to infer warm and wet conditions with
cold months above freezing because the palm family Are-
caceae is found predominately in tropical to sub-tropical
regions with limited to no winter freezing (e.g., Larcher and
Winter 1981; Wing and Greenwood 1993,1995). However,
a considerable limitation of these qualitative NLR approa-
ches is that they do not provide quantitative estimates of
paleoclimate (i.e., a MAT or MAP prediction) and the
accuracy of the paleoclimate interpretation can be strongly
influenced by taphonomic removal of important indicator
taxa as well as the level of taxonomic identification of the
indicator taxa.
Quantitative NLR methods build on the indicator taxa
approach to quantitatively estimate paleoclimate. Originally
quantitative NLR methods used a few selected fossil taxa and
assessed their overlapping climate preferences, which resul-
ted in broad “coexistence intervals”that were used to estimate
paleoclimate (e.g., Grichuk 1969; Hickey 1977; Hickey et al.
1988; Zagwijn and Hager 1987). This approach was modified
independently first as bioclimatic analysis (Kershaw and
Nix 1988; Greenwood et al. 2003,2005), and then as the
coexistence approach (CA) of Mosbrugger and Utescher
(1997). Other NLR methods also use the same fundamental
premise, but assess the climatic ranges and/or distributions
differently (e.g., Thompson et al. 2012; Harbert and Nixon
2015). All NLR methods have four primary assumptions:
(1) there are modern taxa that are systematically closely
related to identified fossil taxa (i.e., their “nearest living rel-
ative”); (2) the climate requirements of the fossil taxon are
similar to those of the nearest living relatives; (3) the climate
tolerances of the nearest living relatives can be derived from
its modern distribution; and (4) the climate data for the nearest
living relatives is good quality (Mosbrugger and Utescher
1997; Utescher et al. 2014).
In CA all identified fossil taxa are conferred to nearest
living relatives. The climatic tolerances of each fossil’s
nearest living relative are then compared to find an interval
where all nearest living relatives could coexist (Fig. 13.4).
The climatic ranges of many vascular plant taxa have been
compiled in the Palaeoflora Database (PFDB; www.
palaeoflora.de), which is the dataset used for CA. Individ-
ual taxa that do not overlap with the rest of the taxa from a
site are considered outliers and are removed from the anal-
ysis. This subjective removal of outliers has come under
criticism (Grimm and Denk 2012; Grimm et al. 2016;
Grimm and Potts 2016). CA has been used to reconstruct
seven different aspects of temperature and precipitation:
MAT, warm month mean temperature, cold month mean
temperature, MAP, monthly minimum precipitation,
monthly maximum precipitation, and warm month mean
precipitation and has a reported precision of 0.1°C and 1 mm
precipitation (Mosbrugger and Utescher 1997). This uncer-
tainty is only the analytical uncertainty of the CA calculation
for the coexisting climatic range, and thus is artificially
precise and considerably underestimates the total uncertainty
of the climate estimate.
Another quantitative NLR method is Bioclimatic Analy-
sis (Kershaw and Nix 1988; Kershaw 1997). Just as in CA,
in Bioclimatic Analysis fossil taxa are conferred the nearest
living relatives, which are compared to determine the climate
interval where all of the nearest living relatives could coexist
originally using the PFDB and more recently using climatic
ranges of taxa using the Global Biodiversity Information
13 Paleoclimate and Paleoecology Using Fossil Leaves 303
Facility (GBIF, http://www.gbif.org/) and the WorldClim
global climate models (Hijmans et al. 2005). However,
unlike CA, which uses the entire climatic range of all taxa to
calculate a climatic variable, Bioclimatic Analysis initially
limited coexistence intervals to the 10
th
(lower limit) to 90
th
(upper limit) or 5
th
to 95
th
percentile values of the total
climatic range for all of the nearest living relative taxa
(Kershaw and Nix 1988; Kershaw 1997; Greenwood et al.
2003,2005). Subsequent iterations of bioclimatic analysis,
however, calculated the estimate as the mean value of the
10
th
and 90
th
percentile of the maximum and minimum
values across all taxa recorded in a sample, with the total
range between these plotted as an “error bar”(e.g., Green-
wood et al. 2005; Eldrett et al. 2009,2014; Pross et al. 2012;
Smith et al. 2012; Reichgelt et al. 2013).
Climate Reconstruction Analysis using Coexistence
Likelihood Estimation (CRACLE) is a new quantitative
NLR method that uses global scale specimen collection data
to infer species climate tolerance by calculating the maxi-
mum joint likelihood of species coexistence (Harbert and
Nixon 2015). CRACLE uses species distribution data from
specimen collections obtained from the GBIF and the
WorldClim global climate models (Hijmans et al. 2005)to
determine species-climate tolerance profiles. When com-
pared to other quantitative NLR methods, MAT estimates
made using CRACLE are more accurate and precise
suggesting this method may offer considerable improve-
ments over other NLR approaches. However, unless modi-
fied, CRACLE is only applicable to the Holocene and late
Pleistocene because it requires modern geographic distribu-
tions of fossil taxa to make climatic estimates. Additionally,
a major limitation of CRACLE is that for it to be applied to
the fossil record, it assumes realized niche stasis, which has
been demonstrated to be invalid (e.g., Jackson and Overpeck
2000; Veloz et al. 2012), and that plant communities are in
equilibrium with climate, which also may not always be a
valid assumption (e.g., Svenning and Sandel 2013; Roka-
taorinivo et al. 2013; Blonder et al. 2015b).
The mutual climatic range (MCR) method estimates
paleoclimate from the present climatic tolerance ranges of
the nearest living relative of fossil taxa and originally was
methodologically nearly identical to CA (Sinka and Atkin-
son 1999; Sharpe 2002). More recently, Thompson et al.
(2012) modified the method and currently MCR uses
factor-analysis based climate space to plot coldest month
temperature against coldest month precipitation and warmest
month temperature against warmest month precipitation and
the estimated temperature and precipitation for a given time
of year (i.e., coldest month, warmest month) is made based
on where the climate space overlaps. The MCR method can
only be applied to the Quaternary and has only been applied
to a few fossil assemblages (Thompson et al. 2012).
Applications to modern coleopteran assemblages demon-
strates the estimates made using MCR are conservative when
compared to meteorological data, but that the method has
systematic biases (winter estimates too warm, summer esti-
mates to cool in Britain) (Thompson et al. 2012). Addi-
tionally, if neither the absence nor proportion of taxa in the
flora can be used to delimit a smaller climatic range, MCR
estimates an overly large climatic range (Thompson et al.
2012).
Shortcomings and limitations of nearest living relative
methods: All NLR methods rely heavily on accurate climate
data for the distribution of the living relative, and precise
taxonomic identification. However, resolution of climate
estimates decreases at higher taxonomic levels: species-level
analyses tend to produce narrow climate intervals, while
family-level analyses tend to produce wide climate intervals.
Additionally, species turnover within a genus that leads to
niche shifts and changes in realized niches through time
(e.g., Veloz et al. 2012), are not captured be this approach.
This is a problem when trying to apply NLR methods to the
fossil record because while younger fossil floras, such as
those from the Quaternary, often have a NLR identified at
the genus or even species level, older fossil floras, such as
those from the Paleogene and Neogene, are often associated
with a NLR at the genus or family level where climate
intervals are larger and assumptions about the relative stasis
Fig. 13.4 Coexistence approach analysis of the latest Eocene Floris-
sant Fossil Beds flora (Baumgartner and Meyer 2014). The
dashed line indicates the estimated coexistence interval using the full
range of climate data, while the solid line indicates the estimated
coexistence interval between the 10–90
th
percentiles. Both coexistence
intervals are limited by Abies and Cathaya; these MAT intervals are
estimated as 13.4–18.0°C (dashed line) and 13.9–16.2°C (solid line).
The age of the flora precludes assigning nearest living relatives at the
species level and some genera (i.e. Betula and Alnus) have relatively
large climatic ranges for MAT. There are no outliers in this example
304 D. J. Peppe et al.
of modern climatic distributions of a genus or family may
not be appropriate. As a result, NLR methods are typically
restricted to Cenozoic floras, though the methods have
sometimes been used on assemblages from the late Creta-
ceous (e.g., Mosbrugger and Utescher 1997; Utescher et al.
2014; Fletcher et al. 2014).
Although NLR methods, and in particular CA, have been
applied to a large number of Cenozoic floras (e.g., Utescher
et al. 2014), they have recently come under major criticism
based on their assumptions about species distributions, the
lack of statistical framework, and how climate variables are
calculated for past plant communities based on modern plant
distributions (Grimm and Potts 2016). Additionally, there
are significant issues with the PFDB, which is used in CA.
Studies by Grimm and Denk (2012), Grimm et al. (2016)
and Grimm and Potts (2016) have shown that the MAT
ranges in the PFDB are incorrect when compared to climate
stations or WorldClim (Hijmans et al. 2005) and even when
using corrected climate data the precision of MAT ranges
was between 5–10°C (Grimm and Denk 2012). These crit-
icisms raise major questions about the ability of quantitative
NLR methods to accurately and precisely reconstruct the
paleoclimate of ancient plant communities.
Leaf Traits and Ecology
In addition to a relationship between leaf traits and climate,
there are a suite of leaf structural and functional traits that are
correlated to each other and to environmental gradients. This
covariation of leaf traits is known as the leaf economic
spectrum (LES; Wright et al. 2004,2005; Reich 2014).
The LES is a framework for organizing leaf functional
ecology, which reflects a series of trade-offs among leaf
traits that are related to carbon balance. Plants on the
“fast-return”end of the spectrum, characterized by rapid
resource acquisition, typically have a short leaf life-span
(less than 12 months), high photosynthetic and respiration
rates, low leaf mass per area (M
A
), high nutrient concen-
trations (principally nitrogen and phosphorus), and fast
growth rates, while the inverse is true for plants with slow
resource acquisition on the “slow-return”end of the spec-
trum (Reich et al. 1997; Westoby et al. 2002; Diaz et al.
2004; Wright et al. 2004,2005; Reich 2014). The interre-
lationships between these traits are largely independent of
phylogeny in seed plants (Ackerly and Reich 1999), have
been observed in all vascular plant groups (Wright et al.
2004; Karst and Lechowicz 2007; Peppe et al. 2014), and
probably reflect evolutionary trade-offs in leaf design to
maximize resource acquisition vs. performance for the
environment in which a plant is living (Shipley et al. 2006).
Recently, researchers have suggested that venation net-
works are linked to the LES either as the underlying deter-
minant of leaf mass per area, leaf life-span, carbon
assimilation rate, and nitrogen concentration (“vein origin”
hypothesis; Blonder et al. 2011,2013; Blonder et al. 2014a)
or that venation networks are only one axis of many that
determine hydraulic conductance and carbon economics, and
in turn the plant’s relative growth rate (“flux trait network”
hypothesis; Sack and Scoffoni 2013; Sack et al. 2013,2014).
