The Theoretical Ecology Lab Teas are designed to be informal meetings for members of the research groups of Simon Levin, Steve Pacala, Henry Horn, and Andy Dobson to give talks on their current research and receive feedback from their audience. The talks are usually 30 minutes, including the question and answer sessions, scheduled on Tuesdays at 2:00 PM. Additionally, other members of the Princeton University community and visitors are welcome to attend and to give presentations.
Talk schedules and email lists are
maintained by Juliet Pulliam and Duncan Menge. Please contact pulliam@princeton.edu
or dmenge@princeton.edu
to have your name added to the labtea email list so that you can
receive reminders about upcoming lab teas.
To view previous schedules and
summaries, go to:
Fall
2002 Spring
2003
Fall
2003
Spring
2004
Fall
2004
Spring
2005
| Tuesday, February 1, at 2:00 PM |
John McNamara |
| Tuesday, February 8, at 2:00
PM |
Heather
Leslie |
| Tuesday, February 15, at 2:00
PM |
Erik Rauch |
| Tuesday, February 22, at 2:00
PM |
Jeremy
Lichstein |
| Tuesday, March 1, at 2:00
PM |
David
Goehring |
| Tuesday, March 8, at 2:00
PM |
Christian
Wirth |
| Tuesday, March 15, at 2:00
PM |
(Spring
break) |
| Tuesday, March 22, at 2:00
PM |
Ben
Strauss |
| Tuesday, March 29, at 2:00
PM |
Mike
Morelli |
| Tuesday, April 5, at 2:00
PM |
Adi Livnat |
| Tuesday, April 12, at 2:00
PM |
Annette
Ostling |
| Tuesday, April 19, at 2:00
PM |
Katie
Hampson |
| Tuesday, April 26, at 2:00
PM |
Stephen
Pratt |
| Tuesday, May 3, at 2:00
PM |
Kelly
Caylor |
| Tuesday, May 10, at 2:00
PM |
Juliet
Pulliam |
| Tuesday, May 17, at 2:00
PM |
Nikolay
Strigul |
Titles and abstracts (posted approximately one
week before the talk):
Juliet Pulliam
Virus Emergence: What factors
allow host jumps? Considering the broad public and scientific interest in infectious
disease emergence, surprisingly little work has been done to
quantitatively describe broad-scale patterns of emergence in light of
the mechanisms that we believe drive them. Specific ecological
processes and pathogen characteristics must affect the probability of
emergence, and until these factors are recognized we will have little
predictive power regarding what pathogens are likely to infect species
other than their current host/s. Focusing specifically on viruses, I
will first describe the steps involved in the emergence process and
discuss several molecular properties that may facilitate spillover. I
will then give an example of how these ideas can be tested and used to
formulate predictive models regarding spillover potential. Finally, I
will describe the structure of a database that I am building to
address questions of emergence potential and, more broadly, of the
determinants of virus host-range. Kelly Caylor Horizontal and vertical
variability
of soil moisture in semi-arid ecosystems Soil moisture is a key hydrological variable that mediates the
interactions
between climate, soil, and vegetation dynamics in water-limited
ecosystems.
