Biocomplexity Blooms in NSF's Research Garden, Science 
The first two scientific revolutions, relativity and
quantum theory were far enough removed from daily experience that
the public accepted them as important scientific breakthrough
based on the endorsement by the experts and based on spectacular
applications like the atom bomb. Complexity theory, however,
seemed to share the fate of general systems, catastrophe, chaos
theory in that it touched directly on personal experiences that
misinterpretation and hype were unavoidable. As a consequence a
counter-swing is certain to follow and it was only three years ago
that John Horgan announced the "End of Chaoplexity". Today we
experience a true renaissance of complex systems research at all
levels.
Rita Colwell, director of the National Science Foundation (NSF)
announced a few weeks ago a $50 million special competition for
research "to better understand and model complexity in biological,
physical, and social systems". It is linked with the National
Ecological Observatory Network (NEON), a collection of outposts
that would provide researchers with the capacity to do high-tech
fieldwork on biocomplexity.
"Biocomplexity is a multidisciplinary approach to understanding
our world's environment," Colwell told a congressional panel
earlier this year. "For generations, scientists have studied parts
of our environmental system--individual species and habitats--in
isolation. Now it is time for a better understanding of how those
parts function together, as a whole."
Biocomplexity
Blooms in NSF's Research
Garden, Jeffrey Mervis,
Science, Volume 286, Number 5447 Issue of 10 Dec 1999,
pp. 2068 - 2069
-
In traditional, reductionistic science understanding of a
complicated system or gadget would correspond to analyzing it,
i.e. taking it apart and looking for internal mechanisms. If one
can put the gadget back together, has some parts left over and the
gadget still works then one could abstract from some unnecessary
details and one has an even more fundamental understanding of the
system. That is, if the system functions under all conditions it
functioned before.
Hutchinson et al. took the simplest living system there is in
the sense that it can independently replicate (viruses don't do
that) the Mycoplasma genitalium a microbe with 517 genes making up
580kb (so the whole genetic sequence information would easily fit
on a floppy disk).
The question for the essential genetic aspect of life was then
attacked by the authors by using Global transposon mutagenesis "to
identify nonessential genes in an effort to learn whether the
naturally occurring gene complement is a true minimal genome under
laboratory growth conditions". By using a "top-down" approach they
estimated which of the genes within the entire genome could be
removed or inactivated and still yield a reproducing cell. A
bigger challenge will be to build a working minimal genome from
scratch in a "bottom-up" approach as it is done e.g. in artificial
life simulations "in silico" instead of the more demanding
conditions of "in vitro".
Hutchinson et al. came to the conclusion that 265 to 350 genes
were essential for the M. genitalium. This figure is very close to
the theoretical minimal value based on the argument that genes
that are conserved across large phylogenetic distances (here M.
genitalium and the other microbe with completely analyzed genome,
Haemophilus influenzae) are likely to be essential.
This work raises a number of ethical questions that are
discussed by Cho et al. Among them are issues of dangers from
genetically engineered organisms in general in terms risks of
bio-toxic effects and unpredictable impact on eco systems. They
come to the conclusion that "The prospect of constructing minimal
and new genomes does not violate any fundamental moral precepts or
boundaries, but does raise questions that are essential to
consider before the technology advances further. How does work on
minimal genomes and the creation of new free-living organisms
change how we frame ideas of life and our relationship to it? How
can the technology be used for the benefit of all, and what can be
done in law and social policy to ensure that outcome?"
Life
Is Pared to Basics. Complex Issues
Arise, Nicholas Wade, New York
Times,
December 14, 1999
Global
Transposon Mutagenesis and a Minimal Mycoplasma
Genome, Clyde A. Hutchison
III, Scott N. Peterson, Steven R. Gill, Robin T. Cline,
Owen White, Claire M. Fraser, Hamilton O. Smith, J. Craig
Venter, Science Volume 286, Number 5447 Issue of 10 Dec
1999, pp. 2165 - 2169
Ethical
Considerations in Synthesizing a Minimal
Genome, Mildred K. Cho,David
Magnus, Arthur L. Caplan, Daniel McGee, and the Ethics of
Genomics Group, Science Volume 286, Number 5447 Issue of
10 Dec 1999, pp. 2087 - 2090
-
Jupiter's moon Europa played a special role in the
science-fiction novel "2001": The astronauts got some mysterious
message to stay away from Europa and it is implied that some
extra-terrestrial form of life is evolving under its icy surface.
