i
,5,
f Le
y o
amp
f C
arch
dm
ey 
w,
ogi
Opinionagainst it have matured. Gone are the days when propo-
nents and opponents of neutral theory could build their
cases on good or bad fits of one particular neutral model
to some empirical species abundance distribution [1?5].
Here, we complement the recent reviews [6?9] and opinions
[10?14] with our own opinion on how neutral theory can aid
progress in ecological research. The opposition to neutral
theory in ecology is not surprising given its radical assump-
tion and we view such criticism as necessary for neutral
theory to grow. The neutral theory of molecular evolution
[15] received similar criticism at first, yet it is now accepted
as a useful tool. We must reiterate that no one believes the
world is really neutral and neutral theory is not a claim that
species (or individuals) are ecologically equivalent [16]. Still,
we find that these myths persist, and even dominate, at least
in informal discussions. Neutral theory is about improving
understanding by making some simplifying assumptions
about complex systems and seeing what can be explained
with the resulting models, a procedure widely accepted
in many branches of science that does not require the
assumptions to be strictly accurate.
reveal process; (ii) should simple or complex ecological
models be developed; and (iii) is ecological drift a process
and is it important? Of these questions, ecological drift is
conspicuously the only issue relating directly to neutral
theory; the others pertain more to philosophy of science in
general and the relative merits of different scientific
approaches. We conclude with our case for the utility of
neutral theory.
Can pattern reveal process?
In ecology, there is rarely a one-to-one relationship between
patterns and processes [22]. The non-spatial species abun-
dance distribution (SAD) in particular does not reveal a
unique process [12,23?26], although it can be informative by
ruling out models that fail to fit it [27,28], with each model
representing a cocktail of processes. This has important
implications for neutral theory, which is often fitted to such
SADs. Neutral theory shows what a neutral world would
look like; unfortunately, many non-neutral models yield
similar results according to popular summary data, such
as SADs. One example is the broken-stick model [29], which
was motivated by the process of partitioning resources into
niches. A variety of other models can produce the same SADsCorresponding author: Rosindell, J. (james@rosindell.org).The case for ecolog
James Rosindell1,2,3, Stephen P. Hubbell4
Rampal S. Etienne9
1 Institute of Integrative and Comparative Biology, University o
2 Institute of Bioinformatics and Evolutionary Studies, Universit
3Division of Biology, Imperial College London, Silwood Park C
4Department of Ecology and Evolutionary Biology, University o
5Center for Tropical Forest Science, Smithsonian Tropical Rese
6Department of Renewable Resources, University of Alberta, E
7SYSU-Alberta Joint Lab for Biodiversity Conservation, State K
510275, China
8Department of Biological Sciences, University of Idaho, Mosco
9Community and Conservation Ecology Group, Centre for Ecol
11103, 9700 CC Groningen, The Netherlands
Ecological neutral theory has elicited strong opinions in
recent years. Here, we review these opinions and strip
away some unfortunate problems with semantics to
reveal three major underlying questions. Only one of
these relates to neutral theory and the importance of
ecological drift, whereas the others involve the link
between pattern and process, the tradeoff between sim-
plicity and complexity in modeling, and the role of
stochasticity and drift in ecology. We explain how neu-
tral theory cannot be simultaneously used both as a null
hypothesis and as an approximation. However, we also
show how neutral theory always has a valuable use in
one of these two roles, even though the real world is not
neutral.
Three key questions that underlie the debate
Understanding of neutral theory has progressed substan-
tially during recent years and the arguments both for and0169-5347/$ ? see front matter  2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.tree.2012.cal neutral theory
 Fangliang He6,7, Luke J. Harmon2,8 and
eds, Leeds, LS2 9JT, UK
f Idaho, Moscow, ID 83844, USA
us, Ascot, Berkshire, SL5 7PY, UK
alifornia, Los Angeles, CA 90095, USA
 Institute, Unit 0948, APO AA 34002-0948, Republic of Panama
onton, Alberta, T6G 2H1, Canada
Laboratory of Biocontrol, Sun Yat-sen University, Guangzhou
 ID 83844, USA
cal and Evolutionary Studies, University of Groningen, Box
There are many different interpretations of what ?neu-
tral theory? really is (Box 1), and this led both proponents
and opponents to ?debate? without being clear what they
were debating about. A formal debate requires a well-
defined motion and this has been lacking in the discussions
so far. We propose the following: ?Neutral theory, an en-
semble of different neutral models of community assembly,
is useful in ecological research? (Box 1). The usefulness of
neutral theory inevitably depends on the context of use.
