Opinion
TREE-1676; No. of Pages 10How does pedogenesis drive plant
diversity?
Etienne Laliberte?1, James B. Grace2, Michael A. Huston3, Hans Lambers1,
Franc?ois P. Teste1, Benjamin L. Turner4, and David A. Wardle5
1School of Plant Biology, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
2US Geological Survey, National Wetlands Research Center, 700 Cajundome Boulevard, Lafayette, LO 70506, USA
3Department of Biology, Texas State University, San Marcos, TX 78666, USA
4Smithsonian Tropical Research Institute, Apartado 0843-03092, Balboa, Ancon, Republic of Panama
5Department of Forest Ecology and Management, Faculty of Forestry, Swedish University of Agricultural Sciences, SE901-83
Umea?, Sweden
Some of the most species-rich plant communities occur
on ancient, strongly weathered soils, whereas those on
recently developed soils tend to be less diverse. Mecha-
nisms underlying this well-known pattern, however,
remain unresolved. Here, we present a conceptual mod-
el describing alternative mechanisms by which pedo-
genesis (the process of soil formation) might drive plant
diversity. We suggest that long-term soil chronose-
quences offer great, yet largely untapped, potential as
?natural experiments? to determine edaphic controls
over plant diversity. Finally, we discuss how our con-
ceptual model can be evaluated quantitatively using
structural equation modeling to advance multivariate
theories about the determinants of local plant diversity.
This should help us to understand broader-scale diver-
sity patterns, such as the latitudinal gradient of plant
diversity.
Links between plant diversity and pedogenesis
One of the most striking global ecological patterns is the
increase in plant diversity from the poles to the equator [1].
This increase in diversity at lower latitudes coincides with
a global gradient in soil development, whereby lower-
latitude soils tend to be older and more strongly weathered
(see Glossary) compared with higher-latitude soils [2,3]
(Box 1). This raises the question: is plant diversity linked to
pedogenesis and, if so, what mechanisms are involved?
Identifying causal links between pedogenesis and plant
diversity across broad latitudinal gradients is problematic,
because many factors co-vary with latitude. However, the
global pattern of increasing plant diversity with pedogen-
esis is reproduced at local scales along well-defined soil-age
gradients from around the world [4,5] (Box 1). We propose
that these ?soil chronosequences? (Box 2) offer considerable
potential for evaluating multiple hypotheses for this in-
crease in plant diversity with soil age.
This article has three sections. First, we present a set of
hypotheses about how pedogenesis might drive local plant
species diversity (Figure 1). Second, we discuss how these
Glossary
Base saturation: proportion of negatively charged soil exchange sites
occupied by base cations [i.e., calcium (Ca2+), magnesium (Mg2+), K+, and
sodium (Na+)].
Horizon: soil layer, parallel to the soil surface, with physical, chemical, or
biological characteristics that are distinct from the layers above or below it,
and usually differentiated based on color, texture, and organic matter.
Humus: relatively stable soil organic matter that does not retain structural
characteristics of the plant or animal tissues from which it originates.
Indicator variable: observed variable that serves as an indirect measure of a
quantity of causal interest in a structural equation model [6].
Kaolinite: aluminosilicate clay mineral with low cation exchange capacity,
common in old, strongly weathered soils.
Mycorrhizal network: network of mycorrhizal fungi that connect plants together
via hyphal links [39].
Natural experiment: field study where variation in a driving factor (e.g., soil
nutrient availability or time since disturbance) is controlled via careful site
selection, rather than by imposed experimental treatments.
Nutrient stoichiometry: the relative abundance of plant elements in the envi-
ronment or in biomass. In particular, the N:P ratio is often emphasized, because
they are the two nutrients that most often limit terrestrial plant productivity.
More broadly, ecological stoichiometry is the balance of multiple chemical
substances that affect, and are affected by, organisms, and that influence
ecological interactions and processes [56].
Oxisol: an order in Soil Taxonomy, the US soil classification system. Oxisols are
defined as soils with an oxic horizon [i.e., containing <10% weatherable miner-
als in the fine sand fraction (50?200 mm) and a low cation-exchange capacity] or
a kandic horizon (clay enriched and low cation-exchange capacity) with >40%
clay in the surface soil. Oxisols represent soils that have undergone extreme
weathering, are low in nutrients, and occur predominantly in stable regions of
the lowland tropics.
Parent material: original material from which a soil is derived.
Pedogenesis: the process of soil formation, generally leading to the formation of
soil horizons. The five major soil-forming factors are: parent material, organ-
isms, topography, climate, and time [57].
Plant?soil feedback: process by which an individual plant modifies its local
abiotic or biotic soil properties, which in turn influences its own performance
and that of other plants.
Resource partitioning: occurs when two (or more) co-occurring individuals use
different resources (or different forms of the same resource), thereby reducing
competition between them.
Soil chronosequence: a series of soils that are derived from similar parent
material but differ in time since initiation of pedogenesis (Box 2). Variation in
other soil-forming factors (i.e., climate, topography, parent material, and organ-
isms) is minimized along a soil chronosequence [50,51]. In most soil chron-
osequences, sites have not experienced constant climate and vegetation
throughout their pedogenic development, but current climate, parent material,
and topography are usually well constrained.
Structural equation modeling (SEM): framework where a (generally multivari-
ate) hypothetical causal model is first specified and then evaluated quantita-
tively against data [6].
Weathering: physical, chemical, and biological processes through which less
stable (e.g., primary) minerals are converted to more stable minerals. In
relatively young soils, weathering is the main source of rock-derived nutrients
(e.g., P).0169-5347/$ ? see front matter
 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tree.2013.02.008
Corresponding author: Laliberte?, E. (etienne.laliberte@uwa.edu.au).Trends in Ecology & Evolution xx (2012) 1?10 1
Box 1. Long-term pedogenesis and local plant species diversity
Although there is a wide diversity of soils in both tropical and
temperate regions, the most strongly weathered soils in the world
(Oxisols) occur almost exclusively in the tropics, and underlie vast
areas of hyperdiverse lowland tropical forest. Conversely, many soils
at higher latitudes have been rejuvenated following recent glacia-
tions. The best known and most extreme cases of high local (i.e.,
alpha) plant diversity on strongly weathered soils come from tropical
rainforests [11] (Figure I). For example, lowland tropical rainforest in
Yasun?? National Park (Ecuador) can contain up to 644 tree species ha?1
[58]. Soils in the Yasun?? region are strongly weathered, very acidic,
kaolinitic, have high exchangeable aluminum (Al) concentrations and
very low base saturation [59]. Another example of hyperdiverse
tropical rainforest on strongly weathered soils in the Asian tropics is
the Lambir Hills National Park (Sarawak, Malaysia), where up to 610
tree species ha?1 can be found, with the greatest diversity on infertile
Ultisols [60].
