From Molecules to Ecosystems with Dynamic Energy Budget Models.
General models describing the acquisition of energy by an individual organism, and its utilization for growth, reproduction and survival, have the potential to link to processes at various levels. The talk will highlight five major areas where such links can occur. First, a successful model based on dynamic energy budgets of individuals (hereafter referred to as a DEB model) can provide insight on sub-cellular processes related to energy use. Second, a DEB model constitutes the energetic basis for the dynamics of populations. Third, the models can make a major contribution to life history theory. Fourth, DEB models can be used to derive inter-specific scaling relationships for physiological rates. Fifth, by using the scaling relationships in combination with population models and some assumptions on stoichiometry, it may be possible to develop tractable models of the flow of elemental matter in communities or ecosystems.
Spatio-Temporal Dynamics of the Human Brain Revealed by Studies in Epilepsy
The human brain, arguably the most complex non-stationary system in nature,
may undergo abnormal, intermittent phase transitions that manifest
themselves as seizures in the electroencephalograms (EEG) of epileptic
patients. Seizures are characterized by spontaneously generated, sustained,
oscillatory discharges of electrical activity at brain sites for a few
seconds to minutes. Based on principles from nonlinear dynamics and
optimization theory, and analysis of days of continuous multi-electrode EEG
recordings from several epileptic patients, the entry (and exit) of the
brain into (and out of) seizures has been shown to be progressive and to
involve a spatial dynamical entrainment (disentrainment) of critical sites
in the cortex and hippocampi [1]. The relevant time and space constants are
really long, in the order of minutes-to-hours-to-days and cm respectively.
We have presented evidence that: 1) early detection of this dynamical
entrainment leads to reliable long-term prediction of epileptic seizures
(average 70 minutes prior to seizure onset with sensitivity close to 90% and
specificity 1 false detection per 7 hours), and 2) seizures dynamically
reset the epileptic brain (at the 95% statistical confidence level) [2, 3].
Similar mathematical analysis of simulation data from coupled chaotic
oscillators supports the above findings and provides further insight into
the mechanisms of these spatio-temporal transitions [4]. The results point
to the feasibility of the development of open-loop (warning) and closed-loop
(control) systems that would monitor and control the route of the brain
towards undesirable, catastrophic transitions. The significance of this
research in other fields of science, medicine and engineering that deal with
a systems-level organization, and an interplay between robustness and
complexity in uncertain environments, will be discussed.
Physiological Measurements along Boreal Forest Chronosequence
Physiological measurements along boreal forest chronosequence We have used a variety of gas exchange techniques to directly measure the ecosystem-atmosphere exchange of carbon, water and energy between five boreal forest sites in central Manitoba. The five sites are in various stages of secondary succession following large stand replacement fires that occurred 11, 19, 37, 70 and 150 years ago. We used the micrometeorological technique eddy covariance to directly measure whole ecosystem scale physiological responses and carbon balance during the 1999 and 2000 growing seasons. I'll discuss ecosystem level patterns in peak midday net CO2 uptake (whole-ecosystem respiration minus whole-ecosystem photosynthesis), whole-ecosystem respiration along the chronosequence. Results from leaf-level gas exchange, respiration rates from boles, and forest floor respiration rates measured to explain ecosystem-level responses will also be discussed.
Adaptive Computation
The field of adaptive computation uses ideas from adaptive systems in nature to develop novel computational techniques and algorithms, and in turn investigates how computational notions such as "information processing" and "computational complexity" can be used to explain the behavior of natural adaptive systems. I will give examples of both types of research: first, using genetic algorithms and cellular automata as biologically inspired, non-traditional computational methods, and second, as methods for investigating information processing in biological systems.
Using Modeling in Lab Classes to Improve Quantitative Skills in Undergraduate Biology Students
A team of biologists and mathematicians designed and implemented a series of laboratory exercises for biology labs at various levels in the biology curriculum. The purpose was to increase student quantitative skills using a combination of laboratory experiments and mathematical modeling. The exercises ranged from simple (requiring little laboratory materials) to complex (requiring extensive laboratory support) and were developed for a range of curriculum levels (freshman to senior). Success, as measured by pre- and post-tests, was variable and sensitive to the instructor's ability to guide students through the modeling process.
