sign in sign up
<<
Home , Ecosystem ecology, Evolutionary ecology, Resource (biology)

PERSPECTIVE PERSPECTIVE Processes and patterns of interaction as units of selection: An introduction to ITSNTS thinking W. Ford Doolittlea,1 and S. Andrew Inkpena,b

Edited by Douglas Futuyma, Stony Brook University, Stony Brook, NY, and approved March 7, 2018 (received for review December 22, 2017)

Many practicing biologists accept that nothing in their discipline makes sense except in the light of , and that is evolution’s principal sense-maker. But what natural selection actu- ally is (a force or a statistical outcome, for example) and the levels of the biological hierarchy (genes, , , or even ) at which it operates directly are still actively disputed among philosophers and theoretical biologists. Most formulations of evolution by natural selection emphasize the differential of entities at one or the other of these levels. Some also recognize differential persistence, but in either case the focus is on lineages of material things: even species can be thought of as spatiotemporally restricted, if dispersed, physical beings. Few consider—as “units of selection” in their own right—the processes implemented by genes, cells, species, or communities. “It’s the song not the singer” (ITSNTS) theory does that, also claiming that evolution by natural selection of processes is more easily understood and explained as differential persistence than as differential reproduction. ITSNTS was formulated as a response to the observation that the collective functions of microbial communities (the songs) are more stably conserved and ecologically relevant than are the taxa that implement them (the singers). It aims to serve as a useful corrective to claims that “holobionts” (microbes and their or hosts) are aggregate “units of selection,” claims that often conflate meanings of that latter term. But ITSNS also seems broadly applicable, for example, to the evolution of global biogeochemical cycles and the definition of . evolution | natural selection | process | persistence |

Our purpose here is to introduce and justify ITSNTS Lewontin (3) summarized the collective wisdom of (“It’s the song, not the singer”) theory as a way of con- neoDarwinian theorists in a three-part recipe for what ceptualizing evolution by natural selection (ENS) that we here are calling “standard ENS.” allows processes as well as things to be “units of selec- tion,” selected for persistence and re-produced but not (i) There is variation in morphological, physio- reproducing (1, 2). We see this as a legitimate and logical, and behavioral traits among members broadly useful application of Darwinian principles in a of a species (the principle of variation). (ii)The broad area ( evolution) currently in want of variation is in part heritable, so that individuals such theory. ITSNTS makes it sensible to talk about resemble their relations more than they re- semble unrelated individuals and, in particular, and functions as relevant to (if not proper- offspring resemble their parents (the principle ties of) multispecies collectives, while standard ENS of heredity). (iii) Different variants leave dif- formulations do not. Whenever what is remarkable ferent numbers of offspring either in immedi- about a biological or cultural phenomenon is the per- ate or remote generations (the principle of sistence of a process or pattern of interactions (a song) differential ). that can be carried out by a variety of cooperating entities (singers), each of these having evolved because Lewontin (3) claimed that these conditions are nec- individual participation has (as in standard ENS) favored essary and sufficient for ENS, and here we take as one its differential reproduction, that is ITSNTS. type of ENS that which will likely occur given them,

aDepartment of and Molecular , Dalhousie University, Halifax, NS B3H 4R2, Canada; and bDepartment of Philosophy, Dalhousie University, Halifax, NS B3H 4R2, Canada Author contributions: W.F.D. and S.A.I. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Published under the PNAS license. 1To whom correspondence should be addressed. Email: [email protected]. Published online March 26, 2018.