High alpha diversity on strongly weathered soils is not, however,
restricted to tropical rainforests (Figure I). Some areas of the Brazilian
cerrado moist savannah biome show relatively high plant alpha
diversity of up to 120 tree or shrub species ha?1 [61]; cerrado soils are
deep, infertile Oxisols [62]. In addition, species-rich fynbos vegetation
of South Africa (up to 114 species 0.1 ha?1) [63] also occurs on ancient,
nutrient-impoverished soils. Similarly, some southwestern Australia
kwongan shrublands show high local species richness (up to 121
species 0.1 ha?1) [64] on heavily leached, infertile sandy soils with
extremely low P availability [65].
Although these examples show that some of the most species-rich
plant communities on Earth occur on strongly weathered, infertile
soils, such broad-scale comparisons do not enable strong causal
inference between pedogenesis and plant diversity, because different
regions can have very different geological and evolutionary histories
(e.g., recent glaciations at higher latitudes). Soil chronosequences, by
contrast, enable comparisons between ecosystems within much
smaller regions, where factors such as climate and parent material
are controlled [50,51]. The available data reveal that total plant
species richness usually increases with soil age across soil chron-
osequences spanning boreal, temperate, subtropical, and Mediterra-
nean climates [4,5,24] (Figure II). Together, these global and smaller-
scale patterns motivate the search for underlying mechanisms that
can explain the greater plant diversity on older soils.
(A) (B)
(C) (D)
(E) (G)(F)
TRENDS in Ecology & Evolution 
Figure I. Examples of species-rich plant communities on strongly weathered, infertile soils. Lowland tropical rainforest in Yasun?? National Park (A) and associated soil
profile (B), showing a strongly weathered kaolinitic Ultisol. Tropical rainforest in Lambir Hills, Malaysia, (C) and Brazilian cerrado (D). Southwestern Australian kwongan
shrublands (E,F) and associated soil profile (G), showing a very thick (0.6 to >1.6 m) and conspicuously bleached E horizon. Reproduced, with permission, from the
Smithsonian Tropical Research Institute (A), Benjamin Turner (B and C), Graham Zemunik (D), and Etienne Laliberte? (E?G).
Opinion Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. x
TREE-1676; No. of Pages 10
2
ose
8 
d va
ges 
ich w
] fo
ic li
 4.28
nd in
cles
TREE-1676; No. of Pages 10a phenomenon termed ?ecosystem retrogression? [9,10].
The large changes in N and P levels during pedogenesis,
and associated changes in primary productivity, are rele-
vant because nutrient availability and stoichiometry are
central to some classical theories of plant species coexis-
tence [11?13]. On the one hand, ?productivity?diversity?
theories suggest that nutrient availability drives primary
productivity, which in turn influences the rate at which
plant species can be displaced competitively from a com-
munity [11,13]. On the other hand, resource-ratio theory
[12] proposes that the balance in the supply of two (or
more) limiting resources determines how many plant
species can coexist, provided that species show trade-offs
in their ability to acquire and tolerate low levels of these
resources.
Following productivity?diversity theories, plant diver-
sity should be low at high productivity sites where compet-
itive exclusion is most rapid, and increase as productivity
declines. Soil chronosequences provide some support
for the productivity?diversity hypothesis, in that total
ratio theory, intraspecific foliar N:P ratio can be used as a
measure of N:P stoichiometry. Higher plant diversity
would be expected at intermediate leaf N:P ratios, indica-
tive of a more balanced nutrient supply [15]. In addition,
empirical tests of resource-ratio theory should involve
nutrient-addition experiments (e.g., [14]), especially given
recent findings that N, P, and K can limit different aspects
of primary productivity in species-rich lowland tropical
forest growing on strongly weathered soils [16].
Diversity of N and P forms
The classical view of resource partitioning is that if different
resources limit the growth of two species, then coexistence is
possible because competition is greater within than between
species [17]. However, N and P occur in a variety of chemical
forms in soils. Plants can take up N in both inorganic (nitrate
and ammonium) and organic (amino acid and small peptide)
forms [18]. Plants take up P as dissolved orthophosphate,
yet soil P occurs in many other inorganic (e.g., pyrophos-
phate and polyphosphate) and organic (e.g., nucleic acids,hypotheses can be tested empirically using long-term (i.e.,
thousands to millions of years) soil chronosequences as
model systems (hereafter simply referred to as ?soil chron-
osequences?) (Box 2). Finally, we describe how our concep-
tual model can serve as a meta-model for quantitative
evaluation using structural equation modeling [6]. In doing
this, we hope to stimulate and guide research along soil
chronosequences on multivariate controls over plant diver-
sity, particularly those that operate belowground.
Potential factors controlling plant diversity
Nutrient availability and stoichiometry
The availability and stoichiometry of nitrogen (N) and
phosphorus (P), the two nutrients that most often limit
terrestrial primary productivity [7], vary strongly during
long-term soil development [8,9]. N is generally absent
from soil parent material and enters ecosystems via N2
fixation and atmospheric deposition, whereas P is derived
predominantly from rock weathering and declines as soils
age [8]. As a result, primary productivity is initially low
and N-limited on young soils, increases and then becomes
co-limited by N and P on intermediate-aged soils, and
eventually decreases as it becomes P-limited on old soils,
1
Pl
an
t 
sp
ec
ie
s 
pl
ot
?1
4 5
10
Glacier BayArjeplog
Chron
15
20
25
30
6
8
10
12
14
2 3 4 5 61 2 3 4 5 6 7 
Figure II. Increases in plant species density with soil age (represented by the ranke
contrasting climates. All of these soil chronosequences include retrogressive sta
which are unpublished (Graham Zemunik). Points represent individual plots, wh
others [5]. Meta-analysis of six chronosequences (these four plus two others in [5
species density with soil age (Poisson linear mixed-effect model with logarithm
random slopes per chronosequence, plot size as an offset; b = 0.120, s = 0.028, z =
in each panel represent the fitted values, taking into account random slopes a
random effect to improve visual clarity. Key: red circles, boreal climate; green cir
Opinionplant species richness generally increases as ecosystem
retrogression proceeds (and productivity declines) [4,5]. By
contrast, according to the resource-ratio theory [12], higher
plant diversity is maintained under equilibrium conditions
if multiple resources limit productivity [14], such that a
positive relation is expected between the number of co-
limiting resources and plant diversity [14]. However, the
available evidence suggests that plant species richness
along soil chronosequences is not greatest on intermedi-
ate-aged soils [5], where multiple nutrient limitation is
most likely [10]. For example, evidence for multiple nutri-
ent limitation [N, P, potassium (K), and/or micronutrients]
along the Jurien Bay chronosequence is only found on
relatively species-poor young to intermediate-aged soils
(<7000 years), whereas P is the only limiting nutrient
on older soils (>120 000 years) [10], where plant diversity
is greatest (Figure II in Box 1).
