Titles and abstracts
Wednesday September 16th at
Exploring phase transitions and collective
animal groups using novel imaging techniques.
To explore collective behavior in animal groups, experiments on fish schools have
been done in
the Couzin Lab. They produced a lot of data to analyze. When the group size is small enough, it
is possible to track individuals and get their trajectories. This provides very precise
information at a very fine level. Unfortunately, these methods loose accuracy when the group size
increases, and it is impossible right now to track more than 300 individuals. However, it is
crucial to be able to make measurements for larger group size. I will present in this Lab Tea a
method based on optical flow estimation for extracting the velocity field, density and alignment
of the fish at a coarse level, which has been successful for analyzing groups of 1000 fish, and
which we can apply even on larger groups. I will also present some preliminary results about
phase transitions and correlation length.
Wednesday September 23rd at
Assigning biological meaning to the parameters of the neutral theory
of biodiversity promotes rigorous confrontation with data.
A key goal of community ecology is to understand the mechanisms that
generate observed global patterns of biodiversity. To achieve this,
we require theoretical mechanistic models based on biologically
meaningful parameters such as dispersal distances and speciation
rates. I will report on our recent successful attempt to relate the
heretofore mysterious immigration paper of Hubbell's spatially
implicit neutral theory to the parameters of a dispersal kernel and
the geometry of the plot defining the community. I will demonstrate
three benefits of this work. Firstly, I will show how the immigration
parameter can now be estimated independently from dispersal data
rather than treated as a free fitting parameter, and I will compare
these new independent estimates with previous fitted values.
Secondly, I will show how the result can be used to make new
predictions about species-abundance distributions and species-area
curves and I will compare these predictions to tropical forest data.
Thirdly, I will discuss how the results can be used to assess the
errors in the mean field approximations associated with the spatially
implicit models. This work is an important step towards promoting
rigorous confrontation of the neutral theory with data and thereby
facilitating comparison with competing models such as niche theories.
Wednesday September 30th at 12.30pm
Animal mediated long distance dispersal of seeds.
Empirical studies suggest that the dispersal of seeds to large distances
can occur much more frequently than a Gaussian kernel would predict. Such
`fat-tailed' long distance dispersal (LDD) is predicted to have
disproportionate impact on population and community dynamics. Animals
form a key biotic vector of seeds and pathogens, but very little is known
about the influence of individual animal behavior that drives LDD. I
present a simple analytical model that links individual animal behavior
with large scale patterns of LDD and makes testable predictions. This
work was done when I was a graduate student (during a summer school) and
in a field am not too familiar with. I am looking forward to receiving
feedback at LabTea.
Wednesday October 7th at 12.30pm
Optimizing reactive responses to outbreaks of immunizing infections - case management vs
When controlling outbreaks of infectious diseases, the two most common approaches are treatment
of infectious cases and mass vaccination campaigns. We formulate a SEIR model that includes
those two control strategies - palliative care and vaccination. Using mostly numerical
techniques, we investigate a range of strategies we can implement after the epidemic has started
and try to find a rule-of-thumb principle for control of outbreaks.
October 12th at 12.30pm
Spiteful and Selfish Behavior through biological warfare.
Francisco Dionisio (Universidade de Lisboa,
Biological warfare, also known as a germ warfare or biological
weapons, is the use of any pathogen or parasite as a means of harming
other individuals. Can individuals use their pathogens/parasites as
biological weapons to harm other individuals? That is, use pathogens
as spiteful or selfish agents? In this seminar, I discuss the
likelihood of this hypothesis.
Wednesday October 14th at 12.30pm
Spatial Networks in Ecology // Coevolution in Networks of Interacting
Miguel A. Fortuna
In this talk I would like to present very briefly what I have done
during my PhD so that you can know my previous work, and also what I am
going to do at Princeton during my postdoc in order to receive feedback.
In the first part of the talk I will show you how the framework of
complex networks (webs consisting on a set of nodes connected by links)
is a useful approach to understand how biodiversity is organized at
different scales. Specifically, I have identified the spatial scale of
ecological processes such as dispersal and migration in fragmented
landscapes, behavioral processes such as social structures in animals,
and evolutionary processes such as gene flow in plant populations. In
the second part of the talk I will introduce the mathematical approaches
to the study of coevolution, that is, reciprocal evolutionary change
between interacting species driven by natural selection. The question to
be explored is to what extent the coevolutionary process shapes the
architecture of ecological networks of interacting species such as food
webs and plant-animal mutualistic networks.
