Is there any evolutionary reason for why some organisms (wrasse) change their sex in adult stage?

In San Francisco, I saw this fish:

and quoting the aquarium's page:

Part of the wrasse family, the California sheephead is a protogynous hermaphrodite. Simply put, all sheepheads are born as females, but eventually transform into males. Due to hormonal changes triggered by environmental and social cues, this fish can go from a reproductively-functional female to a fully-functional male.

I have never heard of something like this for any organism. Is this behavior unique?

But the real biology question is why they do that and not simply retain one gender throughout their life? In other words, what's the evolutionary advantage of such a weird approach?

This is an interesting question. This strategy is common among wrasses, but the Labridae is a rather huge family. I found this article here that I think may help answer your questions:

Evolutionary foundations for cancer biology

New applications of evolutionary biology are transforming our understanding of cancer. The articles in this special issue provide many specific examples, such as microorganisms inducing cancers, the significance of within-tumor heterogeneity, and the possibility that lower dose chemotherapy may sometimes promote longer survival. Underlying these specific advances is a large-scale transformation, as cancer research incorporates evolutionary methods into its toolkit, and asks new evolutionary questions about why we are vulnerable to cancer. Evolution explains why cancer exists at all, how neoplasms grow, why cancer is remarkably rare, and why it occurs despite powerful cancer suppression mechanisms. Cancer exists because of somatic selection mutations in somatic cells result in some dividing faster than others, in some cases generating neoplasms. Neoplasms grow, or do not, in complex cellular ecosystems. Cancer is relatively rare because of natural selection our genomes were derived disproportionally from individuals with effective mechanisms for suppressing cancer. Cancer occurs nonetheless for the same six evolutionary reasons that explain why we remain vulnerable to other diseases. These four principles—cancers evolve by somatic selection, neoplasms grow in complex ecosystems, natural selection has shaped powerful cancer defenses, and the limitations of those defenses have evolutionary explanations—provide a foundation for understanding, preventing, and treating cancer.

The metaphysics of evolution

This paper briefly describes process metaphysics, and argues that it is better suited for describing life than the more standard thing, or substance, metaphysics. It then explores the implications of process metaphysics for conceptualizing evolution. After explaining what it is for an organism to be a process, the paper takes up the Hull/Ghiselin thesis of species as individuals and explores the conditions under which a species or lineage could constitute an individual process. It is argued that only sexual species satisfy these conditions, and that within sexual species the degree of organization varies. This, in turn, has important implications for species' evolvability. One important moral is that evolution will work differently in different biological domains.

1. Introduction: why metaphysics?

Metaphysics is the branch of philosophy that aspires to provide the most general description of reality. Metaphysics aims to say what exists, but at a more general and abstract level than that typical of practical science or, for that matter, everyday life. It may ask, for example, whether there is one kind of being, two (as Descartes believed), or many. It may ask about the relations between very broad categories of entities. Now almost all biologists believe that living beings are made of the same kind of material stuff as the non-living once, however, it was common to suppose that investigating life involved investigating something in certain respects, at least, quite different from the vulgarly material. This, one might say, is an example of progress in metaphysics.

The last example also illustrates that, though they may sincerely deny it, scientists are almost inevitably committed to metaphysical opinions, and that these make a difference to their work. A biologist not committed to the materialist metaphysics mentioned in the last paragraph would not look for the fundamental understanding of life in the properties of specific kinds of matter. Metaphysics can be ignored but not escaped. In the words of theoretical biologist and philosopher Joseph Henry Woodger, ‘physiologists [who] suppose themselves to be above “metaphysics” […] are only a very little above it—being up to the neck in it’ [1, p. 246].

The metaphysics I am interested in is a naturalistic one, providing an ontology, an account of what exists, ultimately grounded in our best science. I have said that scientists cannot avoid metaphysical assumptions, but these need not be explicit. Philosophical analysis of scientific work may help to expose these assumptions. But philosophical reflection on scientific findings may also point to an ontology different in important respects from that originally assumed. One reason it may do so is that philosophical enquiry of this sort is free to range over all domains of scientific enquiry. One motivation for ontological enquiry is thus to explore the consistency of interpretation of scientific results across the sciences or their subfields.

The metaphysical question with which this paper is concerned is an ancient one, the debate whether the world is ultimately composed of things, perhaps eternal and immutable things as was proposed by the Greek atomists, or rather is everywhere in flux, as famously advocated by the Greek philosopher Heraclitus. For process philosophers, enduring things, rather than being the more or less unchanging furniture of the world, are ‘never more than patterns of stability in a sea of process’ [2].

The ontology of things, following the revival of atomism in the seventeenth century, has been the dominant metaphysics for most of the history of modern science. It is closely connected to a further position that underlies conceptions of scientific explanation, mechanicism. For mechanicism, the way to understand or explain a phenomenon is to identify the various constituent things that interact to generate the phenomenon. Arrangements of constituents with particular functions constitute mechanisms. Mechanicism sees living systems as composed of things arranged in a hierarchy of mechanisms. This is a strictly bottom-up perspective, related to, if generally distinguished from, the often criticized but still widely endorsed methodological approach of reductionism. Process ontologists generally reject both mechanicism and reductionism, for they notice that what maintains the patterns of stability in the sea of process is not only the behaviour of the entities that compose the pattern, but also the network of relations between the patterns and their surroundings [3].

The aim of this paper is to explore the relevance of process ontology to evolutionary theory. Of course, no one doubts that evolution itself is a process. Thing (or substance, as a particular, very influential version of the concept is often referred to in philosophical writing [4]) ontologists do not deny that there are processes it is rather that they see processes as generally requiring things as their subjects, as what happens to things. For process ontology, evolution is also, of course, a process, but the organisms and the lineages that are the subject matter of evolution are themselves also processes. I shall try very briefly to justify these claims, and then examine some of the implications they involve for how we should think about evolution.

2. Process metaphysics

What is the difference between seeing some entity as a thing, on the one hand, or as a temporarily stable process, on the other? Consider two paradigm cases: a mountain and a storm. A mountain is naturally thought of as a fairly fixed part of the world's furnishings if any major change befalls it we are entitled to wonder why. In fact, in the context of a more generalized process ontology, on the timescale of tectonics the mountain is very much a stage in a process: for process ontology, being a thing is always relative to a timescale. The mountain will, nonetheless, serve well enough as an intuitive paradigm for a static thing, deriving its stability from inertia. Philosophers have asked how it is even possible for a thing to change and yet remain the same thing through time, and they have generally answered by positing some core of essential properties that must remain fixed regardless of the extent of change.

A storm may also be a very stable element of the world. The Red Spot on Jupiter, for instance, has been observed for several centuries, albeit with gradual changes in shape and size. But unlike the mountain, the Red Spot does not persist because nothing happens to change it, but because a stable pattern is maintained by the very rapid winds that circulate round it. If this activity ceased the Red Spot would dissipate very quickly. Its persistence through time is understood not through unchanging essential properties but through the causal continuity of the processes that maintain the pattern. A process ontology for life starts with the idea that the Red Spot is a more useful paradigm for living systems than the mountain.

Two simple points should be sufficient to confirm the appropriateness of the dynamic, processual perspective for thinking of biological systems. Consider, as paradigm living systems, organisms. The first, decisive reason for taking organisms to be processes is that they are open systems, far from thermodynamic equilibrium. It is an elementary fact of physics that maintaining such a system will require constant interaction with, and intake of matter or energy from, the environment. Its persistence is actively maintained rather than just given. Stasis, for an organism, is death.

Second, organisms undergo developmental cycles. Consider for instance the typical life cycle of an insect, comprising the egg, larva, pupa, and adult. These stages have very different properties. It is unclear what properties could possibly support the claim that these developmental stages were all one and the same thing. What could be an essential property of such a thing? It is sometimes suggested that genome sequence might provide such an essential, continuing property for an organism. I have responded to this idea in detail in [5], but perhaps a sufficient response is to note the work that the cell has to do to sustain a sufficiently accurate sequence [6]: genome sequence is as much the consequence of organismic stability as it is its source. For a process, at any rate, no such constant property is required: persistence is something the organism achieves, not some property or properties that it continues to possess. A process is inherently extended in time, and whatever claims temporal parts of a process have to be parts of one and the same process derive rather from causal connections between these parts.

Let me now mention two reasons why the insistence that living systems are processes rather than things matters. The first is that it motivates a significant shift in emphasis with respect to what stands in need of explanation. The traditional concern for thing-centred ontology is change. I do not expect an explanation of why my desk is very much as I left it when I was last in my office. For a process, on the other hand, persistence requires explanation. Physiology is largely concerned with understanding the multitude of internal processes that enable an organism to stay alive, to maintain its thermodynamic disequilibrium with its environment.

A clarification is needed at this point. When I refer to a process I shall, henceforth unless otherwise stated or obvious, mean an individual process, a process with the sort of coherence and persistence that might suggest treating it as a thing. Organisms, on my view, are paradigms of such coherent individual processes, though less controversially processual entities such as storms or rivers also have good claims to be individuals. Some processes—erosion, inflation, evolution—lack any such coherence. I shall not address philosophical doubts as to whether there even are individual processes that persist through time, though the discussion may give some indication of why such doubts arise and also of why they are misplaced.

The second reason why the processual status of organisms is important is that it places in the proper perspective the search for mechanistic explanation that is often alleged to be central to the contemporary life sciences [7,8]. I take a mechanistic explanation to be, very roughly, one that involves identifying a set of constituents of a phenomenon and showing how their actions and interactions combine to generate the phenomenon. There is no doubt that this has been an enormously productive scientific strategy. Nonetheless, from a process perspective the mechanisms postulated by such explanations must always be abstractions from the wider biological context, and this always poses potential limits on their application. First, the constituents of a biological mechanism are themselves dynamic and more or less transient entities. Mechanistic explanations will be successful only to the extent that the constituents identified are sufficiently stable on the timescale of the phenomenon under investigation. And second, biological processes are typically stabilized not just by the interactions of their parts, but also by interactions of the whole with its wider biological and abiotic context. These limitations do not imply that mechanistic explanations cannot be extremely illuminating they do show that their success should not be taken as a sufficient reason for inferring that the organism really is an interlocking system of mechanisms. It is not.