How to test the productivity?diversity and resource-
ratio theories? Evaluating productivity?diversity theories
requires measurements of soil nutrient availability and
primary productivity to be linked with data on local plant
diversity along soil chronosequences. To test the resource-
30
10
20
30
40
50
60
70
Jurien BayFranz Josef
quence stage
35
40
1 2 3 4 5 6 1 2 3 4 5 67 8 9
TRENDS in Ecology & Evolution 
riable ?chronosequence stage?) for four representative soil chronosequences from
[9,30]. Data are from [5], except data from the Jurien Bay chronosequence [10],
ere 0.01 ha in size for the Jurien Bay chronosequence and 0.03 ha for the three
r which data were available) reveal an overall highly significant increase in plant
nk, random intercepts per chronosequence and ?stage within chronosequence?,
, P 0.0001; no evidence for overdispersion: x2 = 127.2, df = 194, P = 0.999). Lines
tercepts per chronosequence, but omitting the ?stage within chronosequence?
, temperate climate; blue circles, Mediterranean climate.
Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. x3
Box 2. Pedogenic changes along long-term soil chronosequences
Pedogenesis involves a series of processes that includes the
accumulation of organic matter, a decline in pH, the formation of
clays and metal oxides and their leaching from surface to subsurface
horizons, changes in nutrient concentrations, and the formation of
pans or other cemented horizons that restrict the downward growth
of roots. The occurrence and rates at which these processes
contribute to soil formation depend on the soil-forming factors:
climate, organisms, relief, parent material, and time [57]. For
example, soils develop more rapidly under a warm wet climate than
under a cold dry climate, due to more intense weathering and
leaching, and conifers promote rapid soil development compared
with deciduous trees through the acidifying properties of their litter.
Topography influences several factors, such as erosion rate, so that
soil development is retarded on steep slopes compared with flatter
surfaces. Parent material can include a variety of substrates (e.g.,
volcanic lava or ash, glacial moraine, dune sands, and bedrock
exposed by landslides or geologic uplift) and has a pronounced
impact on soil formation. For example, soil development occurs more
slowly on impenetrable bedrock compared with loose and weath-
erable material, such as glacial till. Soil chronosequences isolate the
influence of time as a soil-forming factor by controlling for the
remaining four factors (Figure I).
It is of particular significance for plant diversity that pedogenesis
results in predictable changes in nutrient availability that are
consistent across ecosystems developing under different climates
and on various parent materials [9]. Most parent materials contain
little N but abundant P, which promotes N fixation by symbiotic and
heterotrophic microbes and N accumulation in young soils. Over
time, soil P declines through loss by leaching that exceeds inputs
from bedrock weathering, whereas chemical transformations in the
soil promote the accumulation of organic and recalcitrant inorganic
forms of P [8]. The latter, often termed ?occluded P?, represent P
bound strongly to, or within, metal oxides that are of limited
availability to most plants. The overall effect of these processes is a
long-term reduction in P availability, until a terminal steady state is
reached when outputs in leachate balance P inputs from the atmo-
sphere. This eventually constrains plant biomass and productivity,
which is termed ?ecosystem retrogression? [9].
A
B
C
O
E
B
BC
O
E
B
Iron
pan
Pro?l e 1
(A)
(B) (C) (D) Pro?l e 2 Pro?l e 3
Pro?le 1
?400 years
Pro?le 2
?1800 years
Pro?le 3
?4000 years
Increa sing dun e ag e
TRENDS in Ecology & Evolution 
Figure I. Example of soil development in coastal sand dunes at Haast, on the west coast of the South Island of New Zealand [66] (A). Soils developed under lowland
temperate rain forest in sand deposits formed following periodic earthquake disturbance along the Alpine fault. Pedogenesis is rapid under the warm and wet climate. (B)
Weakly developed soil (Entisol) approximately 400 years old. The profile exhibits limited horizon development, with an organically enriched A horizon on the surface, a
shallow illuvial B horizon showing faint coloration from iron oxides, and a C horizon comprising the parent sand. (C) Moderately developed soil approximately 1800 years
old. The profile shows a thick organic horizon over a bleached eluvial (E) horizon from which iron oxides have been removed, and a B horizon in which iron oxides are
accumulating. (D) Well-developed soil approximately 4000 years old. This soil is a Spodosol and differs from the intermediate-aged soil by its thicker organic horizon, spodic
B horizon containing accumulated metal oxides, and a continuous cemented iron pan immediately below the E horizon that impedes water movement and root penetration.
These changes are accompanied by strong acidification during the early stages, an accumulation of total N, and a marked decline in total P [66]. Reproduced, with
permission, from Andrew Wells (A) and Benjamin Turner (B?D).
Opinion Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. x
TREE-1676; No. of Pages 10
4
Tue) 
 soil
hesi
TREE-1676; No. of Pages 10inositol phosphates, phospholipids and phosphonates)
forms that plants can access via specialized root strategies
(e.g., mycorrhizas, phosphatase enzymes, and carboxylate
exudation) [19,20]. This raises the possibility that partition-
ing of different forms of N or P can contribute to plant species
coexistence, even if all species within a community are
limited by the same nutrient [20,21]. In other words, a
greater diversity of N or P forms in soils might allow a
greater number of plant species to coexist.
