Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
9 How Much Can We Tell About the Pastâand Predict About the Futureâ by Studying Life on Earth Today? Individual organisms are ephemeral, persisting over long time scales only in the form of lineages of ancestors and descendants. Most of the species that have ever lived are extinct, and yet a great deal can be learned about them by examining fossils and by studying the genomes of living descendants. From an individual organismâs genome much can be learned about its parents; at the same time, only the information that passes to the offspring will be available to future generations. Thus, the collective genetic content of all the organisms on Earth represents a treasure trove of historical data and at the same time is the result of a strong winnowing process. Not everything is transmitted to future generations. In addition, extinction has removed countless genetic combinations that were adapted to the environments and communities in which they arose. Thinking about the collective genetic reservoir in this wayâas a record of the past and as the starting point for future evolutionâallows one to ask some intriguing questions. Would it ever really be possible to build a âJurassic Park,â to re-create an ecosystem from the past? Could scientists ever predict what life on Earth will look like 1,000âor 1 millionâyears from now? Future life must evolve from the life now present, and thus, understanding the information about the past that is embodied in current organisms and ecosystems, and understanding how organisms pass that information to future generations, is a fundamental biological question. One of the great triumphs of 20th-century biology was to work out mechanisms of the genetic inheritance system, through which information from evolutionarily successful ancestors, recorded in DNA, is passed on to subsequent generations. The recognition that DNA transmits information 145
146 THE ROLE OF THEORY IN ADVANCING 21ST-CENTURY BIOLOGY across generations and the development of techniques to determine DNA sequence have allowed theory and data to combine elegantly in phylogenetic analyses to describe the evolution of organisms and their component parts, including metabolic, sensory, and developmental pathways, by comparing the DNA sequences of the relevant genes. While much is known about how genetic information is gained and lost through mutation, recombina- tion, conversion, duplication, translocation, selection, and other processes that alter genetic material in individuals and populations, much remains to be learned about the expression and regulation of genome activity that depends on inherited genetic information. The promise of classical genetic theory was the theoretical ability to predict the form and capabilities of an organism by knowing the DNA sequence of its protein-coding genes. A comprehensive understanding of the regulation and interaction of these protein products would explain the process of development, allow predic- tion of the connection of genotype to phenotype (including, for example, the linking of genetic variation to disease susceptibility), and serve as the palette upon which natural selection could act. Research based on this theoretical framework has indeed contributed to the success of biological research in the last few decades and enabled the development of a vibrant biotechnology industry. At a number of levels, observational and experimental data are accu- mulating that suggest that this enormously successful classical framework is ripe for further expansion. This chapter discusses some of the ways in which it is becoming clear that the characteristics of offspring cannot be fully e Â xplained by the genes acquired from their parents. First, an understanding of the roles of noncoding DNA, which makes up the bulk of the genomes of many higher organismsâ genomes, will be required to link genotype to phenotype (see Chapter 3). Also, a number of mechanisms other than DNA sequenceâcollectively designated epigenetic mechanismsâare being shown to represent additional means to pass information from cell to daughter cell, from parent to offspring. Looking beyond the inheritance mechanisms that act within species, increased exploration of the microbial world has profound implications for our understanding of how adaptive mechanisms can be inherited and shared. As introduced in BoxÂ 3-2, Â microbes live in complex multispecies communities where genes can be shared between distantly related organisms. Thus, genetic adaptations can spread across evolutionary lineages. Furthermore, many if not all Âeukaryotic organisms live in intimate association with microbial communities that provide a num- ber of functions from nutrition to host defense, functions that are appar- ently coordinated over evolutionary time scales with the functions encoded by the host organismâs genome. Finally, behavioral, social, and symbolic structures (such as human language) have the potential to be carried from one generation to the next. These characteristics do not exist independently