Indeed, even the critics have often ?made use of neutral
theory? by producing interesting ecological findings directly
through arguing against it [17?19]. Finding data that are
unexplained by neutrality [20,21] is a valuable application
of neutral theory, not a triumph over it. We feel obliged to
discuss semantics here (Box 1), but we do find this a
distraction from the main issues.
We identify three key questions that underlie most
current arguments about neutral theory: (i) can pattern01.004 Trends in Ecology and Evolution, April 2012, Vol. 27, No. 4 203
Box 1. The semantics of neutral theory
The term ?neutral theory? is widely (but regrettably) used in ecology
to mean different things, leading to misconceptions that take the
debates in a less fruitful direction. For some, the term is used purely
interchangeably with ?null model?; to others it refers specifically to
the contents of Hubbell?s book [1]. We use ?neutral theory? to refer to
?an ensemble of different neutral models by various authors?, that
retains the spirit of what most think of as neutral theory without
having too narrow a scope. If ?neutral theory? were taken instead to
be a direct statement that there were really no ecological differences
between organisms, then it would be reduced to a straw man; no
person supports such a ?neutral theory?. Objections relating to the
use of the term ?unified? or ?metacommunity? are again purely
semantic arguments, as long as researchers can all accurately
communicate, the terminology used should not matter.
Clark argues that neutral theory does not in fact describe true
ecological equivalence [10,45] because it is based on the outcome of
a stochastic birth?death process and thus species are in fact
different, if for no other reason than because they were fortunate
in the outcome of stochastic draws. This is also really a semantic
argument because, if it were taken literally, one would be forced to
label only deterministic models assuming ecological equivalence as
?neutral models? and we agree that such models are unlikely to be
interesting. However, there is a clear distinction between models
where the probabilities of reproduction and death for an individual
depend on its species identity, and models where they do not. It
would only add to the confusion if one was forced to invent a new
term to distinguish between these two cases. It thus makes most
sense to persist with the original definition: neutrality is based on
the stochasticity in demographic rates, so fitness equivalence does
Opinionas those resulting from the broken-stick model [30], includ-
ing a neutral model with random fission speciation [28].
It is both easy and dangerous to interpret the verbal
description that motivates a model as being a fundamental
part of the model itself: the same mathematical model and,
consequently, its predicted patterns, can often be equally
motivated by very different ideas. For example, in the
equilibrium theory of island biogeography, MacArthur
and Wilson [31] describe a mainland?island model that
includes species with different dispersal abilities, but one
could replace that description with a neutral interpreta-
tion where species have the same dispersal ability and only
differ in mainland abundances. The logical rules of the
model itself need not change despite this important dis-
tinction [32] (J. Rosindell and L.J. Harmon, unpublished
manuscript). To further complicate the link between
assumptions and processes, the omission of a process from
a model is often not stated as an assumption; rather, it is
simply outside the scope of the model. In this context,
neutrality might not in fact be regarded as an assumption
at all [33]; neutral theory can indeed be regarded as having
few assumptions rather than many [34]. Different process-
es (or assumptions) producing the same pattern is a prob-
lem, but one that occurs much more generally outside the
context of neutral theory; it is simply the problem that
pattern does not equal process.
Once one has accepted that pattern does not equal pro-
cess, it is a small step to accept that individual processes do
not mean sameness but equivalence in the probabilistic sense.
Furthermore, although the stochasticity in neutral models is likely to
be standing in for unknown processes, these might not be rooted in
selection; species can suffer in ways that are independent of their
species identity. One should be aware of these arguments when
interpreting neutral theory, but realize that they are semantic and do
not constitute a reason to abandon the theory.
204not affect pattern in a unique way. Consequently, the full
complement of individual processes might not be explicitly
required to predict patterns accurately with a model. In an
analogy, imagine many children each mixing a large number
of different colors of paint. One can be fairly certain that each
child will mix a similar muddy brown color (one can predict
pattern without explicitly modeling process) even though
each will do so in a different way (process details do not affect
pattern). Although the brown mix does still contain all the
colors (patterns are the outcome of many complex process-
es), it is not possible to tell their proportions just from
looking (pattern does not uniquely define process). If one
really wishes to know the proportions of colors in the mix
(which processes were involved), then one must collect extra
data beyond eyeballing the brown color; for instance,
dynamically observing the paint-mixing process or looking
at multiple patterns (e.g. colors splashed on the child?s
clothing). Returning to an ecological context, this means
that to see differences in the predictions of different models,
one needs to consider more data, different types of data and
probably dynamic data.