Regardless of which hypothesis is correct, both highlight the
importance of vein networks in the LES.
While in modern leaves these LES variables are easily
measured, they cannot be directly measured in fossil leaves.
Instead a series of different proxies based on features that are
measureable in fossil leaves, such as leaf area, petiole width,
and vein density, have been developed to estimate leaf mass
per area M
a
, leaf life-span (LLS), carbon assimilation rate, and
respiration rate (e.g., Brodribb et al. 2007; Boyce et al. 2009;
Brodribb and Feild 2010; Royer et al. 2007,2010,2012;
Blonder et al. 2011; Feild et al. 2011b; Sack and Scoffoni
2013; Sack et al. 2013,2014; Blonder and Enquist 2014;
Peppe et al. 2014). These reconstructed LES variables
can then be used to reconstruct aspects of the paleoecology of
the fossil plant (e.g., leaf life-span, hydraulic conductance), as
well as other features of the ecosystem and environment in
which the plant lived. Since the majority of these methods are
developed based on relationships within the LES, they are
applicable across a variety of plant groups and can be used
throughout the geologic record (e.g., Royer et al. 2007,2010,
2012; Boyce et al. 2009; Brodribb and Feild 2010; Blonder
et al. 2011,2014b; Feild et al. 2011b; Crifo et al. 2014; Peppe
et al. 2014), though some aspects of the LES, such as carbon
assimilation and respiration rates are probably heavily
dependent on past levels of atmospheric CO
2
and O
2
(e.g., De
Boer et al. 2012; McElwain et al. 2016).
Methods of Paleoecological
Reconstruction
Most leaf economic spectrum traits cannot be directly
measured in fossils. However, a proxy for M
a
, which is a
core LES character and also correlates strongly with leaf
thickness and leaf life-span (Wright et al. 2004; Royer et al.
2012), has been developed that is applicable to fossil plants
(Royer et al. 2007,2010; Peppe et al. 2014). The method
uses the scaling relationship between petiole width (PW),
leaf area (A), and M
a
. The relationship between PW and A to
M
a
is based on the biomechanics of the leaves since heavier
leaves need more support, which necessitates larger, thicker
petioles (Royer et al. 2007,2010; Peppe et al. 2014). Leaves
with high M
a
are more commonly evergreen and thicker,
13 Paleoclimate and Paleoecology Using Fossil Leaves 305
which in addition to providing structural support also allows
long-term nutrient transport (Wright et al. 2004).
Proxies for leaf mass per area have been developed for all
woody dicot angiosperms (Royer et al. 2007), herbaceous
angiosperms, broadleaf gymnosperms (Royer et al. 2010),
and ferns (Peppe et al. 2014) (Table 13.2). Each of these
groups has significantly different scaling relationships
between PW and A to M
a
and the relationships in angios-
perms and broadleaf gymnosperms are explained by a dif-
ferent biomechanical model than ferns. The relationships
between area normalized PW and M
a
in angiosperms and
broadleaf gymnosperms are best explained by modeling a
leaf as a pole supporting a singular mass and assumes that
the cross-sectional area of the petiole is proportional to leaf
mass (Royer et al. 2007,2010). The relationships between
PW and A to M
a
in ferns are best explained by a cantilever
beam model that assumes the petiole width and length are
independent, that petiole shape is relatively invariant, and
that the leaf mass is proportional to the flexural rigidity of
the entire petiole (Peppe et al. 2014).
The M
a
of fossil leaves can be used to estimate a number
of paleoecological variables. First, greater M
a
is correlated
with longer leaf life-span and can be used to reconstruct the
leaf life-span of a fossil flora (Royer et al. 2007,2010; Peppe
et al. 2014). Fossil leaves with a M
a
<*87 g m
−2
can be
broadly categorized as deciduous and those with a M
a
>
*129 g m
−2
can be broadly categorized as evergreen, with
the transition between predominately deciduous to predom-
inately evergreen occurring at *111 g m
−2
(Royer et al.
2007). The variance in M
a
and leaf life-span can also be used
to infer seasonality in paleoecosystems with greater variance
in M
a
in plant populations related to greater seasonality
(Wright et al. 2005). Additionally, the distribution of M
a
of a
flora also appears to be environmentally specific (Royer
et al. 2010), suggesting that the distribution of M
a
in a fossil
flora can be compared to modern distributions and used as a
proxy for environment. Finally, M
a
correlates with leaf
nitrogen content (Royer et al. 2005), and could potentially be
used as a proxy for percent nitrogen, which could be used to
infer nutrient cycling in fossil forest ecosystems.
The interaction of plants and their herbivores has a long
geologic record and can be used to infer a variety of eco-
logical variables (Labandeira and Currano 2013). Recon-
structions of M
a
combined with measurements of
insect-plant interactions via their feeding traces can be used
to reconstruct aspects of plant ecology such as nutrient
content and ecosystem disturbance (Royer et al. 2007;
Currano et al. 2008,2011). Leaves with greater M
a
have
lower diversity, lower density, and lower area removed by
insect herbivores when compared to leaves with low M
a
(Coley 1983; Royer et al. 2007). This relationship between
insect damage and M
a
is probably because leaves with high
M
a
are typically thicker and/or tougher, possess a greater
amount and concentration of chemical deterrents, and have
lower leaf nitrogen content, which makes them less nutri-
tious and less palatable for insect herbivores (Coley 1983;
Royer et al. 2007). Additionally, changes in the diversity and
density of insect feeding traces can be used to infer
large-scale changes in ecology following ecological distur-
bances, including changes in climate and recovery after mass
extinction events (e.g., Wilf et al. 2006; Currano et al. 2008,
2011; Donovan et al. 2016).
Recent studies have suggested a link between vein net-
works and the LES (Blonder et al. 2011,2013,2014);
however, this remains controversial (e.g., Sack and Scoffoni
2013; Sack et al. 2013,2014; Li et al. 2015). If there is a
relationship between vein networks and the leaf economic
spectrum, measurements of vein networks in fossil plants
could be used to reconstruct other LES variables, such as
leaf life-span and M
a
, which in turn would allow for pale-
oecological reconstruction. These proxies are in develop-
ment, but offer considerable promise. In addition to its
relationship to the LES, vein networks, and specifically vein
density, is a proxy for water use efficiency and gas exchange,
making it a powerful tool for assessing gas and water
exchange capacity of plants (Sack et al. 2005; Sack and
Frole 2006; Brodribb et al. 2007; Boyce et al. 2009; Bro-
dribb and Feild 2010). For example, angiosperms have much
higher vein densities than plants from all other groups, living
or extinct, which allows angiosperms to have much higher
rates of transpiration (Boyce et al. 2009). This connection
between high vein density and high rates of transpiration in
angiosperms has been used to infer that the evolution of
angiosperms and their increase in maximum leaf vein den-
sity through the Cretaceous enabled the evolution and
expansion of tropical rain forests, because transpiration rates
in other plant groups are too low to maintain the precipita-
tion abundance and evenness necessary for tropical rain-
forest ecosystems (Boyce et al. 2009,2010; Feild et al.
2011a). The distribution of vein densities within a flora
Table 13.2 Leaf mass per area models for woody dicot angiosperms, herbaceous angiosperms, broadleaf gymnosperms, and ferns
Taxonomic Group Model ab r
2
Authors
Woody dicot angiosperms log(M
A
)=a+blog(PW
2
/A) 3.07 0.382 0.55 Royer et al. (2007)
Herbaceous angiosperms log(M
A
)=a+blog(PW
2
/A) 3.015 0.3076 0.44 Royer et al. (2010)
Broadleaf gymnosperms log(M
A
)=a+blog(PW
2
/A) 2.245 0.2204 0.32 Royer et al. (2010)
Ferns log(M
A
)=a+blog(PW
4
/A) 4.207 0.252 0.44 Peppe et al. (2014)
306 D. J. Peppe et al.
differs significantly between canopy and understory plants
and between sites with different forest structures (Crifo et al.
2014). Thus, it is possible to infer ancient forest structure,
and its constituent understory and canopy taxa, from the
range of vein densities within a fossil flora (Crifo et al.
2014).
Shortcomings and Limitations
of Paleoecological Methods
Methods of paleoecological reconstruction are very powerful
tools for assessing aspects of plant ecology (e.g., leaf
life-span, hydraulic conductance) and a ecosystem’s envi-
ronment (e.g., ecosystem disturbance, forest structure);
however, these methods are relatively new and have not
been widely applied. The primary reason for this is that they
require excellent preservation of venation (vein density
proxies), and the entire leaf area and at least part of the leaf
petiole (M
A
proxies). Due to taphonomic constraints, these
proxies cannot be applied to many fossil floras, or in some
cases, can only be applied to the best preserved specimens
within a flora such that the reconstructed ecological variables
may not be representative of the entire flora. As an example,
Blonder et al. (2014b) were only able to analyze <20% of
more than 6,000 collected fossil specimens, suggesting the
potential for significant taphonomic biases in leaf vein
density analyses. Additionally, though there appears to be a
relationship between leaf vein density and forest structure
(Crifo et al. 2014) and M
A
and environment (Royer et al.
2010), these relationships are based on relatively limited
datasets with small sample sizes. Thus, it’s unclear if these
relationships are representative globally and to all
ecosystems.
Application of Paleoclimate
and Paleoecological Proxies
Leaf physiognomic methods for reconstructing paleoclimate
have been applied hundreds, if not thousands of times to fossil
floras from the Cretaceous through the Quaternary, and leaf
trait methods for reconstructing paleoecology have begun to
be applied to fossil plants throughout Earth’s history. The
following is a brief example of an application of these
methods for reconstructing paleoclimate and paleoecology by
Michel et al. (2014) that made it possible to reconstruct the
climate and to determine the climatic, environmental, and
ecological biome of the fossil site. The study of Michel et al.
(2014) documented a diverse ssemblage of fossil vertebrates,
leaves, and tree stump and root casts and used the fossil data
to reconstruct the paleoclimate and paleoenvironment of the
Early Miocene R3 fossil site on Rusinga Island, western
Kenya. The R3 fossil site is of particular interest because of
the co-occurrence of multiple species of Early Miocene
catarrhines, including Dendropithecus macinnesi and
Ekembo heseloni, as well as several vertebrate fossil taxa that
suggest a more closed environment. Prior to the work of
Michel et al. (2014), the paleoenvironment of the site was
relatively poorly constrained. Michel et al. (2014) mapped a
series of fossil tree stump casts of varying diameter preserved
within the same stratigraphic layer across the R3 site. Based
on the density of the tree stump casts across the landscape,
they reconstructed the ecosystem as a dense, closed-canopy,
multi-storied forest. Abundant fossil leaves were found in a
thin mudstone layer stratigraphically directly above the
fossil tree stump casts. The flora was made up of 29 mor-
photypes, which can be considered a rough proxy for bio-
logical species (see Ash et al. 1999 and Peppe et al. 2008 for a
review of the morphotype method), representing 27
dicot and 2 monocotyledonous angiosperms. The dicots
were used to reconstruct the paleoclimate and paleoecology
of the site.