HOW MUCH CAN WE TELL ABOUT THE PAST? 147 of each organismâs genetic and environmental context, so full understand- ing of inheritance will require elucidating the complex interactions among all of these potential mechanisms of transmission of characteristics across generations. TRACES OF EVOLUTIONARY HISTORY The genome of every organism carries many remembrances of events long past, because almost every characteristic of an organism has evolved not ex nihil but instead by modification of preexisting characteristics. F Â ranÃ§ois Jacob (1977), who shared a Nobel Prize for elucidating the mecha- nism of gene regulation, referred to natural selection as a âtinkerer,â rather than a designer, for selection can act only on those mutations and genetic combinations that happen to arise in a population. These need not be the best possible solutions and may well be different in different populations or species, so the construction of adaptations by selection may proceed along different paths, and not to the best possible end. Historical contin- gency, Âunder which the long-term trajectory of change depends on initial geneticâor environmentalâconditions, undoubtedly plays a dispropor- tionately greater role in biology than in other natural sciences. Phylogenetic history has long been recognized as the explanation for otherwise inexplicable morphological features. For example, the human respiratory pathway, from nasal passages through the trachea to the lungs, crosses the digestive pathway in the pharynx. This is explicable not by any functional advantageâindeed, this is why humans may choke while eatingâbut by the evolution of lungs from the air bladder of fishes that did not originally inhale air through nostrils. The brilliant red floral dis- play of poinsettias (Euphorbia pucherrima) that grace many households at C Â hristmas-time is not a red-petaled flower, but instead a circlet of leaves, identical in structure to the normal green leaves below, that surround a cluster of small petal-less flowers. Petals were lost in the ancestor of the entire tribe Euphorbieae, and so the evolution of a red display to attract pollinating birds resulted from selection on available genetic variation, in this case in leaf pigmentation. Historical contingency applies equally in molecular biology. The very term âgenetic codeâ suggests that the correspondence between codons and amino acids is a consequence of early evolutionary history, not of optimal- ity of function. Evolutionarily new functions are performed by proteins that have been modified from ancestral proteins with different functions, and in some cases by proteins that perform a new function without any modification at all. The likely role of historical contingency in this process is dramatically illustrated by the crystalline proteins that compose the eye lenses of diverse animals (Piatigorsky, 2007). In all known vertebrates, the
148 THE ROLE OF THEORY IN ADVANCING 21ST-CENTURY BIOLOGY Î±B-crystallin is not just related to a small heat shock protein (shsp); it is shsp, encoded by the same gene (an example of âgene sharingâ). Diverse other crystallins, that differ among vertebrate groups, serve enzymatic func- tions in other tissues: Î±-enolase in turtles and lampreys; glyceraldehyde- 3-phosphate dehydrogenase in geckos; lactate dehydrogenase in ducks, crocodiles, and hummingbirds; arginosuccinate lyase in reptiles and birds, and NADPH:quinone oxidoreductase in camels and guinea pigs are all crystallin lens proteins and enzymes simultaneously. So far, no convinc- ing functional studies have explained why particular enzymes should have been recruited to serve as lens proteins in these different lineages. What is clear, though, is that any of a great many proteins can be used as crystal- lins and that these proteins have been recruited from preexisting metabolic pathways, rather than evolving de novo. Much of the study of molecular evolution consists of determining the history of origin of functionally new proteins by gene sharing, by duplication and divergence of ancestral genes, and by evolutionary assembly of chimeric proteins from ancestral modules (Graur and Li, 2000). Historical contingency pervades all levels of biological organization. Whether or not the species structure of ecological assemblages may be said to contain a record of past information, it certainly has been profoundly affected by past evolutionary and environmental events. Blood-feeding bats inhabit the American tropics but not Africa, despite the abundance of mammalian prey; marine snakes have evolved in the Indo-Pacific but not the Atlantic Ocean; the ecology of tropical American rain forests is greatly influenced by the abundant water held high above ground in the leaf axils of epiphytic bromeliads, but no comparable plants have evolved in the Old World tropics. The shell-drilling habit (as in modern oyster drills) evolved