A road map for synthesizing the scaling patterns in ecology

Cang Hui

arXiv (arXiv: 1310.2110v1)

Generated on April 19, 2025

Abstract
Ecology studies biodiversity in its variety and complexity. It describes how
species distribute and perform in response to environmental changes. Ecological
processes and structures are highly complex and adaptive. In order to quantify
emerging ecological patterns and investigate their hidden mechanisms, we need
to rely on the simplicity of mathematical language. This becomes especially
apparent when dealing with scaling patterns in ecology. Indeed, nearly all of
ecological patterns are scale dependent. Such scale dependence hampers our
predictive power and creates problems in our inference. This challenge calls
for a clear and fundamental understanding of how and why ecological patterns
change across scales. As Simon Levin stated in his MacArthur Award lecture, the
problem of relating phenomena across scales is the central problem in ecology
and other natural sciences. It has become clear that there is currently a drive
in ecology and complexity science to develop new quantitative approaches that
are suitable for analysing and forecasting patterns of ecological systems. Here
I provide a road map for future works on synthesizing the scaling patterns in
ecology, aiming (i) to collect and sort a diverse array of ecological patterns,
(ii) to present the dominant parametric forms of how these patterns change
across spatial and temporal scales, (iii) to detect the processes and
mechanisms using mathematical models, and finally (iv) to probe the physical
meaning of these scaling patterns. This road map is divided into three parts
and covers three main concepts of scale in ecology: heterogeneity, hierarchy
and size. Using scale as a thread, this road map and its following works weave
the kaleidoscope of ecological scaling patterns into a cohesive whole.

arXiv:1310.2110v1 [q-bio.PE] 8 Oct 2013A road map for synthesizing the scaling patterns in
ecology
Cang Hui
Centre for Invasion Biology, Department of Botany and Zoolog y,
Stellenbosch University, Matieland 7602, South Africa; E- mail: [email protected]
(Dated: July 27, 2021)
Ecology studies biodiversity in its variety and complexity . It describes how species distribute
and perform in response to environmental changes. Ecologic al processes and structures are
highly
complex andadaptive. Inorder toquantifyemerging ecologi cal patternsand investigate their
hidden
mechanisms, we need to rely on the simplicity of mathematica l language. This becomes
especially
apparent when dealing with scaling patterns in ecology. Ind eed, nearly all of ecological patterns
are
scale dependent. Such scale dependence hampers our predict ive power and creates problems in
our
inference. This challenge calls for a clear and fundamental understanding of how and why
ecological
patterns change across scales. As Simon Levin stated in his M acArthur Award lecture, the
problem
of relating phenomena across scales is the central problem i n ecology and other natural sciences.
It has become clear that there is currently a drive in ecology and complexity science to develop
new quantitative approaches that are suitable for analysin g and forecasting patterns of ecological
systems. Here I provide a road map for future works on synthes izing the scaling patterns in
ecology,
aiming (i) to collect and sort a diverse array of ecological p atterns, (ii) to present the dominant
parametric forms of how these patterns change across spatia l and temporal scales, (iii) to detect
the processes and mechanisms using mathematical models, an d finally (iv) to probe the physical
meaning of these scaling patterns. This road map is divided i nto three parts and covers three
main
concepts of scale in ecology: heterogeneity, hierarchy and size. Using scale as a thread, this road
map and its following works weave the kaleidoscope of ecolog ical scaling patterns into a
cohesive
whole.
PACS numbers: 87.23
Ecological patterns are emerging structures observed
in populations, communities and ecosystems. Elucidat-
ing drivers behind ecological patterns can greatly im-
prove our knowledge on how ecosystems assemble, func-
tion and respond to changeand perturbation. Due to the
non-random nature, most, if not all, ecological patterns
change with measurement, characteristic and organiza-
tion scales and exhibit distinct scaling properties. Such
scaling properties can be broadly grouped into patterns
related to heterogeneity, hierarchy and size. The road
map introduces the three groups of scaling patterns. The
emphasishereisnottoprovidesolutionstotheseoutlined
research questions; rather, by grouping relevant scaling
patterns under unique banners, I attempt to highlight
the challenges and connect the emerging clues for build-
ing a unified theory for scaling patterns in ecology in the
near future.
