This essay will look at living systems in the light of
thermodynamics.
Livings systems will be considered as a type of
dissipative structures.
Dissipative structures are dynamical systems with structures
which flatten out energy gradients in the environment - by
degrading sources of potential energy and generating
heat.
Burning bright
The basic idea discussed here is that living systems evolve
in a way that exploit the resources available to them as
rapidly as possible.
Ecosystems tend to evolve by accumulating technology which
assist them in:
identifing sources of potential energy and...
utilising those resources for reproductive ends.
As a result of the evolutionary process, ecosystems
typically become increasingly adept at identifying and
utilising sources of potential energy in their environment.
Self-organising systems dissipate energy
An example of a self-organizing system that is well known
for increasing the rate of entropy increase is the vortex
that forms when draining liquid through a hole.
This example is one of the illustrations of the "gradient
reducing nature of self-organizing systems" used by Kay and
Schneider.
Dribble
Vortex
Anyone who's spent much time getting liquids out of bottles
is already intuitively familiar with the phenomenon
illustrated above.
If you give the liquid a slight spin then a vortex forms -
and the liquid comes out much more rapidly.
The turbulent flow finds that it can best dissipate energy
if it forms an air channel from the surface of the liquid to
the exit hole.
This air channel allows air to flow into the bottle in a
continuous stream - so the pressure difference which would
otherwise act to hold the water in the bottle is
minimised.
In my experiments, creating such a vortex reduced total
drainage times by a factor of three - representing a
substantial rise in the rate of entropy increase after a
slight perturbation has introduced a self-organising
system into the bottle.
As Kay and Schneider point out, much the same phenomena
plays a role in dissipating energy in the atmosphere - in
the form of real tornadoes and hurricanes.
Indeed, there's been quite a bit of work on maximum power
principles in the field of meteorology.
Work in that area is discussed in more detail
here.
Why?
What's the mechanism for the effect? What is it about
self-organising systems that leads to them maximising their
power throughput?
The main reason for thinking that ecosystems might behave in
this way has a conceptually simple theoretical foundation.
This reason the principle applies in biology was first
clearly described by Lotka in 1922.
In a competition between two similar populations the group
with greater ability to locate and utilise sources of energy
is likely to prevail. Consequently we might expect to see a
greater rate of utilisation of resources as time passes -
as organisms get better at using environmental energy
sources - and turning potential energy into offspring.
This essay explores the reason
for the existence of the maximum power principle in some
more depth.
Generality
Biology is quite a bit different from meteorology. How can
the maximum power principle be showing up in in such different
fields?
Maybe that's because it's a law of thermodynamics - that
applies to all self-organising systems.
This essay explores how
widespread the principle is, and the circumstances under
which it applies.
The maximand
The self organising systems under discussion here certainly
look as though their behaviour is the result of
maximising some function. However different authors have
proposed different maximands.
Here I discuss the validity of the various proposals for
exactly what is being
maximised in these systems.
Supporters
Kay and Schneider explored a
number of the ideas discussed here - in 1994.
They wrote, for example:
ecosystems develop and select energetic pathways that
strive to degrade as much of the energy available to
them as possible.
This perspective appears to invert the conventional wisdom
that living systems are attempting to create order - partly
in the form of copies of their genome - rather than
disorder. However - as we will come to see - these
activities are rather fundamentally interrelated.
It responds electrochemically to geochemical and
photochemical tensions on Earth by attempting to resolve
them, remaking itself in the process that it might create a
greater overall disorder. Or in the words of Simon Black
(2000) -"energy uses an organism as a mechanism for self-
dissipation."
Burning bright - or burning long?
The rate at which the resources can be utilised can be a
factor which influences the competitiveness of a population.
If two populations expand into an environment which
has plentiful resources, the population that can most
rapidly translate the energy into offspring is likely to
wind up with a significant weight of numbers - and this can
easily translate into a competitive advantage in any battle
for what remains.
Overall efficiency can also be relevant - in times where
food is scarse, the future may belong - not to the
population that attemps to turn its food into increased
population as rapidly as possible - but rather to those that
maintian a low population size and conserve their food
supplies.
These strategies are usually mutually exclusive - burning
brightly is usually not going to produce a light that lasts
for such a long time.
I believe these strategies can usefully be seen as the result of
optimising over different time periods. The first is more short
sighted - while the latter takes a more long term view.
In the context of this essay the question of whether nature
is "burning bright", or is "burning long", boils down to the
question of whether she is maximising burn rate - or is
creating maximal eventual disorder. It seems likely that
the truth lies somewhere in between these two extremes.
Maximising order - and maximising disorder
Evolutionary biologists often judge the "worth" [1] of biological features, by considering to
what extent they contribute to the copying of the genotype
responsible for that feature.
In thermodynamic terms, that can be seen as a rather simple
approximation of the quantity of order condensed
from the environment into its own tissues.
The approach here suggests another metric - namely the
extent to which a feature contributes to degrading the
various sources of energy in the environment - increasing
overall entropy - and creating disorder.
These metrics are clearly closely related - since
in the process of creating the order present in
those genomes, a great deal of disorder will
inevitably need to be created somewhere.
Attempts to maximise order actually maximise disorder
A thermodynamic analogy may help to see why maximising
order in one place has the effect of maximising
the total disorder:
Imagine you have a fridge, which you want to keep cool. The
colder you make it, the more energy it takes to maintain its
temperature in the face of ambient heat from the
environment. Similarly if you want to make the fridge
larger, it will take more energy to keep it cool - since the
rate of heat loss through the surface will increase - due to
the increased surface area.