in a Triassic gastropod lineage but was lost when this lineage became extinct in the end-Triassic mass extinction and did not originate again for another 120 million years (FÃ¼rsich and Jablonski, 1984). Extinction has left a major imprint on contemporary life. Because the Alps prevented dispersal of many species to low latitudes during Pleisto- cene glaciations, plant diversity in Europe is diminished compared to other northern land areas (Latham and Ricklefs, 1993). Echoing a long history of thought in evolutionary biology, Stephen Jay Gould (1989) argued that the human species would not exist if any of a great many environmental and evolutionary events had been different in the last 500 million years. At all levels, from the molecular to the ecological, a major research challenge is to devise theory and statistical methods that might distinguish the relative roles of historical contingency and optimal functionâof chance and necessity, as Jacques Monod put it. As methods of genetic manipulation develop, heretofore impossible experiments on function will become routine (how do guinea pigs function with a turtleâs lens crystallin?), and historical
HOW MUCH CAN WE TELL ABOUT THE PAST? 149 narrative (to which historians of human events are largely limited) may be supplemented with scientifically testable hypotheses to a greater extent than is now possible. EPIGENETIC MEMORY WITHIN AND BETWEEN GENERATIONS A phenomenon that complicates our ability to predict phenotype from genotype is epigenetics. Epigenetics can help explain, for example, how genetically identical organisms can have phenotypic differences. Epigenetic developmental states record cellular âmemoriesâ of the developmental state of ancestor cells (Jablonka and Lamb, 1995; Turner, 2001). Once cells differentiate, it is often important that their state of differentiationâfor example, into bone, muscle, or nerve cell typesâbe maintained through the rest of development and adulthood. A record of the developmental state of the differentiated parent cell is associated with storing and passing through mitosis a set of epigenetic marks on the DNA and proteins within chromosomes (such as methylation marks on DNA and post-translational modifications of histone molecules). These and other epigenetic memories are ârecalledâ and interpreted in the offspring cell through gene expression, which is regulated by the epigenetic developmental states within the cell. Although development is highly reproducible and usually stable and unidirectional, other epigenetic states are established in a stochastic manner and are plastic, resulting in significant variability between genetically identi- cal individuals. Examples include phenotypic diversity displayed by mono- zygotic twins, stochastic epigenetic silencing of transposable elements that influence adjacent gene expression in plants and animals, position effect variegation in Drosophila (silencing of genes placed next to heterochroma- tin through translocations), and X-chromosome inactivation in mammals. The environment can modulate the establishment and maintenance of par- ticular epigenetic states. One classic example in plants is the flowering response to cold, known as vernalization. In certain species, germination of seeds in spring requires an exposure to cold during winter. Recent work in Arabidopsis has revealed that the cold signal is recorded and remembered through chromatin-level control of key flowering regulatory genes (Sung and Amasino, 2005). The âmemoryâ of winter is used to ensure growth and flowering in the spring and summer. Epigenetic mechanisms can function during development and during the lifetime of the organism but can also be passed on to offspring, resulting in a nonclassical means of inheritance. There are several examples in ani- mals in which a motherâs diet can influence gene expression in the offspring. The best characterized is the Agouti coat color phenotype in the mouse. When pregnant dams are fed methyl-supplemented diets or phytoestrogens, transcription of the Agouti locus is suppressed and this is associated with
150 THE ROLE OF THEORY IN ADVANCING 21ST-CENTURY BIOLOGY increased DNA methylation in a regulatory sequence upstream of the gene (Cooney et al., 2002). Epigenetic inheritance systems that can record and transmit cellular states through meiosis also exist. Some of the best characterized examples come from studies in plants, in which a number of phenomena that involve trans-sensing mechanisms and meiotic heritability of altered epigenetic states have been reported and characterized. These include paramutation, transposon, and transgene silencing (Chandler and Stam, 2004; Matzke and Birchler, 2005). Paramutation is an allele-dependent transfer of epigenetic information, which results in the heritable silencing of one allele by another. A major difference between paramutation and the heritable transmission of silencing associated with transgenes and transposons is that the newly silenced