The relative importance of different forms of N and P
varies greatly during pedogenesis. During the progressive
phase of ecosystem development, there are increases in
nitrification and N mineralization [22]. As a result, peak
productivity coincides with a relatively ?balanced? supply of
Nutrient
availability and
stoichiometry
Soil spaal
heterogeneity
Climate Parent material 
Pedogenic stage
Diversity
of N and
P forms
Organisms
Plant
diversityResource
Resource-rao
model, producvity?
diversity (+/?)
paroning (+)
Niche
theory (+)
Figure 1. General conceptual model representing how soil-forming factors (top, bl
arrows. Blue boxes with broken arrows going to ?Pedogenic stage? indicate the main
[50,51] and, therefore, can be omitted from the model. Relevant theories and hypot
next to the arrows. Abbreviations: N, nitrogen; P, phosphorus.
Opinioninorganic and organic N, potentially generating a greater
opportunity for N partitioning. As retrogression proceeds,
more of the total soil N pool occurs in less available organic
forms (e.g., protein?tannin complexes), N mineralization
rates decrease strongly, and a larger proportion of N is
supplied as amino acids [23,24]. Partitioning of organic N
would still be possible where the majority of the soluble N
supply is in the form of amino acids, given that plant
species can show preferences for specific amino acids or
peptides [18]. However, because productivity is typically P-
limited during the retrogressive phase of ecosystem devel-
opment [10,25], partitioning of organic N is likely to be less
important in promoting plant species coexistence on old
soils. By contrast, although total soil P declines over time,
organic P accumulates during pedogenesis to become a
major fraction of the remaining total P [8], and different
forms of organic P vary in their relative abundance as soils
age [26]. Therefore, partitioning of P could become impor-
tant for coexistence in retrogressive ecosystems where P is
limiting [20]. However, no study has yet quantified the
importance of these mechanisms in driving diversity pat-
terns during long-term pedogenesis, although there is
growing interest in shifts in the belowground traits of
plants and plant functional diversity along soil chronose-
quences [27].How to test the N and P partitioning hypotheses. The N
and P partitioning hypotheses imply that diversity in N
forms could promote plant diversity in younger, N-limited
ecosystems, whereas diversity in P forms would be impor-
tant in older, P-limited ecosystems. The simplest test of
these hypotheses would be to: (i) measure fluxes of differ-
ent N and P fractions along a soil chronosequence, using
established methods [18,26]; (ii) quantify the diversity of N
or P fractions, using standard metrics [e.g., Simpson di-
versity index (1/D)]; and (iii) relate the diversity of N or P
forms to observed plant species diversity and determine
whether these relations depend on soil age.
Soil spatial heterogeneity
opography Time
Aboveground
heterotrophs
Belowground
heterotrophs
Stage-
speci?c
species
pool size
Commonness
of habitat
Abioc
condions
Species pool 
hypothesis (+)
Plant?soil
feedback (+/?)
Herbivory ?
resource
availability (+/?)
Environmental
?ltering (?)
Time-area
hypothesis (+)
TRENDS in Ecology & Evolution 
can drive local plant diversity (bottom, green). Relevant theories are listed next to
-forming factors [57] that are controlled for by using the chronosequence approach
zed effects (+, positive; ?, negative; +/?, nonlinear or context dependent) are shown
Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. xSpatial heterogeneity in nutrient availability or other
environmental factors that regulate it (e.g., soil pH) could
influence the number of plant species that can coexist via
niche partitioning [17]. In other words, if different species
are adapted to different nutrient levels, then a larger
number of plant species might be able to coexist where
there is greater spatial heterogeneity. Despite the long
history of niche theory, the influence of small-scale spatial
heterogeneity in soil properties on plant diversity remains
little studied. In addition, how spatial heterogeneity in soil
properties changes with soil development remains unclear,
although one hypothesis is that more productive (e.g., in
our context, intermediate-aged) ecosystems show lower
spatial heterogeneity. This is expected because productive
ecosystems host larger plants that forage over larger areas,
thereby integrating variation in resource availability and
reducing plant diversity [28].
How to determine the effect of spatial heterogeneity. One
of the few studies to explore the effects of spatial heteroge-
neity on plant diversity during long-term pedogenesis
measured within-community spatial variability in five soil
properties (NH4-N, amino N, PO4-P, microbial biomass,
and litter decomposition rate) and related it to changes in
plant community composition and species richness along
5
TREE-1676; No. of Pages 10the Arjeplog island chronosequence (Figure II in Box 1) in
Sweden [29]. In that system, smaller islands are less
frequently burnt and, therefore, have a thicker humus
layer, which is associated with slower cycling of nutrients
and, thus, lower productivity [30]. Surprisingly, spatial
heterogeneity in soil properties was lower on smaller
islands (i.e., those with older, nutrient-poor soils), which
supported the highest plant species richness [29]. Although
these results suggest that spatial heterogeneity in
resources does not explain differences in plant species
diversity between young and old soils, further tests are
required in other systems.
Role of aboveground heterotrophs
Aboveground herbivores can alter the survival, growth,
and competitive interactions of plants by removing plant
biomass either selectively or nonselectively depending on
whether the herbivores are specialists or generalists. The
effects of herbivores on plant diversity can be positive,
negative, or neutral, and are strongly influenced by herbi-
vore feeding preferences, as well as by soil fertility and the
productivity of the environment, which influences the
ability of plants to replace lost tissue [11,31]. Herbivore
biomass and the intensity of herbivory are generally af-
fected by soil fertility and plant productivity [32], so that
total herbivory is expected to decline with decreasing soil
fertility during long-term pedogenesis.
Herbivory generally promotes plant diversity more in
productive or fertile environments than in unproductive
and infertile areas, because the preferred species in pro-
ductive areas are also more likely to be the dominant
species [31]. With reference to chronosequences, positive
effects of herbivory on diversity are expected to be greatest
on sites where soil fertility and plant productivity are
highest (i.e., in intermediate-aged soils) and plant diversity
is comparatively low. Less positive, or even negative,
effects of herbivory on diversity are expected where pro-
ductivity is low, both on the youngest soils (where diversity
is typically lowest) and on the oldest soils (where diversity
is generally highest). Invertebrate herbivore density and/
or herbivory have been shown to both decrease [33] and
increase [34] with declining fertility as ecosystem retro-
gression proceeds, but the influence of foliar herbivores on
plant diversity patterns observed along soil chronose-
quences remains unstudied.
How to determine the role of aboveground heterotrophs.
Evaluating the effects of aboveground consumers on plant
diversity requires measurements of the intensity of her-
bivory (e.g., the proportion of primary productivity lost to
herbivores), so that the effects of herbivory on plant diver-
sity can be compared at different soil fertility levels [35].