The mean annual temperature and mean annual precipi-
tation of the site was reconstructed using LMA and LAA,
respectively (Fig. 13.5). Mean annual temperature recon-
structed using the LMA regressions of Wolfe (1979), Miller
et al. (2006), Kowalski and Dilcher (2003), and Peppe et al.
(2011) ranged between 22.6°C and 34.5°C (Fig. 13.5).
Mean annual precipitation reconstructed using the LAA
Fig. 13.5 Climatic reconstruction for the Rusinga R3 fossil site from
Michel et al. (2014). Black circle is the average is of all mean annual
temperature and mean annual precipitation estimates. Shaded square is
the range of all climate estimates. Biomes follow Whittaker (1975).
Area shaded in light gray indicates mean annual temperatures higher
than typical conditions on Earth today. The dashed lines represent
projections of biomes outside of modern climate space
13 Paleoclimate and Paleoecology Using Fossil Leaves 307
regression of Wilf et al. (1998), Gregory-Wodzicki (2000),
Jacobs and Herendeen (2004) and Peppe et al. (2011) ranged
from 1,394 to 2,618 mm (Fig. 13.5). Wet month precipita-
tion estimates using the method of Jacobs and Herendeen
(2004) indicate that there was relatively little precipitation
during the dry season. This evidence for seasonality is also
supported by analyses of the paleosols at the R3 site. The
paleoclimate reconstructions for the site indicate a tropical
seasonal forest biome (Fig. 13.5; Whittaker 1975). Interest-
ingly, the forest density evidence for the R3 site are very
similar to modern tropical seasonal forests that support
medium and large bodied primate communities, such as
Barro Colorado Island (BCI), Panama and Ituri Forest,
Democratic Republic of Congo. Only univariate methods
were utilized on the R3 flora because, as discussed above,
the leaf physiognomic space of the fossil flora is not cap-
tured by the modern DiLP dataset (Fig. 13.3). This
demonstrates one of the limitations of the existing multi-
variate methods, and the difficulty of reconstructing the
paleoclimate of African fossil floras.
Leaf mass per area was reconstructed for seventeen of the
twenty seven dicot morphotypes from R3 (Figs. 13.6,13.7).
Leaf mass per area reconstructions for the flora indicate that
the species were predominantly evergreen, but that there was
an important component of deciduous taxa (Fig. 13.6). The
distribution of M
a
at the site is most similar to non-riparian
sites (Fig. 13.7), which is unsurprising given the sedimen-
tological and paleosol evidence from the site. The paleo-
climate reconstructions for the flora are most similar to BCI
site (MAT = 26.5°C, MAP = 2620 mm), and the distribution
of M
a
at the sites is similar, though BCI has a higher pro-
portion of species low M
a
than the R3 flora (Fig. 13.7D).
The R3 flora is also somewhat similar to the non-riparian,
wet tropical rainforest of Buena Vista, Puerto Rico, though
Fig. 13.6 Estimated leaf mass per area (M
a
) for Rusinga R3 flora and comparison to extant vegetation. A, M
a
and leaf lifespan for 678 species of
angiosperms, ferns, and gymnosperms modified from Royer et al. (2010). M
a
and leaf lifespan data from Wright et al. (2004). M
a
bin is 20 g/m
2
.
Gray line at 129 g/m
2
indicates the upper bound of the transition between species whose leaves are predominately deciduous (leaf life-span <12
month) versus those that are predominately evergreen (leaf lifespan >12 months). B, Estimated M
a
for 17 morphotypes from the Rusinga R3 flora
showing that the majority of morphotypes were likely evergreen (10 of 17 reconstructed). Errors are 95% prediction intervals. M
a
data and
morphotypes from Michel et al. (2014)
308 D. J. Peppe et al.
the wet tropics site has a larger proportion of species with
much higher M
a
than reconstructed for the R3 flora
(Fig. 13.7E). Similar to the paleoclimate estimates, the M
a
reconstructions for R3 also suggest a non-riparian tropical
seasonal forest or tropical rainforest biome.
The paleoclimate and paleoecological reconstructions for
the R3 fossil site using fossil leaves indicate a tropical sea-
sonal forest biome and a warm and seasonally wet climate
(Figs. 13.5,13.6,13.7). These results are also in agreement
with reconstructions of the forest density and paleonviron-
mental reconstructions based on the paleosols at the site and
the composition of the vertebrate fauna. Together, the mul-
tiproxy evidence presented by Michel et al. (2014) indicates
that the R3 fossil locality sampled a widespread, dense,
multistoried, closed-canopy tropical seasonal forest set in a
warm and seasonally wet climate. When combined with the
primate fossil record from the R3 site, the paleoclimate and
paleoecological results of Michel et al. (2014) demonstrate
the importance of forested environments in the evolution of
early apes in Africa. The Michel et al. (2014) study is an
example of the power and utility of combining fossil leaf
based estimates of paleoclimate and paleoecology to
Fig. 13.7 Comparison of the reconstructed M
a
distribution of the Rusinga R3 flora with extant sites from varying climate biomes and
environments. M
a
and climate data for extant sites from Peppe et al. (2011). The fossil and extant sites are comprised of exclusively wood dicot
angiosperms. The M
a
bin size is 20 g/m
2
. A, Rusinga R3 (replicated in gray in all panels). B, Connecticut River, near Middletown, Connecticut,
USA. C, Big Hammock Wildlife Refuge Area, Georgia, USA. D, Pee Dee State Park, South Carolina, USA; E, Barro Colorado Island, Panama. F,
Buena Vista, Puerto Rico
13 Paleoclimate and Paleoecology Using Fossil Leaves 309
reconstruct a terrestrial ecosystem, and is a good model for
fossil leaf based research focused on terrestrial ecosystem
and climate reconstruction.
Future Research Directions
Fossil plants, and particularly fossil leaves, are powerful
tools for reconstructing paleoclimate and paleoecology.
However, despite the power of the paleoclimate and pale-
oecology methods that can be applied to fossils, they have
important limitations.
One of the most important potential shortcomings of leaf
physiognomic methods is the possible influence of phy-
logeny and life history on leaf trait-climate relationships.
Recently, the vertebrate paleontology community has begun
to develop a series of trait-based approaches that use rela-
tionships between functional traits and environment to
reconstruct climate and assess the responses of faunal com-
munities to environmental change called ecometrics (e.g.,
Eronen et al. 2010a; Eronen et al. 2010b; Polly et al. 2011;
Lawing et al. 2012,2016; Polly and Sarwar 2014). These
methods have addressed potential issues related to phylo-
genetic influences on functional traits by resampling tech-
niques and by applying phylogenetic corrections to the
datasets (Eronen et al. 2010b; Lawing et al. 2012; Polly and
Sarwar 2014; Lawing et al. 2016). The utilization of these
types of methods when revising and developing new leaf
physiognomic methods for reconstructing paleoclimate has
considerable potential and is an important future research
direction. Additionally, as noted above, the leaf
physiognomic-climate relationships are different between the
Northern and Southern Hemispheres. The development of
new calibration datasets that incorporate more floras from
the Southern Hemisphere, and particularly Africa and South
America, and from tropical regions will add phylogenetic
diversity to the calibration datasets and may make it possible to
ascertain the driver behind the hemispheric differences, espe-
cially if phylogenetic corrections are applied to the datasets.
Leaf life-span also has an important effect on leaf
trait-climate relationships. Royer et al. (2012) found that the
incorporation of leaf thickness, deciduousness, and M
A
into
species-specific leaf physiognomic models reduced the
standard errors of the models. However, this result was
based on a species-based dataset, instead of a site-based
calibration, on which all leaf physiognomic models are
based. Thus, it’s uncertain if these results are broadly
applicable to existing leaf physiognomic models. Using a
dataset of taxa from 17 floras from a range of MATs,
Blonder and Enquist (2014) demonstrated a relationship
between vein density and MAT, which they used to develop
a predictive model for MAT. However, the relationships
between vein density and MAT were different in temperate
and tropical floras, such that they developed one model with
different parameter values for temperature floras and tropical
floras. Taken together, the results of Royer et al. (2012) and
Blonder and Enquist (2014) suggest that the incorporation of
functional traits, such as vein density and M
A
, into leaf
physiognomic models could improve their precision and
accuracy and are an interesting research direction to
pursue.
In addition to potential shortcomings with the methods,
making the measurements necessary to apply paleoclimate
and paleoecological proxies can be time consuming and trait
scoring can be ambiguous. Recently, there has been a push
to develop computer-based algorithms that are automated to
measure and score leaf traits and to identify plants using
leaves (e.g., Cope et al. 2012; Corney et al. 2012a,b; Price
et al. 2012; Green et al. 2014; MacLeod and Steart 2015;
Wilf et al. 2016; Liao et al. 2017). Automated measurements
of leaf size and shape would eliminate ambiguities in leaf
trait scoring, increase scoring throughout, make it easier to
measure continuous variables, and potentially help aid in the
identification of fossil plants. The potential of automated or
semi-automated leaf trait scoring and identification is an
exciting new development that has the potential to revolu-
tionize the way leaf traits are scored and help to significantly
improve leaf trait-climate and ecological proxies.
One of the fundamental shortcomings of leaf-based cli-
mate reconstruction is that we still do not understand the
physiological and/or evolutionary mechanisms that drive the
empirical relationships observed between leaf physiognomy
and climate. As a result, leaf physiognomic proxies are
empirically derived and traits used for extrapolation may not
have a direct functional link to climate or ecology and/or
may be strongly driven by phylogeny and life history. This
might lead to biased paleoclimate and paleoecological esti-
mates, particularly in deep time. Research focused on the
evolution of leaf form and on the genetic underpinnings of
leaf shape and development is making significant strides
towards addressing this shortcoming (e.g., Pérez-Pérez et al.
2010; Schmerler et al. 2012; Edwards and Donoghue 2013;
Chitwood et al. 2014,2016a,b; Dkhar and Pareek 2014;
Rodriguez et al. 2014; Vaughan 2015; Chitwood and Sinha
2016; Coneva et al. 2017). A more complete understanding
of the genetic controls on leaf size and shape and the drivers
of leaf shape evolution will make it possible for us to better
understand how environment and climate influence leaf
physiognomy, as well as provide a basis for understanding
plant species’distributions on climatic gradients. This will
allow for the development of phylogenetically and physio-
logically informed plant-based paleoclimatic and paleoeco-
logical proxies, which has the potential to revolutionize the
310 D. J. Peppe et al.
field and our ability to reconstruct ancient climate and
ecology.