I should note that contemporary mechanicists, or ‘new mechanists’ as they are widely known, are a diverse group with views that diverge in many ways from the very rough summary just offered. Machamer, Darden and Carver [7] acknowledge the ontological importance of processes, but as part of a dualistic ontology very different from that advocated here. Craver & Bechtel [9] explicitly address the question of interlevel explanation, though denying that there is anything properly described as downward causation. Recent work by Bechtel qualifies this sceptical view on downward causation (e.g. [10]) and generally endorses many of the positions here associated with process ontology. Bechtel's status as a new mechanist, however, is a matter of debate (W Bechtel 2017, personal communication). Since this is not, at any rate, a paper about mechanicism, I shall not attempt to explore these divergences and subtleties further.

The organism should not be seen as a hierarchy of interconnected things, but rather as a hierarchy of processes at molecular, cellular, tissue, organ, etc., levels, operating at different interlocking timescales [11]. At each level the more or less stable entities—molecules, cells, organs—are stabilized both by their internal activities and by their interactions with their wider environments. The organism itself, of course, is not the terminus of this hierarchy, but just one further component. The stability of the organism also depends in part on its dynamic relation to its biotic and abiotic environment.

3. What evolves?

Organisms do not, of course, evolve. Evolution relates to the distribution of the properties of organisms over time. What organisms? It is commonly said that the relevant group of organisms should constitute a lineage, and sometimes that the relevant lineage is a species, which can even be made true by fiat as in G. G. Simpson's [12] definition: ‘a lineage (an ancestral descendent sequence of populations) evolving separately from others and with its own unitary evolutionary role’. Technically, it is better to talk of populations, as a species may consist of a number of isolated populations, hence evolving separately, but for present purposes it will do no harm to speak of species. A reason for doing so is that it will be useful to connect with the extensive philosophical literature on the nature of species, reminding ourselves thereby that it is a matter of great uncertainty what constitutes the appropriate kind of coherent lineage. It is popularly supposed, reflecting the lasting influence of Ernst Mayr, that species are interbreeding groups of organisms. But we need only note that the vast majority of species, and all species for the first 80% or so of the history of life, are asexual to see that this account is seriously limited. (Perhaps Mayr's rather dismissive attitude to microbes has helped to direct attention from this embarrassment to his so-called Biological Species Concept.)

A rather different issue has been widely debated by philosophers of biology, namely the question whether species are kinds or, rather, individuals. Philosophers have traditionally taken species terms as paradigmatic classificatory terms, and hence as referring to all the things that satisfy the conditions of membership of the relevant kind. But Michael Ghiselin [13] and David Hull [14] have persuaded the majority of the philosophical community that species are, on the contrary, individuals. Species, according to Ghiselin and Hull, and in accordance with the influential cladistic school of systematics, are properly understood as branches of the phylogenetic tree.

I believe the species as individuals view is partly correct, though with two very important provisos. First, a branch of the phylogenetic tree is a process not a thing. Apart from subsuming the obvious point that any part of the phylogenetic tree is temporally extended and constantly changing, recognition of its processual character immediately addresses some serious concerns that have been raised about the species as individuals thesis. An obvious such objection is that the alleged parts of a species are highly discontinuous. How are they identified as parts? Ruse [15], a prominent critic of the species as individuals thesis, notes that the important point might be integration rather than actual physical connection between the parts of an individual, but then complains that where the only connection between the parts of a supposed individual species is descent, descent begins to look suspiciously like an essential property that serves to define a class. Indeed exactly this view was subsequently defended by Griffiths [16] and others.

For a species-as-individual process view, however, there is no problem to address. A process is necessarily extended in time, and causal relations between temporal stages, or between spatial parts of temporal stages, are required to provide it with whatever integrity it has. Descent is just such a causal connection. A similar problem arises with regard to ambiguity of boundaries. Species have somewhat vague boundaries both synchronically (hybridization,) and temporally (speciation). Again, while this is difficult to align with standard metaphysical accounts of an individual, it is no problem at all for a process. No one expects a thunderstorm or a battle to have precisely delineated boundaries.

In fact, similar problems apply to organisms. Anyone who believes in superorganisms, for example ant colonies, that may include, as well as various castes of ant, domesticated fungi and several essential consortia of microbes, is happy with discontinuous organisms. And the spectrum of degrees of integration with symbionts, from mitochondria, widely thought of as parts of their hosts, through genomically-reduced obligate symbionts such as Wolbachia and Buchnera and obligate but horizontally acquired symbionts to, finally, purely ecological mutualisms, makes it difficult to define unambiguous boundaries to the organism. Processes are more or less well integrated, more or less clearly demarcated. As Hull notes, ‘Most organisms do exhibit more internal organization than most species, but this difference is one of degree, not kind. Most species do not exhibit the internal organization common in vertebrate organisms, but the same can be said for plants as organisms. Most plants do not exhibit the internal organization common in vertebrate organisms’. [17, p. 32].

My second proviso perhaps deviates more strongly from the spirit of the species as individuals thesis. It is that while it is sometimes useful and correct to treat species as individuals, they can also, equally correctly be treated as classificatory terms. In fact, as I shall argue, it may very well be that some species can only be treated in the second way. The point here is that classification is a vital part of any scientific project, and especially vital in a domain with the vast diversity of biology. As I have argued elsewhere in more detail [18,19], the importance of classification provides special desiderata for distinguishing species, and these should not be outweighed by sometimes transient theoretical considerations. In short, species can be units of evolution, units within which evolutionary change takes place and as such should be seen as individual processes but this cannot supplant their equal importance as units of classification (see also [20]).

4. Stabilization of species processes

If (some) species are individual processes, we should ask, as discussed above, what it is that maintains their coherence or integration over time. Note here that while not all processes need have either integration or individual status, to have the latter one must have the former. Geological erosion, for example, is a process with no integration there is no temptation to divide it into distinct individuals. But if Hull and Ghiselin are right, species must be stabilized processes.

A first, and very important part of the answer to what makes species stable, is natural selection. It has often been proposed that most selection is stabilizing selection, and the continued production over sometimes very long periods of time of very similar phenotypes is generally attributed not to the perfection of the reproductive process, but to the greater selective success of a particular phenotype. As Reiss [21] persuasively argues, much of the importance of natural selection is most illuminatingly understood under the rubric of the conditions of existence, a phrase used by Darwin, but more often associated with Georges Cuvier. It is no trivial matter for an organism to satisfy the conditions of existence, and if the areas of morphospace that make this possible are very limited, natural selection will maintain homogeneous species. Darwin also famously observed the production of organisms far beyond the numbers required to maintain a species. Though this is generally remarked as part of the story of adaptive evolutionary change, it is also important that the stability of the species requires overproduction to compensate for the production of inviable individuals and the random losses of pre-reproductive individuals. The latter, in many cases, will constitute the overwhelming proportion of cases. Overproduction, in short, is necessary not just for adaptive evolutionary change, but also for stable maintenance of the lineage.

Natural selection is not, of course, sufficient to stabilize a species over time. Just as an organism must constantly renew the cells of which it is composed, so a species, qua individual, must replace the organisms that are its parts. The Modern Synthesis has understood this process of reproduction as, at its core, replication, and this is a central point of criticism for advocates of an extended, or more radically replaced, understanding of evolution. By replication here I mean exact copying, as is generally understood to occur when a DNA sequence serves as a template for an identical sequence. (For discussion of the distinction between reproduction and replication, see [22].) The quasi-digital nature of this process grounds the claim that this is exact copying, and underlies Richard Dawkins' rather strange claim that genes are immortal [23, ch. 3]: the nucleotide sequence can, in principle, be precisely replicated in perpetuity. With more or less hedging, the Modern Synthesis has taken this to be the overwhelmingly important part of reproduction, more or less explicitly, thereby, assuming that the DNA sequence was sufficient to determine the phenotype.

There is much more to reproduction, however, than replication. Reproduction means, as the etymology suggests, producing again, and there is no reason in principle why the production of a new organism in a lineage should involve the replication of anything. As a matter of fact it appears that terrestrial reproduction always involves nucleic acid sequence replication but, as various contributors to this volume have demonstrated (e.g. Muller on development Stotz on parental effects Jablonka on non-genetic inheritance), it involves much else besides. Moreover as Noble [6] emphasizes, the nucleic acid sequences that are generally thought of as targets of replication are only maintained in a persistent state by elaborate editing and correcting processes in the cell, and thus may themselves be better described as being reproduced.

The stability of a lineage, finally, depends crucially on its relations with the external environment. But rather than this being, as has often been supposed, something achieved by the passive adaptation of the evolving lineage to the demands of the environment, the organisms in a typical lineage do a great deal to adapt the environment to their needs, so-called niche construction [24,25]. This may amount to full-scale engineering of the environment [26], as in the classic examples of beaver dam building or coral reef formation, but may also take more local forms, such as nest building and burrow digging. In fact all organisms have some effect on their environment, and therefore on the conditions of existence that they must satisfy.

Niche construction is often compared to Richard Dawkins' [27] concept of the extended phenotype. For Dawkins the beaver's dam or bird's nest is part of the (extended) phenotype of the beaver or bird, encoded in its genes and expressed as the animal creates the external structure. Niche construction theorists, however, emphasize the bi-directionality of the relation. The altered niche affects the behaviour and ultimately drives the evolution of the organism.