allele is capable of silencing another active allele in subsequent generations (Chandler, 2007). Although the phenomena are best studied in plants, epidemiological studies in humans (Bennett et al., 1997) and recent work in mice (Rassoul- zadegan et al., 2006) suggest similar phenomena occur in mammals. The more frequent observation of meiotically heritable epigenetic states in plants versus animals might be a reflection of developmental differences. Plants do not set aside a germ cell lineage early in development. Instead, cells that will produce gametes differentiate late in development from somatic cells. Thus, mitotically heritable epigenetic states accumulating in plant somatic cells are often transmitted to progeny. Genomic imprintingâallele-specific gene expression depending on whether an allele is inherited from the father or motherâoccurs in both plants and animals. In mice the mechanism in- volves establishment of methylation marks within specific DNA sequences in the parent (there are distinct maternal and paternal marks) that are retained through embryogenesis when most genome-wide methylation is reset (Wood and Oakey, 2006). In the Arabidopsis plant, both alleles are methylated and the maternal allele is demethylated early in embryogenesis via a specific DNA glycosylase (Choi et al., 2002). Several potential roles for and consequences of the transfer of epi- genetic information to progeny can be envisioned. As there are a number of examples in which the environment can modulate the expression state, transferring that state to progeny could be adaptive. To be adaptive, these states would have to be highly heritable, which has been shown for several examples in plants (Melquist et al., 1999; Soppe et al., 2000), including one from a natural population (Cubas et al., 1999). Although there are many examples of highly heritable states, they are potentially reversible at frequencies higher than DNA sequence changes and thus could provide mechanisms for exploring optimum states, which might be later fixed by slower genetic evolutionary processes. High rates of change, responsiveness to environmental inputs, heritability in nongenetic inheritance systems, and
HOW MUCH CAN WE TELL ABOUT THE PAST? 151 their modes of interaction with genetic inheritance call for the expansion of genetic theory to understand the developmental, genetic, ecological, and evolutionary dynamics of living systems in this expanded context. Allele-specific interactions such as paramutation could also contribute to generating functional homozygosity in polyploids and might have evolved from defense mechanisms targeting viruses and other invasive genomes as some mechanistic details, such as a role for RNA-mediated chromatin changes, are shared. Paramutation-like phenomena could contribute to the low penetrance and other aspects of non-Mendelian inheritance frequently observed for genes involved in complex human diseases and the segregation of quantitative characters in other organisms. THE CHALLENGE OF HORIZONTAL GENE TRANSFER AND SYMBIOSIS Contemporary phylogenetic inferenceâinferring the genealogy of spe- cies from records stored in morphology and moleculesâis built on the assumption that life is monophyletic, so that histories of particular groups and of all life are tree-like branching structures that can be traced back to a common ancestor. As introduced in Chapter 3, evidence of wholesale and continuing lateral gene transfer within and among the three major domains of life complicates phylogenetic inferences about the earliest stages of life on Earth (true bacteria, archaea, and eukarya; see Woese, 1998; Doolittle, 1999a, b; Felsenstein, 2004). The Tree of Life Web Project at the University of Arizona acknowledges the theoretical challenge, noting that âthe mono- phyly of Archaea is uncertain, and recent evidence for ancient lateral trans- fers of genes indicates that a highly complex model is needed to adequately represent the phylogenetic relationships among the major lineages of Lifeâ (http://www.tolweb.org). Some argue that new kinds of phylogenetic theory and approaches are needed (Woese, 1998). Woese draws inspiration from physics, developing an analogy between evolution with large amounts of lateral gene transfer and physical annealing. He concludes that âthe uni- versal phylogenetic tree, therefore, is not an organismal tree at its base but gradually becomes one as its peripheral branchings emerge. The universal ancestor is not a discrete entity. It is, rather, a diverse community of cells that survives and evolves as a biological unit [made cohesive by extensive lateral gene transfer]. This communal ancestor has a physical history but not a genealogical one.â Whether Woeseâs radical conclusions are correct or not, it appears that extensions of phylogenetic theory and possibly dif- ferent methods of analysis may be needed for earliest life than for later historical periods. Bacteria live in diverse communities where communication via small molecules and genetic exchange through several mechanisms, including lat-