Acknowledgements We would like to thank Denise Su, Scott Simp-
son, and Darin Croft for inviting us to contribute to the Latest Methods
in Reconstructing Cenozoic Terrestrial Environments and Ecological
Communities workshop and to this special volume. We thank Dana
Royer for many helpful discussions about these topics and review of
this manuscript and David Greenwood for his helpful and constructive
review. This work was supported by the National Science Foundation
(grant EAR-1325552).
References
Ackerly, D. D., & Reich, P. B. (1999). Convergence and correlations
among leaf size and function in seed plants: a comparative test
using independent contrasts. American Journal of Botany, 86,
1272–1281.
Adams, J. M., Green, W. A., & Zhang, Y. (2008). Leaf margins and
temperature in the North American flora: recalibrating the paleo-
climatic thermometer. Global and Planetary Change, 60, 523–534.
Adams, J. S., Kraus, M. J., & Wing, S. L. (2011). Evaluating the use of
weathering indices for determining mean annual precipitation in the
ancient stratigraphic record. Palaeogeography, Palaeoclimatology,
Palaeoecology, 309, 358–366.
Aizen, M. A., & Ezcurra, C. (2008). Do leaf margins of the temperate
forest flora of southern South America reflect a warmer past? Global
Ecology and Biogeography, 17, 164–174.
Ash, A., Ellis, B., Hickey, L. J., Johnson, K. R., Wilf, P., & Wing, S.
(1999). Manual of leaf architecture –morphological description and
categorization of dicotyledonous and net-veined monocotyledonous
angiosperms (65 p.). Washington D. C.: The Leaf Architecture
Working Group, Smithsonian Institution.
Axelrod, D. I. (1948). Climate and evolution in western North America
during middle Pliocene time. Evolution, 127–144.
Axelrod, D. I. (1952). A theory of angiosperm evolution. Evolution,
29–60.
Axelrod, D. I., & Bailey, H. P. (1969). Paleotemperature analysis of
tertiary floras. Palaeogeography, Palaeoclimatology, Palaeoecol-
ogy, 6, 163–195.
Bailey, I. W., & Sinnott, E. W. (1915). A botanical index of Cretaceous
and Tertiary climates. Science, 41, 831–834.
Bailey, I. W., & Sinnott, E. W. (1916). The climatic distribution of
certain types of angiosperm leaves. American Journal of Botany, 3,
24–39.
Baker-Brosh, K. F., & Peet, R. K. (1997). The ecological significance
of lobed and toothed leaves in temperate forest trees. Ecology, 78,
1250–1255.
Baumgartner, K. A., & Meyer, H. W. (2014). Coexistance climate
analysis of the late Eocene Florissant flora, Colorado. Geological
Society of America –Abstract with Programs, 46, 490.
Billings, F. H. (1905). Precursory leaf serrations of Ulmus.Botanical
Gazette, 224–225.
Blonder, B., & Enquist, B. J. (2014). Inferring climate from angiosperm
leaf venation networks. New Phytologist, 204, 116–126.
Blonder, B., Nogués-Bravo, D., Borregaard, M. K., Donoghue, I., John,
C., Jørgensen, P. M., et al. (2015a). Linking environmental filtering
and disequilibrium to biogeography with a community climate
framework. Ecology, 96, 972–985.
Blonder, B., Nogues-Bravo, D., Borregaard, M., Lessard, J.P., Violle,
C., Svenning, J.C., et al. (2015b). Linking environmental filtering
and disequilibrium to biogeography with a community climate.
Ecology, 96, 972–985.
Blonder, B., Violle, C., Bentley, L. P., & Enquist, B. J. (2011).
Venation networks and the origin of the leaf economics spectrum.
Ecology Letters, 14, 91–100.
Blonder, B., Violle, C., Bentley, L. P., & Enquist, B. J. (2014a).
Inclusion of vein traits improves predictive power for the leaf
economic spectrum: a response to Sack et al. (2013). Journal of
Experimental Botany, 65, 5109–5114.
Blonder, B., Royer, D. L., Johnson, K. R., Miller, I., & Enquist, B.
J. (2014b). Plant ecological strategies shift across the
Cretaceous-Paleogene boundary. PLoS Biology, 12, e1001949.
Blonder, B., Violle, C., & Enquist, B. J. (2013). Assessing the causes
and scales of the leaf economics spectrum using venation networks
in Populus tremuloides.Journal of Ecology, 101, 981–989.
Borregaard, M., Lessard, J. P., Violle, C., Svenning, J. C., Rahbek, C.,
& Enquist, B. (2015). Linking environmental filtering and disequi-
librium to biogeography with a community climate. Ecology, 96,
972–985.
Boyce, C. K., Brodribb, T. J., Feild, T. S., & Zwieniecki, M. A. (2009).
Angiosperm leaf vein evolution was physiologically and environ-
mentally transformative. Proceedings of the Royal Society B:
Biological Sciences, 276, 1771–1776.
Boyce, C. K., Lee, J.-E., Feild, T. S., Brodribb, T. J., & Zwieniecki, M.
A. (2010). Angiosperms helped put the rain in the rainforests: the
impact of plant physiological evolution on tropical biodiversity 1.
Annals of the Missouri Botanical Garden, 97, 527–540.
Brodribb, T. J., & Feild, T. S. (2010). Leaf hydraulic evolution led a
surge in leaf photosynthetic capacity during early angiosperm
diversification. Ecology Letters, 13, 175–183.
Brodribb, T. J., Feild, T. S., & Jordan, G. J. (2007). Leaf maximum
photosynthetic rate and venation are linked by hydraulics. Plant
Physiology, 144, 1890–1898.
Burnham, R. J., Ellis, B., & Johnson, K. R. (2005). Modern tropical
forest taphonomy: does high biodiversity affect paleoclimatic
interpretations? PALAIOS, 20, 439–451.
Burnham, R. J., Pitman, N. C. A., Johnson, K. R., & Wilf, P. (2001).
Habitat-related error in estimating temperatures from leaf margins in a
humid tropical forest. American Journal of Botany, 88, 1096–1102.
Burnham, R. J., & Tonkovich, G. S. (2011). Climate, leaves, and the
legacy of two giants. New Phytologist, 190, 514–517.
Canny, M. (1990). Tansley review no. 22 what becomes of the
transpiration stream? New Phytologist, 114, 341–368.
Chaloner, W. G., & Creber, G. T. (1990). Do fossil plants give a
climatic signal? Journal of the Geological Society, 147, 343–350.
Chaney, R. W. (1938). Paleoecological interpretations of Cenozoic
plants in western North America. The Botanical Review, 4, 371–
396.
Chen, W.-Y., Su, T., Adams, J. M., Jacques, F. M., Ferguson, D. K., &
Zhou, Z.-K. (2014). Large-scale dataset from China gives new
insights into leaf margin–temperature relationships. Palaeogeogra-
phy, Palaeoclimatology, Palaeoecology, 402, 73–80.
Chitwood, D. H., Ranjan, A., Martinez, C. C., Headland, L. R., Thiem,
T., Kumar, R., et al. (2014). A modern ampelography: a genetic
basis for leaf shape and venation patterning in grape. Plant
Physiology, 164, 259–272.
Chitwood, D. H., Klein, L. L., O’Hanlon, R., Chacko, S., Greg, M.,
Kitchen, C., et al. (2016a). Latent developmental and evolutionary
shapes embedded within the grapevine leaf. New Phytologist, 210,
343–355.
Chitwood, D. H., Rundell, S. M., Li, D. Y., Woodford, Q. L., Tommy,
T. Y., Lopez, J. R., et al. (2016b). Climate and developmental
plasticity: interannual variability in grapevine leaf morphology.
Plant Physiology, 170, 1480–1491.
Chitwood, D. H., & Sinha, N. R. (2016). Evolutionary and environ-
mental forces sculpting leaf development. Current Biology, 26,
R297–R306.
13 Paleoclimate and Paleoecology Using Fossil Leaves 311
Coley, P. D. (1983). Herbivory and defensive characteristics of tree
species in a lowland tropical forest. Ecological monographs, 53,
209–234.
Comes, H. P., & Kadereit, J. W. (1998). The effect of Quaternary
climatic changes on plant distribution and evolution. Trends in
Plant Science, 3, 432–438.
Coneva, V., Frank, M. H., de Luis Balaguer, M. A., Li, M., Sozzani, R.,
& Chitwood, D. H. (2017). Genetic architecture and molecular
networks underlying leaf thickness in desert-adapted tomato
Solanum pennellii.Plant Physiology, 175, 376–391.
Cope, J. S., Corney, D., Clark, J. Y., Remagnino, P., & Wilkin,
P. (2012). Plant species identification using digital morphometrics,
A review. Expert Systems with Applications, 39, 7562–7573.
Corney, D., Clark, J. Y., Tang, H. L., & Wilkin, P. (2012a). Automatic
extraction of leaf characters from herbarium specimens. Taxon, 61,
231–244.
Corney, D. P., Tang, H. L., Clark, J. Y., Hu, Y., & Jin, J. (2012b).
Automating digital leaf measurement: the tooth, the whole tooth,
and nothing but the tooth. PLoS ONE, 7, e42112.
Couturier, E., Du Pont, S. C., & Douady, S. (2011). The filling law: a
general framework for leaf folding and its consequences on leaf
shape diversity. Journal of Theoretical Biology, 289,47–64.
Couturier, E., Brunel, N., Douady, S., & Nakayama, N. (2012). Abaxial
growth and steric constraints guide leaf folding and shape in Acer
pseudoplatanus. American Journal of Botany, 99, 1289–1299.
Crifo, C., Currano, E. D., Baresch, A., & Jaramillo, C. (2014).
Variations in angiosperm leaf vein density have implications for
interpreting life form in the fossil record. Geology, 42, 919–922.
Christophel, D. C., & Blackburn, D. T. (1978). Tertiary megafossil
flora of Maslin Bay, South Australia: a preliminary report.
Alcheringa, 2, 311–319.
Critchfield, W. B. (1960). Leaf dimorphism in Populus trichocarpa.
American Journal of Botany, 47, 699–711.
Critchfield, W. B. (1971). Shoot growth and heterophylly in Acer.
Journal of the Arnold Arboretum, 52, 240–266.
Currano, E. D., Jacobs, B. F., Pan, A. D., & Tabor, N. J. (2011).
Inferring ecological disturbance in the fossil record: a case study
from the Late Oligocene of Ethiopia. Palaeogeography, Palaeocli-
matology, Palaeoecology, 309, 242–252.
Currano, E. D., Wilf, P., Wing, S. L., Labandeira, C. C., Lovelock, E. C.,
& Royer, D. L. (2008). Sharply increased insect herbivory during the
Paleocene-Eocene Thermal Maximum. Proceedings of the National
Academy of Sciences, USA, 105, 1960–1964.
Davis, M. B. (1986). Climatic instability, time, lags, and community
disequilibrium. In J. Diamond & T. J. Case (Eds.), Community
ecology (pp. 269–284). New York, NY: Harper & Row.
Davis, M. B., & Shaw, R. G. (2001). Range shifts and adaptive
responses to Quaternary climate change. Science, 292, 673–679.
