Generative entrenchment and an evolutionary developmental biology for culture

Inferences from dependencies in development, or from conserved traits, structures, or genes, have become a central and ubiquitous tool in evolutionary developmental biology. An element with many downstream consequences of its operation in an adaptive system should be conserved because its deletion or change would likely have massively deleterious effects. Thus, substantial observed dependencies should predict conservation, and vice versa. Dependencies and conservation are both matters of degree, generating an ordering of rates of evolutionary change. I trace four independent sources for this idea with subtle differences in their application. The origins of the use of genetic pleiotropy to predict conservation is complex and progressed in multiple stages, with partial awareness emerging in the 1960s, and grew erratically in use without becoming a major theme until the 1990s. Since 2000 it has grown to become an essential tool and applied in an expanding number of ways. I distinguish and illustrate “top-down” and “bottom up” approaches, though people now working from one direction often recruit collaborators from the other. Other approaches have had diverse inspirations and points of initial application. Rupert Riedl’s notion of “burden” was applied primarily to morphological traits, with an elaborated account of organizational factors modulating entrenchment. Wallace Arthur explored dependency in “morphogenetic trees,” utilizing as his inspiration cellular descent trees in metazoans. Bill Wimsatt looked for “generative entrenchment” in factors both internal and external to the developing system to analyze innate behavior. Entrenchment was later extended to analyze development in Drosophila and to construct multi-locus population genetic models of gene control networks, yielding new results that illuminate relations between micro and meso-evolution. Applications of generative entrenchment to cultural and technological evolution are also sketched, where development of individuals and other social entities must play a central role.

This article aims to draw a connection between organismic evolution and machine learning as recursive optimization processes. Optimization of complex systems presupposes certain forms or designs of the input-output functions. Recent literatures in evolutionary developmental biology have discussed various design features of the genotype-phenotype mapping, including neardecomposability, generative entrenchment, standardization, plasticity, canalization, and scaffolding as means to solve complex adaptive problems through recursive evolution. I point out similar problems and/or techniques exist in the machine learning literature, and sketch some common features in these two distinct fields.

The innate-acquired distinction has long been a point of connection between evolutionary questions and developmental questions. It is also one of our oldest conceptual tools, dating back at least to the time of Plato. Further, it hs been broadly applied not just in development biology, animal ethology, and evolutionary theory, but also in epistemology, metaphysics, ethics, philosophy of science, cognitive psychology, linguistics, neurophysiology and in theories about the nature and evolution of human sociality, culture, and morality. As befits any concept which has cast such a broad and deep shadow, it has at some times and places been highly honored and at others most deeply dishonored. It has been viewed as an agent of progress and of stagnation, of clarity and of confusion, and of liberation and of bondage.

The question of vertebrate head segmentation has become one of the central issues in Evolutionary Developmental Biology. Beginning as a theory based in comparative anatomy, a segmental theory of the head has been adopted and further developed by comparative embryologists. With the use of molecular and cellular biology, and in particular analyses of the Hox gene complex, the question has been addressed at new levels, but it remains unresolved. In this review, vertebrate head segmentation is reevaluated, by introducing findings from experimental embryology and evolutionary biology. Developmental biology has shown that pattern is generated through hierarchically organized and causally linked series of events. The question of head segmentation can be viewed as a question of generative constraint, that is whether segmentation in the head is imposed by underlying segmental patterns, as it is in the trunk. In this respect, amphioxus appears to be segmented along the entire anteroposterior axis, with myotomes and peripheral nerves repeating with the same rhythm (somitomerism). Similarly, in the vertebrate trunk, the segmental patterns shared by myotomes, peripheral nerves and vertebrae are derived from the somites. However, in the head of vertebrates there is no such mesodermal pattern, although neuromerism and branchiomerism do indicate the presence of constraints derived from rhombomeres and pharyngeal pouches, respectively. These data fit better the concept of dual metamerism of the vertebrate body proposed by Romer (1972), than the traditional head cavity-based segmental model by Goodrich (1930).

Many scientists and philosophers of science are troubled by the relative isolation of developmental from evolutionary biology. Reconciling the science of development with the science of heredity preoccupied a minority of biologists for much of the twentieth century, but these efforts were not corporately successful. Mainly in the past fifteen years, however, these previously dispersed integrating programmes have been themselves synthesized and so reinvigorated. Two of these more recent synthesizing endeavours are evolutionary developmental biology (EDB, or "evo-devo") and developmental systems theory (DST). While the former is a bourgeoning and scientifically well-respected biological discipline, the same cannot be said of DST, which is virtually unknown among biologists. In this review, we provide overviews of DST and EDB, summarize their key tenets, examine how they relate to one another and to the study of epigenetics, and survey the impact that DST and EDB have had (and in future should have) on biological theory and practice.

