Talk:Interdependence

2007/12/14 Request for Volunteers

Everyone, I need some volunteers to help work on a third proto-pattern for the IBE presentation and subsequent JBE paper submission. How individual organisms interact has popped up a number of times recently, such as the Regen Energy article in the December issue of BioInspired! as well as an article on management theory that is planned for the February issue. At the risk of forcing a square peg in a round hole, I would be interested in working on the following principle, exploring how it relates to swarm theory and other ideas about social interaction in natural systems. My sense is that this pattern will tie in to the other two, allowing us to sketch out the very beginning of a pattern network.

I would like to start work on this topic in early January, and would let to start scheduling conference calls. Let me know if you are interested, or if you have another ecological principle that you would like to explore.

Thanks, Norbert

>>>

Ecosystems tend to be made up of interdependent cooperative and competitive relationships

• Emergent effects tend to occur
• Systems tend to be self organising and distributed

2007/12/15 Response from Fil

I'd like to help. Off the top, I see two underlying 'concepts': balance, and evolution.

Preliminary, out-of-my-ear, thoughts:

• Cooperation & competition balance one another, in the context of an environment/context. In a different context, the balance will be different.
  • New/related point: In The Mind of the Market, Michael Shermer writes about how evolutionary economics explains that greed and compassion balance each other and that the market is more like Google's "Do No Evil" model than the Wall Street "Greed is Good" approach. I think greed/compassion fits fairly well with competition/cooperation. --Fil Salustri 13:49, 30 January 2008 (EST)
• Emergent effects are rather like "side-effects" of something else. Thumb-sucking is an emergent property of having an opposable thumb, originally selected for due to the grasping advantage it offered in the original context.
  • Emergent effects stick only if they offer an advantage. If they do, it also reinforces selection for the primary effect.
• Self-organizing also offers evolutionary advantages, but only insofar as the organization renders benefit in a context.

Cheers. Fil

2007/12/26 Response from Eileen

Hi Norbert, I'm interested in helping out as well, but will be out of town through 1/6. Also, I am starting full time work upon my return, so I will face more time constraints (and apologies again for missing the last call at the last minute). Happy new year to everyone; hope all had a great Christmas. Eileen

27 Dec 07: More comments on emergence and self-organization

Two points on emergence:

First, is a behaviour emergent only because we perceive it to be so? That is, if we could measure and model every possible variable and relationship in a system, we could predict very accurately all the possible properties that would arise/emerge from a system. This is a reasonable assumption because we can a-posteriori describe (at least theoretically) how a system's properties emerged. So emergent properties are already in a system, latent or hidden, until conditions are such that the properties naturally manifest themselves. So I suggest we be careful to keep away from the idea of emergence as the creation of new properties.

Second, emergence seems to connect to Don Norman's notion of affordance. An affordance is a property or characteristic of a thing that suggests how it can be used. A vertical door handle suggests pulling it to open the door, whereas a horizontal door handle suggests pushing to open. Behaviours can be said to emerge as a result of affordances. This suggests that there may be things analogous to affordances in natural systems, which lead to emergent natural behaviours. This could be a useful, albeit anthropocentric, approach because we already understand quite a bit about affordance. If we could model natural systems to include affordances, then that could help analogize natural and artificial systems.

Finally, I would argue that naturally self-organizing systems are also naturally distributed because they're constructed bottom-up. Simple lifeforms came first, and systems of those lifeforms evolved (self-organized) over time. Since there's no intent in the self-organization, then the elements must be somehow whole unto themselves, and when they form into a larger self-organized unit, different elements contribute based on what they can do best, given their particular contexts.

Here's a basic problem for designers: design is a top-down process, always has been, and always will be, because it's problem solving with intent. Natural systems are not developed by intentional agents who are solving problems.

Self-organizing systems seem to emerge. And self-organizing systems are more complex than their constituents. So I'd say that a key emergent property is complexity, and that self-organization is either a synonym or subset of complexity itself.

...more grist for the mill.

--Fil Salustri 15:01, 27 December 2007 (EST)

User:Martha Here is what I have come up with on Interdependence after a little brainstorming and reading the Talk. To start, I will use the categories Norbert has put in the Problem or Challenge statement.

