Theoretical Ecology Lab Tea

                                                        Spring 2002

Wednesday, January 9, at 2 PM -- winter recess--
Wednesday, January 16 , at 2 PM Ian Rozdilsky
Wednesday, January 23 , at 2 PM Peter Walsh
Wednesday, January 30 , at 2 PM  (no tea)
Wednesday, February 6 , at 2 PM  (no tea)
Wednesday, February 13 , at 2 PM
Juan Keymer
Wednesday, February 20 , at 2 PM Kai Chan
Wednesday, February 27 , at 2 PM Chris Klausmeier
Wednesday, March 6 , at 2 PM Jonathan Dushoff
Wednesday, March 13 , at 2 PM Eduardo Zea
Wednesday, March 20 , at 2 PM  --- spring break ---
Wednesday, March 27 , at 2 PM David Sumpter (Oxford University)
Wednesday, April 3 , at 2 PM Ben Strauss
Wednesday, April 10 , at 2 PM Colleen Webb
Wednesday, April 17 , at 2 PM Glenn Adelson (Harvard University)
Wednesday, April 24 , at 2 PM Irakli Loladze
Wednesday, May 1 , at 2 PM (no labtea)
Wednesday, May 8 , at 2 PM Erik van Nimwegen (Rockefeller)
Wednesday, May 15 at 12:30 PM Lora Billings (Montclair State University)
Wednesday, May 22 , at 2 PM David Smith (Univ. of Maryland)
Wednesday, May 29 , at 2 PM
Adrew Irwin (Rutgers)


Titles and abstracts most recent first (posted approximately one week before the talk):

Wednesday, January 16 @ 2 pm

Ian Rozdilsky

For the lab tea presentation at 2pm this Wednesday, Ian will be making a
casual presentation on some of his previous work related to stability
properties of dynamical systems, and introduce the background and several
framework models for a project related to the Yellowstone to Yukon (Y2Y)
conservation initiative.

After making a closer analysis of the classical stability / complexity
relationships put forth by May, it was surprising to find that one of the
simplest dynamical models (the multi-species Lotka-Volterra system) exhibits
quite different behavior.  In fact, stability is shown for this example to
increase with greater complexity measured in terms of (high levels of)
connectance.  This observation led to a closer study of "compartmentalized"
food webs, and "compartmentalization" has been found to increase a wide
variety of stability measures.  Then considering more complex dynamical
models, namely multiple stable states, a statistical test has been developed
to determine if time series exhibit this dynamical property. These 3
subjects will be briefly touched upon.

Practical reserve design is a problem plagued by poor quality data and
political infighting.   Over this past summer, a preliminary analysis for
the possibilities of successful reserve design was carried out for the
northern Rocky Mountains. The northern Rocky Mountain ecosystem and some of
the background politics will be introduced, along with the directions of
current conservation efforts. Given this background I will propose several
possible directions that conservation, through reserve design, can proceed
and look forward to comments from the audience.

Wednesday, January 23 @ 2 pm

Peter Walsh

Large Scale Elephant Eanging and Tree Dynamics in a Congo Forest

Large vertebrate species are in the midst of a precipitous worldwide
decline. Particularly hard hit are the tropics, where a booming bushmeat
industry has left many forests structurally intact but devoid of large
animals. An urgent question is, then, whether these ^empty^ forests can
sustain their species diversity in the long term. Can smaller species
easily compensate? Or, do larger vertebrates tend to be critical to the
functioning of tropical forest communities?