I. HETEROGENEITY
This section on the scaling pattern of heterogeneity
aims to investigate how aggregated structures of organ-
isms, diversity and ecosystem service change with mea-
surement scales and which/why biological patterns res-
onate with underlying processes at the same characteris-
tic scales.
Aggregation. Species distributions are not uniform
acrossspace,reflectingtheinterplaybetweenhabitathet-
erogeneity and the underlying nonlinear biotic regulation
[1]. Such non-random, aggregated patterns not only canbe the indicative of non-equilibrium
dynamics (e.g. dur-
ing range expansion [2–4]) but also self-organized pat-
tern emergence (e.g. [5–7]). When ecologists examine
species distributions across scales, the Modifiable Areal
Unit Problem presents itself [8, 9]. The problem can
be described as the change in species distribution char-
acteristics as the unit of measurement changes, both in
terms of size and shape of the sample unit. Such scaling
patterns of aggregation follow three general parametric
forms: logarithmic, power-law and lognormal shape [1].
Following on recent progress of using the Bayesian rule
for cross-scale extrapolation [10, 11], further advance-
ment in this field is to provide a consistent description of
aggregation when scaling up and down, and therefore a
universalbasisofcomparisonfordistributionsindiffering
contexts. A fully-functional model with predictive power
for up- and down-scaling species distribution is needed.
Under certainconditions, this model shouldfurther allow
extrapolating fine-scale occupancy and population densi-
ties from coarse-scale observations (e.g. [12, 13]). Great
potential exists to apply such a predictive model in vari-
ous cross-scale pattern analyses espcially when detection
is imperfect (e.g. [14, 15]).
Space-for-time substitution. The directionality of
community succession is an important concept in conser-
vation biology [16, 17]; it is analogous to the irreversibil-
ity of time in physics that has revolutionised the under-
standingofcomplexadaptivesystems[18]. By definition,
succession is an orderly process of community change af-
ter disturbance [16]. Knowing the directionality of suc-
cession is necessary for (i) distinguishing new from ma-

2
ture communities (i.e. defining the age of a community),
(ii) understanding how communities evolve and respond
todisturbance(e.g. habitatlossandclimatechange),and
(iii) designing more efficient conservation and restora-
tion plans [19]. However, popularising the concept of
directionality in succession is challenging for two rea-
sons. Firstly, acknowledging this concept demands the
acceptance of inherent bias in nature which contradicts
the null hypothesis of a random and isotropic world [18].
Secondly, appropriatelong-termdatarequiredfordetect-
ing the succession direction are scarce, and indices and
analytical methods for such computation are lacking. To
this end designing alternative tests (e.g. the space-for-
time substitution; [17] that can capture the essence of
directionality and irreversibility in community develop-
ment but which can be applied to available data becomes
crucial.
The spatial and temporal scales of ecological processes
are intertwined. Processes that account for the spatial
distribution of species also underpin its temporal dy-
namics. This means that we can potentially forecast
the future or rebuild the history based solely on current
spatial distribution, without resorting to long-term time
series [20, 21]. In other words, the need to wait years
and decades to measure changes in distribution can be
averted through the ability to make sufficiently accu-
rate predictions based only on the spatial distribution
of species at the current time. As the ability to forecast
the temporal trend of a focal species provides crucial in-
formation on its performance and viability [20, 22], the
methodology of space-for-time substitution is extremely
appealing, especiallybecause ourability to obtain spatial
records has been drastically improved. This area of re-
searchcallsforamodelthatcanrelatethescalingpattern
of species current spatial distribution to the near-future
population trend and performance.
Scale resonance. Just as two tuning forks of the
same characteristic frequency resonate, so do ecological
patternsandprocessesworkingatthesamescale. Species
distributions are regulated by a variety of abiotic and bi-
otic processes working in concert but at different scales
[23]. Those processes identified as key biotic drivers us-
ing methods such asmultivariate statistics often resonate
with the scale of the study. That is, information be-
ing picked up represents a product of the measurement
method, rather than the intrinsic cross-scale mechanism.
