Titles and abstracts
Wednesday February 2nd at 12:30pm
Consensus Decision-Making in Animal Groups
Obligatory schooling species often must decide between discrete choices, such as which food patch to visit. There exists
theory in the social science literature about the optimal way to combine individuals' opinions and information
about the quality of the food patches to maximize the groups' probability of choosing the correct patch.
However, it is not clear whether an animal group following simple and local interaction rules can
effectively pool information in this way. I will present simulation results that show that not only can simple swarms
closely match the optimal strategy given the right parameter values, but that a simple, biologically realistic learning rule
allows a swarm to quickly learn these optimal parameter values.
Wednesday February 9th at 1:00pm
"Human-environment Systems in Epidemiology"
Vaccination programs generate herd immunity, which protects unvaccinated individuals. This creates the possibility of dynamic feedback loops between individual vaccinating behaviour and disease dynamics: increased vaccine coverage depresses disease prevalence, which thus reduces the incentive for further vaccination. This can be seen as a class of human-environment interactions, where herd immunity serves as the environmental context for human decisions. I will give an overview of recent work in mathematical modelling of this interaction.
"Global Ecological Change in Forests"
Global forests have reduced in coverage dramatically over the past several millenia due to various causes. But, since just a few millenia, at least in some places in the world, forests have been expanding. Similarly over the past few centuries forests have declined worldwide and yet some regions have seen a "forest transition" with increasing forests with increasing population size, industrialization and urbanisation, contrary to what might be expected. I present some recent related empirical studies and suggest directions for future modelling work.
Wednesday February 16th at 12:30pm
"Optimal seasonal reproduction in birds and butterflies"
Matthew Aardema, Jenny Ouyang, and Allison Shaw
Many organisms exhibit seasonal variation in life history traits such as number of reproductive bouts. In some butterfly populations, individuals will have a single generation per year, whereas in other populations individuals will undergo two generations in one year. Similarly, in some bird populations, individuals vary across populations in the number of clutches they lay within a given season. The variation in strategy for both cases seems to be related to season length (a function of latitude and temperature). We have written a simple analytic model to predict the optimal number of times individuals should breed in a given season, for both birds and butterflies. In the simplest case, we find that individuals in an environment with 'short' seasons should always have one reproductive bout per year whereas individuals in an environment with 'long' seasons should always have two reproductive bouts per year. Additionally, we explore what happens if we add environmental stochasticity to the model, and how the results differ for a scenario with overlapping generations (birds) versus non-overlapping generations (butterflies).
Wednesday February 23rd at 12:30pm (Guyot 100)
"Sustainable Development Hotspots: How to Spend Aid Dollars to Achieve
Environmentally Sustainable Development in Afghanistan and Pakistan"
Recent large increases in aid dollars to Afghanistan and Pakistan
demonstrate US interest for national security but also generate potential for
sustainable development planning in the region. Utilizing natural disaster
data, climate change models, underlying socioeconomic indicators of
populations, and reports of conflict, we identify several development hotspots
where low-risk investments could support strong economic growth. We then
compare our findings with current distribution of development aid. Although
the focus of this paper is on Afganistan and Pakistan, the methodology and
framework can be adapted to other regions and alternative objectives for
Wednesday March 2nd at 12:30pm (Eno 209)
Evolution of a Modular Software Network
I will present some ideas and thoughts about the evolution of modularity in a complex software network with the aim of shedding light on the study of networks of ecological interactions between species. This is an ongoing work in collaboration with Juan A. Bonachela and Simon A. Levin.
Debian GNU/Linux operating system provides a unique opportunity to study simultaneously the evolutionary and ecological processes determining the structure of ecological networks of interacting species as food webs. In the two systems, both processes occur at different time-scales. In the evolutionary time-scale, speciation and extinction can be translated into the appearance and disappearance of software programs (packages hereafter) from one version to the following one. In the ecological time-scale, colonization and local extinction, i.e. community assembly, would be equivalent to the package installation process in a local computer. Dependencies and conflicts between packages mimic predator-prey interactions and competitive exclusion relationships, respectively. Due to them, only a subset of the available packages can be installed in a computer, as only a subset of the species pool can coexist in a local ecological community. Last, there is an interplay between macroevolution and community assembly, because the interactions introduced by the new species (packages) alter the dynamic of the colonization/extinction (installation) in a local community (computer).