Today we know that there is strong evidence for liquid water on
Europa and therefore the possibility of life.
The search for extra-terrestrial life that had been focused on
radio telescopes has increasingly moved underground: Today many
experts agree that life most likely originated not at the surface
but deep in the ocean near some source of geo-thermal heat and
sulfur based energy sources.
The discovery of Lake Vostok, a vast freshwater lake (deeper
than Lake Tahoe, wider than Lake Ontario) 3-4 km beneath the East
Antarctic Ice Sheet provided a unique opportunity to study
extra-terrestrial living conditions on Earth. The question about
the existence of life in Lake Vostok therefore is of uttermost
interest for astro-biologist. Of course the experimental problems
at such an extreme location are considerable: Drilling down
several kilometers in ice is not a problem per se as long as one
can use anti-freeze liquids to keep the borehole from freezing up.
Up to now about sixty tons of anti-freeze have been poured into
the drill hole and of course that causes considerable new problems
if one is interested in finding traces of life in a lake of fresh
water that had been isolated from the rest of the world literally
for millions of years.
To avoid pollution of Lake Vostok by these drilling fluids the
researchers have stopped drilling well above the lake level. It is
still close enough to find ice that is believed to originate from
the lake and not from the covering ice sheet.
Priscu et al did indeed find that ice samples from Lake Vostok
contain bacteria of low bio-diversity and evidence for nutrients
that allow for an active microbial ecosystem. Karl et al. could
confirm the existence of bacteria in the lake ice and additionally
they found evidence for respiratory activity, i.e. these bacteria
are not just frozen but seem to indeed live and thrive in the icy
conditions.
But since biological processes is slowed down at icy
temperatures science fiction fans might have to wait for another
couple of billion years before the ice-bacteria evolve into higher
life forms.
Icy
Life on a Hidden Lake, Warwick
F. Vincent, Science, Volume 286, Number 5447 Issue of 10
Dec 1999, pp. 2094 - 2095
Geomicrobiology
of Subglacial Ice Above Lake
Vostok, Antarctica, John C.
Priscu, Edward E. Adams, W. Berry Lyons, Mary A. Voytek,
David W. Mogk, Robert L. Brown, Christopher P. McKay,
Cristina D. Takacs, Kathy A. Welch, Craig F. Wolf, Julie
D. Kirshtein, Science, Volume 286, Number 5447 Issue of
10 Dec 1999, pp. 2141 - 2144
Microorganisms
in the Accreted Ice of Lake
Vostok, AntarcticaD. M. Karl,
D. F. Bird, K. Björkman, T. Houlihan, R.
Shackelford, L. Tupas, Science, Volume 286, Number 5447
Issue of 10 Dec 1999, pp. 2144 - 2147
Exploring
Lake Is Like Visiting Another
Planet,
National Science/Health,
by The New York Times December 14,
1999
Shopping by Bar Code, Wired News
Imagine you are in an aisle in the supermarket and want
to get some laundry detergent. What are the factors that actually
influence your decision to pick brand A instead of brand B? Do you
actually look at the small print on the label and compare the
figures to determine which brand has the least detrimental effect
on the environment?
There are a considerable number of consumers who go through the
effort to travel physically to the actual producer of vegetables
or chicken to convince themselves about the proper way the
vegetables were grown and the chicken were raised and the same
consumers pay a significantly higher price for those goods than
what they would have had to pay in the supermarket around the
corner.
There are many examples where the information about the product
can completely dominate the price of the product. Art objects are
the extreme example: Even if someone could produce an atom-by-atom
duplicate of a Picasso picture, without the certified information
that it was actually created by the painter its value is
negligible compared to the "original".
In the context of Information Agriculture the concept of food
labeling was discussed back in 1994 within the Cyber Farm project
at the University of Illinois at Urbana Champaign: With the help
of the Internet and modern computers it should be possible to
individually label food product and provide detailed information
about ingredients and history of processing. Now it finally seems
that the technology is there for the first step of consumer level
electronic product identification: BarPoint.com launched a website
that allows to access information encoded as the Universal Product
Code (UPC) in the barcode label of each product. There is a beta
version available for wireless access from the Palm VII and early
next year access from Net-enabled cell-phones will became
available. Barpoint also plans to sell clip-on barcode scanners
for cell phones and Palm OS handhelds by Symbol Technologies with
integrated bar code scanners.