• Substance
  • The things that are interdependent are not always easy for us to see and this poses problems for modelers.
• Structure: Obviously, interdependence is about relationships. I think of the following:
  • niches -- based on strengths as Fil mentioned. Filling or finding niches can serve to reduce competition. But two species with nearly the same strengths, limits, etc. may end up fighting for a defined niche and this fight will not go on indefinitely but will only continue until there is a winner. (Loosing might involve moving elsewhere or dying.) There might be exceptions to this.
  • specialization -- usually leads to greater interdependence among members of a community (I use that word loosely here).
  • symbiotic relationships -- often between specialized members. (So if you were a jack-of-all-trades, you might be so self-reliant that you have no need for this arrangement or much interdependence at all.)
  • Isn't a goal implied? Members come to be interdependent why? Because it serves them somehow. Helps them survive. Helps them thrive (be "successful").
• Space: interdependent members can easily span space. Plants release oxygen in, say, Equador and a respiring organism consumes that oxygen maybe thousands of miles away. Or quite close. Space may be a limiting thing in some contexts: the fungi on plant roots that help the plants absorb more nutrients and water than they could without the fungi -- this depends on the fungi being right on a part of the plant.
• Time: Hmmm, this might be a limiting factor because of the "betweeness." For example, though we are currently dependent on fossil fuels (a product of a process run by organims) the creators of fossil fuels got nothing from us. Thus this is not interdependence. It seems like if time seperating members is great, interdependence just can't describe that relationship (because of the implied goal -- survival of at least one type of member did NOT depend upon the relationship.)
• Scale: again,this might be a limiting thing but I would say there are many examples of interdependence between members that differ in size by magnitudes. Wasn't it E. O. Wilson who said: "Little things rule the world."? Many vertebrates are dependent upon micro-organisms living in their gut and the microbes benefit from having safe habitat with lots of nurishment that just flows by. (So this ties back into -- we can't see the interdependent pattern sometimes until we develop tools that allow us to see the members that live in a scale the naked eye can't access.)

Something else to consider: interdependence as opposed to commensalism. Maybe members look interdependent on first glance but actually their relationship is better described as commensal (one benefits, the other is neither helped nor harmed). And again I think of something from E. O. Wilson: we often need micro-organisms but they don't need us. (After all some of them evolved before we existed and were doing just fine and could do just fine probably without us.)

30 Jan 2008: More comments

--Fil Salustri 13:49, 30 January 2008 (EST)

User:Martha wrote some very interesting stuff. Building (I hope) on that:

"The things that are interdependent are not always easy for us to see...." My take on this has to do with boundaries. I once heard a lecturer suggest that one cannot define where an organism is without a-priori defining the organism's environment, and that one cannot define the environment without a-priori defining the organism. This is a paradox. The lecturer argued against distinguishing between organisms and environments for that reason. But while the paradox is valid as stated, it's not the only way to think about it.

Our brains (and presumably those of other animals) is structured to identify key properties - colour, motion, size, shape - separately and in parallel. My reading of the lit on this leads me to think that the first thing the brain recognizes is a change in something (e.g. a change in colour). A change in something is a boundary. Boundaries needn't be crisp - we engineers use the term boundary layer alot, and it's meaningful here too.

When we don't see things that are interdependent, it's cuz we can't see the boundaries - the changes between the constituent elements. That means we're not looking for the right characteristics/properties. So when modellers can't do their thing, they need to rethink how they're "measuring" the thing they're modelling, think outside the box as it were, and look for different way of distinguishing changes.

I can accept there being a "goal" to becoming interdependent, but it's not a goal that the entities necessarily know. It's built into evolution, the survival instinct. So I resist using the term "goal" which implies intent, which I think is definitely lacking in natural systems. Another problem with "goal" is it implies an end-state. But evolution doesn't have an end-state. It just keeps going no matter what. I think of it as a state-transition thing. There a (hugely complex) set of rules that define how to get from one state to another. A natural system is in a given state. One of the rules is triggered by some event (e.g. a food shortage) and a new state arises. The new state exerts "forces" on the entities that drive them to kinds of behaviours particular to that state. This causes more events, which trigger more rules, leading to another state.... Wash, rinse, repeat.

The matter of time is interesting. The dinosaurs gave us oil without wanting anything in return (can you imagine how rich today's lizards would be if the dinos did?). But I'm not sure there's no counterexamples. Gotta think more about that.

Re: commensal relationships. Sure the micro-organism might not need us, but they'd need something. ...Or am I missing the point?