Because of the importance of water limitation in savannas, a number of
theoretical models of tree-grass coexistence have been developed which
differ in their underlying assumptions about the ways in which trees
and
grasses access and use soil moisture. However, clarification of the
mechanisms that determine structure of savanna vegetation remains a
vexing
problem in both hydrological and vegetation science. A particular
challenge
is the fact that the spatial pattern of vegetation is both a cause and
effect of variation in water availability in semiarid ecosystems. In
particular, at local to landscape scales, the patchy vegetation
structural
mosaic serves to redistribute the availability of soil moisture in
ways that
have important consequences for structural dynamics and community
composition. Here I address the role of patchy vegetation structure
through
the derivation of a landscape-scale model of soil moisture based on
the
statistics of an underlying poisson distribution of individual tree
canopies
and their accompanying root systems. The coupling of individual
pattern to
landscape-scale distribution of soil moisture allows for model
investigations into the role of tree density, tree size, and canopy
root
system structure on the spatio-temporal patterns of soil moisture
dynamics
in savanna ecosystems. Stephen Pratt Can ants count? Drift-diffusion
models of quorum sensing by house-moving
ants In performing collective tasks, social insects often respond to the
local
number or density of their nest mates, but little is known of how they
~Scount~T one another. When ants of the genus Temnothoraxchoose a new
home,
their decision relies on scouts who respond to the number of nest
mates
already at a promising site. At low numbers, they slowly recruit other
scouts to the site, but at high numbers they switch to rapid
recruitment of
the colony~Rs non-scouting majority. This quorum detection depends on
a
scout~Rs rate of encounter with nest mates at the site. To explore how
a
simple cognitive mechanism might use this information, I have adapted
a
model of two-choice decision-making in humans. Quorum sensing by an
individual ant is modeled as a drift-diffusion process with two
absorbing
boundaries, corresponding to the decisions that a quorum has, or has
not,
been attained. When the drift rate is modeled as a saturating function
of
population, the model effectively replicates the population-dependent
recruitment decisions of real ants, as well as the distribution of
decision
times for both kinds of recruitment. Next steps are to fit the model
to
direct measurements of encounter rates, rather than population, and to
determine whether it can predict changes in quorum size as nest area
(and so
encounter rate) is altered. Katie Hampson Multihost-epidemics: empirical
patterns and implications for theory
and control Disease dynamics of multi-host systems have important implications
for public health, livestock husbandry, wildlife conservation and
ecological processes. I explore the infection dynamics of rabies, a
generalist pathogen, during an ongoing outbreak in multi-host
communities of the Serengeti ecosystem. During the last 3 years,
cases of rabies have been recorded in more than 10 species in this
region. Although the vast majority of cases have been domestic dogs,
spatial and temporal clusters of infection in other carnivores
indicate interspecific transmission followed by transient chains of
intraspecific infection. Cross-infection of generalist pathogens such
as rabies is generally assumed to be of little importance to disease
dynamics. However, alternative hosts increase the size of the overall
effective susceptible population and may alter transmission patterns
as well as spatial coupling. Preliminary analysis suggests that
spillover epidemics may act as short-term reservoirs of re-infection,
which can prolong rabies persistence, increase human exposures and
hamper control efforts that target only domestic dog populations.
Annette Ostling Communities under a
recruitment-survival tradeoff: building on a dispersal-assembly theory Some ecologists argue that communities are shaped primarily by
dispersal and stochastic birth and death events, rather than by niche
differences between species. In particular, the recently developed
"neutral theory" of ecology has had some success at predicting the
properties of communities from the starting assumption that
individuals of different species are ecologically equivalent. It's
been proposed that this theory works even if individuals differ in
their traits, as long as they differ along a life-history tradeoff
that equalizes fitness across individuals of different species. I
test this proposal for the simple case of a recruitment-survival
tradeoff, and find this tradeoff leads to a positive correlation
between a species' abundance and its per capita recruitment rate (or
mortality rate, since they are equal under the tradeoff), and to
higher diversity, more rarity and less dominance than the case of
ecological equivalence. It also leads to increased variability in a
species' abundance, and ultimately a shorter time to extinction, with
increasing vital rates. I conclude that there are important
differences between fitness-equalization and ecological equivalence,
but that the neutral theory of ecology provides a useful framework to
build on and elucidate how species differences shape stochastic
variation in the community. The results presented here are also
important because the recruitment-survival tradeoff provides an
alternative to the widely studied competition-colonization tradeoff,
an alternative that is more realistic for many communities and yields
the opposite relative abundance predictions. Adi Livnat Conflict I'll be talking about the emergence of conflict within a system that is
selected as a collective (and only as a collective). The work is motivated
by some unexplained observations of animal and human behavior, but it is
done on an abstract level that makes it applicable in general (like
classical game theory is). The work has some interesting implications. For
example, it follows that we can't say, simply by the observation of
conflict, that the observed system involves independently-selected units
(although in many cases of course it does, this inference is not
necessarily always valid).
Michael Morelli The Impulsive Beverton-Holt
Model We will be looking at how impulses affect the following discrete
Beverton-Holt fisheries model.
X(n+1) = AX(n)/(1+BX(n))
Usually, the above difference equation converges very quickly to its
equilibrium. In our version of the model, we allow periodic
harvesting. Impulses model this harvesting of the fish.