De Boer, H. J., Eppinga, M. B., Wassen, M. J., & Dekker, S. C. (2012).
A critical transition in leaf evolution facilitated the Cretaceous
angiosperm revolution. Nature Communications, 3, 1221.
Diaz, S., Hodgson, J. G., Thompson, K., Cabido, M., Cornelissen,
J. H. C., Jalili, A., et al. (2004). The plant traits that drive
ecosystems: evidence from three continents. Journal of Vegetation
Science, 15, 295–304.
Dilcher, D. (1973). A paleoclimatic interpretation of the Eocene floras
of southeastern North America. In A. Graham (Ed.), Vegetation and
vegetational history of northern Latin America (pp. 39–59).
Amsterdam: Elsevier.
Dilcher, D. L., Kowalski, E. A., Wiemann, M. C., Hinojosa, L. F., &
Lott, T. A. (2009). A climatic and taxonomic comparison between
leaf litter and standing vegetation from a Florida swamp woodland.
American Journal of Botany, 96, 1108–1115.
Dkhar, J., & Pareek, A. (2014). What determines a leaf’s shape?
EvoDevo, 5, 47.
Dolph, G. E., & Dilcher, D. (1980a). Variation in leaf size with respect
to climate in Costa Rica. Biotropica, 12, 91–99.
Dolph, G. E., & Dilcher, D. (1980b). Variation in leaf size with respect
to climate in the tropics of the Western Hemisphere. Bulletin of the
Torrey Botanical Club, 107, 154–162.
Donovan, M. P., Iglesias, A., Wilf, P., Labandeira, C. C., & Cuneo, N.
R. (2016). Rapid recovery of Patagonian plant-insect associations
after the end-Cretaceous extinction. Nature Ecology and Evolution,
1, 12.
Edwards, E. J., & Donoghue, M. J. (2013). Is it easy to move and easy
to evolve? Evolutionary accessibility and adaptation. Journal of
Experimental Botany, 64, 4047–4052.
Edwards, E. J., Spriggs, E. L., Chatelet, D. S., & Donoghue, M. J.
(2016). Unpacking a century‐old mystery: winter buds and the
latitudinal gradient in leaf form. American Journal of Botany, 103,
975–978.
Eldrett, J. S., Greenwood, D. R., Harding, I. C., & Huber, M. (2009).
Increased seasonality through the Eocene to Oligocene transition in
northern high latitudes. Nature, 459, 969–974.
Eldrett, J. S., Greenwood, D. R., Polling, M., Brinkhuis, H., & Sluijs,
A. (2014). A seasonality trigger for carbon injection at the
Paleocene-Eocene Thermal Maximum. Climate of the Past, 10,
759–769.
Eronen, J., Puolamäki, K., Liu, L., Lintulaakso, K., Damuth, J., Janis,
C., et al. (2010a). Precipitation and large herbivorous mammals i:
estimates from present-day communities. Evolutionary Ecology
Research, 12, 217–233.
Eronen, J. T., Polly, P. D., Fred, M., Damuth, J., Frank, D. C.,
Mosbrugger, V., et al. (2010b). Ecometrics: the traits that bind the
past and present together. Integrative Zoology, 5, 88–101.
Feild, T. S., Brodribb, T. J., Iglesias, A., Chatelet, D. S., Baresch, A.,
Upchurch, G. R., et al. (2011a). Fossil evidence for Cretaceous
escalation in angiosperm leaf vein evolution. Proceedings of the
National Academy of Sciences, USA, 108, 8363–8366.
Feild, T. S., Sage, T. L., Czerniak, C., & Iles, W. J. D. (2005).
Hydathodal leaf teeth of Chloranthus japonicus (Chloranthaceae)
prevent guttation-induced flooding of the mesophyll. Plant, Cell &
Environment, 28, 1179–1190.
Feild, T. S., Upchurch Jr, G. R., Chatelet, D. S., Brodribb, T. J.,
Grubbs, K. C., Samain, M.-S., et al. (2011b). Fossil evidence for
low gas exchange capacities for early Cretaceous angiosperm
leaves. Paleobiology, 37, 195–213.
Fletcher, T. L., Greenwood, D. R., Moss, P. T., & Salisbury, S. W.
(2014). Paleoclimate of the Late Cretaceous (Cenomanian–Turo-
nian) portion of the Winton Formation, central-western Queensland,
Australia: new observations based on CLAMP and bioclimatic
analysis. PALAIOS, 29, 121–128.
Flynn, A. G., Peppe D. J., Abbuhl, B., & Williamson, T. E. (2016).
Paleoclimate and Paleoecology of Diverse Early Paleocene Fossil
Flora from the San Juan Basin, New Mexico, USA. XIV Interna-
tional Palynologival Congress –X International Organisation of
Paleobotany Conference.
Fricke, H. C., & Wing, S. L. (2004). Oxygen isotope and paleobotan-
ical estimates of temperature and d
18
O-latitude gradients over North
America during the Early Eocene. American Journal of Science,
304, 612–635.
Gates, D. M. (1980). Biophysical ecology. New York: Springer-Verlag.
Givnish, T. J. (1984). Leaf and canopy adaptations in tropical forests.
In E. Medina, H. A. Mooney & C. Vasquez-Yanes (Eds.),
Physiological ecology of plants of the wet tropics (pp. 51–84).
The Hague: Junk.
312 D. J. Peppe et al.
Green, W. A. (2006). Loosening the CLAMP: an exploratory graphical
approach to the Climate Leaf Analysis Multivariate Program.
Palaeontologia Electronica, 9, 1–17.
Green, W. A., Little, S. A., Price, C. A., Wing, S. L., Smith, S. Y.,
Kotrc, B., et al. (2014). Reading the leaves: a comparison of leaf
rank and automated areole measurement for quantifying aspects of
leaf venation. Applications in Plant Sciences, 2, 1400006.
Greenwood, D. R. (1991). The Taphonomy of plant macrofossils. Ch.
7. In S. K. Donovan (Ed.), The processes of fossilization (pp. 141–
169). London: Belhaven Press.
Greenwood, D. R. (1992). Taphonomic constraints on foliar physiog-
nomic interpretations of late Cretaceous and Tertiary paleoclimates.
Review of Palaeobotany and Palynology, 71, 149–190.
Greenwood, D. R. (1994). Palaeobotanical evidence for Australian
Tertiary climates. In R. S. Hill (Ed.), History of the Australian
vegetation: Cretaceous to recent (pp. 44–59). Cambridge: Cam-
bridge University Press.
Greenwood, D. R. (2005). Leaf margin analysis: taphonomic con-
straints. PALAIOS, 20, 498–505.
Greenwood, D. R. (2007). Fossil angiosperm leaves and climate: from
Wolfe and Dilcher to Burnham and Wilf. Advances in Angiosperm
Paleobotany and Paleoclimatic Reconstruction, 258, 95–108.
Greenwood, D. R., Moss, P. T., Rowett, A. I., Vadala, A. J., & Keefe,
R. L. (2003). Plant communities and climate change in southeastern
Australia during the early Paleogene. Special Papers-Geological
Society of America, 365–380.
Greenwood, D. R., Archibald, S. B., Mathewes, R. W., & Moss, P. T.
(2005). Fossil biotas from the Okanagan Highlands, southern
British Columbia and northeastern Washington state: climates and
ecosystems across an Eocene landscape. Canadian Journal of Earth
Sciences, 42, 167–185.
Greenwood, D. R., Wilf, P., Wing, S., & Christophel, D. C. (2004).
Paleotemperature estimation using leaf-margin analysis: is Australia
different? PALAIOS, 19, 129–142.
Greenwood, D. R., & Wing, S. L. (1995). Eocene continental climates
and latitudinal temperature gradients. Geology, 23, 1044–1048.
Gregory-Wodzicki, K. M. (2000). Relationship between leaf morphol-
ogy and climate, Bolivia: implications for estimating paleoclimate
from fossil floras. Paleobiology, 26, 668–688.
Gregory, K. (1994). Palaeoclimate and palaeoelevation of the 35 Ma
Florissant flora, Front Range, Colorado. Palaeoclimates, 1, 23–57.
Gregory, K. M., & McIntosh, W. C. (1996). Paleoclimate and
paleoelevation of the Oligocene Pitch-Pinnacle flora, Sawatch Range,
Colorado. Geological Society of America Bulletin, 108, 546–561.
Grichuk, V. (1969). An attempt to reconstruct certain elements of the
climate of the Northern Hemisphere in the Atlantic Period of the
Holocene. In M. I. Neishtadt (Ed.), Golosten (pp. 41–57). Moscow:
8th INQUA Congress, Izd-vo Nauka.
Grimm, G. W., & Denk, T. (2012). Reliability and resolution of the
Coexistence Approach—a revalidation using modern-day data.
Review of Palaeobotany and Palynology, 172, 33–47.
Grimm, G. W., & Potts, A. J. (2016). Fallacies and fantasies: the
theoretical underpinnings of the Coexistence Approach for palaeo-
climate reconstruction. Climate of the Past, 12, 611–622.
Grimm, G. W., Bouchal, J. M., Denk, T., & Potts, A. (2016). Fables
and foibles: a critical analysis of the Palaeoflora database and the
Coexistence approach for palaeoclimate reconstruction. Review of
Palaeobotany and Palynology, 233, 216–235.
Hall, J. B., & Swaine, M. D. (1981). Distribution and ecology of
vascular plants in a tropical rain forest: Forest vegetation in Ghana.
Dordrecht: Springer Science+Business Media.
Halloy, S. R. P., & Mark, A. F. (1996). Comparative leaf morphology
spectra of plant communities in New Zealand, the Andes and the
European Alps. Journal of the Royal Society of New Zealand, 26,
41–78.
Harbert, R. S., & Nixon, K. C. (2015). Climate reconstruction analysis
using coexistence likelihood estimation (CRACLE): a method for
the estimation of climate using vegetation. American Journal of
Botany, 102, 1277–1289.
Heer, O. (1870). Die Miocene flora und fauna Spitzbergens. Flora
Fossilis Artica, 1–3, 1–98.
Heer, O. (1878a). Florae fossilis Sachalinensis. Flora Fossilis Artica,
4–5, 1–61.
Heer, O. (1878b). Fossilen flora Sibiriens und des Amurlandes. Flora
Fossilis Artica, 4–5, 1–58.
Heer, O. (1882). Die fossile flora Gronlands, Erster Theil. Flora
Fossilis Artica, 6, 1–147.
Hickey, L. J. (1977). Stratigraphy and paleobotany of the Golden
Valley Formation (Early Tertiary) of western North Dakota.
Geological Society of America Memoir, 150, 1–183.
Hickey, L. J., Johnson, K. R., & Dawson, M. R. (1988). The
stratigraphy, sedimentology, and fossils of the Haughton Formation:
a post‐impact crater‐fill, Devon Island, NWT, Canada. Meteoritics,
23, 221–231.