Suchapatternofscaleresonancehasbeenobservedwhen
synthesizing a series of collaborative works on identify-
ing the factors of the distribution of Argentine ants at
local [24], regional [25] and global scales [26]. This find-
ingbringsintoquestionmanyregionalmanagementplan-
ningpracticesthatarebasedontheupscaleextrapolation
of local-scale studies. Future research needs to explore
the mechanism behind scale resonance in ecology and to
present a statistical remedy for cross-scale inference.
Co-distribution. To exploit resources while miti-
gating conflict, species often partition available habi-
tats, forming co-distribution patterns of association ordissociation. Null models based on
permutation have
been widely applied for detecting signals of association
or dissociation from co-distribution patterns, from which
the type of biotic interactions can be inferred. Fu-
ture research needs to present a model that incorpo-
rates biotic interactions and also captures the transi-
tion from fine-scale dissociation to coarse-scale associ-
ation (e.g. [11]), explaining why this co-distribution
pattern changes across scales and how this scale depen-
dence affects the pairwise measure of species turnover.
It should be reconcile the debate between the Rich-Get-
Richer phenomenon in invasion biology and the opposing
Competitive-Exclusion-Principle.
Biodiversity. Species diversity patterns, such as the
species-area curve [27], endemics-area relationship, dis-
tance decay of similarity and occupancy frequency dis-
tribution [28, 29], are just a few interrelated patterns of
scale dependence emerging from complex ecological sys-
tems. The integration of patterns of species diversity
patterns is central to understanding the processes that
drive species assembly [30]. Changing the measurement
scale will lead to a coordinated change in all diversity
patterns. I envisage a new diversity pattern delta diver-
sity that connects all commonly known diversity scaling
patterns, using delta diversity as building blocks. This
model should be able to further explain Raunkiaers bi-
modal law of frequency [31] and resolves the debate on
the ceiling of species richness in a community.
Ecosystem function and service Ecosystem ser-
vicesareby-productsfromthe function ofecosystempro-
cesses that sustain the basic needs of humans and their
socioeconomic activities. According to the Millennium
Ecosystem Assessment, ecosystem services are generated
frominteractionsrangingfromspecialisttaxatoallbiodi-
versity, and the functional units of the variety of services
rangefromlocalpopulationstoglobalbiogeochemicalcy-
cles. At the local scale, we benefit from services of polli-
nation, pest control, soil fertility and seed dispersal that
are related to biodiversity. At regional scale, we bene-
fit from services of air and water purification, flood and
drought mitigation, and waste decomposition that are
delivered by plants and micro-organisms. At the global
level, we benefit from services of climate stability and
UV protection from plants and biogeochemical cycles.
Understanding how different ecosystem services change
with spatial scales and potentially conceptualizing into a
model for extrapolating the level of service across scales
warrantsgreatattention. Recentstudiesshowedthescale
dependence of acrude ecosystemserviceindicator bioca-
pacity and the resultant sustainability index [32, 33], and
the robustness and invasibility of recipient ecosystems
to biological invasions [34]. Exploiting ecosystem ser-
vices within their maximum sustainable level can ensure
a reliable service provision without triggering a regime
shift. Achieving this balance is a challenge for conserva-
tion management and sustainable development.

3
II. HIERARCHY
The scaling pattern of hierarchydepicts how the struc-
ture and function of asymmetrical ecological systems
emerge and change with the system complexity. Us-
ing ecological networks as the model system, this sec-
tion aims to investigate how cascade interactions affects
the robustness and resilience of networks, how network
architectures, especially nestedness and compartmental-
ization, emerge and function, and the role of network
complexity on the stability of ecological networks.
Cascade. Nodes and edges of a network are a good
proxy of species and their interactions in an ecosystem.