In an ecological context, many of the fundamental
patterns of macroecology have been frequently observed
in systems outside of ecology [35,36]. This might be inter-
preted as supporting the idea that process does not affect
macroecological patterns in ecology, but one could equally
well argue that similar or analogous processes act univer-
sally. In this context, the goal of characterizing the impor-
tance of fundamental ecological processes is distinct from
the goal of predicting and understanding macroecological
patterns.
Should simple or complex ecological models be
developed?
More than four decades ago, Levins [37] pointed out the
inevitable tradeoff between generality, precision and real-
ism of mathematical models in population biology. Al-
though models involving extra processes or exploiting
extra information need not be very complex, there is no
doubt that every extra process added to a model does have
a cost, in terms of understanding the model, its generality
and the ability to test its predictions. In the extreme case,
an impossibly complex beast is created, and little more can
be understood about it than about the real world itself.
There will always be a need for models at a variety of stages
of the complexity scale, but the cost of increasing complex-
ity is in general underappreciated. The development of
computers has enabled the study of extremely complex
models in recent years, but it is tempting to exploit the
technology to add too much complexity at the expense of
generality, predictability and understanding. We suggest
that one should understand the simpler model and its
limitations before progressing to more complex models.
When simple models, such as neutral models, fail, they
do so in informative ways because of something that was
left out and that can often then be identified. This is in
contrast to a more complex model that, if it can be tested at
all, might also fail because of components it erroneously
Trends in Ecology and Evolution April 2012, Vol. 27, No. 4included. One should start with ecological questions and
then think carefully about what level of complexity is
needed to answer them.
The simplicity and tractability of many neutral models
is a great asset to neutral theory, but that alone does not
constitute a reason to value the theory. A good theory need
not be simple, but given several theories that explain the
same phenomena equally well, the simplest is the pre-
ferred (Occam?s razor) [38]. If a cell biologist proposed an
intricate cell-based explanation for all ecology, we ecolo-
gists would not immediately comply and start thinking at
the cell level for all our research; the cell biologist would be
expected to show that those more complex (once scaled to
the community level) explanations do more for us than the
simpler ecological framework. Similarly, neutral theory
says that one does not need to invoke different niches or
habitat types to explain some patterns. The burden of proof
lies with those advocating more intricate explanations for
phenomena to show that they explain more by doing so. We
expect this to mean investigating other types of data, such
as evolutionary and dynamic data, and should ultimately
leave only parsimonious (possibly non-neutral) explana-
tions for new and interesting patterns.
Is ecological drift a process and is it important?
Ecological drift refers to the random fluctuations in popu-
lation size that result from ecological equivalence in the
probabilistic sense: individuals have equal chances of re-
production or death regardless of their species identity
(Box 1). In a recent unifying framework, drift was named
as one of the four key processes in ecology, the others being
speciation, selection and dispersal [11]. Neutral models
encompass any combination of these that excludes selec-
tion.
Clark [10] argues that the stochastic components in
models, including ecological drift, represent noise or an
error term describing what is not yet understood about a
system. He argues that one should aim to convert these
?unknown? (stochastic) components of models into ?known?
(deterministic) components. In this context, ecological drift
appears to be an unsatisfactory ?explanation? for anything
because it is not regarded as a real process but rather is
merely standing in for deterministic processes acting at a
finer scale.
We feel that research goals should not always be to
?explain? stochasticity with determinism. Apart from being
impractical, pursuing this goal can only make models more
complex and is at odds with the need for simple models as
well as complex ones. The goal of simplifying existing
models (possibly by using stochasticity) is equally valid.