While the first generation of product identification will be
based on traditional bar-code information, it is quite feasible
that additional information will soon be encoded as well. Then
special electronic relations between consumers and producers can
be established through electronic profiles: In the example of
figuring out which laundry soap to pick the PDA could rank the
available brands according to the preference profile that the
consumer has provided beforehand. It also can electronically
negotiate with the producer a special discount similar to the
rental car special rates that are available today. Since product
information can be encoded for each individual item this
technology will provide a new opportunity for minority producers
to access a customer base for whom this information is
valuable
Shopping
by Bar Code, Leander Kahney,
Wired News, 12/12/1999
-
If we look at organizations in human societies as complex
adaptive systems then it is not too hard to find out what the
predominant fitness parameter is: It certainly is related to the
amount of money the organization makes or looses. This parameter
is then a good predictor about which self-organized, emergent
structures will be evolutionary fit and which ones will rapidly go
extinct.
When the Internet was about to make the transition from an
academic/military network to a commercial enterprise. Many
predicted that the Internet will not survive unless someone
figures out how to make money on the Net. Well, many have figured
that out and more: According to Business Week some companies use
the Internet more creatively than just simply as on-line shopping
mall. Royal Dutch/Shell Group successfully used the Internet to
coordinate improvements and innovations based on e-mail from
employees. Four out of five of Shell's top business initiatives in
early 1999 emerged from these virtual team efforts known as
GameChangers.
Other companies also increasingly use the internet to stimulate
and manage innovation. These "knowledge markets" have many
features in common with creative Silicon Valley upstarts with the
main difference that there is no single physical location where
these innovative minds get together but they meet in cyberspace,
connected from their computers located sometimes at different
company branches around the world.
In some sense by creating internal, innovation based, start-up
initiatives with access to the company's resources and knowledge
network those companies develop something like a company immune
system against innovative market invaders. Management consultant
Gary Hamel: ''Out there in some garage is an entrepreneur who's
forging a bullet with your company's name on it. You've got one
option now--to shoot first. You've got to out-innovate the
innovators.''
Using
the Net for Brainstorming,
Marcia Stepanek, Businessweek Online : December 13, 1999
Issue
The topography of memory, Nature
A basic question in brain science deals with the
identification of functions of different brain areas. Early
successful brain mappings could be established between low level
sensory input like touch and the mapping of the touch sensors on
the skin to corresponding locations in a homunculus in specific
brain regions. Other successful mappings could be constructed
between retinal patterns and patterns on the visual cortex. Those
mappings are less intuitive since they involve logarithmic
transformations.
Higher cognitive functions and memory are much more difficult
to pin point.
While the experts agree that memory functions (both encoding
and retention) are concentrated in the hippocampus. There have
been, however, different theories about what kind of information
is encoded in what structure. Animal experiments suggest that
spatial memory is encoded there but there is also evidence for the
encoding of non-spatial information in the hippocampus.
Hampson et al. designed some clever experiments with short term
memory tasks for rats that showed in what type of structures in
the hippocampus both types of information is represented. They
could show that neurons that become active when the animal
performs a task with spatial attributes are arranged along spatial
segments. "Within these same segments are ordered arrangements of
neurons that encode the nonspatial aspects of the task appropriate
to those spatial features. Thus, anatomical segregation of spatial
information, together with the interleaved representation of
nonspatial information, represents a structural framework that may
help to resolve conflicting views of hippocampal function."
The
topography of memory, Howard
Eichenbaum, Nature 402, 597 - 599 (1999)
Distribution
of spatial and nonspatial information in dorsal
hippocampus, Robert E.
Hampson, John D. Simeral & Sam A. Deadwyler, Nature
402, 610 - 614 (1999)
Active Hearing, PNAS
Theoretical physicists often turn some empirical common
sense upside down: "It is not that theories have to be developed
to describe experimental findings but experiments have to be built
to test theoretical predictions." The concept of anticipating what
will be observed seems to be a common feature of many complex
systems in many different areas. In ComDig 99.bXX we reviewed work
on how the firing of individual neurons will be not only affected
by the external visual stimulus but also by the anticipatory state
of its cell-assemblies.