Probingprocessesthatcanleadtotheemergenceoflarge-
scale network architectures, e.g. small-world networks
and scale-free networks, is a new wave in science. Scal-
ing laws of food webs depict how a biological network
behaves as a function of its complexity. In reconciling
with Mays stability criterion of complex systems [35],
Cohenscascademodel [36] and its laterdevelopment pro-
vide a phenomenological explanation of some of the scal-
ing laws and scaling invariant patterns. The principle of
Maximum Entropy that identifies the unbiased estimates
under constraints has been widely used in ecology. Eco-
logical networks are efficient energy transporting, non-
equilibrium systems that are adaptive to changes and
disturbances. This calls for the development of models
based on the recently developed principle of Maximum
Entropy Production in non-equilibrium thermodynamics
[37], in an attempt to explain the pattern emergence in
complex adaptive networks, in particular, the cascade in-
teractionsin food webs. Such models will providea phys-
ical understanding on the network emergence and shed
light on solving the complexity-stability debate.
Nestedness. Nested structure has been observed in
many networks, in particular bipartite mutualistic net-
works (e.g. pollination networks and seed-dispersal net-
works [38]). To have models with quantitative accu-
racy and predictive power, one needs to rely on process-
based models. A key feature in ecological networks is the
adaptive and innovative nature of the edges and nodes.
Species are constantly optimizing which partners they
shouldinteractwith formaximumfitness gain, asaccord-
ing to the optimal foraging theory. This can be achieved
at two time scales: 1) At the ecological time scale, in-
teraction switching reflects seeking optimal fitness gain
[39, 40]; 2) At the evolutionary time scale, interaction
switching reflects coevolutionary dynamics between in-
teracting traits.
Compartmentalization. One important hierarchi-
cal structure in ecological networks is compartmentaliza-
tion, i.e. the formation of functional modules, where in-
teractions are most likely to occur within the same mod-
ule, that is to say like is connected to like in a network
[41]. Energy flows directionally in resource-exploitation
networks (e.g. host-parasite networks), forming com-
partmentalized structures. This research area calls for
a process-based model that optimizes energy transportvia adaptive interaction switching partners
for local fit-
ness gain. The strong predictive power requires further
demonstration by comparing the level of compartmental-
ization observed versus predicted for real networks [42].
This, together with the previous two research areas, pro-
vides a physical understanding of pattern emergence in
complex adaptive networks.
III. SIZE
This section on the scaling pattern of size investigates
how form and function change as organisms get larger
or as their traits change. The section aims to investi-
gate how allometric laws of metabolism emerge and how
phenotypes of biological traits form through co-evolution
with other traits, species and the environment and how
this affects the path of evolution and diversification.
Allometry. Allometric scalingis the most salient pat-
ternofhowbiologicalrates, especiallymetabolicrate, are
regulated by organism size [43], known as Kleibers law.
A model, depicting how fractal-structured circulatory or
vascular systems transport energy and matter through-
out the body of the organism, has been proposed for ex-
plainingthe emergenceofallometricscaling, which is fur-
ther summarized succinctly as the Fourth Dimension of
Life [44]. Recently, a not fully developed concept sug-
gests that natural selection can increase the ontogenetic
and population growth rates to a ceiling, above which
the population will crash due to intrinsic instability [45],
in principle similar to the origin of planetary rings that
are only distributed on a thin surface surrounding the
planet, with all other directions, although possible, have
been eliminated through inter-particle collisions. This
progress calls for further development of a conceptual
model for allometric scaling based on the principle of
Maximum Entropy Production and self-organized criti-
cality. This model should be able to simulate the process
of cell differentiation in multicellular organisms.
Trait.This research area focuses on addressing two
questions. First, what are the mechanisms determining
phenotypic traits? Evolution via natural selection relies
on heritable phenotypic variation and has long been re-
gardedas being solely reliant on direct expressionof gene
variation. This assumption may be an exaggeration, as
factors other than genetic variation can also dictate the
outcomes of natural selection and thus evolution. At the
forefront of these alternative platforms is the rapidly ex-
panding field ofepigenetics the biochemicalmodification
of DNA without changes in its sequence that gives rise
to differential gene expression and thus different pheno-
types [46]. A revised Price equation that incorporates
epigeneticmechanismforexplainingphenotypicvariation
is needed. Second, how do traits affect biotic interac-
tions and thus the path of evolution? Biotic interactions,
largely to do with resource competition and exploitation,
arerealised by specific phenotypic traits that areused for
searching and handling resources, for competing and de-

4
fending resources, for memorizing and recognizing ben-
eficial resources, and for escaping from being exploited
as resources by predators or parasites. In this regard,
Adaptive Dynamics is a power tool to explore how trait-
mediated interactions can lead to diverse evolutionary
phenomena [47], from the Red Queen dynamics to speci-
ation [48, 49].