However, the intensity of herbivory can be difficult to
measure and generally requires herbivore-exclusion
approaches (e.g., [36]).
Role of belowground heterotrophs
Long-term pedogenesis leads to major changes in soil biota,
Opinionsuch as bacteria, fungi, and nematodes [37]. Soil biota can
influence plant diversity by promoting resource partitioning
[38], or by facilitating resource sharing via belowground
6pathways (e.g., mycorrhizal networks [39]). The potential for
soil biota to influence plant diversity is high because most
plant species depend on soil microbes for nutrient uptake
[40].
Of particular relevance to plant species coexistence is
the role of soil biota in ?plant?soil feedback?, whereby
plants influence soil biota, which in turn affect plant
performance [41]. Negative feedback occurs when a plant
has stronger negative effects on its own species than on its
competitors, thus promoting coexistence [42]. Positive
feedback, in which a species affects its environment in a
way that favors itself, can lead to dominance [42].
How to determine the role of belowground heterotrophs.
There is evidence that the strength and direction of plant?
soil feedback depend on environmental context, particu-
larly nutrient availability [43,44]. To address this further,
controlled feedback experiments [45] using plant species
and soils collected along soil chronosequences are needed
to determine whether and how plant?soil feedback drives
plant diversity during pedogenesis. Such experiments com-
pare the growth of a plant species in soil conditioned by the
same species versus its growth in soil conditioned by other
co-occurring plant species [41], known as the ?self-other?
approach [45]. Provided that a reasonable number of spe-
cies per community is used, a community-level plant?soil
feedback effect can be estimated [45], which can then
indicate the contribution of plant?soil feedback to the
maintenance of plant diversity. In that case, a stronger
mean negative feedback should be associated with greater
plant diversity [42].
Species pools
Whereas the mechanisms discussed so far involve local
interactions between plants and soils, it has been sug-
gested that evolutionary history can also contribute to
local plant diversity patterns by determining the size of
species pools [46]. The ?species pool hypothesis? proposes
that plant species richness in a local community is directly
influenced by the number of species that can potentially
colonize the site and are physiologically adapted to the
local environmental conditions [46]. For example, it has
been suggested that the higher local plant diversity of older
Hawaiian islands simply reflects their larger regional
species pools, due to more opportunities for colonization
and speciation [4]. However, many chronosequences cover
a sufficiently small area that most species can disperse
easily across the entire area (e.g., [10]). Differences in
abiotic conditions between different pedogenic stages
(e.g., young vs old islands) could still filter out those species
from the regional species pool that have poor local fitness,
leading to a ?stage-specific species pool? (Figure 1).
How to test the species pool hypothesis. Although esti-
mating stage-specific species pool size can be challenging,
one option is to use nonparametric species richness esti-
mators (e.g., Chao-2) [47] on all vegetation samples from a
given pedogenic stage to determine the ?true? asymptotic
Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. xnumber of species present in that habitat [48]. According to
the ?time?area hypothesis?, habitats (and their associated
abiotic conditions) that have been more abundant over
evolutionary time are predicted to have a larger species
pool and, thus, greater local species richness [49]. As a
result, soil chronosequences can enable evaluation of the
relative contribution of local versus ?species pool? effects to
local plant diversity.
Using soil chronosequences to investigate mechanistic
controls over diversity
Pedogenesis operates over thousands to millions of years,
which vastly exceeds the timespan of any manipulative
experiment. Pedogenesis can, however, be studied by using
Box 3. From general theories to quantitative models
Historically, links between theories and empirical tests have been
informally determined, often resulting in unresolved debate (e.g.,
[54]). Recent advances in quantitative modeling provide more
rigorous guidance for evaluating the multiple and simultaneous
underlying mechanisms using data [6]. Furthermore, there has been
an infusion of new ideas into quantitative modeling, such as the
discovery that it is necessary to formalize causal reasoning for
artificial intelligence systems [55]. These ideas are now being
incorporated into SEM, providing a stronger foundation for develop-
ing and testing multivariate models [67].
The SEM process begins by considering the goals of the analysis.
For understanding how pedogenesis drives plant diversity, our
modeling is ?mediation focused?; that is, we wish to determine the
dominant mechanisms that mediate the effects of pedogenesis on
diversity. The general problem then becomes one of hypothesizing
alternative mediating mechanisms, identifying measures that can
serve as ?indicators? for underlying mechanisms, and then testing the
direct and indirect causal implications of models [6]. This sequence of
operations produces both an evaluation of the strengths of hypothe-
sized linkages and the discovery of new linkages not predicted by our
initial theories. Model results can then force the reconsideration of the
original hypotheses and lead to designing additional studies to refine
and generalize our understanding of the system of interacting
processes.
Translating our initial conceptual ideas (see Figure 1 in main text) into
a fully specified SEM (Figure I) requires attention to several issues. First,
theoretical ideas are often only defined at the linguistic level. General
concepts must therefore be carefully specified before they meet the
criterion of being theoretical ?constructs? [6]. Of particular importance is
the question of whether a construct is expected to have general,
unidimensional influences in the system (i.e., behave as one entity) or
whether it is in fact a collection of separate processes that are weakly
correlated with one another (i.e., behave as a collection of entities).
Second, one must consider the level of mechanistic or causal detail
used to interpret relations. Will probabilistic associations (direct links)
be sufficient characterizations, or are there specific causal chains
needed to represent a particular mechanism? Third, what observations
will serve as indicators of the underlying latent mechanisms? In Table I,
we summarize the translation process for latent variables with a direct
effect on plant diversity.
Nutrient
availability
Soil spaal
heterogeneity 
Climate Parent material Topography Time
Aboveground
heterotrophs
Belowground
heterotrophs
Stage-
speci?c
species
pool size
Commonness
of habitat
Nutrient
stoichiometry
Diversity
N forms
Diversity
P forms
Soil
age
Depth to
C horizon
Historical area
pH
?
diversity
Foliar
N:P
Abioc
condionsOrganisms
Simpson
index
N forms
Simpson
index 
P forms
# Liming
nutrients
Soil N
Plant
diversity
Light
Biomass Disturbance
Variance soil
properes
% NPP
consumed
t
ss
%PPFD
Basal
area Time since last
disturbance
Mean plant?
soil feedbackNPP
Phytometer
yield
Pedogenic stage
Producvity
oxe
ing
Opinion Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. x
TREE-1676; No. of Pages 10Rare?ed plan
species richne
Figure I. Structural equation model showing indicator variables (square white b
boxes with broken arrows going to ?Pedogenic stage? represent the main soil-form
therefore, can be omitted from the model. Arrows leading to other arrows indicate int
variables. Abbreviations: #, number of; %PPFD, percent photosynthetic photon flux de
Key: green arrows, positive effect; red arrows, negative effect; orange arrows, nonlinTRENDS in Ecology & Evolution 
s) for the theoretical constructs (colored rounded boxes, bold characters). Blue
 factors [57] that are controlled for by the chronosequence approach [50,51] and,
eractive effects, whereby one variable influences the relation between two other
nsity at ground level; N, nitrogen; NPP, net primary productivity; P, phosphorus.
ear effect; blue arrows, context-dependent effect.