Furthermore, complex deterministic models can yield cha-
otic behavior that appears random and, conversely, sto-
chastic systems can have deterministic components; for
example, their expected behavior is deterministic. Stochas-
ticity in a model is much more than the ?unexplained?
component: stochastic processes encompass knowledge
in their distributions. For example, a dispersal kernel
contains valuable information [39] about the chance of a
dispersing seed reaching certain distances from its origin;
however, it must be stochastic because the deterministic
alternative is impractical. In the case of ecological drift,
OpinionRicklefs [14] suggests that ?deterministic influences of
specialized pathogens? are the cause of patterns resem-
bling those arising from stochastic neutral theory. Weagree with the essential idea that neutral drift can result
from deterministic processes that are complex and not well
understood. In this context, two distinct research goals
emerge: first, to use neutral models to approximate and
predict these apparently neutral processes in the most
parsimonious way possible and, second, to develop more
specific fine-scale models involving factors, such as the
specialized parasites suggested by Ricklefs, to help under-
stand how and why a non-neutral world might appear
neutral in some cases.
Ecological drift must be caused by something at a lower
level, but it could still be thought of as a process at a higher
explanatory level. When asking whether drift constitutes a
?process?, one must be aware that the answer is influenced
strongly by perspective. The world appears to be arranged
in a hierarchy of different levels of detail and, consequent-
ly, so is science. Those interested in processes at a high
level usually find processes lower down on the hierarchy
irrelevant because of the dominance of emergent proper-
ties at higher levels. Conversely, those interested in lower-
level phenomena find higher-level processes ?process free?.
Whether the real world is, in the strictest sense, funda-
mentally stochastic or deterministic is another interesting
question, but is outside the scope of this article. What is
important is how to interpret stochasticity in models and
how best to model elements that may be unknown but
intractably complex, outside the scope of interest, or acting
on a different scale.
A key question is how strongly ecological drift contrib-
utes to community assembly. Neutral theory has not only
heightened awareness for drift [1], but also shows that it is
difficult to quantify drift versus selection merely by study-
ing community summary data, and detailed experiments
are often required. For example, Seipielski et al. [40]
conducted manipulative experiments and were able to
show that two Enallagma damselfly species do indeed
appear to be ecologically equivalent. They argue more
generally for a layered view of community structure that
integrates both niche and neutral ideas. We agree that
ideas from niche theory and neutral theory should be
integrated, but we also argue that neutral theory remains
useful on its own.
The utility of neutral theory
As null model or an approximation but not both
The predicted abundance distribution of neutral theory is
very robust, even to the breaking of neutrality [41?43] and
of other assumptions [16,44]. Some have used this to argue
that neutral theory is not useful as a null model because it
gives the same results as niche models and, thus, cannot
detect neutrality [10,45]. However, this argument is essen-
tially just a restatement of the pattern?process problem
and is specific to certain data types. If data cannot tell two
models apart based on sound statistical methodology, then
there is an insufficiency in the data. The models detect the
problem but do not cause it. To solve the problem, more and
different types of information are needed. Models incapa-
ble of making predictions beyond those used for testing
Trends in Ecology and Evolution April 2012, Vol. 27, No. 4restricted data can be criticized on these grounds, but
neutral theory makes many testable predictions beyond
the non-spatial SAD that receive little or no attention [7,8].
205
from being a reality. Stochasticity can be very helpful for
predicting the survival of species [57]. Much work on the
conservation biology of entire areas has emerged from
simply extrapolating a power law species?area curve
[58]. This practice has been criticized for missing many
important factors [59] and for being mathematically incor-
rect [60], but one point that even the critics acknowledge
[59] is that species?area relationship extrapolation has the
advantage of not needing to have species-specific details,
about which unfortunately so little is known. Neutral
theory has the same advantage while potentially including
more biological details than the extrapolation of simple
species?area relationship extrapolation. We consequently
expect versions of neutral theory to be useful as tools in
some conservation applications although we do, of course,
caution for neutral models to be used and interpreted
alongside the results from other methods.
As a foundation for other models
Neutral theory is useful as a foundation or motivation for
other models. For example, it has been used as a starting
that over timescales larger than 3000 years, the pace at which
communities changed was slower than expected from a neutral
benchmark.
A neutral model provided the first mechanistic explanation for the
full S-shaped species?area relationship with three phases, the
second of which follows the classic power law [48]. It also showed
the importance of long-distance dispersal in the system because
without this, the gradient of the most frequently observed second
phase was too shallow on logarithmic space [59,73]. Furthermore,
long-distance dispersal events can stand in for speciation in this
model [49], which might be true much more generally.
