Theory: Biological systems organize to maximize entropy production subject to information and biophysicochemical constraints Grant uri icon

abstract

  • This project seeks to answer the question: What is the governing principle that determines how energy and matter flow through biological systems composed of independent but interacting individual organisms, such as occurs in ecosystems? Surprisingly, no predictive theory exists for such a fundamental question. The theory of evolution by natural selection provides a mechanism for self-organization of complex biological structures, but is indeterminate in regards to the emergent properties biological systems follow, if any. As a consequence, the flow of energy and mass through biological systems is often attributed to the chance composition of the community at any instance in time, which is currently unpredictable. This project takes the perspective that biological systems evolve and organize in a manner that is, in a sense, independent of community composition. In the field of nonequilibrium thermodynamics a provisional proof on the theory of maximum entropy production (MEP) has recently been proposed, which posits that steady state systems with sufficient degrees of freedom will organize to maximize the rate of entropy production; that is, the rate of energy dissipation. While organized structures decrease the entropy of a system, they are maintained by external entropy production and have a higher probability of persistence if their presence increases overall entropy production. However, the configuration of structures that generate entropy, and dissipate energy, are constrained by system resources from which the structures must be synthesized from. Hence, biophysicochemical constraints (i.e., elemental resources, organic chemistry, etc.) limit the complexity of dissipative structures. Hurricanes that dissipate thermal energy between the atmosphere and ocean are examples of such dissipative structures. This project proposes that evolution by natural selection produces biological systems that tend to follow a pathway of maximum entropy production by dissipating high temperature radiation and chemical potential. Consequently, an ecosystem composed of organisms that produce entropy at a high rate has a greater probability of persistence and occupation than an ecosystem under the same constraints that produces entropy at a lower rate. While MEP theory does not distinguish between abiotic and biotic systems, biological systems differ from abiotic ones in one key way: biological systems store information within their metagenome. Therefore, it is proposed that abiotic systems maximize entropy production instantaneously, while information stored within the metagenome allows biological systems to produce entropy along pathways that can increase entropy production when averaged over time. For instance, by storing internal energy, biological systems can maintain entropy production and persist during periods when external energy inputs cease. Based on MEP theory, it is hypothesized that biological systems with greater information content will have higher entropy production rates than biological systems with lower information content. To test these hypotheses, the project will use flow through microcosms (i.e., chemostats) as experimental systems inoculated with natural microbial communities. Changes in chemical composition will be used to determine entropy production and massively parallel 454 pyrosequencing applied to hypervariable regions in rRNA genes will provide a direct measure of the information content of complex microbial communities. The project will demonstrate that 1) community composition changes to maximize entropy production, 2) loss of information due to decreases in biodiversity results in lower entropy production and 3) communities organize to maximize entropy when averaged over time. In addition to experimental tests, the project will develop a mathematical framework based on MEP theory to model biogeochemistry orchestrated by biological systems using a distributed metabolic network representation. Computational models and experimental results from this project, including educational outreach activities, will be posted on the project's web site: http://ecosystems.mbl.edu/MEP

date/time interval

  • September 1, 2009 - August 31, 2013

total award amount

  • USD 758000

sponsor award ID

  • 0928742