4006–4014 | PNAS | April 17, 2018 | vol. 115 | no. 16 www.pnas.org/cgi/doi/10.1073/pnas.1722232115 Downloaded by guest on September 25, 2021 regardless of the nature of the evolving entity. But as commonly If p = 1, it is meaningful to think of communities as reproducing understood, these principles entail differential reproduction of (ar- and forming parent–offspring lineages. The pool has a single guably) material entities producing parent–offspring lineages. contributing parent community and is the exclusive contributor to ITSNTS questions this requirement, seeing Lewontin’s conditions one or more “offspring” communities. The situation would be like as sufficient but not necessary for ENS, which we take to be any that of a lichen, whose dispersal structures include individuals reiterated causal process resulting in greater relative representa- from both fungus and phycobiont or, as we envision here, of a

tion at time tn later than t1 of certain types of entities by virtue of multispecies biofilm that “reproduces” by budding off a fragment those types’ possession of certain traits, those traits for that reason containing a random sample of the organisms present in it, this to being adaptations (2). reattach and grow at a new site. If fragments are large enough to Godfrey-Smith (4) describes the more limited, traditional view overcome “ecological drift” (10), each offspring community will when he writes that: likely resemble its parent or siblings in species composition more than it resembles unrelated communities. Lewontin’s conditions Evolution by natural selection is change in a ow- would apply and community-level properties, such as the ten- ing to variation, heredity and differential reproductive dency to form more robust propagules more often (community- success. This is usually seen as a micro-evolutionary process reproductive traits that might depend on species make-up) could acting on organisms, but the criteria required are abstract; be selected for. genes, cells, social groups and species can all, in principle, However, suppose instead that all parental communities re- enter into change of this kind. For any objects to be units of lease the individual organisms they contain into a common pool, selection in this sense, however, they must be connected by parent–offspring relations; they must have the capacity to from which n are randomly chosen to form each community in the reproduce. Units of selection in this sense can be called next generation of communities. In this case, the species com- Darwinian individuals. An evolutionarily relevant case of positions of the offspring communities will reflect only that of the reproduction can take many forms. There need not be rep- common pool and will, sampling error aside, resemble each other. lication, the faithful production of copies. Replicators are By the “law of large numbers,” such intercommunity resemblance Darwinian individuals with high-fidelity, asexual reproduc- will get closer and closer as n increases, just as the frequency of tion, and it is possible to have evolution by natural selection heads flipped with an honest coin gets closer and closer to 50% on units where reproduction is sexual and heredity is weak the longer one flips. (emphasis ours). Of necessity, n cannot be less than p, the number of parents Although sexual heredity is weaker because each parent’s con- contributing at least one individual to a new community. In many tribution is diluted by half, children usually resemble parents and real microbial communities assembled by de novo siblings more than they do “unrelated individuals,” and this is all from multiple sources, such as multispecies biofilms, the micro- that Lewontin’s second condition actually demands. But the more biomes of many and , and environmental sites like parents, the more the dilution, and Godfrey-Smith (4) cautions hydrothermal vents or whale carcasses on the sea bottom, p and n that “... there must be some parent–offspring similarity, and the will both often be very large—dozens to thousands, perhaps— clarity of a ‘parent-offspring’ relation of the relevant kind is in- and new recruits can also come from lineages of organisms pre- versely related to the number of parents—if there are too many viously only loosely associated with circumscribable communities. parents there are no parents at all.” (It is important to remember that it is the number of parent communities that contribute individuals to each offspring com- How There Can Be “Too Many Parents” munity that matters in this model, not the number of different What ITSNTS theory aims to undermine and supplant are claims species that these individuals might represent.) that multispecies groups comprise material units of selection even So, the relevance of Godfrey-Smith’s (4) “too many parents” when in violation of Lewontin’s recipe, in violation because there warning is this. When p is large, n is as large and usually larger. It are too many parents. Such claims (5–9) made with respect to might in principle be possible to identify all parental communities microbial communities (symbioses, “holobionts,” biofilms, and contributing at least one individual to any particular – circumscribed ecosystems) initially motivated ITSNTS, but the offspring community: complex parent offspring relations could then still be defined. But individual offspring communities will not theory is more broadly applicable, as will be argued later. resemble their “parents” or relations significantly more than they Consider the following simple model. Imagine a constant resemble “unrelated” communities. Lewontin’s second condition number, c, of communities, each formed by the coming together does not hold and ENS as normally understood cannot affect at random (recruitment) of n individual asexual organisms (for communities as communities. So it seems appropriate to say that example, bacteria) representing s species, where s << n. After at some value of p there come to be too many parents. Past this formation of communities, individual organisms can replicate point we should consider that, although the collection of com- within them. We stipulate that no species’ lineage has a repro- munities is re-produced (created again), communities do not in- ductive advantage over another during such in-community repli- dividually reproduce. Communities cannot be units of selection in cation, so that differences in species composition at the end of a the sense of Godfrey-Smith (4). community’s life compared with its beginning are stochastic. ’ In this model, at the end of a community s life the individual The Hole in Holobiosis organisms contained in it are released into a pool, and n of these Proponents of holobiosis, however, often do consider animals and are chosen randomly from such a pool to form each community in plants together with their respective microbiota to be units of the next generation of communities. What matters to the argu- selection (5–9), even when the latter are randomly recruited (re- ment here is how many original communities (call them parental produced) at each reproductive generation of the former. There is communities) contribute to the pool from which a new one is some ambiguity in this literature about whether designating an formed. Call that number p. entity as a unit of selection means that it takes part in an ENS