The difference in these perspectives nicely illustrates the difference between a thing- and a process-centred ontology. The extended phenotype concept extends the boundaries of the object (organism), but these boundaries are still fully determined by that object's internal, intrinsic properties, and the lineage is just the sum of these objects. Seeing the organism, or in this case the lineage, as a process, on the other hand, we should expect its limits to be maintained by activities at its boundaries, as a living membrane actively transports numerous molecules to maintain the chemical discontinuity it marks, or the surrounding flows maintain a whirlpool. This is just the difference the niche construction perspective signals from the extended phenotype.

If species are processes of this kind, then evolution is the change within such processes. Stabilization of a process is always limited, so some such change is to be expected, as has been extensively discussed in accounts of drift. Where does adaptive change come from? A trivial but sometimes obfuscated point is that it never comes from natural selection. Selection cannot occur unless some other process provides alternatives to select from. It follows that any thesis about the power of natural selection to generate change implicitly presupposes a thesis about a process or processes that generate selectable change. A distinctive thesis in the Modern Synthesis is that the overwhelmingly predominant source of selectable change is small random mutations, and consequently views about the power of natural selection have sometimes smuggled in assumptions about the ability of cumulative small mutations to generate almost arbitrary degrees of phenotypic change. Contributors to this special issue describe various other sources of variation, and indeed of adaptive variation, so questions about the efficacy of particular sources including random mutation should be seen as open. I shall turn very briefly to enumeration of some sources of adaptive variation towards the end of this paper.

5. Kinds of lineage and degrees of integration

More or less stable, coherent lineages are not necessary for evolutionary change. The first 2.5 billion years of solely unicellular life were apparently characterized by asexual reproduction and promiscuous lateral transfer of genes between sometimes distantly related individuals. It is hard to see why there should be any well-distinguished, species-like sub-processes within this evolving whole. To the extent that there are strong divisions between kinds, this is likely to be because natural selection favours and disfavours particular areas of morphospace. Put differently, the combinations of traits that satisfy the conditions of existence occupy discontinuous regions of trait space. (This does rather oversimplify the matter, as the conditions of existence depend on what other organisms concurrently exist. But this should not significantly affect the main point.)

Sexual reproduction introduces something quite new, internal integration of the lineage. Sex involves both horizontal and vertical connections between members of a species: horizontal between sexual partners and vertical between parents and offspring. Boundaries between species reflect not merely the contingencies of adaptation, but the fact that species have more or less effective means of policing their boundaries. The importance of this policing was particularly stressed by Paterson's [28] mate recognition species concept, defining species in terms of the ways that members were distinguished from non-members for reproductive purposes. Surely this overestimates the effectiveness of this boundary-preserving activity and underestimates the frequency of hybridization and, for that matter, its important role in speciation [29,30]. But as already noted, vague boundaries are no problem or surprise between processes.

I suggest that the invention or emergence of sex is also the emergence of species as individuals. Without sex there are no horizontal relations between the members of a species and they are connected only by their ancestry. But unless every individual, or at least every individual with a minimal novelty (e.g. a point mutation), is the ancestor of a new species there must be some horizontal connections that establish a group of individuals as an appropriate set of ancestors to found a species, and we appear to be launched on an infinite regress. If there existed species-like processes prior to sexual reproduction, these lacked any coherence or integration that could qualify them as processual individuals with persistence as such through time. This proposal also puts Mayr's familiar biological species concept in a slightly different light. Reproductive connections are indeed fundamental to the existence of species as individuals.

Sex is a minimal condition for a species to form as a coherent individual. In many, perhaps most, sexual species it provides all the coherence that there is. This is generally the case, at any rate, for those species that ecologists have described as r-selected, species, that is to say, that produce very large numbers of offspring of which a tiny fraction will survive. (The distinction between r- and K-selected species has been largely abandoned by ecologists, in recognition of the fact that there is a continuum of intermediate cases. Here I use the terminology only to indicate the extremes of this spectrum.) In such species there is minimal parental investment in offspring, and little opportunity for the emergence of culture or sociality. Frequently the contact between sexual partners is also minimal, sometimes in great danger of slipping into the relation of predator and prey. (I shall return shortly to those great niche constructors, the social insects.)

It is true that fairly r-selected species may well affect their niches, and may do so in ways that are advantageous to themselves. An excellent example are the earthworms studied in great detail by Charles Darwin [31]. The typical earthworm is, in many ways, more adapted to an aquatic than to a terrestrial life. But by its manipulation of the soil, notably the constant introduction of decaying organic matter, it keeps the soil wet enough to meet its adaptive requirements. It is unclear whether this is properly seen as a species-maintaining activity. There are many species of earthworm, so there is no species-specific benefit to their alterations of the environment. It is an interesting speculation that such processes of niche construction can create partially coherent supra-specific lineages at a much higher level than the reproductively connected lineage. But I shall not pursue that thought here. It seems likely that the kind of local and focused niche construction exemplified by beavers or nest-building birds is not found except where there is major parental investment in offspring, though I certainly do not rule out the possibility that more broadly directed kinds of niche construction may make important contributions to species coherence.

With K-selection, the strategy of producing much smaller numbers of offspring and investing heavily in their development, new forms of integration become possible. While some extragenetic maternal effects, mediated by molecules transferred to the oocyte, are possible even for strongly r-selected species, substantial periods of child-rearing allow far greater possibilities for parental, most commonly maternal, influence on the developing phenotype. The widely recognized phenomenon of phenotypic plasticity [32] provides ample opportunities for the mother to divert the offspring's development into directions that are adaptive in the context of perceived environmental conditions. Wolf and Wade [33] define maternal effects as a causal connection between some aspect of the mother's genotype or phenotype, and the phenotype of the offspring. Clearly the extended period of parental care in many vertebrate species provides many opportunities for such causal connections, and processes that allow parents to direct development in adaptive directions will be strongly selected. It seems likely a priori that such opportunities would be exploited, and the evidence supports this expectation [34].

One such process is epigenetic modification of the offspring's genome. Some kind of epigenetic system seems inevitable for a multicellular organism with highly differentiated cell lineages. The existence of such a system, in turn, provides a set of levers by which the parent (or any other aspect of the developmental environment) can influence the developmental trajectory of the organism. It again seems a priori plausible that parents would come to exploit these levers in adjusting the development of their young to changeable environmental conditions. And again this appears to have happened. A classic instance is the study of maternal care and its effect on the behavioural dispositions of rat pups by Meaney and colleagues [35,36].

Parental care provides opportunities for highly targeted niche construction, targeted, specifically, on the immediate environment of the offspring. Birds' nests provide a paradigm of this sort of activity, but social insect colonies remind us that this kind of niche construction is not necessarily tied to the kind of intergenerational relations found in vertebrates. This is becoming a familiar aspect of current evolutionary thinking [24,25] though the profound significance of replacing a picture in which the evolving lineage reacts passively to the environment, with one in which the lineage simultaneously shapes the environment to which it adapts, is not always sufficiently appreciated.

Parental care also provides unparalleled opportunities for enculturation, and hence for the evolution more generally of culturally transmitted behaviour. Such behaviour may also have physiological effects, for example mediated by epigenetic modifications. There is no reason in principle why a strongly r-selected species might not develop some kind of culture, and for all I know there may be examples of this. Nonetheless, it seems unlikely that there could be any very complex culture in the absence of the systematic collocation provided by parental care. Culture, in any case, provides a new channel for both horizontal and vertical transmission and evolution of behavioural traits.

A further crucial feature that adds a new dimension of integration to many K-selected lineages is sociality, the development of various more or less cooperative relations between individuals beyond parents and offspring. Though this is a complex and controversial subject, the existence of sociality is an indisputable empirical fact. It is widely though not universally believed that sociality creates supra-organismic level entities that can be selected [37].

In most social species it is assumed that social groups are disjoint: every individual is a member of at most one social group and it could be argued that in that case sociality does not add to the integration of the species, but only adds an intervening level of organization between the organism and the species. This is patently not the case, however, for humans. Typical humans are involved in numerous social groups, more or less cooperative and more or less significant to the course of their lives. Humans belong simultaneously to families, organizations, companies, clubs, churches, political parties, etc., and thus the species is connected by a mass of criss-crossing and overlapping links and supra-organismic level entities.

This kind of social integration may be unique to humans, indeed even to modern humans in complex civilizations. It is perhaps part of the reason why some (e.g. [38]) have thought that humans exemplify the very special kind of sociality known as eusociality. The paradigms for eusociality are the social insects, numerous species of Hymenoptera (ants, bees and wasps) and Isoptera (termites). It is also said to be found in two mammalian species (of mole rats) a few other insect species, and a few crustaceans. The most distinctive feature of eusociality is the division of reproductive from non-reproductive labour, with specialist reproducers and communal care of the young by non-reproducers. There is often much further division of labour into so-called castes. Such systems provide a highly effective context for shaping the development of the young in various behaviourally modulated ways. While humans certainly do not have a distinct reproductive caste, they do have a more elaborate division of labour by far than any other species. So although eusocial species have the most clear cut supra-organismic level of organization, it is equally clearly a disjoint division into social wholes. Humans may be unique in having a species-wide network of cooperative and group-forming relations, and may therefore reasonably be claimed to be the most fully integrated species we know.

A central aspect of the move from a mechanistic thing ontology to a process ontology is that the commitment to strictly bottom up causal influences, from parts to wholes, is replaced with a recognition that whole systems can contribute to determining the properties of their parts. It is, therefore, likely that the emergence of the species as an integrated individual will affect the behaviour of organisms, its parts. The most obvious relevant examples come from niche construction, and the most obvious specific case is that of Homo sapiens. Modern humans live in a constructed niche that is necessary for a large proportion of the behaviour they undertake, and acquire the capacities they have in a constructed developmental niche including hospitals, schools, and a great deal else besides. It is not, of course, the species as a whole that produces these resources, but they are made possible by numerous distributed parts of the species, generating a remarkable degree of effective cooperation.

In sum, although any lineage may be said to be a process of a sort, the degree of integration of these processes is very varied. And hence the degree to which these processes may count as persistent individuals, or continuants, is very varied. Pace Hull and Ghiselin, not all species are individuals. It seems plausible that the kind of process that constitutes a particular lineage may have important implications for the evolutionary processes that it is liable to undergo.