2 Feb 2008: comments

Martha Love Yes, I can agree with the boundary thing. Our perception of boundaries is crucial. So when we can't see the fungi on the plant roots (can't see where a fungal membrane contacts an epithileal layer of the plant root), we can't see that there is a relationship and exchange between them. And with "boundaries needn't be crisp," I think of something I read about the layer of bacteria on our skin -- where does "you" end? Conceptions of our boundary (our skin) actually includes the microbes -- they are, in a way "you" also.

When I use the word "goal" for natural systems, I don't mean to suggest conscious goals, but I understand how the word itself triggers a qualminess. In this context, I think of goals as related to natural laws. I have been influenced by reading this book Into the Cool (by Schneider and Sagan). So, for example, nature does have a "goal" in that "nature abhors a gradient" and events come about, things and systems evolve, "in order to" dissipate gradients. I looked ahead in the book and the end actually addresses this directly with a segment on "Purpose". Here again, we scientists get qualms -- the creationists are going to think we are playing into their hands because the words "goal" and "purpose" begin to suggest "intelligence". But I would say, some Intelligent Design proponenets need to think outside of the box on the word "intelligence" and what that can be. Schneider and Sagan actually include segments on the ID writers.

I, too, was trying to think of cournter-examples about the time part of interdependence. But I admit I haven't devoted a lot my time to it. Also, after I sumbitted those comments, I started seeing maybe a confusion of perspectives on my part: We can look at "interdependence" on a very broad level, where we acknowledge that yes, these living entities are dependent on something in the system (like the microbes that don't need us do indeed live on something). And on another level, like in using the "commensal" concept, I zoomed into a specific relationship of interdependence. All those words like "mutualistic", "symbiotic", and "commmensal" describe a relationship between just a couple or a few entities. So again, context comes up. (Hmm, it seems like context makes it hard to get to general patterns.)

Ecological Networks

As promised during the last conference call, I uploaded a paper discussing food webs and ecological networks in general. Filing this work about material, energy and information exchanges between organisms beneath the Interdependence pattern seems appropriate to me. If others believe it fits better in a different location, I will not object to having it moved. --John 12:25, 14 February 2008 (EST)

--Norbert 15:41, 19 February 2008 (EST) John, the "Ecological networks and their fragility" starts well, talking about the presence of well defined patterns, differences from non-ecological networks, and the paradox that complexity should lead to fragility (at least in the mind of mathematicians and computer simulations). However, I was not able to tease a lot of additional insights - every time I thought I came across something I could understand, it turned to fog. For example, does paragraph 2two on (real) page 263 imply that removing a top predator has a greater impact on species two levels down, than removing a lower level predator? Paragraph three seems to imply that food webs are resilient in the face of random species extinction, but vulnerable if well-connected species are lost. This seems to contradict statements early in the article that "each species [is] typically closely linked to all others...". Maybe a distance of two or three links does not justify 'close'? Any insights would be appreciated!

--John 11:06, 20 February 2008 (EST)Norbert, the authors supposedly wrote the paper to, "...compare ecological networks to non-ecological networks, and consider their similarities, differences and underlying causes." I believe they failed to achieve this goal. They spend most of the paper discussing ecological networks, metrics meant to measure properties of ecological networks and a few common properties of said networks. Networks of technological significance such as the communication network, electrical grid, or even a municipal water supply receive scant, if any, mention. However, they provide operational descriptions of a number of network metrics relevant to ecology, and they include discussions of food webs with less studied ecological networks such as mutualist networks. The also reference interesting papers by May on stability and by Watts on "small-world" networks. I chose this paper not because it tells a complete and compelling story but because it contains a wealth of connections to many interesting stories about ecological networks.

--Norbert 22:00, 19 February 2008 (EST) It just occurred to me that fragility or stability has two aspects: loss of species or loss of the niche which those species inhabit. I suspect that the two are related, in that a niche incorporates both physical and biological elements. However, my guess is that a niche is less prone to disappearing compared to a species - if vacated, it is likely that the niche will be re-occupied by a similar species to the one that went extinct. That species may subsequently fulfill similar 'food web' relationships, giving the appearance (over the right time scale) of a resilient food web.