X(1) = AX(0)/(1+BX(0))
X(2) = AX(1)/(1+BX(1))
X(3) = AX(2)/(1+BX(2))
X(4) = AX(3)/(1+BX(3))
Every fourth year we allow, say, sixty percent of the fish to be
harvested with an additional 10 sent to the EPA for testing. Hence we
have
X_plus(4) = (1 - 0.60)X(4) - 10
We use the impulsed population to determine the fish population the
next year.
X(5) = AX_plus(4)/(1 + BX_plus(4))
We continue the model in this fashion.
Some questions that arise include, "How much fishing should we
allow and how often to avoid killing off all the fish in the pond?"
and "How much time is required to allow the fish population to reach
equilibrium?"
Also, if there is time, we shall look at the impulsive version
of E.C. Pielou's discrete delay equation
X(n+1) = AX(n)/(1 + BX(n-k))
The delay of "k" years in the model allows the possibility of
oscillations and periodic behavior to occur.
Ben Strauss Tracking and modeling dispersal
in snails I will present data and two unconventional models of snail dispersal
across a range of
different stream environments. The first model is coarse and hinges on
dispersal kernels
which change as a function of environment (dispersal mean, variance and
tail length depend
on the average force of the stream). The second model is individual-based
and incorporates
a finer level of environmental detail. Work in progress.
Christian Wirth Wildfires up north: Detecting and
modeling the species imprint on fire regimes and succession in boreal
forests In the first part of the talk I will present results from a
meta-analysis where I show (1) that there are striking differences in the
fire regimes between the boreal forests of North America and Eurasia (i.e.
it burns more often in Eurasia and fires are less severe), (2) that this
can be explained by the fire adaptation strategies of the dominating tree
species, and (3) why this has important consequences for biome-scale
carbon dynamics. In the second part I will focus on fire-vegetation
interactions in Siberian forests and ask how changes in burn frequency may
alter successional pathways and cause a switch between alternate stable
states (light versus dark taiga). To answer this question I have
parameterized the forest simulator SORTIE for Siberian tree species and
added a fire and fuel production model. I will briefly introduce the main
model components and how they were parameterized from a wide array of
field data. Although this is work in progress I might dare to show some
premature first model results.
David Goehring Goal-Oriented Reserve Management --
Ecosystem Service Haloes: Patterns of Propagation
As we enter the twilight of de novo reserve creation, the need is
intensifying for theory to help optimize our ecosystem management and
restoration efforts. I present preliminary work and early thinking in
this area, emphasizing spatial and dynamic contexts. The challenge is
to understand the interaction among diverse, sometimes conflicting,
objectives that must be considered in an effective management
program. The goals which I have begun pondering are:
population/species persistence, maintenance of ecosystem services,,
learning/information, and resilience. I will address some inchoate
thoughts on these goals, and present the model described below.
Researchers have begun asking how to optimally manage ecosystems with
the goal of maintaining their economically tractable services. There
have been some recent stimulating studies that have demonstrated
interesting empirical properties of service haloes around patches of
natural landscape. If proximity to natural land increases the
benefits of certain services, we'll see some interesting landscape
patterns among stakeholders and their land-use strategies. The
situation is further complicated by the fact that the diverse suite
of services will likely have different propagation (e.g., that of
erosion, pollination, or water quality). The challenge is to properly
incentivize conservation for stakeholders, given their spatially
contingent interactions with resources - a dilemma with significant
consequences for the fate of comanagement and community management. I
have formulated a simple toy model which considers static, spatial
incentives for stakeholders in artificial landscapes, in the form of
service haloes around natural landscape patches. I will present
preliminary descriptive results regarding the role of the pattern of
fragmentation and the scale of property rights, with the goal of
getting input on future directions.
Jeremy Lichstein Drift, determinism, and
intraspecific variation in a simple model of plant
competition
Interspecific tradeoffs lead to multi-species coexistence in a variety
of
models, and empirical studies support the existence of such tradeoffs.
Jim
Clark et al. (2003) used super-duper Bayesian hierarchical modeling to
show
that traditional statistical analyses underestimate intraspecific
variation,
and, therefore, overestimate interspecific differences in growth rates.