6. Implications

Evolutionary change requires sources of novelty. Although the debate over the current status of the Modern Synthesis is often presented as a debate about the importance of natural selection, this is misleading. As I have noted, natural selection cannot create anything. When theorists applaud the power of natural selection, what they are really doing is remarking on the poverty of the sources of change with which selection has to work, these being restricted to small random changes in the genome. In the debate over the adequacy of the Modern Synthesis, questions arise whether certain kinds of change happen or not, notably changes with some inherent tendency to be adaptive (Lamarckianism) and also whether kinds of changes that are acknowledged to happen are available to evolution by natural selection. The latter question tends to revolve around the adequacy of the modes of inheritance that are supposed to embed the relevant changes in a lineage.

Numerous sources of evolutionary novelty have been proposed (here I do not mean by ‘novelty’ any particular exceptional degree of novelty). The Modern Synthesis typically restricts these to genetic changes, notably mutation and recombination, but in principle also lateral acquisition of genetic material, though often this last is argued to be of relatively small importance. (Even very occasional lateral acquisition could be disproportionately important, however, as it might come, as is familiar in bacteria, with pre-packaged functionality. The vast numbers of viruses and similar entities in the biosphere provide a plausible means for such acquisition.)

In the microbial world, where the processes I have been discussing that account for the emergence of species, or lineages, as individuals do not occur, it is plausible that pretty much the standard Modern Synthesis model of genetic change and selection is sufficient to account for evolutionary change. As microbial evolution is all there was for 80% of the history of life, this is no minor concession. It is again important, however, to note the potential significance of lateral acquisition of genetic material. Microbes evolved in a context in which a far wider pool of genetic resources was potentially available than merely those in their own lineage, narrowly conceived. On the other hand, the price paid for this, one might say, was the impossibility of establishing higher-level entities, integrated lineages.

The emergence of sex in eukaryotes, at least 1.2 billion years ago [39], made possible the appearance of species as persisting individuals. Rescher [2] remarks, ‘For process philosophy, what a thing is consists in what it does’, so if sexually integrated species are indeed individual processes, we might wonder whether there is anything they do, beyond just persisting through time. The answer to this question might even offer a fresh perspective on the long debated question of why sex evolved at all.

The immediate answer to the question what species (or strictly, as noted earlier, populations) do is, of course, evolve. But the capacity to evolve preceded the appearance of sex, so what we should consider is whether the species as individual provides enhanced evolvability. Moreover, since sexual reproduction provides a boundary to the species, and a barrier to the acquisition of external genetic material, it appears prima facie to reduce evolvability. So if evolvability is indeed an advantage that partly explains the persistence and increasing dominance of sexual species, we might expect the gains in this regard to be substantial. The ability of advantageous genetic features to spread more rapidly through a species, and the ability, through recombination, of several advantageous alleles to be selected simultaneously, are sometimes proposed as decisive advantages of sex. However, this does little to explain the evolution of K-selected sexual species, where such advantages seem only a minor compensation for the great losses in this respect due to slow reproductive processes and small numbers of offspring.

If the highly integrated species is indeed a vehicle for greater evolvability, it is surely because it provides new sources of selectable variability. And indeed there are many familiar phenomena, already discussed above and in other essays in this volume, that offer to provide just this.

First, integrated species appear to offer a much more favourable environment for the transition from intra-specific competition to cooperation, as exemplified in the very high levels of cooperation found in eusocial species and in humans. In the former case, especially in the eusocial insects, it is widely accepted that the integrated colonies are a kind of organism (or ‘superorganism’) and clearly they have capacities far beyond those of their constituent individuals. The striking success of these insects and indeed of humans testifies to the evolutionary success of this kind of cooperation.

The extended care found in K-selected species provides an opportunity for a developmental system with multiple inputs in addition to the material of reproduction [40,41]. These include the environmental inputs made possible by the niche constructing activities of previous and present conspecifics and a wide variety of parental effects. They also provide an opportunity for the transmission of sometimes complex cultural traditions. All of these aspects of the developmental system are in principle entirely heritable, and thus provide potential pathways of evolutionary change. Niche construction and maintenance activities, or parenting activities can be learned and passed down the generations, and culture may be passed down through this and other routes in a more widely social species. This evolution may be solely behavioural, but it may also be physiological through the epigenetic direction of developmental plasticity.

It is hard to deny, though there is a very powerful ideological tendency to do so, that much evolutionary change through these pathways has the potential to be both acquired and adaptive. At the most uncontroversial end is human culture. We can argue, of course, whether it is a good thing, but that innovations in food production, say, are introduced because they produce more food, is uncontroversial. Much behavioural innovation that has been observed in other primates—food washing, termite fishing, and so on—has a similar character. How widespread this is is not something I shall discuss here. The point is only that a more integrated species does indeed provide multiple new evolutionary pathways that have in demonstrable instances resulted in adaptive evolutionary change.

There is a curious tendency to dismiss all such evolutionary pathways on the grounds that they are too transient and allegedly less durable than genetic change. Perhaps this tendency has been encouraged by Dawkins's already remarked appeal to immortality [23, ch. 3] in his argument for the overwhelming evolutionary importance of DNA. It is at any rate extraordinary that one should require the explanation of a changing process to be grounded in unchanging causes, and perhaps can be seen as a paradigm of the misleading effects of a substance- rather than properly process-based ontology.

One further key point is the following. Species are a diverse category. Arguably they are ontologically diverse, encompassing both processes and kinds, as profound a diversity as imaginable. More prosaically, even as concrete entities, they differ in very significant respects. If species are what evolve, we should not, for this reason, expect quite general accounts of evolution. The Modern Synthesis, specifically, may be more or less true for some kinds of species, but quite inadequate for others. If species have evolved new forms of evolvability, this is surely to be expected. Evolvability of many populations may just be a summative property of organism properties, but as species become integrated processes it is plausible that evolvability might emerge as a specific capacity of lineages.

This leads me to a more speculative final thought. There is a philosophical tradition of seeing organisms as a kind of agent, as beings in some way autonomously pursuing their own goals or interests. Denis Walsh [42] argues that this is a vital part of an organism-centred view of evolution of the kind championed by Darwin, and as opposed to contemporary molecule-centred views. Substance- (or thing-) based thinking has struggled with the idea of organisms as agents, and has often considered that at most humans achieved this rarefied status. For a process, intrinsically dynamic, and dynamic in ways that conduce to the persistence of the process, agency is a much more natural attribution. Hence process thinkers, such as the mid-twentieth century organicists [43–45] thought agency a quite general feature of organisms. If some species are themselves living processes, might they themselves have a kind of agency, inherent tendencies to change (act) in ways that promote their survival? If we take seriously the claim that species are individuals then this is at least a possibility worth investigation.

Evolved from a Can of Worms: Evolution and the Culture of Death

It was Veterans’ Day in Washington, D.C., and evolution was the farthest thing from my mind. I had accompanied my friend, Fr. Jack Murphy, an Army veteran, to provide prayer support for him and for other veterans who were standing up for life at a D.C. abortion mill. A large group of pro-abortion hecklers had turned out to harass us, and the police cordoned off the parking lot and forced us all into one small area. A young man in his twenties held up a poster of a preborn child with the caption “Does this look like a blob of tissue?” Two young women who looked like college students mocked him. “Didn’t these people take high school biology?” one of them asked the other. “If they knew anything about evolution, they would know that the fetus isn’t human until the third trimester.” The other said something about the baby in the poster going through “the fish stage.” Another woman added that the souls of the “fetuses” were better off being aborted, since they would be reincarnated in better circumstances.

A Radical Rejection of God’s Revelation

The Fathers rejected not only the idea of the pre-existence of souls, but also the notion that Adam’s body was formed before his soul, or that a human body could pre-exist a human soul. According to St. Gregory of Nyssa:

[A]s man is one, the being consisting of soul and body, we are to suppose that the beginning of his existence is one, common to both parts, so that he should not be found to be antecedent and posterior to himself, if the bodily element were first in point of time, and the other were a later addition[.] … For as our nature is conceived as two-fold, according to the apostolic teaching, made up of the visible man and the hidden man, if the one came first and the other supervened, the power of Him that made us will be shown to be in some way imperfect, as not being sufficient for the whole task at once, but dividing the work, and busying itself with each of the halves in turn. [2]

Sacred Scripture teaches that Jesus was a man like us in all things but sin and that He was already fully human in the womb of the Blessed Virgin a few days after the Incarnation, when His Mother visited her cousin St. Elisabeth. The Sacred Liturgy affirms the full Humanity of Jesus from the moment of the Incarnation on March 25, just as it affirms the sinless humanity of the Blessed Virgin from the moment of her Immaculate Conception. Thus, the Church’s teaching concerning the first Adam and the first Eve perfectly complements her teaching concerning the New Adam and the Second Eve. In both cases, a human body and soul were created together, not the soul before the body or the body before the soul.

This teaching on the creation of Adam and Eve has been the common teaching of all of the fathers, doctors, popes and councils since the time of the Apostles. However, recent popes, while not abrogating that teaching – which would be impossible – have held back from affirming it unequivocally for one simple reason. Since Darwin, they have been afraid to rule out the possibility that natural science might discover irrefutable evidence for human evolution.

In one sense, their hesitancy is understandable. It appears to follow from the Augustinian principle (affirmed by Leo XIII in his encyclical Providentissimus Deus) not to deviate from the plain and obvious sense of Scripture, except when reason dictates or necessity requires. In Humani generis, Pope Pius XII asked Catholic scholars to weigh the evidence for and against the hypothesis of human evolution, while defending many elements of the traditional interpretation of Genesis. To this day, the holy father’s request has not been heeded by the community of Catholic scholars, although there are three reasons why this request should long since have led to a definitive rejection of the human evolution hypothesis. The first reason has to do with the limitations of natural science, the second with the actual state of the scientific evidence, and the third with the obvious harm that this hypothesis has done and is doing to souls.