7
od
ind
be
ote
d ar
apt
s [4
 a r
at t
ive
 for
tion
 the
TREE-1676; No. of Pages 10Table I. The translation process from a general theoretical m
variables with a direct effect on plant diversity (Figure I)
Construct Indicator(s) Justifications for 
Stage-specific
species pool size
g diversity of each stage Estimates the num
species that can p
colonize a site an
physiologically ad
the soil condition
Diversity of N forms Simpson diversity index
(1/D) of N forms
Simpson index is
diversity index th
into account relat
abundance
Diversity of P forms Simpson diversity index
(1/D) of N forms
Same as of for N
Nutrient stoichiometry Foliar N:P ratio Intraspecific varia
ratio can indicate
Opinionspace-for-time substitution [50] (Box 2). Numerous soil
chronosequences driven by pedogenesis have been thor-
oughly investigated. By ensuring that confounding factors
(e.g., elevation, aspect, rainfall, parent material, and dis-
turbance) are minimized [9], these chronosequences have
provided most of our understanding of how soils develop
[8,9,51]. However, changes in community-level properties
along soil chronosequences, such as plant diversity [5],
have received less attention [9].
Soil chronosequences are less suitable for studying
vegetation successional dynamics, because vegetation on
older sites cannot be assumed to have gone through the
same stages or number of successional cycles as that on
younger sites [52]. However, chronosequences can be used
to compare spatial differences in present-day community
and ecosystem properties (e.g., local plant species diversi-
ty) across sites of different ages if the potentially confound-
ing effects of succession are avoided [50].
and strength of nutri
limitation
Number of (co)limiting
nutrients
Represents how man
(co)limiting resources
compete most strong
[14]
Soil spatial
heterogeneity
Variance soil properties Represents spatial
heterogeneity in soil
properties, directly re
to niche theory [17,2
Belowground
heterotrophs
Mean plant?soil feedback Role of belowground
heterotrophs on plan
diversity is expected 
strongly driven by pl
feedback [38]
Aboveground
heterotrophs
Percentage of NPP
consumed
Direct measure of pri
production lost to
consumers [35]
Biomass Basal area Good proxy for stand
biomass, at least in f
Disturbance Time since last
disturbance
Ecologically relevant
of disturbance
Light Percent photosynthetic
photon flux density at
ground level
Direct measure of lig
availability relevant t
photosynthesis
8el (Figure 1, main text) to a quantitative model for latent
icator How to measure Justifications for effect on plant
diversity
r of
ntially
e
ed to
8]
Nonparametric species
richness estimator (e.g.,
Chao-2) [48]
Larger species pool leads to
greater local plant diversity [48]
obust
akes
N fractionation [18] Diversity in N forms increases
plant diversity in younger, N-
limited ecosystems [21]
ms P fractionation [26,68] Diversity in P forms increases
plant diversity in older, P-limited
ecosystems [20]
 in N:P
 type
Leaf [N] and [P] Plant diversity is greatest at
intermediate N:P ratios [12,69]
Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. xDespite their potential, there are limitations to the use
of soil chronosequences for examining factors regulating
plant diversity. For example, there can be difficulties in
holding some environmental controls over community and
ecosystem properties constant, even through careful site
selection [50]. Moreover, potential drivers of local plant
diversity (e.g., soil pH, nutrient availability, water-logging,
and disturbance frequency) can co-vary during pedogene-
sis, complicating causal interpretations. However, we be-
lieve that appropriate statistical methodologies can
alleviate these problems, as we discuss in the next section.
Multivariate controls over diversity
The maintenance of plant species diversity is a multivari-
ate problem [53]. Consequently, significant progress is
more likely to arise from evaluating the relative impor-
tance of multiple drivers rather than focusing on individual
ones [53,54]. Analytical approaches, such as structural
ent
y
 plants
ly for
Nutrient-limitation
experiments [10,14,25]
Plant diversity is greater when
more resources are co-limiting
[14]
levant
9]
Multivariate spatial
dispersion [70] of soil
properties
More niches, more species [17]
t
to be
ant?soil
Community-level mean
plant?soil feedback [45]
Stronger negative feedback
leads to greater diversity [38]
mary Herbivore exclusion
approaches [36]
Effects of aboveground
consumers on plant diversity
depend on nutrient availability
[31]; in productive
environments, the preferred
species are generally the
dominant species, thus
increasing diversity [31]
ing
orests
Standard field methods Effects on diversity other than
light competition
 aspect e.g., previous aerial
photos, fire scars
Direct effects on diversity not
mediated by light competition
ht
o
Light meters Light competition is strongly
asymmetric and reduces
diversity [11]
TREE-1676; No. of Pages 10equation modeling (SEM), are well suited to this task,
because they enable explicit linkage of theory to data
and can test all direct and indirect causal implications
that are derived from a conceptual model [6] (Box 3). A key
challenge, however, is how to translate a theoretical model
into one that can be evaluated quantitatively [6]. Manipu-
lative experiments can also be used to establish causality
for specific mechanisms [55], although this is beyond the
scope of this review. We argue that confronting a priori
multivariate models derived from theory against data
through the use of SEM [6] (Box 3) will strengthen the
soil chronosequence approach and provide valuable
insights into controls on plant diversity. To understand
how long-term pedogenesis affects plant diversity, we sug-
gest ways to translate our general theoretical model
(Figure 1) into one that can be evaluated quantitatively
(Figure I and Table I in Box 3).
Concluding remarks
In this review, we present a set of older [11?13] and more
recent [20,21,38,49] hypotheses that describe how long-
term pedogenesis can drive local plant species diversity.