It has been known for a while that the auditory system has a
similar mechanism in place to selectively respond to anticipated
sensory input in the cochlea. In this context of hearing it is the
sub-microscopic (nano) movement of sensory hair cells that is used
to selectively amplify sound signals. Martin and Hudspeth were
able to directly observe this amplification mechanism by
stimulating hair bundles of a frog -not used by the frog for
hearing but for detecting the smallest ground vibrations- with a
flexible glass probe. The oscillation amplitude the hair bundle
was observed to be as "great" as 40nm with frequencies in the
infrasonic range of 5-40Hz. The stimulation amplitude can be as
low as 5nm to reliably trigger the stimulated oscillation of the
hair bundle. The mechanical response of the hairs was that of an
anti-damped oscillator, thereby selectively compensating the
friction losses in the surrounding endolymph fluid.
Earlier attempts to demonstrate this active hair movement
failed, presumably because the ambient liquid plays an essential
role and the mechanism will not became active in standard extra
cellular fluid solutions. The oscillatory movement of the hair
bundles was highly non-linear and of the integrate-and-fire type
of neuronal oscillations: a slow buildup of about 30ms and a fast
release within about one millisecond.
The exact details of how exactly the hairs are moved and how
they convert chemical in mechanical energy is not completely
clear. The mechanism itself might have some interesting
application in nano robotics.
Active
hair-bundle movements can amplify a hair cell's response
to oscillatory mechanical
stimuli, Pascal Martin and A.
J. Hudspeth, PNAS, Vol. 96, Issue 25, 14306-14311,
December 7, 1999
-
Ovulation is one of the almost periodic, dynamic
phenomena in the female organism. The timing of which is triggered
by a number of hormonal and other factors. Many cases of
infertility are based on chronic anovulation which can be treated
by different forms of medication. For many mainstream therapies it
seems that timing of treatment and medication especially in
relation to chrono-biological factors of the organisms (" internal
clocks") and their phases are surprisingly under represented. The
standard prescription of taking medication at a specified time of
the day instead of at a specific phase of a physiological process
can be interpreted from a complex dynamical systems perspective as
"periodic driving of a non-linear oscillator with the resulting
possibility for bifurcations and chaos. Alternatively one can
apply nonlinear resonant stimulation that matches the intrinsic
phases of the driven oscillator.
Along these lines Buckler et al. studied if a "recombinant
follicle stimulating hormone (FSH) (Puregon®) can be
administered less frequently and at lower doses during ovulation
induction than is current practice. "They treated young, infertile
women who were resistant to previous medical treatments with low
doses of Puregon® but on alternate days only.
The treatment was highly successful ("there were six
pregnancies, five from Puregon® (100 IU) and one from
Puregon® (50 IU); four pregnancies proceeded to term") and the
authors conclude "that low dose alternate day Puregon®
treatment is suitable for this difficult patient group."
Ovulation
induction with low dose alternate day recombinant
follicle stimulating hormone
(Puregon), H.M. Buckler1, W.R.
Robertson, A. Anderson1, M. Vickers1 and A. Lambert,
Human Reproduction, Vol. 14, No. 12, 2969-2973, December
1999
What are Natural Antibodies Good For?, Science
The immune system is a complex adaptive system whose
agents learn to recognize and attack antigens. Most antibody
agents are formed as a response to antigens and they will remain
part of the immune system that will specifically respond to the
antigen to which it was originally linked. That is the basic idea
behind immunization. There are, however, natural antibodies
present before any immune response of the organism has taken
place. Therefore these antibodies do not respond to specific
antigens. They are often dismissed as background. According to
Ochsenstein et al. the natural antibodies might play an important
general role in the first stage of defense against an antigen
attack: Natural antibodies appear to be crucial in trapping
antigens in secondary lymphoid organs before they can spread to
vital organs.
The researchers prepared mice that did not have the background
of natural antibodies and then infected them with various viruses
and bacteria. They found that the number of antigens in peripheral
organs like kidney and brain were 10 to 100 times higher than in
those mice that had their natural antibodies intact. On the other
hand, the mice with intact natural antibodies had a 10 to 100
times higher concentration of antigens in their secondary lymphoid
organs. Although this result is quite convincing evidence for the
role of natural antibodies their more specific role in host
protection still leaves a number of open questions. For instance
it seems that they are also involved in the induction or
prevention of auto-immune diseases.
Control
of Early Viral and Bacterial Distribution and Disease by
Natural Antibodies, Adrian F.