Three Reasons to Reject Human Evolution

Nowadays, it seems unfashionable in many circles to suggest that natural science has limitations. But the Catholic doctors who laid the foundation for the positive development of the natural sciences during the past 800 years recognized and articulated these limitations. The spirit of the great medieval doctors is well expressed by the twelfth-century French scholastic philosopher William of Conches, who wrote:

I take nothing away from God. He is the author of all things, evil excepted. But the nature with which He endowed His creatures accomplishes a whole scheme of operations, and these too turn to His glory since it is He who created this very nature. [3]

Implicit in this enthusiastic attitude toward the scientific investigation of nature was the understanding that the origin of the order of nature and of the natures of living things could not be explained by natural processes, or, to use the words of St. Thomas Aquinas, “[i]n the works of nature, creation does not enter, but is presupposed to the work of nature” [4]. Thus, St. Thomas and William of Conches knew for certain that the origin of human nature – the creation of Adam and Eve – lay beyond the sphere of natural science. While natural scientists could learn many things about the structure and functioning of the human body, it was obvious to the medieval doctors that scientific research could no more shed light on how God formed the body of Adam from the dust of the earth than it could shed light on how Jesus changed water into wine at the wedding of Cana. The great doctors distinguished between the order of creation, when God created the different kinds of creatures by His Word, and the order of providence, which began only after the creation of Adam and Eve.

Modern natural science has almost completely abandoned this distinction between the order of creation and the natural order, or the order of providence. Ironically, however, 21st-century natural science has amply confirmed the reasonableness of this distinction. For example, in the field of genetics, natural scientists have learned a great deal about the transmission and variation of genetic information, but no scientist has observed the spontaneous appearance of a new genetic program, such as would be needed to produce a new organ, like an eye or an ear, in an organism that lacked such an organ. Instead, 21st-century genetics has revealed that, far from evolving or increasing in functionality, genetic information degrades and devolves over time, at a rate that, in the words of one geneticist, places “a limit on the length of vertebrate lineages” – a limit much lower than the ages assigned to them by evolutionary theory [5]. Indeed, the discoveries of 21st-century genetics have been fatal to all current hypotheses of human evolution, as they demonstrate that it would be impossible for a common ancestor of chimpanzees and men to acquire the necessary “beneficial mutations” without acquiring a greater number of deleterious mutations – a number that would lead to extinction long before human evolution was achieved!

In short, not only does the hypothesis of human evolution collide with the unanimous teaching of the fathers of the Church and with nineteen hundred years of authoritative magisterial teaching, but it has also come into fatal conflict with the findings of natural science. Indeed, there is no doubt that if the balanced examination of the evidence called for in Humani generis were undertaken today, the hypothesis of human evolution would be rejected.

Embryonic Recapitulation: Devaluing the Human Embryo

Tragically, most Catholic intellectuals have not had the opportunity to study the evidence against evolutionary theory and continue to embrace the theory in spite of the harm that it has done – especially to respect for the pre-born child. Faith in the truth of the evolutionary hypothesis has repeatedly led scientists and medical researchers to believe that organs of the human body that have no apparent function are “vestigial” and expendable. The full extent of the danger inherent in this unsubstantiated assumption emerged soon after the publication of Origin of Species with the popularization of the concept of embryonic recapitulation by Darwin’s disciple, the German medical doctor and professor of anatomy Ernst Haeckel (1834-1919).

Darwin had argued that similarities in structure among diverse life forms indicate that they all evolved from a common ancestor. According to Haeckel, the existence of similarities in embryos of various kinds of organisms proves that the higher life forms “recapitulate” their evolutionary history before birth and that they had descend from a common ancestor. To make this “proof” more compelling for his contemporaries, Haeckel doctored drawings of the embryos of fish, salamanders, chickens, turtles, rabbits, pigs, and human beings to exaggerate their similarities and minimize their differences [6]. Although Haeckel’s fraud was discovered and exposed during his lifetime, the evolutionary hypothesis demanded common descent, and the concept of embryonic recapitulation continued to exert a profound influence on the study of embryology for many decades.

According to Jane Oppenheimer in her work Essays in the History of Embryology and Biology, Haeckel’s influence on embryology was considerable, “act[ing] as a delaying rather than an activating force[,] and … was stifling to immediate progress” [7]. One of the leading lights in the study of embryology in the twentieth century, Gavin R. de Beer, wrote that “Haeckel’s theory of recapitulation … thwarted and delayed the introduction of causal analytic methods into embryology,” since “if phylogeny was the mechanical cause of ontogeny as Haeckel proclaimed, there was little inducement to search for other causes” [8]. De Beer’s observation implies that Haeckel’s influence had come to an end by the 1950s – but this was far from the case. To this day, biology textbooks all over the world argue that similarities among embryos of fish, amphibians, reptiles, humans, and lower mammals constitute evidence for the evolutionary hypothesis. Typical of examples too many to cite is the caption that accompanies drawings of embryos of various life forms from a widely used American biology textbook published in 2002. Entitled “Embryonic development of vertebrates,” it states:

Notice that the early embryonic stages of these vertebrates bear a striking resemblance to each other, even though the individuals are from different classes (fish, amphibians, reptiles, birds, and mammals). All vertebrates start out with an enlarged head region, gill slits, and a tail regardless of whether these characteristics are retained in the adult. [9]

Although Haeckel’s distorted drawings do not accompany this caption, the statement gives the impression that human embryos – as members of the vertebrate phylum – possess gill slits. But this is patently false. The pharyngeal arches in human embryos have no connection with gill slits whatsoever rather, they develop into the outer and middle ear, and into the neck bones, muscles, nerves, and glands.

Moreover, after the discovery of DNA, confidence in the truth of the evolutionary hypothesis led many evolutionary biologists to predict that similar body parts in diverse organisms would be controlled by the same genes. This, however, proved to be false, as embryologists have discovered that the realization of the same body plan – such as five-digit extremities – in diverse organisms (such as whales and humans) is controlled by different genes and is achieved through totally different embryonic pathways [10].

Indeed, not only did the idea of embryonic recapitulation lead embryonic researchers down the wrong pathways – it has also led to a denigration of the unborn child. All over the world, abortion advocates have used the alleged similarity between human and lower animal embryos to trivialize abortion in the early stages of pregnancy. For example in Germany, pro-abortion activists (emphasis added):

… skillfully exploited the disunity of the German Catholic intellectuals to bring their demands for the legalization of abortion to the legislature. … Karl Rahner … wrote in Naturwissenschaft und Theologie (brochure 11, page 86, 1970): “I think that there are biological developments which are pre-human, but these developments are still aimed in the direction of man. Why cannot these developments be transferred from phylogeny to ontogeny?” [11]

With these words, the most influential theologian in the German-speaking world formulated a Haeckelian evolutionary rationale for abortifacient contraception and abortion long after Gavin de Beer had claimed that Haeckel’s influence had disappeared. In reality, in the “year of Darwin,” the implicit message of most high school biology textbooks is still clear: human embryos pass through a “gill slit” stage. These are “developments in the direction of man,” to use Fr. Rahner’s phrase. Therefore, to accord the human embryo the dignity of a human being from conception is biological nonsense.

In reality, of course, the development of the human embryo is quite distinct from that of the other vertebrates in Haeckel’s drawings, and there is no empirical evidence to support the claim that he passes through any stage that is not fully human, in the biological sense of the word. However, Fr. Rahner’s misguided faith in evolution continues to erode the faith of Catholics in the humanity of the unborn child.

An Abortionist Meets St. Thomas Aquinas

Ours is not the only period in Church history when the conventional wisdom of Catholic scholars has been influenced by a false hypothesis in natural science. Soon, the Catholic Church will celebrate the feast of St. Thomas Aquinas. Anyone the least bit familiar with the writings of St. Thomas knows how deeply he revered the Word of God. However, with regard to the time of human ensoulment, St. Thomas allowed Aristotelian natural science to overshadow the plain sense of the Word of God. Under Aristotle’s influence, St. Thomas wrote that human life begins forty days after fertilization. In contrast, the Eastern fathers of the Church, who spoke the language of Aristotle, were much less likely than St. Thomas to let “the Philosopher” determine their interpretation of God’s Word. St. Maximus the Confessor exemplified the attitude of many Eastern fathers when he held (in II Ambigua 42) that Jesus was a man like us in all things but sin and that therefore His assumption of our humanity from the moment of the Annunciation signified that we, too, become fully human from the moment of our conception.

The international pro-life community rightly rejoiced over the recent conversion of Serbian abortionist Stojan Adasevic through an apparition of St. Thomas, but scant attention has been paid to Adasevic’s interpretation of St. Thomas’s heavenly visitation. Educated in communist schools, Adasevic had been thoroughly indoctrinated in evolutionism and had regarded the unborn child in the womb as nothing more than a blob of tissue. Before his conversion, Adasevic performed 48,000 abortions, as many as 35 per day. Then St. Thomas Aquinas came to him in a dream and showed him the souls of the unborn babies he had aborted. Although he resisted at first, Adasevic finally renounced abortion and embraced Christianity. He became Eastern Orthodox, but he also studied the writings of St. Thomas Aquinas and was struck by the Angelic Doctor’s mistaken views on ensoulment. The former abortionist then concluded that the saint might have visited him “to make amends for his error” [12].

Nowadays one often hears that such and such a holy priest or bishop or even pope believed in evolution, so how could it be a dangerous doctrine? But Adasevic’s visitation suggests that if even a saint and doctor of the Church could be wrong about a hypothesis in natural science – with deadly results – how much more could modern Church leaders be deceived by a more far-reaching theory, with far deadlier consequences?

The High-Stakes Debate on Origins

There is a lot at stake for the pro-life movement in the origins debate.