Neutral theory and its failures to explain the mean lifetimes of
species [69] have given rise to the concept of protracted speciation
[17]: a simple speciation model that includes the concept that
speciation is a gradual process. This concept was later used in the
birth?death model of diversification, where it provided a new
explanation for observed slowdowns in diversification rates [74].For example, it can predict beta diversity [46], spatial
structure [47], species?area curves [48,49], population
and community dynamics [50,51], phylogenetic tree shape
and branch lengths [52,53], endemicity of species on
islands [54] and much more [7,8]. In our opinion, future
work should do more to test these patterns. We do not
expect the equivalence of niche and neutral models [42] to
hold true against all further tests. For example, the spatial
patterns of a pure niche structure can be different from
those caused by pure dispersal limitation, but appear the
same (as in [42]) when everything is well mixed in space.
Niche theory as a whole might make more predictions than
neutral theory [55], but in this context ?niche theory? refers
to almost all ecological models that are not neutral. Similar
to all theory, neutral theory does have a scope and some
predictions simply fall outside this.
As argued by Gotelli and McGill [56], if neutral models
are being used as null models, then an alternative hypoth-
esis must be stated. The appropriateness of neutral theory
as a null will thus depend on the hypothesis that it is a null
for. One new possibility is to regard neutral models as null
models, not for the actual existence of niches, but rather for
the detectability of niches in different empirical data sets.
Of course data collected for the explicit purpose of finding
niches will succeed in that goal, but niches might not be
detectable in more general summary data. If neutral theo-
ry fits a particular data set, it does not replace existing
explanations for those data, but it should cause scrutiny of
the current explanations that are sufficient, but not neces-
sary to explain the data. Neutral theory as a null model
does not provide the solution to the problems it uncovers in
this capacity, but the same is true of any null model.
The argument that neutrality is not useful as a null
model, because it makes predictions that are essentially
the same as those emerging from non-neutral theory,
implies conceding that the non-neutral explanations are
not necessary to make accurate predictions. It follows
logically that neutrality must be useful as an approxima-
tion and its significant tractability advantages over the
alternatives should make it extremely powerful in this
role. Conversely, stating that neutral theory is not a good
approximation implies arguing that it fails to fit data and,
hence, in failing the model, yields useful information as a
null model. If neutral theory never succeeded in fitting any
data, then its use would of course be diminished, but we
already know this is not the case. We expect that, for some
macroecological data, neutral models will prove useful as
an approximation; for other types of data, they will prove
useful as a null model. Neutral theory cannot be useful in
both capacities simultaneously because being useful as an
approximation requires it to succeed in its predictions,
whereas being useful as a null model requires it to fail
(at least some of the time). In Box 2, we give some examples
where neutral theory has served ecology in these roles by
failing (or succeeding) to fit various data sets.
In conservation
It is appealing to say that conservation decisions should be
Opinionmade from a perspective of full knowledge of all species and
processes going on in a system [10]. However, such full
knowledge or the ability to use it tractably in models is far
206Box 2. Examples of the use of neutral theory
In the main text, we discussed several possible uses of neutral
theory. In particular, we focused on how neutral models have a
utility in cases where they succeed in fitting data and in cases where
they fail. Here, we illustrate this with some examples.
Although neutral theory with a basic spatial structure consisting
of a well-mixed metacommunity and a separate local community
can fit species abundances in a single forest plot, it cannot
simultaneously fit species abundances in three distinct forest plots
with the same parameters as obtained for a single plot [71]. This
highlights a role not only of niche differentiation, but also of spatial
structure that the spatially implicit model misses.
A spatially explicit neutral model that respects the network
structure of a river and its tributaries captures the species richness
of riverine fish very accurately. This implicates a key role of
dispersal and the spatial structure of the river network in explaining
riverine biodiversity [34].
A neutral model fails to explain the increase in abundance of
some tree species for the amount of time they have been present:
they increased their abundance faster than chance allowed and thus
are expected to have had a competitive advantage of some type
[6,7,51].
A neutral model provides a quantitative baseline against which
the pace of change in ancient communities, as indicated by fossil
data, can be compared [72]. In this capacity, neutral theory showed
Trends in Ecology and Evolution April 2012, Vol. 27, No. 4point or inspiration for more complex non-neutral models.
An important example is the development of nearly neu-
tral theory in ecology [61?63], and work calling for, or
attempting, a unification of niche and neutrality
[24,42,43,55,64?67]. Purves and Turnbull [13] argue that
a mechanism to maintain precisely equal fitness between
individuals (neutrality) over long periods is very difficult to
envisage. Both selection and ecological drift are indeed at
play in most ecosystems; so further development of nearly
neutral models is needed.