IV. EPILOGUE
Nature never fails to amaze us. It coordinates the
interplay of numerous organisms in their environments,
formingacomplexfunctionalsystemthat sustainsus and
many other species via ecosystem service. As the ulti-
mate goal of natural sciences, quantifying emerging pat-
terns in nature and understanding hidden mechanisms
arethe pinnacle ofscience. An importantphenomenonin
quantifying ecological patterns is that nearly all of themare scale dependent. This scale
dependence creates prob-
lems in our inference, yet also simultaneously provides
opportunities for us to pry into the core of how nature
assembles, organizes and functions. Once again, we need
to relate these research areas back to the overarching re-
search philosophy in studying natural systems: (i) what
patterns exist in nature (i.e. using statistical methods to
measure and quantify ecological patterns); (ii) how such
patterns emerge (i.e. proposing mathematical models to
unveil mechanisms driving the ecological pattern forma-
tion; (iii) why nature organizes itself in such a way (i.e.
using physicallawsto revealthe adaptiveand/oroptimal
nature of ecological systems). This has been eloquently
presented as the Ouroborosof scales in Just Six Numbers
by Martin Rees. William Blake also expressed the role
of scales in his Auguries of Innocence in 1803, ‘To see a
world in a grain of sand, and a heaven in a wild flower;
Hold infinity in the palm of your hand, and eternity in
an hour’.
[1] Hui, C., Veldtman, R., McGeoch, M.A. (2010) Measures,
perceptions and scaling patterns of aggregated species
distributions. Ecography , 33: 95-102.
[2] Ramanantoanina, A., Hui, C., Ouhinou, A. (2011)
Effects of density-dependent dispersal behaviours on
the speed and spatial patterns of range expansion
in predator-prey metapopulations. Ecological Modelling ,
222: 3524-3530.
[3] Hui, C., Roura-Pascual, N., Brotons, L., Robinson, R.A. ,
Evans, K.L. (2012) Flexible dispersal strategies in native
and non-native ranges: environmental quality and the
good-stay, bad-disperse rule. Ecography , 35: 1024-1032.
[4] Berthouly-Salazar, C., van Rensburg, B.J., le Roux, J.J .
van Vuuren, B.J., Hui, C. (2012) Spatial sorting drives
morphological variation intheinvasivebird, Acridothere s
tristis.PLoS ONE , 7: e38145.
[5] Zhang, F., Hui, C. (2011) Eco-evolutionary feedback and
the invasion of cooperation in prisoners dilemma games.
PLoS ONE , 6: e27523.
[6] Hui, C., Yue, D.X. (2005) Niche construction and poly-
morphism maintenance in metapopulations. Ecological
Research , 20: 115-119.
[7] Hui, C., Li, Z.Z., Yue, D.X. (2004) Metapopulation dy-
namics and distribution, and environmental heterogene-
ity induced by niche construction. Ecological Modelling ,
177: 107-118.
[8] Hui, C. (2009) A Bayesian solution to the modifiable
areal unit problem. In: Hassanien, A.-E., Abraham, A.,
Herrera, F. (eds.) Foundations of Computational Intel-
ligence Volume 2: Approximate Reasoning , pp. 175-196.
Springer-Verlag, Berlin.
[9] Openshaw, S. (1984) The modifiable areal unit problem .
GeoBooks.
[10] Hui, C., McGeoch, M.A., Warren, M. (2006) A spatially
explicit approach to estimating species occupancy and
spatial correlation. Journal of Animal Ecology , 75: 140-
147.[11] Hui, C. (2009) On the scaling pattern of species spatial
distribution and association. Journal of Theoretical Biol-
ogy, 261: 481-487.
[12] Hui, C., McGeoch, M.A., Reyers, B., le Roux, P.C.,
Greve, M., Chown, S.L. (2009) Extrapolating population
size from the occupancy-abundance relationship and the
scaling pattern of occupancy. Ecological Applications , 19:
2038-2048.