If God created the first man and woman body and soul from the first moment of their existence – and the “new Adam” and the “new Eve” body and soul from the first moment of their conception – then we can confidently hold that:

– Human life is sacred from the beginning.

– Abortion at any stage is murder.

– The human soul is the form of a particular human body.

But what if a subhuman primate could “evolve” to the point where it could “receive” a human soul?

This would mean that the same body that housed a human soul was the body of a modified brute whose animal soul was replaced by a rational human soul. This would seem to give plausibility to reincarnation – the transmigration of souls – and to the equally pernicious idea that ensoulment takes place at some point after conception.

What if the “parents” of the body that became the “fine tuned” body of Adam were themselves “brutes”?

This would mean that the bodies of brute animals would be deserving of honor as the ancestors, in a real sense, of all mankind and would give credibility to Peter Singer’s proposal to give chimpanzees the same legal rights as human beings.

What if the body of the first human being was the fruit of the sexual union of two brute animals?

This would mean that human sexuality comes up from the lower, irrational animals, rather than down from above, as a finite reflection of the love of the Most Holy Trinity.

What if the animal ancestors of Adam and Eve (and of us all) practiced promiscuity, polygamy, polyandry, or adultery?

This would mean that such behavior is “natural” and certainly not to be condemned as a crime “against nature.”

On the other hand: What if the common message of all of the Church fathers, doctors, popes, and council fathers in their authoritative teaching on the creation of Adam and Eve were boldly proclaimed from every pulpit in Christendom?

Then the faith of all Catholics in the dignity of the human person from the first moment of life would be strengthened, and it would no longer be possible for Catholics to use evolution to trivialize abortion and sexual perversion as some do now.

Therefore, the time has come for the pro-life community to recognize the strong link between evolution and the culture of death and to work and pray for a restoration of the traditional Catholic doctrine of creation.

[1] ST. JOHN OF DAMASCUS, On the Orthodox Faith 2:12.

[2] ST. GREGORY OF NYSSA, On the Making of Man 28-29.

[3] Quoted in THOMAS WOODS, How the Catholic Church Built Western Civilization (Washington, D.C.: Regnery, 2005), p. 87.

[4] ST. THOMAS AQUINAS, S.Th. I. q. 45, a. 8.

[5] ALEXEY KONDRASHOV, Journal of Theoretical Biology, 1995, 175:583.

[6] Cf. MICHAEL K. RICHARDSON ET AL Anatomy and Embryology, “There is no highly conserved stage in the vertebrates implications for current theories of evolution and development,” Vol. 196, No. 2, Springer Verlag, Heidelberg, Germany, 1997, pp. 91-106.

[7] JANE OPPENHEIMER, Essays in the History of Embryology and Biology, MIT Press, 1967, p. 154.

[8] GAVIN DE BEER, Embryos and Ancestors, Third Edition, Clarendon Press, Oxford, 1958, p. 172.

[9] PETER H. RAVEN and GEORGE B. JOHNSON, Biology, 6th ed,, McGraw Hill, 2002, p. 1229.

[10] GAVIN DE BEER, quoted in “Homology: A Theory in Crisis” JONATHAN WELLS and PAUL NELSON (accessed 3-08-09).

[11] ALFRED HAUSSLER, The Betrayal of the Theologians, Human Life International, 1982, p. 2.

[12] In fairness to the Angelic Doctor, if St. Thomas were living on Earth today, he would be the first to reject the Aristotelian view of ensoulment in light of the scientific evidence – just as he would be the first to reject theistic evolution, on theological and scientific grounds.

Hugh Owen is a convert to the Catholic Faith (since 1972) and the son of Sir David Owen, the first secretary general of International Planned Parenthood Federation. He and his wife Maria have nine living (on Earth) children and sixteen grandchildren (so far), including two daughters who are members of a Benedictine Abbey. Hugh is a professional writer and editor and the director of the Kolbe Center for the Study of Creation, which provides a forum for Catholic theologians, philosophers, and natural scientists all over the world who defend the traditional Catholic doctrine of creation and who expose the fatal flaws in the molecules-to-man evolutionary hypothesis in both its theistic and atheistic forms.

Temporal isolation

Populations may mate or flower at different seasons or different times of day. Three tropical orchid species of the genus Dendrobium each flower for a single day the flowers open at dawn and wither by nightfall. Flowering occurs in response to certain meteorological stimuli, such as a sudden storm on a hot day. The same stimulus acts on all three species, but the lapse between the stimulus and flowering is 8 days in one species, 9 in another, and 10 or 11 in the third. Interspecific fertilization is impossible because, at the time the flowers of one species open, those of the other species have already withered or have not yet matured.

A peculiar form of temporal isolation exists between pairs of closely related species of cicadas, in which one species of each pair emerges every 13 years, the other every 17 years. The two species of a pair may be sympatric (live in the same territory), but they have an opportunity to form hybrids only once every 221 (or 13 × 17) years.


General Program Design and Overview

Rather than trying to observe ongoing evolutionary change in the classroom (e.g., Bordenstein et al., 2010), students can instead investigate adaptively divergent populations of organisms that represent the outcome of past evolutionary change, much like the examples used by Charles Darwin in presenting his case for evolution by natural selection (Darwin, 1872). Populations of living organisms that differ from each other in obvious ways, because they have become locally adapted to divergent environments, provide a powerful empirical framework to explore the conditions necessary for natural selection to occur. In particular, by quantifying variation within and between populations, with foundational knowledge of inheritance, locally adapted divergent populations can be used to illustrate four important concepts that students must grasp to understand evolution by natural selection: (1) variation exists within and among populations (2) much of that variation is inherited through genes (3) selection determines which individuals pass on their genes to the next generation and (4) over time this leads to genetic changes in a population, or evolution. These concepts are often reduced to four words—Variation, Inheritance, Selection, and Time—and are commonly represented by the abbreviation VIST ( An authentic science approach allows students to make observations and conduct experiments to discover ideas 1–3, and can conclude with thought exercises about outcomes over time, idea 4, and how this leads to the observed differences among the populations before them.

The first objective of the program is to engage the students in the scientific process, which begins with observation. Before introducing students to the terms “evolution” or “natural selection,” they can be engaged by making observations of a plant or animal system and noting differences within and among populations. Live organisms are particularly engaging (Allen, 2004), and it is important to maximize engagement in order to overcome belief persistence (Dole & Sinatra, 1998 Nelson, 2008). Almost any plant or animal system can be used as long as there are obvious, quantifiable phenotypic differences that can be intuitively connected to a source of natural selection (Figure 1). Whenever possible, it is helpful to use a familiar or local species, or to even use domesticated organisms, since students are more likely to engage if they feel a connection to the content because of past experiences (Dole & Sinatra, 1998).

Schematic showing how VIST can be applied to a hypothetical plant and animal system.

Schematic showing how VIST can be applied to a hypothetical plant and animal system.

Once students are familiarized with the contrasting environments where the organisms live (e.g., sunny vs. shady, with and without predators Figure 1), they can begin to ask questions and propose explanations for why differences might exist among the populations. They can then design observational or experimental studies to test their hypotheses. These studies should quantify the variation in traits of interest within and among populations (Figure 1a), explore the evidence for inheritance (Figure 1b), and demonstrate how selection acts on variation via differential survival or reproduction in response to environmental variation (Figure 1c).

At this stage, students have engaged in authentic science by generating hypotheses and predictions for why populations might differ in traits of interest, and have observed and collected data on how traits vary within and between populations. Students can then use their own data and observations to test their own hypotheses within the VIST framework:

Students can explore the concept of variation (V) by examining plots of their data (e.g., frequency distributions of measured traits like plant height or fish coloration) and visualizing that there is variation both among individuals within a population and between populations (Figure 1a). Sources of genetic and environmental variation can also be introduced and discussed (e.g., Broder et al., 2016). More generally, discussions of the observed results should emphasize that variation within and between populations is commonly found in almost all organisms, and this can be illustrated with familiar visual examples of other plants and animals.

Students can then apply their knowledge of inheritance (I) by focusing on heritable variation (Figure 1b) using traits from the system being studied as well as familiar examples (e.g., domestic animals). A review of Mendelian genetics is important since an emphasis on genetics in evolution programs can promote conceptual change (Kampourakis & Zogza, 2009).

Students can then be introduced to the idea of selection (S), that not all individuals are equally likely to survive and reproduce depending on the environment. If students have carried out experiments in class, they can discuss how certain traits increase or decrease the likelihood of surviving and reproducing in various environments. For example, students might explore how variation in light shapes leaf size or survival in low-light conditions in plants, or how predation regime influences color patterns in fish (e.g., background matching may provide a selective advantage to prey fish if predators are present Figure 1c). However, if such experiments are not possible, then students can discuss familiar examples of natural selection in the wild (disease, drought, etc.).

Finally, students can consider the consequences of these findings over time (T), that this variation in survival and reproduction leads to changes in allele frequencies across generations (Figure 1d). There are many ways to make this last point, including online tools, computer simulations, and thought exercises. The main conclusion the students should draw is that genetic change over time is an inevitable outcome when selection acts on heritable variation.

Following a discussion of variation, inheritance, selection, and time, the final step is to describe this process as evolution by natural selection. By defining the process of evolution as a change in allele frequencies within a population over time—an idea that students just discovered during the VIST activities—students must reconcile their preconceived ideas of evolution with the conclusions they reached examining their own results. We feel this point is critical. Students are more likely to overcome preset beliefs through asking their own questions and engaging in the scientific process (Nelson, 2008), collecting their own data (Slusher & Anderson, 1996), and drawing and defending their own conclusions (Tetlock, 1983) about evolution before learning the definition from instructors. This approach should lead to increased acceptance and understanding of evolution.