Evolutionary biology is currently experiencing an emergence of several research topics that transcend the boundaries of the Modern Synthesis, which was the last major conceptual integration in evolutionary biology [1]. The Modern Synthesis used the concepts of population genetics to integrate Mendelian genetics with evolution by natural selection [2]. Pigliucci [3, and citations within] identified several major areas of innovation that transcend the Modern Synthesis: epigenetics, evolvability, phenotypic plasticity, evolution on adaptive landscapes, evolutionary developmental biology, and systems biology. Integrating these new ideas with the Modern Synthesis will form a new conceptual framework of evolution, which they termed the Extended Synthesis, as it will extend, rather than refute, the Modern Synthesis [3]. This subject has been the focus of much recent work, and an excellent description is provided in the book Evolution—The Extended Synthesis [2].

Michael Akam has been awarded the 2007 Kowalevsky medal for his many research accomplishments in the area of evolutionary developmental biology. We highlight three tributaries of Michael’s contribution to evolutionary developmental biology. First, he has made major contributions to our understanding of development of the fruit fly, Drosophila melanogaster. Second, he has maintained a consistent focus on several key problems in evolutionary developmental biology, including the evolving role of Hox genes in arthropods and, more recently, the evolution of segmentation mechanisms. Third, Michael has written a series of influential reviews that have integrated progress in developmental biology into an evolutionary perspective. Michael has also made a large impact on the field through his effective mentorship style, his selfless promotion of younger colleagues, and his leadership of the University Museum of Zoology at Cambridge and the European community of evolutionary developmental biologist.

In the 1960s the new technique of gel electrophoresis revealed what was then considered to be an astonishing amount of molecular variation in natural populations, with about 30% of genetic loci being polymorphic. It was thought at the time that natural selection the dominant principle in evolutionary biology could not account for such high levels of variability the costs of selection would be too high. This prompted the proposal of the neutral theory of molecular evolution (Kimura 1968; King & Jukes 1969; Kimura 1983), which postulated that most molecular evolution did not involve natural selection at all: in this model, selectively neutral mutations arose and their frequencies simply fluctuated at random, as is inevitable in a finite population. Tropical forests pose a similar puzzle for ecologists. The enormous numbers of species seem to challenge our classical notions that coexistence requires that each species has its own unique niche: how can so many species construct unique niches from such a small number of requirements sun, water, a patch of ground? The same puzzle in a different context became known as the 'paradox of the plankton' (Hutchinson 1961). As in molecular evolution, so too in ecology a neutral theory has been proposed which may resolve our puzzle (Bell 2001; Hubbell 2001) and is attracting much attention. The theory's advocates have more in mind than tropical trees, but I will restrict myself to trees to make the discussion specific. This was the context in which the theory was first proposed (Hubbell 1979) and one of the main biological areas where we have a puzzle apparently requiring a radical solution. Hereafter I will refer to the neutral theory of biodiversity as NTB. The underlying stochastic theory is the same in the two areas, although it has been studied for much longer in population genetics, as neutral advocates point out (Hubbell 2001; Volkov et al. 2003). This means we can import many results from the population genetics literature, simply interpreting the biology in a different way. In this paper I will import an important result concerning the time-scale of the neutral process, which raises serious difficulties for the NTB as an explanation of tropical forest diversity. This difficulty has been noted before (Leigh 1999; Lande et al. 2003), but does not appear to be as widely known as it should. Nonetheless, as I will discuss, neutral theory can still provide useful null models for the interpretation of data. In the fir t section, I gather together some basic theoretical results that are common to both population genetics and NTB. This is for two reasons. First of all, it serves to emphasize that we really are talking about the same theory, giving us the confidence to bring into the body of NTB an important result from population genetics concerning the time-scale of the neutral process. This will be done in the third section. Secondly, people may find it useful to have these basic results gathered toget er in one place they are currently somewhat scattered about. Having emphasized the underlying identity of the theory, I then briefly explore the implications of the biological differences between the worlds of molecular evolution and biodiversity for the likely future development of the theory in the biodiversity context.

The Evolution of Gene Regulation Underlies a Morphological Difference between Two Drosophila Sister Species

Evolutionary developmental biology (evo–devo) is that part of biology concerned with how changes in embryonic development during single generations relate to the evolutionary changes that occur between generations. Charles Darwin argued for the importance of development (embryology) in understanding evolution. After the discovery in 1900 of Mendel’s research on genetics, however, any relationship between development and evolution was either regarded as unimportant for understanding the process(es) of evolution or as a black box into which it was hard to see. Research over the past two decades has opened that black box, revealing how studies in evo–devo highlight the mechanisms that link genes (the genotype) with structures (the phenotype). This is vitally important because genes do not make structures. Developmental processes make structures using road maps provided by genes, but using many other signals as well—physical forces such as mechanical stimulation, temperature of the environment, and interaction with chemical products produced by other species—often species in entirely different kingdoms as in interactions between bacteria and squid or between leaves and larvae (Greene Science 243:643–666, 1989). Not only do genes not make structures (the phenotype), but new properties and mechanisms emerge during embryonic development: genes are regulated differentially in different cells and places; aggregations of similar cells provide the cellular resources (modules) from which tissues and organs arise; modules and populations of differently differentiated cells interact to set development along particular tracks; and organisms interact with their environment and create their niche in that environment. Such interactions are often termed “epigenetic,” meaning that they direct gene activity using mechanisms that are not encoded in the DNA of the genes. This paper reviews the origins of evo–devo, how the field has changed over the past 30 years, evaluates the recognition of the importance for development and evolution of mechanisms that are not encoded in DNA, and evaluates what the future might bring for evo–devo. Although impossible to know, history tells us that we might expect more of the same; expansion of evo–devo into other areas of biology (ecology, physiology, behavior); absorption of evo–devo by evolution or a unification of biology in which evo–devo plays a major role.