[13] Hui, C., McGeoch, M.A. (2007) Capturing the ‘droopy
tail’ in the occupancy-abundance relationship. Eco-
Science, 14: 103-108.
[14] Hui, C., Foxcroft, L.C., Richardson, D.M., MacFadyen,
S. (2014) A cross-scale approach for abundance estima-
tion of invasive alien plants in a large protected area. In:
L.C. Foxcroft, D.M. Richardson, P. Pysek, Genovesi, eds.
Plant Invasions in Protected Areas: Patterns, Problems
and Challenges . Springer, Berlin. DOI 10.1007/978-94-
007-7750-7-5
[15] Hui, C., Foxcroft, L.C., Richardson, D.M., MacFadyen,
S. (2011) Defining optimal sampling effort for large-scale
monitoring of invasive alien plants: a Bayesian method
for estimating abundance and distribution. Journal of
Applied Ecology , 48: 768-776.
[16] Odum E.P. (1969). The strategy of ecosystem develop-
ment. Science, 164: 262-270.
[17] Pickett S.T.A., McDonnell M.J. (1989). Changing Per-
spectives in Community Dynamics - a Theory of Succes-
sional Forces. Trends in Ecology and Evolution , 4: 241-
245.
[18] Jorgensen S.E., Fath B.D., Bastianoni S., Marques J.C. ,
Muller F., Nielsen S.N., Patten B.C., Tiezzi E., Ulanow-
icz R.E. (2007). A new ecology: systems perspective . El-
sevier, Amsterdam.
[19] Roura-Pascual, N., Krug, R.M., Richardson, D.M., Hui,
C. (2010) Spatially-explicit sensitivity analysis for con -
servation management: exploring the influence of deci-
sion in invasion alien plant management. Diversity and
Distributions , 16: 426-438.

5
[20] Hui, C. (2011) Forecasting population trend from the
scaling pattern of occupancy. Ecological Modelling , 222:
442-446.
[21] Hui, C., Boonzaaier, C., Boyero, L. (2012) Estimating
changes in species abundance from occupancy and ag-
gregation. Basic and Applied Ecology , 13: 169-177.
[22] Hui, C., Richardson, D.M., Robertson, M.P., Wilson,
J.R.U., Yates, C.J. (2011) Macroecology meets invasion
ecology: linking the native distributions of Australian
acacias to invasiveness. Diversity and Distributions , 17:
872-883.
[23] McGill B.J. (2010) Matters of Scale. Science, 328, 575-
576.
[24] Roura-Pascual, N., Bas, J.M., Hui, C. (2010) The spread
of the Argentine ant: environmental determinants and
impacts on native ant communities. Biological Invasions ,
12: 2399-2412.
[25] Roura-Pascual, N., Bas, J.M., Thuiller, W., Hui, C.,
Krug, R.M., Brotons, L. (2009) From introduction to
equilibrium: reconstructing the invasive pathways of the
Argentine ant in a Mediterranean region. Global Change
Biology, 15: 2101-2115.
[26] Roura-Pascual, N., Hui, C., Ikeda, T., Leday, G.,
Richardson, D.M., Carpintero, S., Espadaler, X., Gmez,
C., Gunard, B., Hartley, S., Krushelnycky, P., Lester,
P.J., McGeoch, M.A., Menke, S.B., Pedersen, J.S., Pitt,
J.P., Reyes, J., Sanders, N.J., Suarez, A.V., Touyama,
Y., Ward, D., Ward, P.S., Worner, S.P. (2011) The rel-
ative roles of climate suitability and anthropogenic in-
fluence in determining the pattern of spread in a global
invader. Proc. Natl Acad. Sci. USA , 108: 220-225.
[27] Hui, C. (2008) On species-area and species accumulatio n
curves: A comment on Chong and Stohlgrens index. Eco-
logical Indicators , 8: 327-329.
[28] Hui, C., McGeoch, M.A. (2007) Modeling species dis-
tributions by breaking the assumption of self-similarity.
Oikos, 116: 2097-2107.