Answers to these questions may lie in the role of large vertebrates as
seed predators. Tropical trees display two contrasting strategies for
escaping predation. "Mast" fruiting species set seed at long intervals to
depress predator population size and highly synchronously to satiate
predators alive at the time of seed set. Other species escape in space,
dispersing their seed widely so that seeds and seedlings are either too
far from conspecifics for host-specific pathogens and predators to find or
too rare for generalist predators to profitably search for. Many spatial
escape species even convert potential predators into mutualists by
providing a fruit reward in return for seed dispersal. Here I argue that a
major axis of competition in tropical forests may be between the guild of
spatial escape trees that cooperate to maintain populations of their
shared seed dispersers and the guild of time escape trees that seek to
suppress predators that include the seed dispersers of their spatial
escape competitors. Vertebrate body size may influence this competition
through its affect on ranging scale, with larger animals both providing
spatial escape trees a suite of long distance dispersal advantages and
acting as particularly efficient seed predators.

In this talk I will present data collected in Northern Congo supporting an
effect of vertebrate ranging scale on tree dynamics. In particular, I will
show that:
1)Elephants disperse large quantities of seed from many tree species over
unprecedented distances.
2)Elephant dispersed species are less aggregated in space than species
dispersed by other animals, wind, or gravity.
3)Similarity in species composition decays less rapidly in space for
elephant dispersed species than for other species.
4)Within species fruiting synchrony shows a different pattern of spatial
autocorrelation in elephant dispersed species than other species.

If Josh let^s me get this far, I will then briefly discuss plans for a
coupled tree-animal model that will be used to examine how removal of
large vertebrate species will affect the stability of tropical forest

Wednesday, February 13 @ 2 pm

Juan Keymer

Adaptive dynamics and particle mechanics:
Is the Mean-field an ESS?

I will be talking about the unit of selection problem and fitness measures
in spatially explicit models as well as evolutionary game theory and its
relationships (assumptions about genetic architectures) with adaptive
dynamics of continuous traits. These traits (fecundity, mortality, and
dispersal) specify the rules of particle mechanics of evolving agents
playing the "space-time-capture" game (preemtive competition for space).
The topology of spatio-temporal interactions is evolved among strategies
otherwise equivalent (same R0) under mean-field settings. In these cases,
the nature of the "interactors" (in the sense of Hull) lead some authors
(van Baalen and Rand) to define eigenvectors or pair-densities as "the
unit of selection". In the model I consider however, interaction takes
place at multiple timescales (making the pair approximation to break-down)
and thus the unit of selection (or interaction) becames diluted among
those scales--Indeed the spatio-temporal scales of interaction are
themselves evolving. The role of multiple scale pair-approximations is
discussed to derive fitness measures and detect units of evolutionary
activity at multiple scales.

Wednesday, February 20 @ 2 pm

Kai Chan

Phylogenies--as histories of the divergence of lineages--provide a wealth
of information critical to the study of diversification.  Variation in
rates of diversification between lineages leaves characteristic patterns
in phylogenies: asymmetric distributions of species among lineages, and
skewed distributions of branch lengths (periods between diversification
events). Consequently, we can detect shifts in diversification rate by
evaluating the symmetry of trees, or their distribution of branch lengths.
I will summarise work that Brian Moore and I have been doing to advance
the study of diversification based on tree symmetry.  We have (1)
developed three new symmetry metrics; (2) demonstrated these metrics using
recently obtained supertrees for bats, carnivores, lagomorphs, and
primates; (3) implemented the statistics in a program for analysing trees
against the expectations of a commonly applied null model; and (4)
assessed their power to shifts in diversification rate by simulating tree
growth under several alternative models.

Other important factors can also affect tree shape.  The null model of
cladogenesis implicitly assumes that there is no gene flow between
lineages. Violations of this assumption are common, but the implications
for phylogenetics have been largely ignored.  I would like to get input on
a model that I am working on that is intended to assess the
macroevolutionary implications of microevolutionary processes.

Wednesday, February 27 @ 2 pm

Chris Klausmeier

Nonequilibrium Foodwebs

Planktonic communities are subject to environmental forcing at frequencies
that span many orders of magnitude.  First, I'll talk about the impact of
fluctuating light on phytoplankton competition.  Using a combination of
numerical and analytical approaches, we (Litchman & Klausmeier 2001) found
that fast (daily) fluctuations can alter the identity of the superior
competitor, but not lead to stable coexistence; slow (seasonal)
fluctuations can also alter the competitive dominant as well as permit the
stable coexistence of multiple competitors.  Second, I'll present
work-in-progress on understanding seasonal succession in a more
complicated food web consisting of a basal resource, two phytoplankton
species, and a generalist herbivore.