One of the central issues in evolutionary developmental biology is how we can formulate the relationships between evolutionary and developmental processes. Two major models have been proposed: the 'funnel-like' model, in which the earliest embryo shows the most conserved morphological pattern, followed by diversifying later stages, and the 'hourglass' model, in which constraints are imposed to conserve organogenesis stages, which is called the phylotypic period. Here we perform a quantitative comparative transcriptome analysis of several model vertebrate embryos and show that the pharyngula stage is most conserved, whereas earlier and later stages are rather divergent. These results allow us to predict approximate developmental timetables between different species, and indicate that pharyngula embryos have the most conserved gene expression profiles, which may be the source of the basic body plan of vertebrates.

Developmental Plasticity and Evolution . by Mary Jane West-Eberhard. Oxford University Press, New York, 2003. 814 pp. $100, £79.50. ISBN 0-19-512234-8. Paper, $49.95, £31.99. ISBN 0-19-512235-6. Discussing a wide range of ideas and empirical studies that support them, West-Eberhard focuses on showing what recent findings in development, physiology, and behavior suggest about the origins of evolutionary novelties.

The main topic of the paper is a discussion of the ways through which the theory of evolution remakes itself, changes and grows, keeping alive and reinforcing its Darwinian explanatory core. The theory shows a 150 years old history of theoretical and empirical extensions and revisions, without any apparent radical change of “paradigm” and without a rival Research Programme able to replace it. The ongoing transition from the Modern Synthesis (MS) to a so-called “Extended Evolutionary Synthesis” (ES) is here interpreted through the Methodology of Scientific Research Programmes, proposed by the epistemologist Imre Lakatos and updated. The current situation in evolutionary biology could be represented by a “progressive” shift of the Darwinian research programme, moving from the quite rigid theoretical framework of the standard version of Modern Synthesis (gradualism, extrapolationism, adaptationism) to the more inclusive and pluralistic “core” and “protective belt” of the Extended Synthesis. Promising and advanced researches – like those concerning evolutionary developmental biology (Evo-Devo), epigenetics, multiple ways of speciation and the role of structural internal constraints – find in this perspective a realistic interpretation as theoretical and empirical novelties with huge implications, nevertheless not incoherent with an extended Neo-Darwinian explanatory core. A Neo-Lakatosian approach seems useful when we discuss the extension of evolutionary models in non biological fields, avoiding the application of just metaphorical forms of “ultra-Darwinism.” This analysis in terms of a rational and continuous dynamics of growth of biological thought seems much needed also for a critical examination of some popular and radicalized controversies about the health of a no better defined “Darwinism or Neo-Darwinism.”

The evolutionary dynamics of quantitative traits in a species depends on demographic structure (e.g. age‐ and size‐dependent vital rates, migration across subpopulations, mate preferences, and the presence or absence of overlapping generations), which in turn can depend on interspecific interactions with other evolving species. Furthermore, biologists increasingly appreciate that evolutionary change in a single species can substantively affect broader ecological processes, such as community assembly and ecosystem functioning. Models integrating insights from classical population and quantitative genetics, on the one hand, and community ecology theory, on the other, are therefore critical to elucidating the interplay between ecological and evolutionary processes. However, few modelling approaches integrate ecological and genetic details into a comprehensive framework. Such models are needed to account for the reciprocal feedback between evolutionary change and ecological dynamics, and develop effective responses to anthropogenic disturbances on natural communities, improve agricultural practices and manage global health risks such as emerging pathogens. Here we introduce an open‐source, multi‐species forward‐time population genetics simulator and framework for rapidly simulating eco‐evolutionary dynamics. Our approach permits building models that can account for alternative genetic architectures, non‐standard population dynamics and demographic structures, including those driven by interspecific interactions with other evolving species and the spatial dynamics of metacommunities. By integrating these processes into a common framework, we aim to facilitate the development of models that can further elucidate eco‐evolutionary dynamics. While multi‐species, forward‐time population genetic models can be computationally expensive, we show how our framework leverages relatively economical graphics cards installed in modern desktops. We illustrate the versatility and general applicability of our approach for two very different case studies: antagonistic coevolution in space, and the evolution of life‐history traits in response to resource dynamics in physiologically structured populations. We find that our analyses can run up to c. 200 times faster on a single commodity graphics card than on a single CPU core, comparable to the performance gains on small‐to‐medium sized computer clusters. Our approach therefore substantively reduces implementation difficulties to integrating ecological and evolutionary theory.

The Evolution of Development Thinking