--Martha 11:43, 23 February 2008 (EST) Initially, I was going to suggest looking into "keystone" species roles in food webs but maybe that's 'old hat'. Not every food web has to have a keystone species, I'm guessing. For example, if there is a lot of redundancy in the food web, removal of one species may not have a drastic effect on the composition of the community or the environment, which is a way to recognize who is a keystone species. The paper was pretty advanced (for me) and I was struggling to understand what "compartments" in food webs were (the diagram in the paper helped me though). I don't think they mentioned keystone species, at least not using that term. But the discussion of how removing a predator ends up decreasing the numbers of a species 2 links down reminded me of keystone species. If the removal of the predator caused a population explosion of the prey 1 link down and this in turn decimated the population of its prey, the drastic change that this might cause in the environment would earn the top predator mentioned the label "keystone species", as I understand the concept.

--Fil Salustri 11:23, 26 February 2008 (EST): I'm getting a sense from this discussion that (a) our models suggest complex systems should be fragile, but (b) they're not in reality. Could it be we have the wrong model? Specifically, could the complexity be only a perception of ours? I know it's a wacky idea, but when there's a conflict between theory and observation, one must question one's assumptions. I will read the paper "real soon" and if it makes sense to me, will post info.

--Martha 19:00, 26 February 2008 (EST) I'm not yet buying that complex systems should be fragile (but I am not well-read in this area). This sounds a lot like the debate of "Does greater diversity lead to greater stability?". (The general biology textbook I taught from two years ago, said that essentially the jury is still out on that.) What I'm thinking is that this is something that is hard to make generalization about, because 1) "complex systems" can be so different (unless one make some rigid definition of "complex system") -- the argument could go that one complex system with more redundancy could be less fragile than another complex system and 2) again, semantic in nature, what does "fragile" mean? Complete collapse? Ability to bouce back? If so, how quickly? That last one brings in the dimension of time, which seems to confound me on several on my thought excursions related to this pattern language project. Related to interdependence, but changing the subject slightly, I looked up "Why big fierce aminals are rare?" which I mentioned in the conference call in regard to food webs. I thought it was going to be an article but it's a book, which I haven't purchased yet. (So Norbert, here is yet another assignment for all of us students who find ourselves back at university thanks to this project.) The blurb about Why Big Fierce Animals Are Rare: An Ecologist's Perspective by Paul A. Colinvaux, on the Princeton University Press website said: "A mind-tingling survey...of the many factors involved in the interrelationships of all living things..." It was published in 1979 and is a classic. Perhaps some of you have read it already.

--Norbert 22:31, 27 February 2008 (EST) Fil/Martha, I think you are on to something. We need to quantify what 'complexity' and 'fragility' mean. Another aspect is that things we see as complex may have simple origins (based on complexity theory) - it may be that the perceived complexity is maintained or restored because simple processes are rebuilding it.

I suspect scale are also important. For a while, we thought mature ecosystems were stable, but that was a figment of looking at the wrong time frame and spatial scale. Mature forests are constantly changing, yet appear to maintain a degree of continuity if you pick the right scale.

Boundaries may also be critical. I can see a food web on an island being much more vulnerable to events that significantly disrupt the interrelationships within that web. Food webs that are not isolated and suffer local damage can be 'repaired' by migration from outside the area of damage.

A diagram showing interrelationships between species does not take these factors into account. Time to do some more reading (:-).

--Martha 13:03 29 February 2008 (EST) I think I'm simply confused by contradictory things I've read. On the one hand, someone says well, the Potato Famine was caused by having homogeneous crop (lack of comlexity, it would seem). It was a fragile community. On the other hand, it does seem somewhat intuitive that a complex system should be fragile. Yes, I agree that some of these apparent contradictions might dissolve if we get a grasp of scale.

--Norbert 13:56, 29 February 2008 (EST) Martha, does it make sense to think about 'weak links'? In the case of the potato famine, large parts of the population where heavily dependent on a single crop. That crop was struck by the blight, and people starved (although I am told there is a lot more to the story). Many of our complex systems contain numerous weak links, such that the risks and/or impacts are additive - a problem in any one of these links has far-reaching and potentially disastrous results. A simple example is a watch (analog or digital). There are numerous components that all need to work, otherwise the watch stops.

Natural systems may be equally complex but contain sufficient redundancy and flexibility so that one or more weak links can fail without system failure, or the scope of damage is limited (an example is nacre, where the soft layers prevent crack propagation, and in addition heal the damage). I may be using poor terminology - if a failing 'weak link' has limited effect, is it really a 'weak link'?