They conclude that (1) species are not as different from each other as we
typically assume, and (2) the broad overlap in growth rates between
competing species implies a more prominent role for stochastic drift and a
less prominent role for interespecific tradeoffs than most people, besides
Steve Hubbell, presently believe. Using a simple two-species model of
plant
competition, I will show that these conclusions are not generally correct.
I examine a model in which two species have the same fecundity (per-adult
seed production) but differ in their distributions of seed quality. When
fecundity is high, the species with the highest potential seed quality
always wins (i.e., it always persists, while the other species always goes
extinct), even if the two distributions of seed quality are nearly
identical. In contrast, when fecundity is low, the two species have about
the same chance of winning, even if the distributions of seed quality are
non-overlapping. I will also show that, depending on the fecundity, a
species whose distribution of seed qualities has a relatively high mean
but
low variance can beat, get beaten by, or coexist with a competing species.
I apologize to anyone has read this entire abstract. I did not intend for
it to be so long.
Erik Rauch Patterns of genetic diversity
Within-species (genetic) diversity is important in a wide range of
biological systems. For species of conservation interest, genetic
diversity
is necessary for their survival in the face of environmental changes and
disease. For pathogens, diversity plays an important role in their
population dynamics and control. In this talk I will show that genetic
diversity is distributed unevenly. Using simple genealogical models, I
show
that genetic distinctiveness has a power-law distribution. This property
implies that much of the diversity is concentrated in small
sub-populations.
These theoretical results agree with genetic data on the distribution of
diversity in global samples of Pseudomonas bacteria. I will also present a
new method of estimating the genetic diversity of a population from a
sample. Finally, I show that the total genetic diversity in a population
depends strongly on the size of its habitat, much more strongly than does
biodiversity as measured by the number of species.
Heather Leslie Managing for resilience in
coastal social-ecological systems Humans impact almost every aspect of ecological systems, and threaten
the continued production of valued ecosystem services. Forecasting how
human activities alter the structure and functioning of ecosystems, and
thus the production of ecosystem services, is fundamentally a biological
problem. But biology alone will not be sufficient as ecological and
social systems are coupled. In particular, we must better understand the
responses of coupled social-ecological systems to disturbance.
Resilience (i.e. the amount of disturbance that a system can withstand
before its structure and functioning change significantly) theory and
related concepts (robustness, vulnerability, adaptive capacity) have the
potential to contribute to more effective coastal marine management and
conservation. Core research questions include: What are the important
variables, linkages, and drivers of change that define coupled coastal
systems? What are the sources of resilience in these coupled systems? I
will introduce how the developing knowledge of resilience can be applied
to advance sustainability at the regional scale, by focusing on the
coupled social-ecological systems of Chesapeake Bay.
John McNamara A critique of how game theory is
used; based on models of conflict and cooperation
Game theoretical models in behavioural ecology often make simplifying
assumption without thought as to their consequences. It is usually
assumed that the choice of action by contestants is simultaneous. In other
words individuals ignore others in making their choice. This assumption
not only conflicts with empirical evidence, but crucially affects the
outcome of the game. I argue that we need instead to carefully model the
process (over time) by which decisions are reached. Model often ignore
differences between individuals. It may seem that this simplifying
assumption is innocuous when differences are not large, but differences
can fundamentally change the nature of a game. Finally, theoretical
analyses of extensive form games usually assume that individuals are
completely flexible and can always take the best action given circumstance
(subgame perfection). Restricting flexibility can again fundamentally
change predictions. All these three point are inter-related and will be
illustrated by simple games involving conflict and cooperation.
Tuesday,
May 3 @ 2:00 PM
Tuesday,
April 26 @ 2:00 PM
Tuesday,
April 19 @ 2:00 PM
Tuesday,
April 12 @ 2:00 PM
Tuesday,
April 5 @ 2:00 PM
Tuesday,
March 29 @ 2:00 PM
Tuesday,
March 22 @ 2:00 PM
Tuesday,
March 8 @ 2:00 PM
Tuesday,
March 1 @ 2:00 PM
Tuesday,
February 22 @ 2:00 PM
Tuesday,
February 15 @ 2:00 PM
Tuesday,
February 8 @ 2:00 PM
Tuesday,
February 1 @ 2:00 PM
Last updated 01/05/05
pulliam@princeton.edu