As a tool for integrating ecology and evolution
A further possible use of neutral theory is to approximate
macroevolutionary patterns and, in doing so, perhaps help
to unite ecology and evolution, which is a key goal of
modern biology [68]. However, a purely neutral model
demonstrates slow dynamics compared with those ob-
served in the natural world [18,19,69] and thus makes
unrealistic predictions about timescales; for example,
when applied to phylogenetic trees [52]. Environmental
stochasticity provides a possible neutral solution [70];
alternatively nearly neutral models [61] might help. Neu-
tral models might still be successful at describing macro-
ecology without invoking macroevolution; for instance, if
deep time has little effect (or a fast decaying effect) on
present-day macroecological summary data. This area
needs further investigation with testing against empirical
data. We agree that deep-time evolutionary predictions are
a significant problem for neutral theory as it stands, but
this is an interesting finding in its own right and it does not
yet represent an insurmountable problem or a cause to
abandon the theory and its other predictions.
Concluding remarks
Ecological inquiry is sufficiently diverse that researchers
will continue to need a multiplicity of approaches and
models [37]. Neutral theory must ultimately fall into place
as part of the spectrum of different tools that are available.
To achieve this will require moving beyond semantics, with
each position being clearly defined. Neutral theorists must
acknowledge the limitations of the theory and opponents
must acknowledge its uses. The time has now come to move
forward and embrace the neutral theory of ecology as one of
several tools that help advance understanding of biodiver-
sity. A concerted effort should be made to integrate the
ideas of both the proponents and opponents to converge on
new concepts that draw from the foundations laid down by
classic neutral theory and niche theory alike.
Acknowledgments
We thank Bart Haegeman, Omar Al Hammal, Egbert Leigh Jr, Albert
Phillimore and Robert Ricklefs for very useful comments on the
manuscript. We also thank Jonathan Eastman for interesting and
useful discussions. JR was funded by an EPSRC overseas postdoctoral
research fellowship at the life sciences interface (grant EP/F043112/1)
and is now funded by a NERC fellowship; RSE was funded by a VIDI
grant from NWO-ALW, LJH by NSF DEB-0919499 and FH by NSERC
(Canada).
References
1 Hubbell, S.P. (2001) The Unified Neutral Theory of Biodiversity and
Biogeography, Princeton University Press
2 McGill, B.J. (2003) A test of the unified neutral theory of biodiversity.
Nature 422, 881?885
Opinion3 McGill, B.J. et al. (2006) Empirical evaluation of neutral theory.
Ecology 87, 1411?14234 Volkov, I. et al. (2003) Neutral theory and relative species abundance in
ecology. Nature 424, 1035?1037
5 Etienne, R.S. and Olff, H. (2005) Confronting different models of
community structure to species-abundance data: a Bayesian model
comparison. Ecol. Lett. 8, 493?504
6 Leigh, E.G. (2007) Neutral theory: a historical perspective. J. Evol.
Biol. 20, 2075?2091
7 Leigh, E.G. et al. (2010) Unified neutral theory of biodiversity and
biogeography. Scholarpedia 5, 8822
8 Rosindell, J. et al. (2011) The unified neutral theory of biodiversity and
biogeography at age ten. Trends Ecol. Evol. 26, 340?348
9 Kopp, M. (2010) Speciation and the neutral theory of biodiversity.
Bioessays 32, 564?570
10 Clark, J.S. (2009) Beyond neutral science. Trends Ecol. Evol. 24, 8?15
11 Vellend, M. (2010) Conceptual synthesis in community ecology. Q. Rev.
Biol. 85, 183?206
12 McGill, B.J. and Nekola, J.C. (2010) Mechanisms in macroecology:
AWOL or purloined letter? Towards a pragmatic view of mechanism.
Oikos 119, 591?603
13 Purves, D.W. and Turnbull, L.A. (2010) Different but equal: the
implausible assumption at the heart of neutral theory. J. Anim.
Ecol. 79, 1215?1225
14 Ricklefs, R.E. (2010) Applying a regional community concept to forest
birds of eastern North America. Proc. Natl. Acad. Sci. U.S.A. 108,
2300?2305
15 Dietrich, M.R. (1994) The origins of the neutral theory of molecular
evolution. J. Hist. Biol. 27, 21?59
16 Rosindell, J. et al. (2010) Protracted speciation revitalizes the neutral
theory of biodiversity. Ecol. Lett. 13, 716?727
17 Leigh, E.G. et al. (1993) The decline of tree diversity on newly isolated
tropical islands ? a test of a null hypothesis and some implications.