152 THE ROLE OF THEORY IN ADVANCING 21ST-CENTURY BIOLOGY eral gene transfer and bacteriophage infection, are important to community behavior and survival. Until recently, the tools to study these complex inter- actions have been limited, and both data and theory needed to understand the rules of bacterial community interactions are inadequate. However, the realization that communities of single-celled organisms have mechanisms for storing information about the past has immediate relevance. Antibiotic resistance provides an example of a phenotype that can emerge in one bacterial species, be maintained in the absence of the antibiotic, and re- emerge and be shared with other bacterial species under selective pressure. Further development of theories explaining how genetic exchange within and among species provides the means to access a memory bank in different bacterial communities would have numerous applications. Are all the meta- bolic, defensive, and communication adaptations developed in any bacteria available to all others, making the global bacterial genetic pool a repository of all surviving past adaptations? Or are some genetic pathways, species, or communities isolated or limited in their capacity to share or be shared? It is also becoming increasingly clear that many, if not all, eukaryotes live in close association with more or less complex communities of bacteria and archaea. How these communities assemble, the degree to which their composition is inherited, and the roles they play in the fitness of their host are only beginning to be imagined, much less described. These phenomena also seem to suggest that memory can be created at a different levelâthat of the community of gene-exchanging units. Lateral gene transfer early in the history of life and throughout the history of the domain of true bacteria and archaea, as well as the prevalence of symbioses in eukarya, cloud the genealogical record of biochemical pathways. Even if there remain gene- trees, these phenomena of information exchange, distributed storage, and sharing complicate current methods of phylogeny reconstruction and raise the possibility that the extension of evolutionary theory will be needed to take these phenomena into account. ARE THERE INHERITANCE SYSTEMS NOT YET DISCOVERED? Many types of âmemoryâ phenomena involve the interplay of organ- ism and environmental states. Just as studies of paramutation and heritable epigenetic change are leading to a new appreciation of the complexity of inheritance and the variety of memory phenomena, attention to ways in which environmental changes induced by organisms feed back to behavior, development, selection, and inheritance of organism traits may lead to the discovery of new kinds of inheritance systems. If organism-environment interactions result in feedback loops and there are sufficient combinatorial states of both organism and environment, then there is the potential for novel forms of nongenetic inheritance. If organisms alter their environments
HOW MUCH CAN WE TELL ABOUT THE PAST? 153 via traits that are passed to offspring, and environments are correlated, then offspring states (behavioral preferences, morphologies, and internal states) acquired in development and learning will record information about the historical environment in a way that potentially exposes the organism- environment interaction itself to natural selection, even if the traits are not genetically determined (Oyama et al., 2001). For example, food preferences of parents can expose offspring to a par- ticular range of tastes, smells, and sights associated with food (BilkÃ³ et al., 1994; Jablonka and Lamb, 2005; see also Chapter 4). The preference for certain foods records the historical information that their rearing parent(s) preferred such foods and that such foods were present in the parentâs en- vironment (and the offspringâs). Historical records of this type span many levels of biological organization, from the molecular level where odiferous molecules are transmitted to offspring (through motherâs milk in lactating mammals; Galef and Henderson, 1972) to parent-offspring social interac- tions that bring offspring into contact with food, to organism-environment interactions including exogenous effects such as the persistence and decay of odors due to weather. In general, the coupling of parent-offspring trait correlations (heritabil- ity) with ancestor-descendant environment correlations through organismsâ alteration of their own selective environments produces a transgenerational feedback loop. In cases where the relevant traits are genetically determined, the evolution of such correlated structures is called âniche constructionâ (Odling-Smee et al., 2003). If traits are epigenetically (co)determined, how- ever, an expansion of current evolutionary theory will be required to fully characterize the dynamics. Inconsistencies between observation and