A Physico-Evolutionary Approach to Biological Development
The earliest multicellular organisms to emerge in evolution
consisted of cells whose genes had evolved mainly to serve single-cell
functions rather than the global multicellular coordination seen in modern
multicellular development. The forms assumed by ancestral multicellular
organisms were thus likely to have been governed by inherent physical
properties of multicellular aggregates in interaction with the physical
environment ("self-organization") rather than by hierarchies of gene
regulation ("programs"). Physical and self-organizing processes such as
differential adhesion-driven cell sorting, biochemical oscillation and
multistability, and reaction-diffusion coupling, can account for most of the
structural features of animal body plans and organ forms, including tissue
multilayering (gastrulation), segmentation, and differentiation. Such
physically-generated morphologies provided templates for stabilizing
evolution, leading to genetic assimilation and integration (the "Baldwin
effect"), and eventual genetic regulatory hierarchies for the realization of
body plans and organ forms. This physico-evolutionary perspective, by which
genetic alterations may follow and consolidate, rather than originate,
morphological innovation, has proved useful in understanding the character and
generation of the metazoan body plan, the prevalence of segmentation across
metazoan phyla, and the underlying mechanisms of vertebrate limb development.
Subcellular Level Eigenmode Methods for Diagnosing and Controlling Large-Scale Cardiac Rhythm Patterns
While a standard method in physics and engineering, eigenmode analysis is rare and generally unknown as an investigative tool to the biological community. In our research, we have combined eigenmode concepts and eigenmode-based computer algorithms with time-dependent simulations to understand how the dynamics at the ion channel level determines cardiac rhythm at the tissue and/or organ level. This approach has enabled us to devise novel methods which may be effective in eliminating undesirable components in abnormal cardiac rhythms.
In Silico Research and Development of Large-Scale, Dynamical Models of Biological Systems
Cindy has been travelling and it is Jim Powell's fault that he didn't get a
title and abstract from her earlier. Her expertise includes:
Gene-Environment Interactions and Recurrent Airway Disease: From Cellular Events to Population Dynamics
In genetically predisposed subjects, environmental exposures to common pollutants, allergens and viral infections can cause recurrent episodes of airway inflammation and bronchoconstriction. The sequence of interactions between the immune system and the environment during early infancy can play a critical role in the subsequent development of asthma. To characterize the dynamics of gene-environment interaction in early infancy, we assessed respiratory symptoms by weekly interviews in a prospective birth cohort study of 78 unselected normal Caucasian infants in their first year of life. We analyzed the periods between two consecutive episodes, the inter episode intervals (IEI). This population of infants was homogeneous in terms of triggers and past history of episodes and the immune system of the infants were assumed to be in a similar state at birth. Thus, data were pooled from all infants to construct a single distribution function of IEI. We found that the long term temporal dynamics of recurrent airway disease in this population of infants display a complex intermittent pattern, and the distribution of IEI follows a power law. We interpret the data using a model of the dynamics of attack episodes in which cellular build-up of bronchoconstrictor mediators due to exposure to randomly varying hazardous agents continuously competes with relaxation due to cellular repair processes. Eventually, this competition leads to maximal muscle contraction which in turn triggers a spatially propagating avalanche of airway collapse with measurable clinical symptoms. The model produces an intermittent pattern of episode sequence with the distribution of IEI following a power law. We conclude that our model based integrative approach can bridge the gap between cellular events and organ level behavior as well as predict some aspects of population dynamics.
Transposable Elements: The Biochemistry of Evolution.
Polar opposites among biologists have been thrown together by the phenomenon of transposable elements (TEs). Biochemists are very interested in TEs because of their self-contained ability to catalyze complex DNA reactions. Developmental biologists and geneticists see TEs as causes of variability. Theoretical population and evolutionary biologists see them as the ultimately simplified selfish genes. In spite of this, there is remarkably little integration between these disciplines in the study of TEs. The mathematical problems are difficult, requiring a detailed stochastic analysis of mutation, selection, population subdivision, and demographics. The biological complexities are no less daunting, as the diversity of TEs is vast. Finally, the limitations of and cultural differences between scientists interested in different parts of this puzzle make long-term productive collaborations difficult to establish and maintain. The study of TEs is a microcosm of the problems faced by integrative biologists. Models at the appropriate level of generality and detail exist, they can account for most observed patterns of TE evolution, they have made predictions that were proved correct, and point out areas where information is critical but lacking. Thus it is a field poised for integration. The question is, how?