Evol. Ecol. 7, 76?102
18 Ricklefs, R.E. (2006) The unified neutral theory of biodiversity: do the
numbers add up? Ecology 87, 1424?1431
19 Nee, S. (2005) The neutral theory of biodiversity: do the numbers add
up? Funct. Ecol. 19, 173?176
20 Wootton, T.J. (2005) Field parameterization and experimental test of
the neutral theory of biodiversity. Nature 433, 309?312
21 Dornelas, M. et al. (2006) Coral reef diversity refutes the neutral theory
of biodiversity. Nature 440, 80?82
22 Levin, S.A. (1992) The problem of pattern and scale in ecology. Ecology
73, 1943?1967
23 McGill, B.J. et al. (2007) Species abundance distributions: moving
beyond single prediction theories to integration within an ecological
framework. Ecol. Lett. 10, 995?1015
24 Adler, P.B. et al. (2007) A niche for neutrality. Ecol. Lett. 10, 95?
104
25 Magurran, A.E. (2005) Species abundance distributions: pattern or
process? Funct. Ecol. 19, 177?181
26 Du, X. et al. (2011) Negative density dependence can offset the effect of
species competitive asymmetry: a niche-based mechanism for neutral-
like patterns. J. Theor. Biol. 278, 127
27 Etienne, R.S. et al. (2007) Modes of speciation and the neutral theory of
biodiversity. Oikos 116, 241?258
28 Etienne, R.S. and Haegeman, B. (2011) The neutral theory of
biodiversity with random fission speciation. Theor. Ecol. 4, 87?109
29 MacArthur, R.H. (1957) On the relative abundance of bird species.
Proc. Natl. Acad. Sci. U.S.A. 43, 293?295
30 Cohen, J.E. (1968) Alternative derivations of a species?abundance
relation. Am. Nat. 102, 165?172
31 MacArthur, R.H. and Wilson, E.O. (1967) The Theory of Island
Biogeography, Princeton University Press
32 Hubbell, S.P. (2009) Neutral Theory and the Theory of Island
Biogeography. In The Theory of Island Biogeography Revisited
(Losos, J. and Ricklefs, R.E., eds), pp. 264?292, Princeton University
Press
33 Wennekes, P.L. et al. The neutral niche debate: a philosophical
perspective. Acta Biotheor, doi:10.1007/s10441-012-9144-6
34 Muneepeerakul, R. et al. (2008) Neutral metacommunity models
predict fish diversity patterns in Mississippi?Missouri basin. Nature
453, 220?222
Trends in Ecology and Evolution April 2012, Vol. 27, No. 435 Warren, R.J. et al. (2011) Universal ecological patterns in college
basketball communities. PLoS ONE 6, e17342
207
36 Nekola, J.C. and Brown, J.H. (2007) The wealth of species: ecological
communities, complex systems and the legacy of Frank Preston. Ecol.
Lett. 10, 188?196
37 Levins, R. (1968) Evolution in Changing Environments: Some
Theoretical Explorations, Princeton University Press
38 Friedman, M. (1966) The Methodology of Positive Economics,
University of Chicago Press
39 Nathan, R. and Muller-Landau, H.C. (2000) Spatial patterns of seed
dispersal, their determinants and consequences for recruitment.
Trends Ecol. Evol. 15, 278?285
40 Siepielski, A.M. et al. (2010) Experimental evidence for neutral
community dynamics governing an insect assemblage. Ecology 91,
847?857
41 Allouche, O. and Kadmon, R. (2009) A general framework for neutral
models of community dynamics. Ecol. Lett. 12, 1287?1297
42 Chisholm, R.A. and Pacala, S.W. (2010) Niche and neutral models
predict asymptotically equivalent species abundance distributions in
high-diversity ecological communities. Proc. Natl. Acad. Sci. U.S.A.
107, 15821?15825
43 Etienne, R.S. and Haegeman, B. (2011) Independent species in
independent niches behave neutrally. Oikos 120, 961?963
44 Etienne, R.S. et al. (2007) The zero-sum assumption in neutral
biodiversity theory. J. Theor. Biol. 248, 522?536
54 Rosindell, J. and Phillimore, A.B. (2011) A unified model of island
biogeography sheds light on the zone of radiation. Ecol. Lett. 14, 552?