Wednesday October 21st at
Tree coexistence in a forest metacommunity
The factors that control tree coexistence, and thus diversity, are still poorly understood. The
recently developed quasi-realistic, yet tractable, perfect plasticity approximation (PPA) forest
model is a tool that can potentially provide some answers. I will discuss a metacommunity version of
the model, in which trees coexist via successional and edaphic (soil-related) tradeoffs. An
analytical approximation for the late successional equilibrium (i.e., a forest that has not been
disturbed for a very long time) suggests that even slight edaphic specialization results in the
locally dominant species occupying nearly the entire habitat. However, the transient dynamics are
slow, so that single-species dominance is rarely observed. The competitive dynamics between
late-successional species are qualitatively similar in the presence or absence of early-successional
species. The results suggests that source-sink dynamics may be an important factor contributing to
local (alpha) tree diversity.
Wednesday October 28th at 12.30pm
Effects of demography and vaccination on the spatiotemporal dynamics of rotavirus epidemics
in the United States and beyond.
The seasonal geographical spread of rotavirus in the United States has been a hitherto
mysterious pattern. Rotavirus epidemics have historically begun in the southwest, where
epidemics peaked in December-January, and spread towards the northeast, where epidemics
peaked in March-April. In recent years, however, this trend has diminished as epidemics have
been occurring later in southwestern states. Standard explanations such as traveling waves
or environmental drivers cannot account for these patterns. Using a mathematical model of
rotavirus transmission calibrated against state-specific data on disease incidence, we show
that spatiotemporal variations in U.S. birth rates can explain the apparent traveling wave
of epidemics. The recent large-scale introduction of rotavirus vaccination provides a means
of validating our model predictions. Allowing for this artificial reduction in susceptible
recruitment generates a predicted pattern of reduced and lagged epidemics closely matching
the observed response to vaccination. Armed with this validated model, we explore the
relative importance of direct and indirect (herd immunity) protection, a key issue in
determining the worldwide benefits of vaccination.
11th at 12:30pm
Adaptive Dynamics of Virulence: Mycoplasma conjuctivitsis in House Finch Populations and
I will present some preliminary theoretical results on virulence evolution when disease
transmission occurs before major disease mortality or morbidity in single and two-species
host communities. My work is motivated by lab and field data collected by collaborators
on Mycoplasma gallisepticum (MG) infecting House Finch, American Goldfinch, and other bird
populations across the United States. We have exciting new results consistent with
geographic variation in virulence evolution during the emergence of this disease. I would
welcome any feedback on our interpretation of these empirical results and on my models of
18th at 12:30pm
A Hidden Dimension to Decision Making: Animals Balance Speed, Effort and Accuracy
Adrian de Froment
All animals need to make the best decisions possible in order to maximize their Darwinian fitness. A
widespread view in biology is that animals face a simple trade-off between speed and accuracy, which they
balance according to the costs and benefits involved. For example, a swarm of bees will spend less time
choosing a new nest site if the cost of remaining in the open increases, at the same time accepting a higher
chance of choosing a suboptimal home. However this picture ignores a third dimension to the trade-off:
animals can also vary the effort per-unit-time that they invest in reducing their uncertainty about their
options. This is important because the costs of time and effort can vary independently. Using a standard
model of decision-making, I examine whether perfectly adapted animals should make use of this to improve
their decision-making. I compare individuals who can adjust both the time and the effort they invested in a
decision with a second group who can adjust only the time. As the cost of time rises, the animals in the
both scenarios make quicker decisions, but those in the first group also increase their level of effort, and
so are able to maintain a higher level of accuracy. The model makes testable predictions about how this
three-way trade-off between speed, effort and accuracy should manifest itself in nature. This will be
important to researchers interested in decision-making in biology and neuroscience.
19th at 12:30pm in Guyot Hall 100
Robust ecological pattern formation induced by demographic noise
We demonstrate that demographic noise can induce persistent spatial
pattern formation and temporal oscillations in a simple model due to
Levin and Segel of predator-prey model for plankton-herbivore
population dynamics. Although the model exhibits a Turing instability
in mean field theory, demographic noise greatly enlarges the region of
parameter space where pattern formation occurs. To distinguish between
patterns generated by fluctuations and those present at the mean field
level in real ecosystems, we calculate the power spectrum in the
noise-driven case and predict the presence of fat tails not present in
the mean field case. These results may help account for the prevalence
of large-scale ecological patterns, beyond that expected from
traditional non-stochastic approaches.