We believe that soil chronosequences can be used to quan-
tify the relative importance of the different mechanisms
that affect plant species diversity, and how they depend on
environmental context. We propose that analyzing data
from soil chronosequences using structural equation
modeling will advance the development and testing of
multivariate theories of plant species coexistence. In par-
ticular, we suggest ways to translate our conceptual model
(Figure 1) into one that can be evaluated quantitatively
(Box 3). Much empirical research remains to be done on
multivariate controls over plant diversity, particularly in
relation to belowground mechanisms. Exploration of these
issues should help us understand why some ecosystems
with ancient, strongly weathered soils support an incredi-
bly high number of plant species, whereas younger, more
fertile ecosystems are often dominated by considerably
fewer species.
Acknowledgments
This paper originated from a research workshop held at The University of
Western Australia (UWA) on 20?25 February, 2012. Funding for this
workshop was provided by UWA, specifically by the School of Plant
Biology (40%), the Pro Vice-Chancellor (Research) (40%), and the Faculty
of Natural and Agricultural Sciences (20%). We thank G. Zemunik for
providing data and photos, and A. Wells for photos. E.L. was supported by
a DECRA (DE120100352) from the Australian Research Council. M.A.H.
was supported by NSF OPUS Grant 0918927. H.L. and F.T. were
supported by an ARC Discovery grant. D.A.W. was supported by a
Wallenberg Scholars award.
References
1 Hillebrand, H. (2004) On the generality of the latitudinal diversity
gradient. Am. Nat. 163, 192?211
2 Huston, M.A. (2012) Precipitation, soils, NPP, and biodiversity:
resurrection of Albrecht?s curve. Ecol. Monogr. 82, 277?296
3 Sa?nchez, P.A. (1976) Properties and Management of Soils in the
Tropics. Wiley
4 Crews, T.E. et al. (1995) Changes in soil phosphorus fractions and
ecosystem dynamics across a long chronosequence in Hawaii. Ecology
76, 1407
5 Wardle, D.A. et al. (2008) The response of plant diversity to ecosystem
Opinionretrogression: evidence from contrasting long-term chronosequences.
Oikos 117, 93?1036 Grace, J.B. et al. (2010) On the specification of structural equation
models for ecological systems. Ecol. Monogr. 80, 67?87
7 Elser, J.J. et al. (2007) Global analysis of nitrogen and phosphorus
limitation of primary producers in freshwater, marine and terrestrial
ecosystems. Ecol. Lett. 10, 1135?1142
8 Walker, T.W. and Syers, J.K. (1976) The fate of phosphorus during
pedogenesis. Geoderma 15, 1?19
9 Peltzer, D.A. et al. (2010) Understanding ecosystem retrogression.
Ecol. Monogr. 80, 509?529
10 Laliberte?, E. et al. (2012) Experimental assessment of nutrient
limitation along a 2-million year dune chronosequence in the south-
western Australia biodiversity hotspot. J. Ecol. 100, 631?642
11 Huston, M.A. (1994) Biological Diversity. Cambridge University Press
12 Tilman, D. (1982) Resource Competition and Community Structure.
Princeton University Press
13 Grime, J.P. (1973) Competitive exclusion in herbaceous vegetation.
Nature 242, 344?347
14 Harpole, W.S. and Tilman, D. (2007) Grassland species loss resulting
from reduced niche dimension. Nature 446, 791?793
15 Braakhekke, W.G. and Hooftman, D.A.P. (1999) The resource
balance hypothesis of plant species diversity in grassland. J. Veg.
Sci. 10, 187?200
16 Wright, S.J. et al. (2011) Potassium, phosphorus, or nitrogen limit root
allocation, tree growth, or litter production in a lowland tropical forest.
Ecology 92, 1616?1625
17 Chesson, P. (2000) Mechanisms of maintenance of species diversity.
Annu. Rev. Ecol. Syst. 31, 343?366
18 Hill, P.W. et al. (2011) Vascular plant success in a warming Antarctic
may be due to efficient nitrogen acquisition. Nat. Clim. Change 1, 50?53
19 Lambers, H. et al. (2008) Plant nutrient-acquisition strategies change
with soil age. Trends Ecol. Evol. 23, 95?103
20 Turner, B.L. (2008) Resource partitioning for soil phosphorus: a
hypothesis. J. Ecol. 96, 698?702
21 McKane, R.B. et al. (2002) Resource-based niches provide a basis for plant
species diversity and dominance in arctic tundra. Nature 415, 68?71
22 Vitousek, P.M. et al. (1983) Soil development and nitrogen turnover in
montane rainforest soils on Hawai?i. Biotropica 15, 268?274
23 Northup, R.R. et al. (1995) Polyphenol control of nitrogen release from
pine litter. Nature 377, 227?229
24 Wardle, D.A. et al. (2012) Linking vegetation change, carbon
sequestration and biodiversity: insights from island ecosystems in a
long-term natural experiment. J. Ecol. 100, 16?30
25 Vitousek, P.M. and Farrington, H. (1997) Nutrient limitation and soil
development: experimental test of a biogeochemical theory.
