Dear Kees,

A very interesting question. The orthodox (and correct in my opinion) answer is that, at any moment of the evolutionary process, there is no reason why complexity should increase rather than decrease. Yet, if we look back to the course of evolution a trend toward increased complexity appears to be obvious. So what is going on? Insufficience in the (Darwinian) theoretical frame?, inherent trend toward complexity? Distorted perception of reality?

I must say that I have nothing against enriching the Darwinian theory whenever it is justified, but the last explanations is certainly relevant. Firstly, the dilemma between increased and decreased complexity is an asymetrical one because complexity cannot be less than 0. More generally, once a certain degree of complexity is attained, environmental constraints will not allow too drastic a reduction of complexity, which will usually result in loss of vital functions. Only very specific conditions will permit it. The most obvious case is obligatory parasitism or symbiosis, especially endosymbiosis. The case of mitochondria and chloroplasts is interesting because according to what you look at, you can see a decrease in the complexity of the genome of the organella and an increase in the complexity resulting in the appearence of the eucaryotic cell.

The difficulty with concepts like complexity, information, entropy is that they are highly dependant on the point of view of the observer.

There is an answer in Steve Gould's (1996) book "Full house" and it goes like this:

There is a threshold of minimum complexity for full-functioning organisms which was crossed (once?) when life originated but never during the latter course of biological evolution (otherwise a living organism would have degraded to non-living matter).

Starting from a set of organisms with the lowest possible complexity, steps of evolution lead equally often to (a) a more complex organism, (b) an organism of similar complexity, or (c) a less complex organism (if possible).

If you now run a simulation of evolution, e.g. a Monte Carlo experiment, and implement the boundary conditions mentioned above, you will find that over time an extremely right-skewed complexity distribution arises with the bulk of organisms (e.g. bacteria) close to the threshold of lowest complexity and a few organisms becoming increasingly complex (maximum complexity increases passively over time).

Gould says that human view on evolution is biased and that we (wrongly) percieve an overall complexity increase over time because we tend to neglect the bulk of organisms that has stayed close to the threshold of lowest complexity over billions of years.

Of course Gould's answer was not intended to be final. Today organisms of lowest complexity (bacteria + archeans) are superior in their number of individuals, but is that also true for the amount of biomass (Is there really much more bacterial than plant biomass? - this question was also discussed in Gould's book) or measures of diversity?

And also: Is focussing on individual complexity reasonable or should we rather look at the overall complexity of ecological relations or the functioning (e.g. energy and material flux) of ecosystems and biosphere as a whole?

There has been some discussion about special thermodynamic conditions on earth that induce a global long-term evolutionary trend, see e.g. RG thread here:

https://www.researchgate.net/post/Is_Darwins_theory_of_evolution_through_natural_selection_complete_and_consistent

and papers by Karo Michaelian about that matter.

Luc and Michael, thank you for your insights. I shall consult the readings that you suggested to get a better feel for the current understanding of whether there is a trend in complexity over uninterrupted evolutionary time.

Luc, I like your three options:1. Insufficient (Darwinian) theoretical framework? 2, inherent trend toward complexity? 3. Distorted perception of reality? option 3 is certainly a possibility since it is we who impose constructs on natural phenomena to help us understand them. The insufficient theoretical framework is this limited to biology? Has the abiotic part of the universe not become more complex since its origins from a singularity to clusters of galaxies, black holes and other amazing structures? If the abiotic part has become more complex over time, might not the same principles apply to the biotic part of the universe?

Michael, I agree wholeheartedly with your point about how and where we look for complexity. The scale question is extremely important. I think that we should be looking at different scales, but make sure that we state very clearly at which scale we are working. For example, adult parasites may become morphologically less complex than their non-parasitic ancestors, but their life cycles may become more complex with each stage of development having its own specific adaptations to its mode of life.

There is no reason that, in the course of evolution, there was no reduction in the complexity. Evolution is not targeted.