Hijmans, R. J., Cameron, S. E., Parra, J. L., Jones, P. G., & Jarvis, A.
(2005). Very high resolution interpolated climate surfaces for global
land areas. International Journal of Climatology, 25, 1965–1978.
Hinojosa, L. F., Armesto, J. J., & Villagran, C. (2006a). Are Chilean
coastal forests pre-Pleistocene relicts? Evidence from foliar phys-
iognomy, palaeoclimate, and phytogeography. Journal of Biogeog-
raphy, 33, 331–341.
Hinojosa, L. F., Pesce, O., Yabe, A., Uemura, K., & Nishida, H.
(2006b). Physiognomical analysis and paleoclimate of the Ligorio
Márquez fossil flora, Ligorio Márquez Formation, 46°45′s, Chile.
In H. Nishida (Ed.), Post-Cretaceous floristic changes in southern
Patagonia, Chile (pp. 45–55). Tokyo: Chuo University Press.
Hinojosa, L. F., Pérez, F., Gaxiola, A., & Sandoval, I. (2011).
Historical and phylogenetic constraints on the incidence of entire
leaf margins: insights from a new South American model. Global
Ecology and Biogeography, 20, 380–390.
Holdridge, L. R. (1967). Life zone ecology. San Jose, Costa Rica:
Tropical Science Center.
Huff, P. M., Wilf, P., & Azumah, E. J. (2003). Digital future for
paleoclimate estimation from fossil leaves? Preliminary results.
PALAIOS, 18, 266–274.
Jacobs, B. F. (1999). Estimation of rainfall variables from leaf
characters in tropical Africa. Palaeogeography, Palaeoclimatology,
Palaeoecology, 145, 231–250.
Jacobs, B. F. (2002). Estimation of low-latitude paleoclimates using
fossil angiosperm leaves: examples from the Miocene Tugen Hills,
Kenya. Paleobiology, 28, 399–421.
Jacobs, B. F., & Deino, A. L. (1996). Test of climate-leaf physiognomy
regression models, their application to two Miocene floras from
Kenya, and
40
Ar/
39
Ar dating of the Late Miocene Kapturo site.
Palaeogeography,Palaeoclimatology, Palaeoecology,123, 259–271.
Jacobs, B. F., & Herendeen, P. S. (2004). Eocene dry climate and
woodland vegetation in tropical Africa reconstructed from fossil
leaves from northern Tanzania. Palaeogeography, Palaeoclimatol-
ogy, Palaeoecology, 213, 115–123.
Jackson, S. T., & Overpeck, J. T. (2000). Responses of plant
populations and communities to environmental changes of the late
Quaternary. Paleobiology, 26, 194–220.
Jacques, F. M., Su, T., Spicer, R. A., Xing, Y., Huang, Y., Wang, W.,
et al. (2011). Leaf physiognomy and climate: are monsoon systems
different? Global and Planetary Change, 76, 56–62.
Jones, C. S., Bakker, F. T., Schlichting, C. D., & Nicotra, A. B. (2009).
Leaf shape evolution in the South African genus Pelargonium l’her.
(Geraniaceae). Evolution, 63, 479–497.
Jordan, G. J. (1997). Uncertainty in palaeoclimatic reconstructions based
on leaf physiognomy. Australian Journal of Botany, 45, 527–547.
13 Paleoclimate and Paleoecology Using Fossil Leaves 313
Jordan, G. J. (2011). A critical framework for the assessment of
biological palaeoproxies: predicting past climate and levels of
atmospheric CO
2
from fossil leaves. New Phytologist, 192, 29–44.
Karst, A. L., & Lechowicz, M. J. (2007). Are correlations among foliar
traits in ferns consistent with those in the seed plants? New
Phytologist, 173, 306–312.
Kelly, A. E., & Goulden, M. L. (2008). Rapid shifts in plant
distribution with recent climate change. Proceedings of the National
Academy of Sciences, USA, 105, 11823–11826.
Kennedy, E. M. (1998). Cretaceous and Tertiary megafloras from New
Zealand and their climate signals. Ph.D. Dissertation, Open
University.
Kennedy, E. M., Arens, N. C., Reichgelt, T., Spicer, R. A., Spicer, T.
E., Stranks, L., et al. (2014). Deriving temperature estimates from
Southern Hemisphere leaves. Palaeogeography, Palaeoclimatol-
ogy, Palaeoecology, 412, 80–90.
Kennedy, E. M., Spicer, R. A., & Rees, P. M. (2002). Quantitative
palaeoclimate estimates from the late Cretaceous and Paleocene leaf
floras in the northwest of the South Island, New Zealand. Palaeo-
geography, Palaeoclimatology, Palaeoecology, 184, 321–345.
Kershaw, A. (1997). A bioclimatic analysis of early to middle Miocene
Brown Coal floras, Latrobe Valley, south-eastern Australia.
Australian Journal of Botany, 45, 373–387.
Kershaw, A., & Nix, H. (1988). Quantitative palaeoclimatic estimates
from pollen data using bioclimatic profiles of extant taxa. Journal of
Biogeography, 589–602.
Khan, M. A., Spicer, R. A., Bera, S., Ghosh, R., Yang, J., Spicer, T.
E. V., et al. (2014). Miocene to Pleistocene floras and climate of the
eastern Himalayan Siwaliks, and new palaeoelevation estimates for
the Namling-Oiyug Basin. Tibet. Global and Planetary Change,
113, 1–10.
Kobayashi, H., Kresling, B., & Vincent, J. F. (1998). The geometry of
unfolding tree leaves. Proceedings of the Royal Society of London
B: Biological Sciences, 265, 147–154.
Kovach, W. L., & Spicer, R. (1996). Canonical correspondence
analysis of leaf physiognomy: a contribution to the development of
a new palaeoclimatological tool. Paleoclimates, 2, 125–138.
Kowalski, E. A. (2002). Mean annual temperature estimation based on
leaf morphology: a test from tropical South America. Palaeogeog-
raphy, Palaeoclimatology, Palaeoecology, 188, 141–165.
Kowalski, E. A., & Dilcher, D. L. (2003). Warmer paleotemperatures
for terrestrial ecosystems. Proceedings of the National Academy of
Sciences, USA, 100, 167–170.
Küchler, A. W. (1964). Potential natural vegetation of the contermi-
nous United States. New York: American Geographical Society.
Labandeira, C. C., & Currano, E. D. (2013). The fossil record of
plant-insect dynamics. Annual Review of Earth and Planetary
Sciences, 41, 287–311.
Larcher, W., & Winter, A. (1981). Frost susceptibility of palms:
Experimental data and their interpretations. Principes, 25, 143–152.
Lawing, A. M., Eronen, J. T., Blois, J. L., Graham, C. H., & Polly, P. D.
(2016). Community functional trait composition at the continental
scale: The effects of non-ecological processes. Ecography, 13, 1–13.
Lawing, A. M., Head, J. J., & Polly, P. D. (2012). The ecology of
morphology: the ecometrics of locomotion and macroenvironment
in north american snakes. In J. Louys (Ed.), Paleontology in
ecology and conservation (pp. 117–146). Heidelberg: Springer.
Li, L., McCormack, M. L., Ma, C., Kong, D., Zhang, Q., Chen, X., et al.
(2015). Leaf economics and hydraulic traits are decoupled in five
species-rich tropical-subtropicalforests. Ecology letters, 18, 899–906.
Liang, M.-M., Bruch, A., Collinson, M., Mosbrugger, V., Li, C.-S.,
Sun, Q.-G., et al. (2003). Testing the climatic estimates from
different palaeobotanical methods: an example from the middle
Miocene Shanwang flora of China. Palaeogeography, Palaeocli-
matology, Palaeoecology, 198, 279–301.
Liao, F., Peng, J., & Chen, R. (2017). LeafletAnalyzer, an automated
software for quantifying, comparing and classifying blade and
serration features of compound leaves during development, and
among induced mutants and natural variants in the legume
Medicago truncatula.Frontiers in plant science, 8, 915.
Lielke, K., Manchester, S., & Meyer, H. (2012). Reconstructing the
environment of the northern Rocky Mountains during the
Eocene/Oligocene transition: Constraints from the palaeobotany
and geology of south-western Montana, USA. Acta Palaeobotanica,
52, 317–358.
Little, S. A., Kembel, S. W., & Wilf, P. (2010). Paleotemperature
proxies from leaf fossils reinterpreted in light of evolutionary
history. PLoS ONE, 5, e15161.
MacGinitie, H. D. (1953). Fossil plants of the Florissant Beds.
Colorado: Carnegie Institution of Washington.
MacLeod, N., & Steart, D. (2015). Automated leaf physiognomic
character identification from digital images. Paleobiology, 41, 528–
553.
Malhado, A., Malhi, Y., Whittaker, R., Ladle, R., Ter Steege, H.,
Phillips, O., et al. (2009). Spatial trends in leaf size of Amazonian
rainforest trees. Biogeosciences, 6, 1563–1576.
McDonald, P. G., Fonseca, C. R., Overton, J. M., & Westoby, M.
(2003). Leaf-size divergence along rainfall and soil-nutrient gradi-
ents: Is the method of size reduction common among clades?
Functional Ecology, 17, 50–57.
McElwain, J. C., Montañez, I., White, J. D., Wilson, J. P., & Yiotis, C.
(2016). Was atmospheric CO
2
capped at 1000 ppm over the past
300 million years? Palaeogeography, Palaeoclimatology, Palaeoe-
cology, 441, 653–658.
Merkhofer, L., Wilf, P., Haas, M. T., Kooyman, R. M., Sack, L.,
Scoffoni, C., et al. (2015). Resolving Australian analogs for an
Eocene Patagonian paleorainforest using leaf size and floristics.
American Journal of Botany, 102, 1160–1173.
Merriam, C. H. (1894). Laws of temperature control of the geographic
distribution of terrestrial animals and plants. National Geographic
Magazine, 6, 229–238.
Michaletz, S. T., Weiser, M. D., Zhou, J., Kaspari, M., Helliker, B. R.,
& Enquist, B. J. (2015). Plant thermoregulation: energetics, trait–
environment interactions, and carbon economics. Trends in Ecology
& Evolution, 30, 714–724.
Michel, L. A., Peppe, D. J., Lutz, J. A., Driese, S. G., Dunsworth, H.
M., Harcourt-Smith, W. E., et al. (2014). Remnants of an ancient
forest provide ecological context for early Miocene fossil apes.
Nature Communications, 5, 3236.
Miller, I. M., Brandon, M. T., & Hickey, L. J. (2006). Using leaf margin
analysis to estimate mid-Cretaceous (Albian) paleolatitude of the
Baja BC block. Earth and Planetary Science Letters, 245, 95–114.
Mooney, H. A., Björkman, O., & Collatz, G. J. (1978). Photosynthetic
acclimation to temperature in the desert shrub, Larrea divaricata I.