theory might be resolved if the possibility that environmental conditions (or their interactions with organisms) can themselves be heritable is taken into consideration. For example, the explanation of sex ratios was traditionally analyzed as a problem of the evolution of sex-determining genes, but recent studies of environmental sex determination in vertebrates hint that an ex- panded theoretical approach that includes organism influences on environ- ments plus feedback to both genetic and epigenetic inheritance mechanisms may be required to fully answer this unsolved problem of evolutionary theory (see Box 9-1). Certainly such a possible feedback loop should be considered when studying the coupling of microbial community metabo- lism and nutrient cycles in the atmostphere and ocean, since environmental changes may affect which metabolic pathways are activated, which species predominate, and which genes are laterally transferred. Behavioral states carry information about what has been previously experienced or learned. It is not enough, however, to record information in order to serve the inheritance function. The information must be âusedâ and must play a role in the growth, development, or maintenance of the
154 THE ROLE OF THEORY IN ADVANCING 21ST-CENTURY BIOLOGY Box 9-1 Temperature-Dependent Environmental Sex Determination (TSD) and the Evolution of Sex Ratios Classical sex allocation theory predicts an equilibrium 1:1 ratio between the sexes, assuming equal energy allocation to offspring of each sex (Fisher, 1930). Past explanations of unequal allocation appealed to a variety of genetic factors (Hamilton, 1967; see also Freedberg and Wade, 2001). Historical scenarios as- sumed environmental sex determination (ESD) was primitive and genetic sex determination (GSD) is derived (Ohno, 1967; see Bull, 2004), but failed to predict that highly derived groups such as vertebrates might include a mix of taxa with ESD and GSD. The discovery of ESD in vertebrates, such as temperature-dependent sex determination (TSD) in some reptiles, challenges standard population genetics theory. ESD shows how environmental inputs to development can alter evolution- ary outcomes from classical expectations (Gilbert and Bolker, 2003). The cycling of information between gene regulatory states and sex ratios to behavior, environ- ment, and hormones and back again could result in âheritabilityâ of the behavior variations, such as that manifested in transgenerational correlations of nest-site choice and sex ratioâa form of ânon-Mendelianâ behavioral or cultural inheritance (Freedberg and Wade, 2001). Given the adaptive significance of sexual reproduction and strong selection against unequal sex ratios, theory predicts that mechanisms producing equal sex ratios should be conserved and those producing deviant ratios should be absent, yet a wide variety of ESDs are known in vertebrates (Janzen and Krenz, 2004). Bull (2004) recounts how the facts were long resisted, then accepted only as exotic exceptions, and only recently acknowledged as challenges to conventional theory. The greatest promise in solving this long-standing problem in evolutionary biology lies in integrating physiology and molecular biology with developmental biology, ecology, and evolution across many levels of organization (Valenzuela, 2004). Current approaches rely on indirect comparison of genes discovered in tractable mammal and fish systems. New approaches, such as RNA interference studies of reptilian embryos, might be needed (Bull, 2004). Progress could depend on new theory as well, as suggested by new models of sex-ratio evolution in rep- tiles based on alternative inheritance mechanisms, such as cultural inheritance of nest sites (Freedberg and Wade, 2001). Gilbert (2005) identifies several pathways for environmental signal trans- duction into genomic regulatory responses. The endocrine system is known to be such a transducer, for example, through temperature-sensitive expression of steroidogenic factor, Sf1 (homologous to Fushi tarazu Factor 1 in Drosophila), in TSD turtle species. Sf1 is temperature insensitive in a GSD turtle species ( Â Valenzuela et al., 2006). Since maternal behavior, which is also affected by hormones, is a factor in nest choice and egg laying in turtles (Bull et al., 1988), behavior could complete an inheritance cycle linking genes, hormones, behavior, and ecology (see figure below). Gene regulatory networks could affect (1) hormone-conditioned maternal behavior, which can control (2) exposure of eggs to environmental temperature in the nest, inducing (3) steroidogenic factors to regulate genes, so as to produce