Ochsenbein, Thomas Fehr, Claudia Lutz, Mark Suter, Frank
Brombacher, Hans Hengartner, Science, Volume 286, Number
5447 Issue of 10 Dec 1999, pp. 2156 - 2159
How chaotic is the Weather and Where?, PNAS
The butterfly effect introduced by meteorologist Ed
Lorenz served as a demonstration of the non-linear nature and
chaotic behavior of atmospheric phenomena like weather and climate
(According to Lorenz: "Climate is what you expect and weather is
what you get.")
In the early years of non-linear time series analysis a number
of attempts have been made to demonstrate that the weather is
indeed chaotic. These attempts -mostly based on calculating
fractal dimensions- generally failed. In hindsight one can argue
that the atmosphere has a much higher complexity than what can be
described by a small number of differential equations.
The degree of chaos in the atmosphere also changes in time and
also in geographical location. Qualitatively this can be seen from
the accuracy of local weather forecasts. More explicitly this has
been done by Pierrehumbert and coworkers some years ago by
estimating the locally instability, i.e. the magnification of a
perturbation caused by a butterfly.
Sugihara et al. basically ask the some question in a very
elegant way: they expected the effect of non-linearities to be
small in the tropics and increase at higher latitudes where
distinct cooperative phenomena self-organize into structures that
not only have names for their category, e.g. hurricanes, but each
individual, transient pattern is also given its distinct male or
female human name. At higher latitudes the structures loose this
individualized treatments and are lumped into generic categories
like cold-front, storm system, etc.
The authors analyzed daily barometric pressure values at four
different locations in Australia in tropical, sub-tropical and two
temperate zones. Their method to quantify the degree of
non-linearities in the data is based on Residual Delay Maps (RDM)
which basically indicate when the forecasting error of a linear
model shows non-random structures (the authors call it "V"-shaped
but it also looks a little like the "smile" of econo-physicists).
Since the linear model forecast mainly fails for small and high
pressure values (the "wings" of the "V") it seems to be clear that
this agrees with common sense that low and high pressure systems
should be mainly associated with non-linearities. The authors
could go even further with their analysis and explain an asymmetry
in the two wings of the V by looking at the data at different time
scales: low pressure systems appear to be more ephemeral, small,
and intense whereas high pressure systems tend to be more
persistent and large.
The authors then continue to analyze the data from a model of
the European Center for Medium-range Weather Forecast (ECMWF): it
was not a surprise that the primarily linear model produced
simulated weather patterns that lacked these non-linear features.
The authors predict that by taking these non-linearities into
account one could significantly improve accuracy of weather
forecasts.
We want to add that for any form of weather control (in the
sense of dynamic control of chaos) it certainly would be most
important to identify regions of high non-linearity and
sensitivity to external control forces (such as localized cloud
seeding).
Residual
delay maps unveil global patterns of atmospheric
nonlinearity and produce improved local
forecasts, George Sugihara,
Martin Casdagli, Edward Habjan, Dale Hess, Paul Dixon,,
and Greg Holland, PNAS, Vol. 96, Issue 25, 14210-14215,
December 7, 1999
El Niño Affects on Greenhouse Gas Emissions and Tree Reproduction, Science
The increasing number of environmental sensors both on
site and remote provide an increasingly complete picture about
what is happening within the global earth system. Instead of
isolated observations within any of the classical disciplines
physics, chemistry and biology it becomes possible now to tell
some more integrated stories about the state of the planet and
especially about the intricate interactions among the subsystems.
The latest 97/98 El Niño event (the largest such event
in this century) has been studied by Chavez
et al. from an interdisciplinary perspective: Wind and
temperature changes are correlated with changes in chemical
CO2 exchange between ocean and atmosphere as well as
with the abundance of biological phytoplankton and chlorophyll.
The authors find that an amazingly complex chain of events
followed both the warming during El Niño and also the
cooling during El Niña. The increase in sea surface
temperature cause a dramatic weakening of the trade winds which
then lead to a reduction in upwelling of cold, nutrient rich
water. This leads to an almost complete halt of the CO2
flux from the ocean to the atmosphere.
Note, that the El Niño region of the pacific is the
largest oceanic source of natural CO2 , almost 1015g
(that is about 50% of the 1996 US emissions of CO2)
have not been emitted by the ocean because of the El Niño
event. Consequently production of phytoplankton and chlorophyll
has been reduced by more than 50%.