An example of such phenomena is the total loss of genes, involved in the stomach functions, which occurred in many groups of animals (i.e. platypus):

L.F.C. Castro (2013) "Recurrent gene loss correlates with the evolution of stomach phenotypes in gnathostome history". DOI: 10.1098/rspb.2013.2669

The awful problematic of the question is the concept of complexity. But I think we can avoid with the question:

Why the evolutionary reduction, e.g the body plan simplification of Psilotum genera, is less common that the conservation or specialization of morphological traits? Is it just a contingency?

The challenge here is to determine if there is a theoretical reason as to why complexity should increase over uninterrupted evolutionary time. I think Micheal's referral to Karo Michaelian's papers is a significant step in increasing our understanding of the issue. The Second Law of Thermodynamics (non-equilibrium dynamics) drives the trend to increasing complexity because increased complexity increases the rate at which entropy is produced. The single celled organisms have not disappeared because their ecological function creates the foundations upon which more complex life forms can evolve and be supported. The complex multicellular organisms have communities of bacteria in their guts to digest their food and access nutrients contained therein.

Aimer implicit in your statement is the importance to get a shared understanding of the definition of complexity because without it we shan't make much progress. Although often it is useful to change the frame of reference, in this case avoiding the concept of complexity is not the better approach. I say this because we can use our understanding of the causes of complexity to increase the interconnectivity between biology and the realm of chemistry and physics.

I don't recall McShea and Brandon addressing non equilibrium thermodynamics as the theoretical basis for the first law of Biology. They claim there is this law or tendency but they do not provide a thermodynamic basis for this tendency. I shall read the book again just to make sure. But I do agree with them that the tendency to increase in complexity occurs with or without natural selection. Non-equilibrium thermodynamics drives the tendency to increasing complexity and this is aided by matter's characteristic to self-organise. The same principles that create increased complexity in chemical and physical entities also create increased complexity in biological systems. So an energy gradient is necessary to drive the tendency to increased complexity.

Hi Emily thanks for that. I use Holling et al.'s Panarchy Theory to help my understanding of ecological systems at different levels of organisation ,i.e. different temporal and spatial scales. That having been said, Panarchy does not explain why systems become more complex over uninterrupted evolutionary time. It does explain why disturbance does decrease complexity. 

Increasing complexity is related to adaptation. Systems are in an environment and have to steadily adapt to challenges. This must be done in due time at all levels of systems and sub-systems. There is a kinetic aspect. Adaptations are done on base of internal organization and available resources as consequences of positive and negative feedback loops of any order appearing in the system.

For living organisms, environmental challenges will induce internal changes, e.g. genome mutations, which are random by default but can be challenge oriented and depend on the available resources (see reaction to temperature, … radiations, ….. chemical or physical stresses, ….., lack of resources, all operons – e.g. lactose operon, tryptophane operon in E-coli, etc…etc…etc…). Evolution in a given local environment will favor individuals with the best adaptation solutions. It has been demonstrated that favorable mutations at a given locus can happen under specific stress conditions (environmental changes) much more often than in neutral conditions (B. G. Hall « Spontaneous point mutations that occur more often when advantageous than when neutral » Genetics, 126 (1990), pp. 5-16). Even what one sees as damage at a given level might induce increasing differentiation and complexity at an higher level. Note that adaptation is also at the root of punctuated equilibria at smaller scales which give evolutionary patterns towards complexity at higher scales (= « uninterrupted evolutionary time » as you put it in your question).

There is no general trend towards complexity. Only more complex solutions may sometimes arise from local adaptations to challenges with the time going (which is available at leisure on evolutionary scales). This is like « left censored random walk » as described by S. J. Gould in the book mentioned by Michael Buchwitz (« Full house », Ed. Harmony Books, New York, 1996 or – in French - « L’éventail du vivant » Ed. du Seuil, Paris, 1997). Like an erratic drunkard which sometimes walks a bit farther to the right instead of frequently bumping upon the wall existing on the left….