560
55 Chase, J.M. (2005) Towards a really unified theory for metacommunities.
Funct. Ecol. 19, 182?186
56 Gotelli, N.J. and McGill, B.J. (2006) Null versus neutral models: what?s
the difference? Ecography 29, 793?800
57 Lande, R. et al. (2003) Stochastic Population Dynamics in Ecology and
Conservation, Oxford University Press
58 Tilman, D. et al. (1994) Habitat destruction and the extinction debt.
Nature 371, 65?66
59 Lewis, O.T. (2006) Climate change, species?area curves and the
extinction crisis. Philos. Trans. R. Soc. B 361, 163?171
60 He, F. and Hubbell, S.P. (2011) Species?area relationships always
overestimate extinction rates from habitat loss. Nature 473, 368?371
61 Zhou, S.R. and Zhang, D.Y. (2008) A nearly neutral model of
biodiversity. Ecology 89, 248?258
62 Lin, K. et al. (2009) Demographic trade-offs in a neutral model explain
death-rate?abundance-rank relationship. Ecology 90, 31?38
63 He, F. et al. (2012) Coexistence of nearly neutral species. J. Plant Ecol.
5, 72?81
64 Gravel, D. et al. (2006) Reconciling niche and neutrality: the continuum
hypothesis. Ecol. Lett. 9, 39?409
65 Leibold, M.A. and McPeek, M.A. (2006) Coexistence of the niche
Opinion Trends in Ecology and Evolution April 2012, Vol. 27, No. 445 Clark, J.S. et al. (2007) Resolving the biodiversity paradox. Ecol. Lett.
10, 647?659
46 Chave, J. and Leigh, E.G. (2002) A spatially explicit neutral model of
beta-diversity in tropical forests. Theor. Popul. Biol. 62, 152?168
47 Chave, J. et al. (2002) Comparing classical community models:
theoretical consequences for patterns of diversity. Am. Nat. 159, 1?23
48 Rosindell, J. and Cornell, S.J. (2007) Species?area relationships from a
spatially explicit neutral model in an infinite landscape. Ecol. Lett. 10,
586?595
49 Rosindell, J. and Cornell, S.J. (2009) Species?area curves, neutral
models and long distance dispersal. Ecology 90, 1743?1750
50 Keil, P. et al. (2010) Predictions of Taylor?s power law, density
dependence and pink noise from a neutrally modeled time series. J.
Theor. Biol. 265, 78?86
51 Gilbert, B. et al. (2006) Can neutral theory predict the responses of
Amazonian tree communities to forest fragmentation? Am. Nat. 168,
304?317
52 Davies, T.J. et al. (2011) Neutral biodiversity theory can explain the
imbalance of phylogenetic trees but not the tempo of their
diversification. Evolution 65, 1841?1850
53 Jabot, F. and Chave, J. (2009) Inferring the parameters of the neutral
theory of biodiversity using phylogenetic information and implications
for tropical forests. Ecol. Lett. 12, 239?248208and neutral perspectives in community ecology. Ecology 87, 1399?
1410
66 Holt, R.D. (2006) Emergent neutrality. Trends Ecol. Evol. 21, 531?
533
67 Holyoak, M. and Loreau, M. (2006) Reconciling empirical ecology with
neutral community models. Ecology 87, 1370?1377
68 McInnes, L. et al. (2011) Integrating ecology into macroevolutionary
research. Biol. Lett. 7, 644?646
69 Ricklefs, R.E. (2003) A comment on Hubbell?s zero-sum ecological drift
model. Oikos 100, 185?192
70 Allen, A.P. and Savage, V.M. (2007) Setting the absolute tempo of
biodiversity dynamics. Ecol. Lett. 10, 637?646
71 Etienne, R.S. (2007) A neutral sampling formula for multiple samples
and an ?exact? test of neutrality. Ecol. Lett. 10, 608?618
72 McGill, B.J. et al. (2005) Community inertia of Quaternary small
mammal assemblages in North America. Proc. Natl. Acad. Sci.
U.S.A. 102, 16701?16706
73 Pigolotti, S. and Cencini, M. (2009) Speciation-rate dependence in
species?area relationships. J. Theor. Biol. 260, 83?89
74 Etienne, R.S. and Rosindell, J. (2012) Prolonging the past counteracts
the pull of the present: protracted speciation can explain observed
slowdowns in diversification. Syst. Biol. 61, 204?213