  Wednesday, March 6 @ 2 pm

Jonathan Dushoff


The human Influenza A virus displays immunologically significant
variation on two scales:  continual, gradual changes (drift) and more
abrupt changes, associated with reassortment with non-human viruses.
The extent, and population-scale implications, of immunological
cross-reactions between strains of the virus is poorly understood.

Scaling arguments show that (in contrast to SIS models) even the
simplest stochastic SIR models can have extremely large critical
community sizes, implying that it is not safe to ignore stochasticity in
SIR modeling.  I will discuss how stochastic modeling might shed light
on the evolutionary ecology of influenza, and present a small amount of
work (in progress, as foreshadowed) on this topic.

  Wednesday, March 13 @ 2 pm

Eduardo Zea

Linking plant morpho-physiology and ecohydrological theory

Ecohydrological theory characetrizes plant water use strategies in terms
of a few summary features of plant water use (e.g. maximum daily
transpiration, wilting point, stress soil moisture). Using this high level
of abstraction the theory is able to deal analitically with stochasticity
in soil water availability. In this study we ask the question: how do real
plants "implement" these summary features of water use? We use a detailed
model of plant photosynthesis and water transport to explore how different
trait combinations determine different summary water use features. Then we
analize the effect of different trait combinations on the probabilistic
structure of soil moisture, assimilation and transpiration (i.e pdf,

Work done in collaboration with Edoardo Daly and Ignacio R-Iturbe.

Wednesday, March 27 @ 2 pm

David Sumpter


 Wednesday, April 3 @ 2 pm

Ben Strauss


Wednesday, April 10 @ 2 pm

Colleen Webb


Wednesday, April 17 @ 2 pm

Glenn Adelson

Axes of Diversity

"'Diversity' is a term with no essential philosophical, political, or aesthetic content."

The purpose of my talk is to explore whether Menand could have added "scientific" to his list, and if the
answer is "yes," what are the implications for biodiversity studies and conservation?

My initial response is that the answer is, in fact, "yes," but that such an answer opens up as opposed to
closes off useful discussion of diversity in biology.  Diversity has no essential scientific content
because there are too many kinds of diversity, which I will call axes of diversity, and only after you
choose which axis --  species richness, structural diversity, genetic diversity within populations,
genetic diversity between populations, taxonomic dispersion, etc, -- that you can begin to provide

An interesting aspect of the conceptualization of diversity in this manner is that you notice that it is
often the case that when diversity increases on one axis, it decreases on another.  A well-known example
is that when genetic diversity between populations increases, it can often be the direct result of
genetic diversity within populations decreasing.  I will provide a number of examples like this, and ask
how we can formalize the increase and decrease in diversity, especially with an eye toward conservation.

Wednesday, April 24 @ 2 pm

Irakli Loladze


Wednesday, May 15 @ 12:30 pm

Lora Billings

Noise Induced Chaos in the SEIR model

Population models are fascinating dynamical systems which capture all
ranges of behavior, from periodic to chaotic time series. The SEIR model
is a commonly used base model to predict epidemic outbreaks in many
diseases. Our work identifies the global mechanism to induce chaos by
stochastic perturbations, or population noise, where chaos does not
naturally occur. This mechanism combines both random fluctuations and
global topology to create a stochastic version of a heteroclinic
connection of the system. To refine the possibility of epidemic control,
we then identify the stochastic transport between basins. This is joint
work with Ira Schwartz and Erik Bollt.

Wednesday, May 22 @ 2 pm

David Smith

Wednesday, May 29 @ 2 pm

Andrew Irwin

[back too

[back too schedule ]