In this analogy, the cool fridge represents the order in a
living system. Attempts to increase this appear to
inevitably result in more disorder elsewhere.
It seems that any dynamical system that attemps to create
order in one region will inevitably wind op doing so by
increasing the total entropy. The more order it attempts to
maintain, the greater rate at which it depletes the
available energy sources.
The thermodynamic view
There are advantages to adopting a thermodynamic stance. It
better allows comparisons between individuals from different
species - than counting genes does.
Often, counting genes, genomes - or any other reproductive
unit - would be a rather pointless exercise - since
comparions between species would make little sense.
To those unfamiliar with the idea, it might seem odd to
describle the "purpose" or "function" of a "designoid"
system in thermodynamic terms - rather than in terms of
making more copies of genes - but it seems to me that such a
description more accurately represents what is fundamentally
going on.
Such thermodynamic metrics - in common with the more
familiar reproductive ones - can both be applied over
different timescales.
Since different elements of a system may be attempting to
optimise over different timescales - constructing a single
function that is being optimised may prove challenging.
The relationship between created order and environmental
disorder suggests a flip-side to approaches involving measuring
reproductive success.
Rather than attemping to estimate the extent of the order
created by a living system - you could see to what
extent it has made use of the available energy sources.
Looking at the state of sources of environmental energy may
sometimes be more practical than attemping to assess the
order created by an ecosystem.
Is life creating order - or disorder?
Is the creation of disorder by ecosystems a "by-product" of
their striving towards order?
Or can we follow [Kay and Schneider] - and say that ecosystems "strive to
degrade as much of the energy available to them as
possible"?
It seems established that attempts to create order
will necessarily increase overall entropy.
Similarly it seems that one of the best ways to increase
entropy - in the long term - is to introduce a complex
living system into the environment.
The thesis that living systems are attempting to maximise
total entropy is "paradoxically" very similar to the one that
they are attempting to create maximum local order (by making
many copies of their genome).
Overall, I'm inclined to state the goal of living systems in
overall thermodynamic terms - and thus to talk about
creating disorder - rather than (say) maximising the number
of copies of genes created.
Though very similar, the two ideas expressed above are
different from one another - and will make different
predictions.
I think the thermodynamic expression will eventually be seen
as more accurate description of what's going on under such
circumstances.
Does life really maximise disorder?
It is not easy to show that life is maximally
effective at increasing entropy. However it seems likely
that it will be much more effective than most of the
alternatives - provided not too short term a view is taken.
Nuclear explosions might well create a rapid local entropy
increase faster than a typical living system would manage -
but the effect is localised - and short lived. A living
system ought to be able to do a much better at
increasing entropy in the long term than such an
explosion.
Inefficiencies
The idea that ecosystems maximise their rate of energy
dissipation does not necessarily mean that they are
coming to utilise their energy more efficiently.
Indeed, perhaps rather the opposite: it's been argued that
the process of degrading available energy sources as rapidly
as possible will inevitably involve a substantial degree of
waste - since utilising energy rapidly and utilising it
efficiently are different goals that will inevitably
conflict.
However, it can be argued that such "maximum power
principles" will result a certain sort of efficiency of a
kind - and avoidance of some sorts of "pointless" waste.
One possible objection to the idea that populations
come to utilise their resources with increasing efficiency
is that populations are fundamentally lacking in harmony -
by virtue of being divided into discrete individuals with
conflicting objectives.
The tragedy of the commons, greedy predation, "driving
genes", within-species conflict and runaway sexual
ornamentation can all be seen as being counter-productive -
from the point of view of overall efficiency - and it might
be that barriers such as these will prevent energy from ever
being utilised very efficiently.
While I don't regard this objection as very serious, I
discuss it in more detail here.
More to life than power?
Isn't there more to living systems that the power they generate?
What about the possibilty of organisms defeating other organisms
on other grounds than their ability to rapidly utilise the available
energetic resources?
These are reasonable questions - and they are discussed further
here.
Between order and chaos
I hope the ideas discussed here will throw some light on the
popular notion of "life evolving to the edge of chaos".
I have a separate document discussing this matter -
here.
Progress
Does the tendency of living systems to get better at
utilising energetic resources result in a sort of
progress.
The picture painted here of living systems inexorably
increasing their power consumption might suggest images of
future living systems as explosions - expanding through
unoccupied regions at great speed and causing complete
destruction in their wake.
I don't mean to suggest that in the future there will also
be such emphasis on maximum speed in the short term.
Life has been rather short-sighted in the past.
Part of the reason for the historical short-sightedness is
that until relatively recently, saying much about the
future was very difficult because the equipment available
for building models of the future were very primitive.
Now living systems are becoming much better at understanding
their environment and predicting the future - though
unfortunately a rapid pace of change is acting
simultaneously to make the task more difficult.
Living systems have demonstrated that they can conserve
resouces when that's appropriate. Many of the adaptations
that become evident when organisms are calorie restricted
are evidence of this. However, organisms normally only
react to current shortages, and don't do a good job of
anticipating future resource shortages.
I hope in the future living systems will do a better job of
managing their resources based on future expectations.
It might well sometimes turn out to be a better stratgey for
living systems to save resources for later - in the hope of
persisting for longer.
If "maximum speed" does continue to remain
desirable then I do not know how rapidly future generations
will be able to expand their horizons - or to what extent
they will succeed in utilising the resources that lie in
their path - but I fully expect that very impressive feats
on these fronts are possible.
![[dribble]](graphics/dribble.jpg)
![[vortex]](graphics/vortex.jpg)