The NASA approach is to building backup systems and systems to backup the backup systems. Simple duplication is not enough - this approach can limit the impact of an internally-caused failure, but an external force can either wipe out the primary and backup, or exploit a flaw that both exhibit (I have battle scars to back up this up). My impression is that in natural systems, numerous 'nodes' can 'fill in' if one of the nodes fails, not because they are clones, but because they are multi-functional and are able to fill an empty niche. Few predators are so specialized that they cannot switch to other prey if the need arises (although possibly at reduced efficiency/effectiveness). This links back to the Kay/Schneider work on measuring the impact of stress on a system - the system may survive but run degraded for some period of time.

At the same time, redundancy and flexibility incur costs (reference NASA). Can natural systems teach has how to balance redundancy/flexibility with cost? One of the big challenges in industry is that redundancy that is not regularly used is quickly cut from the budget when times are tough. Having too much redundancy can end up attracting attention to itself.

It might be worthwhile creating a number of scenarios using the structured narrative approach suggested by Kay and looking at a range of human and natural systems. Interestingly, the Internet is often portrayed as a very complex system that is virtually impossible to fully comprehend. Yet the design of the Internet makes it surprisingly resilient in some cases, such as the cable cuts in the Mediterranean. Even though there was an impact, alternative routes were established relatively quickly. In contrast, the Internet can also be vulnerable - when Pakistan decided to block YouTube access, an error by one of the ISPs 'infected' the Internet routing structure and took YouTube down for a large part of the world for several hours.

--Martha 15:15 (Daylight Savings) 29 March, 2008 Sorry for the time gap in responding. I'm not an expert on this topic but, in answer to your question, Norbert, I would say yes, weak links would be a pretty big factor in networks, ecological or otherwise. I like the point you made about how in nature generally 'nodes' can fill in because of multi-functionability. So...where to go with this? Perhaps this can be used in some definition or quantification of "fragility", as you suggested we needed. Like, if somehow we know that such-n-such ecological network has x number of weak links (and some number of possible replacement 'nodes') that network can be described as "very fragile", or "somewhat fragile", etc. Do ecologists look for weak links? If so, I don't know their methods. Intuitively it would seem acquring knowledge of weak links and multi-functionability would require a lot of time in one particular ecosystem and watching it closely over a range of conditions.

It just now occurred to me that (if we want an example scenario that might lead to insights that are hopefully general to networks and interdependence) pathobiology could be useful, unless we just want to focus on ecology in making the interdependence pattern. (Though ecology is interesting to me, metabolic networks are my favorite and I'd like to think that I can still use my training and carve out a little niche for me in an unlikely place.) In school, I had to take a course on the pathobiology of inherited diseases. These were often characterized by one (or if recessive, two) defective allele(s) causing a horrible disease. That was a whole semester of studying weak links. These metabolic pathways could be compared to ones that rarely malfunction (because the latter have less fragility, i.e. redundancy, adaptibility). I could look into this and hope it leads to something fruitful (applicable) and not a useless exercise.

--Norbert 12:25, 1 April 2008 (EDT) Martha, it would be great if you could follow up on weak links in metabolic pathways. What is intriguing about the work we are doing is the applicability across a wide range of scales. You may discover something at the metabolic level that provides insights at the ecological level.

Ecology: Recent Nature Articles

Ecology: Destabilized fish stocks p825

Fishing of natural populations increases the variability of fish abundance. A unique data set from the southern California Current has allowed an evaluation of three hypotheses for why that should be so. Nils Chr. Stenseth and Tristan Rouyer

doi:10.1038/452825a

Full Text | PDF

See also: Editor's summary

Why fishing magnifies fluctuations in fish abundance p835

Increased volatility of exploited fish stocks is due to amplified nonlinear behaviour caused by fishing. This paper shows how selective harvesting can alter the basic dynamics of exploited populations, and lead to unstable booms and busts that can precede systematic declines in stock levels. Christian N. K. Anderson et al.

doi:10.1038/nature06851

Abstract | Full Text | PDF | Supplementary information

See also: Editor's summary | News and Views by Stenseth and Rouyer

Ecological Network Literature

At one point, Norbert asked for literature on food webs and other ecological networks. I prepared a literature summary on ecological networks as part of my ongoing dissertation research. It provides a brief overview of both older and cutting edge publications in this area and is germane to discussions logged on this wicki. Enjoy!