Wednesday March 9th at 12:30pm
Different diseases call for different treatment: modeling the influence of
pathogen biology on treatment strategies to contain resistance
Hospital-acquired infections contribute substantially to global
morbidity and mortality. The rise of resistance together with the shortage of
new broad-spectrum antibiotics lead to the question of how available drugs
can be optimally used to minimize disease burden. This question has been
answered differently depending on the infectious disease. Mathematical
models are useful for pointing out factors leading to diverging treatment
recommendations as well as for guiding resource-intensive clinical studies. We
analyze three strategies for coordinating empirical usage of two drugs in a
hospital ward: population wide combination therapy, random assignment to
different drugs (Mixing) and rotating first-line therapy (Cycling). We find that
the long-term population-wide benefit is expected to depend mainly on three
clinically relevant and accessible factors: the prevalence of resistance among
incoming patients, the relative transmissibility of resistant pathogens as well
as the pathogen turnover rate. Specifically, combination therapy is expected
to be the superior strategy unless the community prevalence of doubly
resistant strains is high and the doubly resistant pathogens are very easily
transmitted. This benefit of combination therapy is particularly pronounced in
the small and therefore highly stochastic patient populations characteristic of
hospital wards. Cycling may reduce inappropriate treatment if the cycling
period is optimized. However, the benefit when employing a period adapted to
the turnover rate is always lower than the loss when the cycling period is too
long. Taken together, we find that depending on the biological and clinical
specifics of the pathogen and on the clinical setting each of the three
strategies can be optimal. However, despite this variability, combination
therapy is for most settings optimal from the point of view of minimizing
Wednesday March 23rd at 12:30pm
Towards Mathematical Models of Cultural Dissemination
The enormous success of mathematical descriptions of biological
evolution begs the question of whether similar concepts could be used to
quantify the evolution of cultural objects, for instance the techniques of the fine
arts. In this talk we consider ways of approaching such modeling, with a focus
on the evolution of musical forms, making analogies and drawing from methods
in population genetics, stochastic processes on graphs, and the theory of
language evolution. Both the possible insights that could be gained, as well as
the new difficulties in constructing such a theory will be evaluated.
Thursday March 31st at 12:30pm (Guyot 100)
Leaf size scaling of plant trait relationships
Leaf traits such as maximum photosynthetic rate (A), nitrogen
content (N), respiration rate (Rd), and phosphorus content (P) are of
fundamental importance across a range of scientific disciplines. Although they
may be expressed using per-leaf-area normalizations, per-leaf-mass
normalizations are often preferred because they tend to show much stronger
inter-trait relationships. Here, we develop a simple method that partitions, in
an unbiased way, trait data from a large global database into area- and mass-
based components. We describe the most comprehensive examination to date
of leaf size scaling of physiological traits and its impact on the structure of
trait relationships in order to reconcile discrepancies between per-unit-area
and per-unit-mass patterns. Consistent with the idea that leaves are
constructed to perform area-based light interception and gas exchange, we
find that A and P are almost entirely area-based, whereas N and Rd are
largely, but not entirely, area-based. We go on to show that apparently strong
mass-based relationships are largely due to normalization-induced correlations
that do not reflect the underlying biological relationships between the traits.
We introduce statistical tools to remove leaf size scaling effects and present a
cursory analysis of the size-independent leaf economics spectrum. These
analyses will greatly improve the effectiveness and accuracy of the use of
plant trait relationships in important applications ranging from predicting
agricultural yields to the response of the biosphere to global change.
Wednesday April 6th at 12:30pm
Wednesday April 13th at 12:30pm
Wednesday April 20th at 12:30pm
Wednesday April 27th at 12:30pm
Wednesday May 4th at 12:30pm
Tuesday August 2nd at 10:00am
Colonization-competition tradeoff promotes species packing
Adam Lampert www.weizmann.ac.il/home/felamper
Many species are subject to a metabolic tradeoff between quick reproduction and better
resource utilization (colonization-competition tradeoff). Quicker reproducers (colonizers)
are advantageous when resource is abundant, whereas better resource competitors are
advantageous when resource is limiting. It is well understood that this tradeoff may lead to
a coexistence of many species, each of which is characterized by its colonization ability q,
provided that the resource is spatially extended and is subject to fluctuations. However, it
is unknown how evolution shapes the abundance of species as a function of q. In this talk,
I will show that the colonization-competition tradeoff inherently leads to “species packing”
in which several values of q are widespread and exhibited by many species, while values in
between these packs are relatively rare. Such patterns were indeed observed in body-size
distributions of multi-cellular organisms.
Links to previous schedules