This formula for teaching evolution aligns with the Next Generation Science Standards for middle school. Specifically, this program teaches middle school disciplinary core ideas LS4.B (natural selection) and LS4.C (adaptation) while allowing students to accomplish MS-LS4-4, “Construct an explanation based on evidence that describes how genetic variations of traits in a population increase some individuals' probability of surviving and reproducing in a specific environment” (NGSS Lead States, 2013). Additionally, the final step of our program asks students to predict how different scenarios would affect the frequencies of alleles and consequently the frequency of traits in a population, supporting MS-LS4-6, where students should “use mathematical representations to support explanations of how natural selection may lead to increases and decreases of specific traits in populations over time” (NGSS Lead States, 2013).

Specific Program Details

We developed and implemented a program based on the above framework for 7th grade students at Windsor Middle School in Windsor, Colorado, and Severance Middle School in Severance, Colorado. Evolution by natural selection was illustrated using live Trinidadian guppies (Poecilia reticulata). Guppies are a model system in evolutionary biology for studying natural selection in the wild (Reznick et al., 1990 1997), and are familiar to many students via the pet trade. We used guppies sourced from three populations: (1) wild guppies collected from a stream where predation from larger fish is very high, and individuals exhibit a suite of genetically based morphological traits (e.g., reduced male coloration) and behavioral traits (e.g., faster escape responses) known to reduce predation risk (2) wild guppies collected from a stream where most predators are absent, and individuals exhibit traits that reflect reduced predation risk (e.g., increased male coloration, more courtship displays, slower escape responses) and (3) domesticated guppies from a local pet store that exhibit exaggerated traits (e.g., ornate colors, elongated fins and tails, naive behaviors toward predators) that have been artificially selected. Students generated hypotheses to explain differences they observed among guppy populations (e.g., male body coloration) and designed two experiments: (1) observations of mating behavior to test that male color provided a mating advantage, and (2) a predator encounter experiment to test if dull colored males were less visible to predators.

We strived to make the experience as authentic as possible based on Chinn & Malhotra's distinction between simple versus authentic science (2002). We guided the students' questions and experiments toward reproduction and survival because of our content goals, but allowed students to design details. For example, we encouraged students to allow guppies to interact with a predator to learn something about antipredator behavior and survival. They chose to place three male guppies (one from each of the three populations) in a tank with a predator, allowed an acclimation time before removing a barrier, and decided which antipredator behaviors to record. Students could have designed this experiment differently (e.g., used females instead of males or allowed one guppy to interact with the predator at a time). Students also helped design data sheets and came to a consensus on the operational definitions of behaviors. For example, in the mating trials, students observed videos of mating behavior, and agreed on definitions of a courtship display and a forced copulation attempt. Students then divided into pairs and observed mating behavior of one male and one female guppy from the same population. Though each pair of students collected data on a single mating pair, they also switched tanks with classmates to observe pairs of guppies from the other two populations. We compiled the data from the entire class, and each student made figures summarizing both experiments and evaluated the results in light of their hypotheses. We did not know the outcomes of the experiments in advance, and students often had to justify anomalous results—an important part of the authentic science experience (Chinn & Malhotra, 2002).

We followed the student experiments with discussions of their data and experiences to introduce our four concepts:

Variation: The coloration and number of courtship displays performed by male guppies was variable both within and among populations.

Inheritance: A Mendelian Punnett-square approach illustrated how genes for bright coloration and high rates of courtship are passed on to offspring (Kane et al., 2018, this Issue).

Selection: Results from the predator encounter experiment, where domestic guppies were six times more likely to be depredated than males from the two wild populations, illustrated a mechanism of natural selection. Results from the mating experiments, where females were more interested in bright males, demonstrated a mechanism of sexual selection. We also discussed how selective breeding for the pet trade produced the exaggerated coloration of domestic guppies, to explain artificial selection. Finally, we discussed how particular traits should affect fitness in the three environments.

Time: A thought exercise, where students imagined that males could have alleles that code for bright (A) or dull (a) coloration, allowed them to predict how the ratio of A to a alleles might change over time in the population under different scenarios (see Kane et al., 2018, this Issue, for details).


We implemented this program in April 2012 and April 2013 at Severance and Winsor Middle Schools. All participants were 7th grade students associated with two teachers (SW and KDK), and the guppy program described above replaced their regular unit on evolution by natural selection. We administered pre- and post-program assessments in 2013 to five of KDK's classes at Severance and four of SW's classes at Windsor (n = 204 total students). To estimate knowledge of evolution by natural and artificial selection, we used the seven-question, multiple-choice test administered each year to Windsor Middle School 7th grade students ( Appendix). These questions were written by one of the teachers (SW) based on the suggested learning outcomes for microevolution education in the Colorado Academic Standards (2009). The presenters of the program (EDB, CKG, and LMA) did not alter these questions, ensuring that the questions were not influenced by the program. We were unable to supplement this with a published assessment, because of a lack of tools in the literature appropriate for this age group at the time of this study. Other assessments exist for high school and college students, including “The Knowledge of Evolution Exam” (Moore et al., 2009) and the “Measure of Understanding of Macroevolution” (Nadelson & Southerland, 2010), but they were deemed inappropriate given their Flesch-Kincaid grade-level scores of 8.9 and 9.3, respectively (Flesch, 1948).

To measure acceptance of evolution, we selected four questions from the MATE Instrument, which uses a Likert scale to indicate agreement with various statements (Rutledge & Sadler, 2007 Appendix). We recognize the limitations of using only a subset of the MATE Instrument, and our results cannot be compared to other studies that used the full MATE instrument. We had to exclude questions that mentioned religion or beliefs because the teachers (SW and KDK) felt that it was a violation of their teaching agreement to include such questions. This assessment was granted exemption by the Colorado State University Human Subjects Approval Board (IRB ID 038-14H).

To analyze the results, we performed two repeated measures ANOVAs the first used the average scores from the seven multiple-choice questions (knowledge), and the second used the average scores from the four Likert scale questions (acceptance). In both analyses, the test (pre- or post-) was a fixed effect. We also included random effects of individual student (n = 204) nested within classroom (n = 9) nested within teacher (n = 2). We excluded students who did not have both a pre- and post-program assessment.

Access options

Get full journal access for 1 year

All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.

Get time limited or full article access on ReadCube.

All prices are NET prices.

Elemental Signatures of Evolution in Classic Traits

Because the elemental phenotype emerges from variation in classic traits, elemental phenotypes consequently should be affected by both genes and the environment and shaped by the same mechanisms and constraints that underlie evolution in classic traits. Natural selection on the elemental phenotype may be direct, if environmental availability of elements constrains elemental budgets of organisms, or indirect, if adaptive modification of classic phenotypes alters elemental demands. Resulting evolutionary shifts in the elemental phenotype can affect ecology. For example, altered supply of elements such as P due to excessive application of fertilizer has resulted in microevolutionary shifts toward decreased P use efficiency of crops (Evans 1993 ), indicating a shift in ecological function from a P sink to a P source.

It is likely that evolution of phenotypes that have distinctive elemental demands translates into differential use of several of the

25 elements required for life. All classical traits require more than one element. As such, an evolutionary shift in one trait should affect demand and processing of suites of elements. For example, reductions in the number of bony lateral plates in populations of threespine stickleback (Gasterosteus aculeatus see Box 2) in freshwater ecosystems (Bell and Foster 1994 ) should not only result in altered P content and processing compared to marine counterparts with more lateral plates, but also in changes of calcium content and processing because bone is

1 part P (Kramer and Shear 1928 ). Furthermore, such evolutionary shifts in body armor can affect processing of sodium because the cellular physiology of P transport into cells is dependent on sodium (Werner et al. 1998 ). The ecological consequences of such correlated shifts in the use of multiple elements (e.g., Kaspari et al. 2009 ) will be important for a rigorous exploration of how evolutionary shifts impact ecology.

Box 2. Illuminating the potential ecological consequences of phenotypic evolution in sticklebacks using the stoichiometric framework

Armored (oceanic, top) and unarmored (freshwater, bottom) threespine stickleback (Gasterosteus aculeatus). Loss of bony lateral plates, made of calcium and phosphorus, should lead to lower Ca and P content in unarmored form compared to armored ancestors.

No differences in Ca and P content between the two forms indicate reallocation of Ca and P from lateral plates to other traits in the freshwater form. Such shifts in allocation of Ca and P within the individual would be ecologically neutral within the stoichiometric framework, although accommodative shifts in the content of other elements (e.g., sodium) could alter the elemental phenotype, with potentially unique shifts in ecological functions.

Alternatively, a difference in the content of Ca and P between the two forms would indicate altered rates at which Ca and P are acquired, assimilated, and/or excreted, with potentially distinct consequences at the community and ecosystem level, as shown for benthic and limnetic freshwater forms (Harmon et al. 2009 ).

Similar experiments with stoichiometrically explicit data collection and analyses have great potential to rigorously answer the questions furnished in Fig. 3. Photo courtesy: RDH Barrett (Barrett et al. 2008 ).

Can Evolutionary Quantitative Genetics Provide a Bridge Between SET and the EES?

Laland et al. (2015) reviewed and compared the structures, assumptions and predictions of the EES and contrasted these against the MS. Among the core assumptions of the MS that they identifed were “The pre-eminence of natural selection” and “Gene-centred perspective” (their Table 1). They further criticized the “blueprint”, “program” and “instruction” metaphors in genetics and the MS. In their criticism of MS and SET, Laland et al. (2015) wish to extend the domain of reciprocal causation from the interaction between ecological and evolutionary processes (as discussed in this article) to the domain of organismal development, or what they call “constructive development”. I will not dwell too deeply in to this here, due to space limitations, except that I note that there is of course no a priori reason why reciprocal causation and dialectical thinking should not be possible to apply also to development. However, constructive development is also perhaps the aspect of the EES that is most controversial and which has sofar been met with most resistance. Nevertheless, the increasing interest in epigenetic inheritance is certainly justified and will most likely lead to new empirical insights. Clear cases of epigenetic inheritance now exists (Dias and Ressler 2014) and it is now mainly an empirical issue to understand the importance of such effects and how widespread they are (Charlesworth et al. 2017). Here, I take issue with some of the claims by Laland et al. (2015), and I argue that their characterization of MS and SET provides a wrong, or at least a very biased, picture of the state-of-the-art research in modern evolutionary biology. I also suggest that Laland et al. (2015) have underestimated the flexibility and scope of evolutionary genetics, particularly evolutionary quantitative genetics.