HOW MUCH CAN WE TELL ABOUT THE PAST? 155 (4) a particular sex ratio in offspring. In producing female offspring, a given tem- perature would also set particular behavioral preferences that could in turn lead to more (or less) female offspring in the next generation. It is known that the brain plays a role in the sex determination pathway of some TSD reptiles, as the locus of transduction of temperature into hormonal Box 9-1 signal. Aromatase, which converts testosterone to estradiol, is differentially ex- pressed in the brain of the red-eared slider turtle, Trachemys scripta elegans, in a pattern that explains sex determination data in laboratory studies (Willingham et al., 2000). Thermal factors have been shown to act very early in developmentâfor example, Sf1 acts before the temperature-sensitive period in Chrysemys pictaâ and that might explain why traditional genetic theory failed to predict field observa- tions that temperature prior to the sensitive period can alter sex ratio (Valenzuela et al., 2006). If temperatures experienced early in development depend on ma- ternal behavior, then behavior together with environment would close the loop in the determination of offspring sex by hormone transduction of environmental signals regulating gene expression. This possible behavioral inheritance system is not only of academic interest because TSD turtle species could be indicators of global warming. Rising average temperatures may lead to female-biased sex ratios in these species and eventual extinction (Janzen, 1994). Although speculative, this kind of scenario points to the need to consider whole-organism behavior as well as developmental gene regulation if novel regu- latory inheritance systems comparable to the genetic inheritance system are to be discovered. Behavior might be the missing link in the formation of inheri- tance cycles from causal chains leading from environment to hormones to gene regulation to phenotype. If so, then there may be many more types of behav- ioral inheritance systems than just those based on sophisticated forms of social learning. Just as developmental biology has added considerably in recent years to evolutionary biology, some argue that evo-devo (evolutionary developmental biology) needs supplementation to consider ecological aspects of development, or âeco-devo-evoâ (Gilbert and Bolker, 2003). If behavioral feedback can create inheritance systems as suggested in the TSD scenario, then theorists should explore âetho-eco-devo-evo.â SOURCE: Courtesy of James R. Griesemer, copyright 2007.
156 THE ROLE OF THEORY IN ADVANCING 21ST-CENTURY BIOLOGY offspring that inherit the information (along with its carriers), and thereby expose inheritance system variants of a population to evolutionary pro- cesses such as natural selection and drift. If these nongenetic adaptations, epigenetic, behavioral, or symbolic variants are to be considered true inheritance systems, part of organismsâ evolutionary legacies, they must contribute to fitness differences. Major challenges to extending investigations of nongenetic inheritance to an evo- lutionary context include development of new experimental tools and meth- ods to distinguish genetic from nongenetic variation, methods of measuring fitness costs and benefits, and theory development to predict and explain evolutionary dynamics when more than one inheritance system is operat- ing. Dual inheritance theories designed to handle cultural inheritance (e.g., Boyd and Richerson, 1985) only begin to scratch the surface of the types of inheritance systems and transmission rules involved. CONCLUSION Extending the concept of inheritance to include biotic and social rela- tions implied by epigenetic mechanisms, social learning, symbolic com- munication through language, and interactions with environments raises questions about whether there might be a general theory of transgenera- tional âmemoryâ for living systems. That is, is there a theory of biological conditions and mechanisms that record system states with the potential for closed information loops and an ontogeny and evolution of informa- tion? Just as growing awareness that genomes are dynamic complicates the simple concept that inheritance flows from genes to phenotypes and back to genes in the next generation, discoveries of other information loops (from behavior to environmental modification and back to behavior in niche construction and from symbols to social change and back to symbols in cul- tural evolution) present theoretical challenges as formidable as those faced by Mendel. What sorts of regular structures and mechanisms in behavior might there be to suggest transmission ârulesâ? What social mechanisms govern the production, manipulation, and propagation of symbols that can be captured in theories of cultural evolution? It took half of the 19th century to move from the most elementary understanding of the hereditary consequences of cross-breeding to Mendelâs theory and most of the 20th century to link the implications of Mendelâs theory to an understanding of the molecular mechanisms of the genetic system of inheritance in popula- tion and evolutionary processes. Scientists are only beginning to explore the mechanisms and theoretical implications of other inheritance systems and how these might interact in an expanded evolutionary dynamic.