The subsequent cooling event El Niña caused an increased
emission of CO2 from the pacific ocean into the
atmosphere. The tremendous amounts of CO2 that are
involved in this exchange dramatically demonstrate the importance
of the oceans on the atmospheric CO2 balance. An
increasingly complete network of Earth observing systems will make
it possible to directly visualize geo-system phenomena at global
scales and also the impact of human activity on those
phenomena.
A very dramatic effect of El Nino has been observed by
Curran
et al. in a family of canopy trees in Borneo: El Nino triggers
the synchronous dispersal of their seeds. It is known from complex
systems that synchronization events often require a critical,
minimal size. Because of logging in surrounding forests it seems
that even in a protected national park the trees fail to reproduce
even in the presence of a strong El Nino event.
Biological
and Chemical Response of the Equatorial Pacific Ocean to
the 1997-98 El Niño, F.
P. Chavez, P. G. Strutton, G. E. Friederich, R. A. Feely,
G. C. Feldman, D. G. Foley, M. J. McPhaden , Science,
Volume 286, Number 5447 Issue of 10 Dec 1999, pp. 2126 -
2131
Impact
of El Niño and Logging on Canopy Tree Recruitment
in Borneo, L. M. Curran, I.
Caniago, G. D. Paoli, D. Astianti, M. Kusneti, M.
Leighton, C. E. Nirarita, H. Haeruman , Science Volume
286, Number 5447 Issue of 10 Dec 1999, pp. 2184 - 2188
The role of stress transfer in earthquake occurrence, Nature
Earthquake prediction is one of the ancient challenges to
human ingenuity that iss still not solved to any level of
satisfaction. The number of catastrophic earthquakes in the past
year alone demonstrates its importance in terms of saving human
lives. While in ancient times cultures like China worked on
methods based on altered animal behavior (for instance geese) for
short-term predictions of earthquakes modern science has made
considerable progress in data acquisition and modeling to improve
the accuracy of identifying areas where most likely earth quakes
will happen in the intermediate future.
Ross S. Stein of the US Geological Survey describes how the
analysis of stress distributions along and perpendicular to fault
lines and especially their changes can be used to predict at a
given moment where and when most likely new earthquakes will
happen. He presents some very intuitive computer maps of the area
around San Francisco that illustrate the connection between stress
distribution and earthquake frequency. For instance in the 75
years prior to the 1906 great St. Andreas fault earthquake there
were 14 strong earthquakes in the area (magnitude > 6)
occurring along all major faults. In the 75 years following that
earth quake the number dropped down to just one magnitude 6
earthquake. That indicates that there is indeed a coupling between
earthquakes happening at different fault lines in the same region.
The author then discusses Coulomb stress transfer as "one
interaction criterion that promises a deeper understanding of
earthquake occurrence, and a better description of probabilistic
hazard."
While the analysis of static stress buildup has been used for
many years to identify probable locations of future earthquake,
the more recent work focuses on dynamic changes in stress
distribution that can be orders of magnitude larger than static
stress. Besides earthquakes (and underground nuclear explosions)
the author also mentions tidal forces that can create considerable
changes in fault line stress. It is not clear if this helps to
predict the time of the day an earth quake is likely to happen
The
role of stress transfer in earthquake
occurrence, Ross S. Stein,
Nature 402, 605 - 609 (1999)
How the Biosphere is Organized, Science
"Why are there roughly 700 species of birds that breed in
North America, rather than 7 or 70,000? And why, in comparison,
only 200-odd in Britain? For the first question, we are a long way
from having the basic understanding that could begin to answer it.
For the second, we have an empirical rule and the beginnings of
fundamental explanation, but we still have much to learn.
(...) Fragile Dominion is organized around six fundamental
questions: What patterns exist in nature? What are the relative
roles of historical accident versus environment determinism? How
do ecosystems assemble themselves? How does evolution, acting on
individuals, shape assemblies? What is the relation between an
ecosystem's structure and its function? And does evolution favor
resilient systems? (The central question these are all derived
from is, "Why is this organism different from other organisms?")
"
How
the Biosphere is Organized?
Robert May, Science Volume 286, Number 5447 Issue of 10
Dec 1999, p 2091
Review of Fragile Dominion
Complexity and the Commons, Simon Levin, Helix
(Perseus), Reading, MA, 1999