With respect to Laland et al’s (2015) claim of the pre-eminence of natural selection in contemporary evolutionary biology, it must be emphasized that most evolutionary biologists today, including many molecular population geneticists would strongly disagree (see Welch 2016 and Charlesworth et al. 2017 for further discussion). On the contrary, leading molecular population geneticists are highly critical of what they consider an excessive adaptationist research programme in some areas of evolutionary and behavioural ecology. Some leading evolutionary biologists would instead argue that random processes such as genetic drift should more often be used as a null modell and point of departure, before invoking natural selection (Lynch 2007 Charlesworth et al. 2017). Historically, and from the very beginning of the MS, the non-adaptive process of genetic drift was considered to have a much more powerful evolutionary role than it perhaps deserved to have, something which only became clear after extensive empirical investigations in both the field and in laboratory studies (Provine 1986).

With respect to Laland et al.’s (2015) further characterization of the MS as gene-centred, many organismal biologists and evolutionary ecologists would strongly disagree (see also Futuyma 2017). Evolutionary quantitative genetics focus on whole organisms and use measurements of phenotypic traits (variances and covariances) as its point of departure, and thereby ignores underlying molecular genetic and developmental mechanisms behind these traits (Lynch and Walsh 1998). This might be perceived as a weakness with the evolutionary quantitative genetics approach, but it can also be perceived as a strength (Steppan et al. 2002). Through this procedure, quantitative genetics become liberated from the tyranny of genetic details in classical population genetics, as argued forcefully recently by Queller (2017).

Laland et al’s (2015) call for more appreciation of constructive development is certainly compatible with evolutionary and quantitative genetics theory and methods. For instance, gene expression is often strongly environment-dependent (e.g. Lancaster et al. 2016) and that such environment-dependent gene expression is also often likely to be heritable. Likewise, it is not controversial that genes, environmental conditions, gene–gene interactions (epistasis) and gene-by-environment interactions (GEI:s) all influence the development of the adult phenotype (Lynch and Walsh 1998). Moreover, the trait variance decomposition approach in quantitative genetics would work equally well in a non-DNA world with non-genetic inheritance, as long as there is trait heritability, i.e. this mechanism-free approach is general and flexible. For instance, the Price Equation does not assume that heredity is based on DNA, but is based on the phenotypic resemblance between relatives, such as parents- offspring covariance (Frank 1995, 1997). Thus, the quantitative genetic approach does already present a substantial extension of classical population genetics from which it grew out from, and could potentially be extended further to account for various forms of non-genetic inheritance, such as ecological inheritance (see Helanterä and Uller 2010 for discussion). Quantitative genetics does therefore already partly take constructive development in to account by modelling not only additive genetic variances and covariances, but also environmental components, dominance variation, epistasis and GEI:s (Lynch and Walsh 1998). Few evolutionary biologists and quantitative geneticists today would argue that the genotype-phenotype map is perfectly linear, that all genetic variation is additive and few would deny that genes interact with other genes and with environments during organismal development.

The possibilities of genetic assimilation and genetic accommodation that have been put forward in criticisms of SET by proponents of EES (Laland et al. 2015) as well as by West-Eberhard (2003) have actually already been successfully modelled using quantitative genetic approaches (Price et al. 2003 Lande 2009). Evolutionary quantitative genetics can be used to model reaction norm evolution, canalization and phenotypic plasticity, e.g. by treating slopes and intercepts of reaction norms as separate traits, which can be connected through genetic correlations (Chevin et al. 2010). Furthermore, developmental bias, put forward by EES-proponents as a challenge to SET does not by any means provide any major conceptual or methodological difficulty for contemporary evolutionary theory. Instead, such developmental bias can be viewed as the mechanistic basis of genetic trait correlations, which can bias evolution along “genetic lines of least resistance” (Schluter 1996). Proximate (“mechanistic”) and ultimate (“evolutionary”) explanations are therefore complementary to each other, rather than being mutually exclusive (Sinervo and Svensson 1998). When both these forms of explanations are considered jointly, they provide a richer understanding of organismal biology compared to when each type of explanation is considered alone. For instance, life-history trade-offs can be studied either by estimating genetic correlations between traits or dissecting the mechanistic basis of such trait correlations, by combining quantitative genetics with experimental manipulations of hormonal pleiotropy (Sinervo and Svensson 1998). Mechanistic and evolutionary perspectives therefore complement each other and little conceptual insight would therefore be gained by abandoning the distinction between proximate and ultimate causation (Futuyma 2017), in contrast to the claims made by Laland et al. (2011). Given previous successful attempts to combine mechanistic and evolutionary biology through evolutionary quantitative genetic and experimental approaches, there is therefore ground for optimism that eventually new insights from evolutionary developmental biology and epigenetics will become successfully integrated in to modern evolutionary biology research (Futuyma 2017).

Moreover, the different variance components in quantitative genetics are not static, but are dynamic and can evolve. For instance, after population bottlenecks, epistatic variance can be converted to additive genetic variance (Meffert et al. 2002) and models of the Fisherian Runaway process of sexual selection have revealed that genetic covariances can evolve through a dynamic feedback between the selective environment (female choice) and male secondary sexual traits (Kirkpatrick 1982). It is also worth emphasizing that natural selection can be viewed as both an ultimate and proximate explanation, as argued recently by Gupta et al. (2017). The process of natural selection has actually nothing to do with genetics, and questions about the causes of selection are also questions about ecological selective agents, which have their origin in the external environment (Wade and Kalisz 1990). Therefore, in the evolutionary quantitative genetics research tradition, genes are certainly not the main causal agents explaining evolution by natural selection it is instead the selective environment that is the main causal agent (cf. Brandon 1990 Wade and Kalisz 1990).

In their call for an EES Laland et al. (2015) asked for greater appreciation for reciprocal causation in evolutionary biology, but argued that:

However, reciprocal causation has generally been restricted to certain domains (largely to direct interactions between organisms), while many existing analyses of evolution, habit- or frequency-dependent selection are conducted at a level (e.g. genetic, demographic) that removes any consideration of ontogeny. Such studied do capture a core structural feature of reciprocal causation in evolution—namely, selective feedback—but typically fail to recognize that developmental processes can both initiate and co-direct evolutionary outcomes (p. 7. Laland et al. 2015).

Laland et al. (2015) thus admit that reciprocal causation is and has often been studied by evolutionary biologists, but they argued that ontogeny and development should be incorporated in such analyses. I hardly disagree here, and I think incorporating the role of development and ontogeny in studies of (say) frequency-dependent selection, eco-evolutionary dynamics, co-evolution and analyses of selection is likely to yield many novel and important insights. However, the reason that development has not been incorporated in that many previous studies in this field is not that the researchers in question rely on an outdated and simple view of unidirectional causation. The reason is more likely a practical one: it is extremely difficult and empirically challenging to understand and study reciprocal causation even at single ontogenetic level, such as among adults. I therefore disagree with Laland et al. (2011, 2013) in their suggestion that the lack of consideration of development in past studies is due to the lasting legacy of Ernst Mayr’s proximate-ultimate dichotomy, and that evolutionary biologists in general adher to an outdated view of unidirectional inheritance. Rather, the lack of studies of this kind reflect legitimate and difficult empirical challenges. I am not convinced that the EES-framework alone can solve these problems, unless some more concrete novel methodological or analytical tools are provided.

Moreover, evolutionary geneticists and evolutionary ecologists have actually paid attention to the interplay between ontogeny and selection. For instance, researchers have modelled and investigated how selection pressures change both in magnitude and sign during the organism’s life cycle (Schluter et al. 1991 Barrett et al. 2008). Moreover, there is much interest and ongoing theoretical and empirical research aiming to integrate and model the interaction between interlocus sexual conflict at the adult stage over the reproductive interests of males and females, with intralocus sexual conflict experienced earlier in ontogeny (Rice and Chippindale 2001 Chippindale et al. 2001 Barson et al. 2015 Pennell et al. 2016). There is also an increased appreciation of how alternative reproductive strategies shape ontogenetic trajectories, and how the same ontogenetic trajectories in turn affect adult phenotypes (Neff and Svensson 2013), another example of reciprocal feedback during development.


(1) Although camouflage is predominantly thought of as an anti-predator defence mechanism, many unique strategies are also observed in predators, and those expressed in both groups are likely to be driven by different selective forces.

(2) In ambush predators, three main strategies have evolved that act to minimise the prey's ability to detect or identify the predator before an attack: aggressive mimicry (which may involve a generalised or specialised lure), aggressive masquerade, and aggressive crypsis. In pursuit predators, four main strategies have evolved that minimise the prey's ability to detect or identify the predator: motion camouflage, motion masquerade and dynamic crypsis viabackground-matching or disruptive camouflage.

(3) Two evolutionary explanations for camouflage differences between predators and prey are the ability of predators to control when an attack occurs, and size differences between the two groups. There is a case for these being key drivers in the evolution of unique predatory camouflage strategies, however, few studies have addressed this and it remains an avenue for future research.

(4) Based on data in other biological systems, the camouflage strategy a predator adopts is likely to be affected by both their prey and their environment. Future research should shift towards applying empirical and theoretical frameworks to how these driving forces lead to the evolution of different predatory camouflage strategies. In particular, focus should be given to understanding how the position of a predator in the food chain impacts the camouflage strategy they use, as this will provide information not only on why a trait has evolved, but also on whether camouflage is used differently when avoiding or initiating an attack.

(5) The evolutionary forces driving camouflage in predators have received little attention. The development of new technologies is opening up avenues for future research projects on predators, and we hope that this review will stimulate interest in this area.