Community Ecology
My early work on stochastic community models
(McKane
et al 2000)
and metapopulation models
(Alonso
and McKane 2002)
establishes a stochastic mathematical framework to analyze and model complex
ecological interactions ("The
Stochastic Nature of Ecological
Interactions", PhD Thesis, Alonso 2004).
My approach emphasizes the discrete nature of individuals and
their inherent random interactions. My
most influential contributions in
community ecology are rather recent. In
collaboration with
Rampal
Etienne form the
University of
Groningen,
we have developed a general sampling theory which takes into account the
sampling, ecological and evolutionary processes that
ultimately determine the number and relative frequencies of the
species (or strains) arising in a typical sample from a given
community
(Etienne
et al 2007).
The Neutral Theory of
Biodiversity
Neutral theory explores the
theoretical consequences of
assuming that communities are
evolutionarily shaped by random speciation
and ecologically assembled by
random dispersal and random
ecological drift
(Hubbell 2001, Thomson
Special
Topics,
Etienne
and
Alonso
2007).
In the context of this theory, we have shown
that samples from any local community can be seen as a result of
a dispersal-limited sampling
from a larger biogeographical
area
(Alonso
and
McKane
2004,
Etienne
and
Alonso
2005,
Alonso
et al
2006). This new
formulation of sampling has applications both to community
ecology and population genetics and provides a
unifying theoretical framework where other factors beyond
neutrality can be easily considered.
Simple models for food web structure
The topology of food webs is a
major determinant of
ecosystems dynamics and
is ultimately responsible
for their responses to human
impacts (Allesina
et al
2008).
Several simple models have
been proposed for the intricate food webs
observed in nature. However,
none of these previous models
is fully compatible with food
web data. In collaboration
with
Stefano
Allesina
and
Mercedes
Pascual, we
have developed a method to identify
observed trophic links that are
incompatible with a given
model and a likelihood approach to compare different
simple models based on the
full structure of the network.
These results motivated a new
general model that is able to
generate all empirical data
sets and does so with the
highest likelihood. We hope
that the derivation of the
likelihood for simple models
of network structure will
contribute to a better
understanding of the
relationship between
structure and dynamics of
food webs (Allesina
et al
2008).