Biogeochemistry 37, 63?75
26 Turner, B.L. et al. (2007) Soil organic phosphorus transformations
during pedogenesis. Ecosystems 10, 1166?1181
27 Holdaway, R.J. et al. (2011) Species- and community-level patterns in
fine root traits along a 120 000-year soil chronosequence in temperate
rain forest. J. Ecol. 99, 954?963
28 Tilman, D. and Pacala, S. (1993) The maintenance of species richness
in plant communities. In Species Diversity in Ecological Communities
(Ricklefs, R.E. and Schluter, D., eds), pp. 13?25, University of
Chicago Press
29 Gundale, M.J. et al. (2011) Resource heterogeneity does not explain the
diversity?productivity relationship across a boreal island fertility
gradient. Ecography 34, 887?896
30 Wardle, D.A. et al. (2004) Ecosystem properties and forest decline in
contrasting long-term chronosequences. Science 305, 509?513
31 Hillebrand, H. et al. (2007) Consumer versus resource control of
producer diversity depends on ecosystem type and producer
community structure. Proc. Natl. Acad. Sci. U.S.A. 104, 10904?10909
32 McNaughton, S.J. et al. (1989) Ecosystem-level patterns of primary
productivity and herbivory in terrestrial habitats. Nature 341, 142?144
33 Gruner, D.S. (2007) Geological age, ecosystem development, and local
resource constraints on arthropod community structure in the
Hawaiian Islands. Biol. J. Linn. Soc. 90, 551?570
34 Crutsinger, G.M. et al. (2008) Ecosystem retrogression leads to
increased insect abundance and herbivory across an island
chronosequence. Funct. Ecol. 22, 816?823
35 Laliberte?, E. and Tylianakis, J.M. (2012) Cascading effects of long-term
Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. xland-use changes on plant traits and ecosystem functioning. Ecology
93, 145?155
9
36 McNaughton, S.J. et al. (1996) How can net primary productivity be
measured in grazing ecosystems? Ecology 77, 974?977
37 Williamson, W.M. et al. (2005) Changes in soil microbial and nematode
communities during ecosystem decline across a long-term
chronosequence. Soil Biol. Biochem. 37, 1289?1301
38 Bever, J.D. et al. (2010) Rooting theories of plant community ecology in
microbial interactions. Trends Ecol. Evol. 25, 468?478
39 Simard, S.W. et al. (2012) Mycorrhizal networks: mechanisms, ecology
and modelling. Fungal Biol. Rev. 26, 39?60
40 Van der Heijden, M.G.A. et al. (2008) The unseen majority: soil
microbes as drivers of plant diversity and productivity in terrestrial
ecosystems. Ecol. Lett. 11, 296?310
41 Mangan, S.A. et al. (2010) Negative plant-soil feedback predicts tree-
species relative abundance in a tropical forest. Nature 466, 752?755
42 Bever, J.D. (2003) Soil community feedback and the coexistence of
competitors: conceptual frameworks and empirical tests. New Phytol.
157, 465?473
43 Kardol, P. et al. (2006) Temporal variation in plant?soil feedback
controls succession. Ecol. Lett. 9, 1080?1088
44 Manning, P. et al. (2008) Nitrogen enrichment modifies plant
community structure via changes to plant?soil feedback. Oecologia
157, 661?673
45 Kulmatiski, A. and Kardol, P. (2008) Getting plant?soil feedbacks out
of the greenhouse: experimental and conceptual approaches. Prog. Bot.
54 Adler, P.B. et al. (2011) Productivity is a poor predictor of plant species
richness. Science 333, 1750?1753
55 Pearl, J. (2011) The causal foundations of structural equation
modeling. In Handbook of Structural Equation Modeling (Hoyle,
R.H., ed.), pp. 1?36, Guilford Press
56 Sterner, R.W. and Elser, J.J. (2002) Ecological Stoichiometry: The
Biology of Elements from Molecules to the Biosphere. Princeton
University Press
57 Jenny, H. (1941) Factors of Soil Formation. McGraw-Hill
58 Valencia, R. et al. (2004) Yasun?? forest dynamics plot, Ecuador. In
Tropical Forest Diversity and Dynamism: Findings from a Large-Scale
Plot Network (Losos, E.C. and Leigh, E.G., eds), pp. 609?628,
University of Chicago Press
59 Korning, J. et al. (1994) Characters of three Udults and their relevance
to the composition and structure of virgin rain forest of Amazonian
Ecuador. Geoderma 63, 145?164
60 Lee, H.S. et al. (2002) Floristic and structural diversity of mixed
dipterocarp forest in Lambir Hill National Park, Sarawak,
Malaysia. J. Trop. Forest Sci. 14, 379?400
61 Ratter, J.A. et al. (1997) The Brazilian cerrado vegetation and threats
to its biodiversity. Ann. Bot. 80, 223?230
62 Furley, P.A. and Ratter, J.A. (1988) Soil resources and plant
communities of the Central Brazilian Cerrado and their
development. J. Biogeogr. 15, 97?108
63 Cowling, R.M. (1990) Diversity components in a species-rich area of the
Opinion Trends in Ecology & Evolution xxx xxxx, Vol. xxx, No. x
TREE-1676; No. of Pages 1069, 449?472
46 Zobel, M. (1997) The relative role of species pools in determining plant
species richness: an alternative explanation of species coexistence?
Trends Ecol. Evol. 12, 266?269
47 Gotelli, N.J. and Colwell, R.K. (2001) Quantifying biodiversity:
procedures and pitfalls in the measurement and comparison of
species richness. Ecol. Lett. 4, 379?391
48 Grace, J.B. et al. (2011) Local richness along gradients in the Siskiyou
herb flora: R.H. Whittaker revisited. Ecology 92, 108?120
49 Zobel, M. et al. (2011) The formation of species pools: historical habitat
abundance affects current local diversity. Global Ecol. Biogeogr. 20,
251?259
50 Walker, L.R. et al. (2010) The use of chronosequences in studies of
ecological succession and soil development. J. Ecol. 98, 725?736
51 Vitousek, P.M. (2004) Nutrient Cycling and Limitation: Hawai?i as a
Model System. Princeton University Press
52 Johnson, E.A. and Miyanishi, K. (2008) Testing the assumptions of
chronosequences in succession. Ecol. Lett. 11, 419?431
53 Grace, J.B. (1999) The factors controlling species density in herbaceous
plant communities: an assessment. Perspect. Plant Ecol. Evol. Syst. 2,
1?2810Cape Floristic Region. J. Veg. Sci. 1, 699?710
64 Hnatiuk, R.J. and Hopkins, A.J.M. (1981) An ecological analysis of
kwongan vegetation south of Eneabba, Western Australia. Aust. J.
Ecol. 6, 423?438
65 Lambers, H. et al. (2010) Plant mineral nutrition in ancient
landscapes: high plant species diversity on infertile soils is linked
to functional diversity for nutritional strategies. Plant Soil 334,
11?31
66 Turner, B.L. et al. (2012) Soil nutrient dynamics during podzol
development under lowland temperate rain forest in New Zealand.
Catena 97, 50?62
67 Grace, J.B. et al. (2012) Guidelines for a graph-theoretic implementation
of structural equation modeling. Ecosphere 3, 1?44 art73
68 Hedley, M.J. and Stewart, J.W.B. (1982) Method to measure microbial
phosphate in soils. Soil Biol. Biogeochem. 14, 377?385
69 Cardinale, B.J. et al. (2009) Separating the influence of resource
?availability? from resource ?imbalance? on productivity?diversity
relationships. Ecol. Lett. 12, 475?487
70 Anderson, M.J. et al. (2006) Multivariate dispersion as a measure of
beta diversity. Ecol. Lett. 9, 683?693