Carbon dioxide exchange characteristics of intact leaves. Plant
Physiology, 61, 406–410.
Monteith, J., & Unsworth, M. (2007). Principles of environmental
physics, Academic Press.
Mosbrugger, V., & Utescher, T. (1997). The coexistence approach –a
method for quantitative reconstructions of tertiary terrestrial
palaeoclimate data using plant fossils. Palaeogeography, Palaeo-
climatology, Palaeoecology, 134, 61–86.
New, M., Lister, D., Hulme, M., & Makin, I. (2002). A high-resolution
data set of surface climate over global land areas. Climate research,
21, 1–25.
Parkhurst, D. F., & Loucks, O. (1972). Optimal leaf size in relation to
environment. The Journal of Ecology, 60, 505–537.
Peppe, D. J., Hickey, L. J., Miller, I. M., & Green, W. A. (2008).
A morphotype catalogue, floristic analysis, and stratigraphic
description of the Aspen Shale flora (Cretaceous-Albian) of
314 D. J. Peppe et al.
southwestern Wyoming. Bulletin of the Peabody Museum of
Natural History, 49, 181–208.
Peppe, D. J., Lemons, C. R., Royer, D. L., Wing, S. L., Wright, I. J.,
Lusk, C. H., et al. (2014). Biomechanical and leaf-climate
relationships: a comparison of ferns and seed plants. American
Journal of Botany, 101, 338–347.
Peppe, D. J., Royer, D. L., Cariglino, B., Oliver, S. Y., Newman, S.,
Leight, E., et al. (2011). Sensitivity of leaf size and shape to climate:
global patterns and paleoclimatic applications. New Phytologist,
190, 724–739.
Peppe, D. J., Royer, D. L., Wilf, P., & Kowalski, E. A. (2010).
Quantification of large uncertainties in fossil leaf paleoaltimetry.
Tectonics, 29, TC3015.
Pérez-Pérez, J. M., Esteve-Bruna, D., & Micol, J. L. (2010). QTL
analysis of leaf architecture. Journal of Plant Research, 123, 15–23.
Polly, P. D., Eronen, J. T., Fred, M., Dietl, G. P., Mosbrugger, V.,
Scheidegger, C., et al. (2011). History matters: Ecometrics and
integrative climate change biology. Proceedings of the Royal
Society of London B,Biological Sciences, 278, 1131–1140.
Polly, P. D., & Sarwar, S. (2014). Extinction, extirpation, and exotics:
Effects on the correlation between traits and environment at the
continental level. Annales Zoologici Fennici, 51, 209–226.
Price, C. A., Wing, S., & Weitz, J. S. (2012). Scaling and structure
of dicotyledonous leaf venation networks. Ecology letters, 15, 87–95.
Pross, J., Contreras, L., Bijl, P. K., Greenwood, D. R., Bohaty, S. M.,
Schouten, S., et al. (2012). Persistent near-tropical warmth on the
Antarctic continent during the early Eocene epoch. Nature, 488,
73–77.
Rakotoarinivo, M., Blach-Overgaard, A., Baker, W. J., Dransfield, J.,
Moat, J., & Svenning, J.-C. (2013). Palaeo-precipitation is a major
determinant of palm species richness patterns across Madagascar: a
tropical biodiversity hotspot. Proceedings of the Royal Society of
London B, Biological Sciences, 280, 20123048.
Reich, P. B. (2014). The world-wide ‘fast–slow’plant economics
spectrum: A traits manifesto. Journal of Ecology, 102, 275–301.
Reich, P. B., Ellsworth, D. S., Walters, M. B., Vose, J. M., Gresham,
C., Volin, J. C., et al. (1999). Generality of leaf trait relationships: a
test across six biomes. Ecology, 80, 1955–1969.
Reich, P. B., Walters, M. B., & Ellsworth, D. S. (1997). From tropics to
tundra: global convergence in plant functioning. Proceedings of the
National Academy of Sciences, USA, 94, 13730–13734.
Reichgelt, T., Kennedy, E. M., Mildenhall, D. C., Conran, J. G.,
Greenwood, D. R., & Lee, D. E. (2013). Quantitative palaeoclimate
estimates for Early Miocene southern New Zealand: evidence from
Foulden Maar. Palaeogeography, Palaeoclimatology, Palaeoecol-
ogy, 378, 36–44.
Rodriguez, R. E., Debernardi, J. M., & Palatnik, J. F. (2014).
Morphogenesis of simple leaves: regulation of leaf size and shape.
Wiley Interdisciplinary Reviews: Developmental Biology, 3, 41–57.
Royer, D. L. (2012a). Climate reconstruction from leaf size and shape:
New developments and challenges. The Paleontological Society
Papers, 18, 195–212.
Royer, D. L. (2012b). Leaf shape responds to temperature but not CO
2
in Acer rubrum.PLoS ONE, 7, e49559.
Royer, D. L., Kooyman, R. M., Little, S. A., & Wilf, P. (2009a).
Ecology of leaf teeth: a multi-site analysis from an Australian
subtropical rainforest. American Journal of Botany, 96, 738–750.
Royer, D. L., McElwain, J. C., Adams, J. M., & Wilf, P. (2008).
Sensitivity of leaf size and shape to climate within Acer rubrum and
Quercus kelloggii.New Phytologist, 179, 808–817.
Royer, D. L., Meyerson, L. A., Robertson, K. M., & Adams, J. M.
(2009b). Phenotypic plasticity of leaf shape along a temperature
gradient in Acer rubrum.PLoS ONE, 4, e7653.
Royer, D. L., Miller, I. M., Peppe, D. J., & Hickey, L. J. (2010). Leaf
economic traits from fossils support a weedy habit for early
angiosperms. American Journal of Botany, 97, 438–445.
Royer, D. L., Peppe, D. J., Wheeler, E. A., & Niinemets, U. (2012).
Roles of climate and functional traits in controlling toothed
vs. untoothed leaf margins. American Journal of Botany, 99,
915–922.
Royer, D. L., Sack, L., Wilf, P., Lusk, C. H., Jordan, G. J., Niinemets,
U., et al. (2007). Fossil leaf economics quantified: calibra-
tion, Eocene case study, and implications. Paleobiology, 33,
574–589.
Royer, D. L., & Wilf, P. (2006). Why do toothed leaves correlate with
cold climates? Gas exchange at leaf margins provides new insights
into a classic paleotemperature proxy. International Journal of
Plant Sciences, 167, 11–18.
Royer, D. L., Wilf, P., Janesko, D. A., Kowalski, E. A., & Dilcher, D.
L. (2005). Correlations of climate and plant ecology to leaf size and
shape: Potential proxies for the fossil record. American Journal of
Botany, 92, 1141–1151.
Sack, L., & Frole, K. (2006). Leaf structural diversity is related to
hydraulic capacity in tropical rain forest trees. Ecology, 87,
483–491.
Sack, L., & Scoffoni, C. (2013). Leaf venation: structure, function,
development, evolution, ecology and applications in the past,
present and future. New Phytologist, 198, 983–1000.
Sack, L., Scoffoni, C., John, G. P., Poorter, H., Mason, C. M.,
Mendez-Alonzo, R., et al. (2013). How do leaf veins influence the
worldwide leaf economic spectrum? Review and synthesis. Journal
of Experimental Botany, 64, 4053–4080.
Sack, L., Scoffoni, C., John, G. P., Poorter, H., Mason, C. M.,
Mendez-Alonzo, R., et al. (2014). Leaf mass per area is independent
of vein length per area: avoiding pitfalls when modelling pheno-
typic integration (reply to). Journal of Experimental Botany, 65,
5115–5123.
Sack, L., Scoffoni, C., McKown, A. D., Frole, K., Rawls, M., Havran,
J. C., et al. (2012). Developmentally based scaling of leaf venation
architecture explains global ecological patterns. Nature Communi-
cations, 3, 837.
Sack, L., Tyree, M. T., & Holbrook, N. M. (2005). Leaf hydraulic
architecture correlates with regeneration irradiance in tropical
rainforest trees. New Phytologist, 167, 403–413.
Salvucci, M. E., & Crafts-Brandner, S. J. (2004). Relationship between
the heat tolerance of photosynthesis and the thermal stability of
Rubisco activase in plants from contrasting thermal environments.
Plant Physiology, 134, 1460–1470.
Schmerler, S. B., Clement, W. L., Beaulieu, J. M., Chatelet, D. S.,
Sack, L., Donoghue, M. J., et al. (2012). Evolution of leaf form
correlates with tropical–temperate transitions in Viburnum (Adox-
aceae). Proceedings of the Royal Society B, Biological Sciences,
279, 3905–3913.
Scoffoni, C., Rawls, M., McKown, A., Cochard, H., & Sack, L. (2011).
Decline of leaf hydraulic conductance with dehydration: relation-
ship to leaf size and venation architecture. Plant Physiology, 156,
832–843.
Sharpe, S. E. (2002). Constructing seasonal climograph overlap
envelopes from Holocene packrat midden contents, Dinosaur
National Monument, Colorado. Quaternary Research, 57, 306–313.
Shipley, B., Lechowicz, M. J., Wright, I., & Reich, P. B. (2006).
Fundamental trade-offs generating the worldwide leaf economics
spectrum. Ecology, 87, 535–541.
Sinka, K., & Atkinson, T. (1999). A mutual climatic range method for
reconstructing palaeoclimate from plant remains. Journal of the
Geological Society, 156, 381–396.
Smith, R. Y., Basinger, J. F., & Greenwood, D. R. (2012). Early
Eocene plant diversity and dynamics in the Falkland flora,
Okanagan Highlands, British Columbia, Canada. Palaeobiodiver-
sity and Palaeoenvironments, 92, 309–328.
Spicer, R. A. (2007). Recent and future developments of CLAMP:
building on the legacy of Jack A. Wolfe. Courier Forschungsin-
stitut Senckenberg, 258, 109–118.
13 Paleoclimate and Paleoecology Using Fossil Leaves 315
Spicer, R. A. (2016). CLAMP, climate leaf analysis multivariate
program, The Open University.
Spicer, R. A., Herman, A. B., & Kennedy, E. M. (2004). Foliar
physiognomic record of climatic conditions during dormancy:
climate leaf analysis multivariate program (CLAMP) and the cold
month mean temperature. The Journal of Geology, 112, 685–702.
Spicer, R. A., Herman, A. B., & Kennedy, E. M. (2005). The sensitivity
of CLAMP to taphonomic loss of foliar physiognomic characters.
PALAIOS, 20, 429–438.
Spicer, R. A., Valdes, P. J., Spicer, T. E. V., Craggs, H. J., Srivastava,
G., Mehrotra, R. C., et al. (2009). New developments in CLAMP:
calibration using global gridded meteorological data. Palaeogeog-
raphy, Palaeoclimatology, Palaeoecology, 283, 91–98.