[29] Hui, C., McGeoch, M.A. (2007) A self-similarity model
for occupancy frequency distribution. Theoretical Popu-
lation Biology , 71: 61-70.
[30] Hui, C., McGeoch, M.A. (2008) Does the self-similar
species distribution model leadtounrealistic prediction s?
Ecology, 89: 2946-2952.
[31] Hui, C. (2012) Scale effect and bimodality in the fre-
quency distribution of species occupancy. Community
Ecology, 13: 30-35.
[32] Yue, D.X., Guo, J.J., Hui, C. (2013) Scale dependency
of biocapacity and the fallacy of unsustainable develop-
ment.Journal of Environmental Management , 126: 13-
19.
[33] Yue, D.X., Xu, X.F., Hui, C., Xiong, Y.C., Han, X.M.,
Ma, J.H. (2011) Biocapacity supply and demand in
Northwestern China: aspatial appraisal of sustainability .
Ecological Economics , 70: 988-994.[34] Gaertner, M., den Breeyen, A., Hui, C., Richardson,
D.M. (2009) Impacts of alien plant invasions on species
richness in Mediterranean-type ecosystems: a meta-
analysis. Progress in Physical Geography , 33: 319-338.
[35] May, R.M. (1973) Stability and complexity in model
ecosystems . Princeton University Press.
[36] Cohen, J.E., Luczak, T., Newman, C.M., Zhou, Z.M.
(1990) Stochastic structure and nonlinear dynamics of
food webs: qualitative stability in a Lotka-Volterra cas-
cade model. Proceedings of the Royal Society of London,
Series B , 240: 607-627.
[37] Dewar, R.C. (2003) Information theory explanation of
the fluctuation theorem, maximum entropy production
and self-organized criticality in non-equilibrium statio n-
ary states. Journal of Physics A: Mathematical and Gen-
eral, 36: 631-641.
[38] Bascompte, J., Jordano, P., Melian, C.J., Olesen, J.M.
(2003). The nested assembly of plantanimal mutualistic
networks. Proc. Natl Acad. Sci. USA , 100, 93839387.
[39] Zhang, F., Hui, C., Terblanche, J.S. (2011) An interac-
tion switchpredictsthenestedarchitecture ofmutualisti c
networks. Ecology Letters , 14: 797-803.
[40] Zhang, F., Hui, C. (2013) Recent experience-driven be-
haviour optimizes foraging. Animal Behaviour , in press.
[41] Hui, C., Richardson, D.M., Pysek, P., Le Roux, J.J.,
Kucera, T., Jarosik, V.(2013) Increasingfunctional mod-
ularitywithresidencetimeintheco-distributionofnativ e
and introduced vascular plants. Nature Communications ,
4: 2454.
[42] Nuwagaba, S., Zhang, F., Hui, C. (2014) The emer-
gence of compartmentalization in antagonistic networks
via adaptive interaction switch. Submitted .
[43] Peters, R.H. (1983) The Ecological Implications of Body
Size. Cambridge University Press, Cambridge.
[44] West, G.B., Brown, J.H., Enquist, B.J. (1997) A general
model for the origin of allometric scaling laws in biology.
Science, 276: 122-126.
[45] Ginzburg, L.R., Burger, O., Damuth, J. (2010) The
May threshold and life-history allometry. Biology Letters ,
DOI:10.1098/rsbl.2010.0452
[46] Genereux, D.P., Miner, B.E., Bergstrom, C.T., Laird,
C.D. (2005) A population-epigenetic model to infer site-
specific methylation rates from double-stranded DNA
methylation patterns. Proc. Natl. Acad. Sci. USA , 102:
5802-5807.
[47] Dieckmann, U., Law, R. (1996) The dynamical theory
of coevolution: a derivation from stochastic ecological
processes. Journal of Mathematical Biology , 34: 579612.
[48] Hui, C., McGeoch, M.A. (2006) Evolution of body size,
range size and food composition in a predator-prey
metapopulation. Ecological Complexity , 3: 148-159.
[49] Zhang, F., Hui, C., Pauw, A. (2013) Adaptive divergence
in Darwins race: how coevolution can generate trait di-
versity in a pollination system. Evolution , 67: 548-560.