25th at 12:30pm
Multiscale, adaptive representations of microbial community structure and biogeochemical cycles in the oceans
The development of predictive models of microbial community composition in the oceans for various regimes of
disturbance and nutrient availability is central to our understanding of how these communities drive, and respond
to, global biogeochemical cycles. Progress has been hampered by three major difficulties. First, experimental
and field information about the relative distribution of biogeochemically relevant metabolic functions and
trade--offs is limited to a handful of organisms that can be cultured; although new cultivation techniques,
together with environmental genetic sequencing (i.e. metagenomics) seem promising, significant challenges remain.
The second problem is theoretical. How should one construct a model that is capable of adaptation, in the sense
that types that occur at low abundances under some environmental conditions, discarded in traditional aggregated
models, are allowed to become dominant when the right conditions prevail? Third, experiments typically target
fully mixed environments, but fluid environments in the oceans are highly structured (layers, turbulent eddies).
My current research targets these problems by developing computational experiments based on an algal cell IBM
for the first case , flexible, trait based representations of community composition in the second, and variants
of fractional kinetics in the third.
Work in progress, in collaboration with S. Dutkiewicz (MIT), M. Follows (MIT), John Dunne (GFDL), Charlie
Stock(GFDL), Juan Bonachela (Levin Lab) and S. Levin
Wednesday December 2nd at 12:30pm
Efficient stochastic searches and methods to analyze 'worm' movement behaviors
The study of animal foraging behavior has extreme importance in the field of ecology and exemplifies the wider scientific problem of optimizing search strategies. There is empirical evidence
that the type of searching strategies employed by animals while foraging (systematic or random) depends on the amount of information that is involved in the search process. When information is
lacking, the random search is a possible strategy. In this case, an important question to ask is what statistical strategy will be the most efficient to find food? To be able to answer it, it
is important to know which are the behavioral mechanisms that allow animals to perform efficient stochastic searches.
During labtea, I will be focusing on the behavior mechanisms that the nematode C. Elegans use during its searching process and on the methods to automatically identify these mechanisms.
Wednesday December 9th at 12:30pm
Leaf longevity: understanding community-level patterns in forests by scaling-up individual-level physiology in the context of light and nitrogen competition
Leaf life spans of forest trees range from a few months to nearly a decade. Three generally accepted relationships pertain to any explanation of this conspicuous fact: Compared to shorter-lived leaves, longer-lived leaves
are... (1) thicker, with lower maximum rates of photosynthesis; (2) more nitrogen-use-efficient because the nitrogen used to build each leaf is retained for longer; and (3) slower to decompose. Innumerable verbal arguments and
a few models have integrated these facts to explain observed patterns, but none of them take a whole-plant, community-level perspective and none are quantitative. The Pacala Lab’s PPA model of height-structured light
competition is an ideal modeling framework to fill this void (and I do hereby swear that my review of it will be brief). I may have some results by Wednesday, but my main goal for this Lab Tea is to lay out the problem and our
proposed approach in the hope that you can help identify profitable routes we hadn’t considered or demons to avoid.
Wednesday December 16th at 1:30pm
A fairly simple model of the evolution of bacterial quorum sensing
Many bacteria possess quorum-sensing (QS) systems that are thought to detect population density in the local environment. They accomplish this by secreting small signal molecules, monitoring
their concentration with specialized receptors, and adjusting their behavior in response to changes in signal abundance. Numerous phenotypes fall under QS control, including basic metabolic
activity, sporulation, competence (the ability to take up extracellular DNA and incorporate it into one's own genome), surface attachment, biofilm formation, and the secretion of digestive
enzymes and many other molecules. QS is a type of social behavior at heart, and social evolutionists have shown great interest in exploring it theoretically and empirically. However, their
analyses often do not distinguish between the QS regulatory mechanism and the phenotypes that it regulates. They also tend only to consider a small subset of QS-regulated phenotypes: cooperative
public goods like extracellular enzymes. My goal is here is to start simpler, thinking about why and when quorum sensing provides a net benefit as a buffer against environmental uncertainty.
Links to previous schedules