Spicer, R. A., & Yang, J. (2010). Quantification of uncertainties in fossil
leaf paleoaltimetry: Does leaf size matter? Tectonics, 29, TC6001.
Steart, D. C., Spicer, R. A., & Bamford, M. K. (2011). Is southern
Africa different? An investigation of the relationship between leaf
physiognomy and climate in southern African mesic vegetation.
Review of Palaeobotany and Palynology, 162, 607–620.
Stranks, L., & England, P. (1997). The use of a resemblance function in
the measurement of climatic parameters from the physiognomy of
woody dicotyledons. Palaeogeography, Palaeoclimatology,
Palaeoecology, 131, 15–28.
Su, T., Xing, Y. W., Liu, Y. S., Jacques, F. M. B., Chen, W. Y., Huang,
Y. S., et al. (2010). Leaf margin analysis: a new equation from
humid to mesic forests in China. PALAIOS, 25, 234–238.
Svenning, J.-C., & Sandel, B. (2013). Disequilibrium vegetation
dynamics under future climate change. American Journal of
Botany, 100, 1266–1286.
Teodoridis, V., Mazouch, P., Spicer, R. A., & Uhl, D. (2011). Refining
CLAMP—investigations towards improving the Climate Leaf
Analysis Multivariate Program. Palaeogeography, Palaeoclimatol-
ogy, Palaeoecology, 299, 39–48.
ter Braak, C. J. F. (1986). Canonical correspondence analysis: A new
eigenvector technique for multivariate direct gradient analysis.
Ecology, 67, 1167–1179.
Thompson, R., Anderson, K., Pelltier, R., Strickland, L., Bartlein, P., &
Shafer, S. L. (2012). Quantitative estimation of climatic parameters
from vegetation data in North America by the mutual climatic range
technique. Quaternary Science Reviews, 51, 18–39.
Traiser, C., Klotz, S., Uhl, D., & Mosbrugger, V. (2005). Environ-
mental signals from leaves–a physiognomic analysis of European
vegetation. New Phytologist, 166, 465–484.
Upchurch, G., & Wolfe, J. (1987). Mid-Cretaceous to Early Tertiary
vegetation and climate: Evidence from fossil leaves and woods.
In E. M. Friis, W. Chaloner & P. R. Crane (Eds.), Origins of
angiosperms and their biological consequences (pp. 75–106).
Cambridge: Cambridge University Press.
Utescher, T., Bruch, A. A., Erdei, B., François, L., Ivanov, D., Jacques,
F. M. B., et al. (2014). The coexistence approach—theoretical
background and practical considerations of using plant fossils for
climate quantification. Palaeogeography, Palaeoclimatology,
Palaeoecology, 410, 58–73.
Vaughan, J. (2015). Investigating the genetic basis of natural leaf shape
variation in Arabidopsis thaliana, Ph.D. Dissertation, University of
York.
Veloz, S. D., Williams, J. W., Blois, J. L., He, F., Otto-Bliesner, B., & Liu,
Z. (2012). No-analog climates and shifting realized niches during the
late Quaternary: implications for 21
st
-Century predictions by species
distribution models. Global Change Biology, 18, 1698–1713.
Vogel, S. (1970). Convective cooling at low airspeeds and the shapes of
broad leaves. Journal of Experimental Botany, 21, 91–101.
Vogel, S. (2009). Leaves in the lowest and highest winds: temperature,
force and shape. New Phytologist, 183, 13–26.
Von Humboldt, A., & Bonpland, A. (1807). Essai sur la géographie
des plantes : accompagnéd'un tableau physique des régions
équinoxiales, fondésur des mesures exécutées, depuis le dixième
degréde latitude boréale jusqu'au dixième degréde latitude
australe, pendant les années 1799, 1800, 1801, 1802 et 1803. Paris:
Chez Levrault, Schoell et compagnie. Translated by Sylvie
Romanowski as Essay on the geography of plants. In S. T. Jackson
(Ed.). (2009). Essay on the geography of plants (pp. 57–143).
Chicago: University of Chicago Press.
Wahid, A., Gelani, S., Ashraf, M., & Foolad, M. R. (2007). Heat
tolerance in plants: an overview. Environmental and Experimental
Botany, 61(3): 199–223.
Webb, L. J. (1959). A physiognomic classification of Australian rain
forests. Journal of Ecology, 47, 551–570.
Webb, L. J. (1968). Environmental relationships of structural types
Australian rain forest vegetation. Ecology, 49, 296–311.
West, C. K., Greenwood, D. R., & Basinger, J. F. (2015). Was the
Arctic Eocene ‘rainforest’monsoonal? Estimates of seasonal
precipitation from early Eocene megafloras from Ellesmere Island,
Nunavut. Earth & Planetary Science Letters, 427, 18–30.
Westoby, M., Falster, D. S., Moles, A. T., Vesk, P. A., & Wright, I.
J. (2002). Plant ecological strategies: some leading dimensions of
variation between species. Annual Review of Ecology and System-
atics, 33, 125–159.
Whittaker, R. H. (1975). Communities and ecosystems. New York:
Macmillan.
Wiemann, M. C., Manchester, S. R., Dilcher, D. L., Hinojosa, L. F., &
Wheeler, E. A. (1998). Estimation of temperature and precipitation
from morphological characters of dicotyledonous leaves. American
Journal of Botany, 85, 1796–1802.
Wilf, P. (1997). When are leaves good thermometers? A new case for
leaf margin analysis. Paleobiology, 23, 373–390.
Wilf, P. (2008). Fossil angiosperm leaves: paleobotany’s difficult
children prove themselves. Paleontological Society Papers, 14,
319–333.
Wilf, P., Wing, S. L., Greenwood, D. R., & Greenwood, C. L. (1998).
Using fossil leaves as paleoprecipitation indicators: an Eocene
example. Geology, 26, 203–206.
Wilf, P., Wing, S. L., Greenwood, D. R., & Greenwood, C. L. (1999).
Using fossil leaves as paleoprecipitation indicators: an Eocene
example: Reply. Geology, 27, 91–92.
Wilf, P., Labandeira, C. C., Johnson, K. R., & Ellis, B. (2006).
Decoupled plant and insect diversity after the end-Cretaceous
extinction event. Science, 313, 1112–1115.
Wilf, P., Little, S. A., Iglesias, A., del Carmen Zamaloa, M., Gandolfo, M.
A., Cuneo, N. R., & Johnson, K. R. (2009). Papuacedrus (Cupres-
saceae) in Eocene Patagonia: a new fossil link to Australasian
rainforests. American Journal of Botany, 96(11): 2031–2047.
Wilf, P., Zhang, S., Chikkerur, S., Little, S. A., Wing, S. L., & Serre, T.
(2016). Computer vision cracks the leaf code. Proceedings of the
National Academy of Sciences, USA, 113, 3305–3310.
Wing, S., Bao, H., & Koch, P. L. (2000). An Eocene cool period?
Evidence for continental cooling during the warmest part of the
Cenozoic. In B. T. Huber, K. G. MacCleod & S. Wing (Eds.), Warm
climates in Earth history (pp. 197–237). Cambridge, UK: Cam-
bridge University Press.
Wing, S., & Greenwood, D. R. (1993). Fossils and fossil climate: the
case for equable continental interiors in the Eocene. Philosophical
Transactions of the Royal Society of London Series B, 341, 243–252.
Wing, S. L., Herrera, F., Jaramillo, C. A., Gomez-Navarro, C., Wilf, P.,
& Labandeira, C. C. (2009). Late Paleocene fossils from the
Cerrejon Formation, Colombia, are the earliest record of neotropical
rainforest. Proceedings of the National Academy of Sciences, USA,
106, 18627–18632.
316 D. J. Peppe et al.
Wolfe, J., & Hopkins, D. (1967). Climatic changes recorded by
Tertiary land floras in northwestern North America. In K. Hatai
(Ed.), Tertiary correlations and climatic changes in the
Pacific(pp. 67–76). Tokyo, Japan: Eleventh Pacific Science
Congress.
Wolfe, J. A. (1971). Tertiary climatic fluctuations and methods of
analysis of Tertiary floras. Palaeogeography, Palaeoclimatology,
Palaeoecology, 9, 27–57.
Wolfe, J. A. (1978). A paleobotanical interpretation of Tertiary climates
in the Northern Hemisphere: data from fossil plants make it
possible to reconstruct Tertiary climatic changes, which may be
correlated with changes in the inclination of the Earth’s rotational
axis. American Scientist, 66, 694–703.
Wolfe, J. A. (1979). Temperature parameters of humid to mesic forests
of eastern Asia and relation to forests of other regions in the
Northern Hemisphere and Australasia. United States Geological
Survey Professional Paper, 1106, 1–37.
Wolfe, J. A. (1990). Palaeobotanical evidence for a marked temperature
increase following the Cretaceous/Tertiary boundary. Nature, 343,
153–156.
Wolfe, J. A. (1993). A method of obtaining climatic parameters from
leaf assemblages. United States Geological Survey Bulletin, 2040,
1–71.
Wolfe, J. A. (1995). Paleoclimatic estimates from Tertiary leaf
assemblages. Annual Review of Earth and Planetary Sciences, 23,
119–142.
Wolfe, J. A., & Spicer, R. A. (1999). Fossil leaf character states;
multivariate analyses. In T. P. Jones & N. P. Rowe (Eds.), Fossil
plants and spores: Modern techniques (pp. 233–239). London,
United Kingdon: Geological Society.
Wright, I. J., Reich, P. B., Cornelissen, J. H. C., Falster, D. S., Groom,
P. K., Hikosaka, K., et al. (2005). Modulation of leaf economic
traits and trait relationships by climate. Global Ecology and
Biogeography, 14, 411–421.
Wright, I. J., Reich, P. B., Westoby, M., Ackerly, D. D., Baruch, Z.,
Bongers, F., et al. (2004). The worldwide leaf economics spectrum.
Nature, 428, 821–827.
Wright, I. J., Dong, N., Maire, V., Prentice, I. C., Westoby, M., Díaz,
S., et al. (2017). Global climatic drivers of leaf size. Science, 357,
917–921.
Yang, J., Spicer, R. A., Spicer, T. E., Arens, N. C., Jacques, F., Su, T.,
et al. (2015). Leaf form–climate relationships on the global stage:
an ensemble of characters. Global Ecology and Biogeography, 24,
1113–1125.
Yang, J., Spicer, R. A., Spicer, T. E., & Li, C.-S. (2011). ‘CLAMP
online’: a new web-based palaeoclimate tool and its application to
the terrestrial Paleogene and Neogene of North America. Palaeo-
biodiversity and Palaeoenvironments, 91, 163–183.
Zagwijn, W., & Hager, H. (1987). Correlations of continental and
marine Neogene deposits in the south-eastern Netherlands and the
lower Rhine District. Mededelingen van de Werkgroep voor
Tertiaire en Kwartaire Geologie, 24, 59–78.
13 Paleoclimate and Paleoecology Using Fossil Leaves 317