Doolittle and Inkpen PNAS | April 17, 2018 | vol. 115 | no. 16 | 4007 Downloaded by guest on September 25, 2021 process or is simply an assertion that it (like an organism) is a cir- translate to the individuation of processes: Is a given set of bio- cumscribable thing subject to environmental forces (11, 12). But it chemical or developmental steps performed under (even very is in the former sense that holobiosis theory has attracted most slightly) different conditions to be counted as the same or different? interest and criticism (13–16). However, when intermediates are freely and globally diffusible (as Sometimes, of course, unit of selection claims might be sen- is the case for many intermediates in major biogeochemical cycles), sible. The value of p could be 1.0, for example when we are it makes no sense to attempt any such differentiation. It seems dealing with an obligate maternally transmitted , or li- reasonable to regard redundancy as real and irreducible, and the chens and biofilms as imagined above. The value of p could also stability and change over time of the more broadly defined pro- be fairly low for symbioses of coevolved species in which re- cesses implemented by redundant taxa as that which a complete cruitment of new partners is local (as for and their pets) theory of ENS needs to encompass and explain. (17). But, as noted above, the value of p (and thus n) could be very ITSNTS is cartooned in Fig. 1. Consider a long-standing large for complex ephemeral communities, such as those that interconnected metabolic interaction, the individual steps or establish themselves independently on the sea bottom or in our components of which are implemented by individuals from differ- guts after a course of antibiotics. Thus, Lewontin’s recipe should ent microbial or multicellular taxa recruited from the environment, not apply. For something like the “community” of microbes that, as none being able to perform all of the steps. These individuals might free-living lineages or symbionts, collectively perform the be physically associated or exchange metabolites over long dis- cycle, p is incalculably large, if a community can be defined at all. tances and time periods. The bobtail squid and its recruited vibrios And it is around the problematics of collective differential repro- (23) meet our criteria, but again the global nitrogen cycle may be a duction that invocations of even more all-embracing global better example (24). No individual of any species involved in that organism-like evolved entities, such as “Gaia,” have foundered (2, 18). cycle performs its role self-sacrificially. Each has coevolved with It is for the same reason problematic to speak of of other participants serving as its environment, or against a back- ground of the products of these other participants. No alleles have multiparental communities formed by recruitment or re-production. been fixed just because they promote continuance of the nitrogen Sober and Wilson (19) recommend what they call “William’s cycle as a cycle, for all that alleles that have been fixed may do that: Principle”—that “adaptation at a level requires that there was the time scales are too long and there are “too many parents.” selection at that level”—noting that “[t]he fact that a trait now Thus, a completed nitrogen cycle (or the squid–vibrio symbiotic benefits groups does not entail that it evolved because it was relationship) is only a “fortuitous benefit” of the of beneficial to groups.” Indeed, it is not clear that any property can separate species, not an adaptation of some interspecies aggre- be considered “beneficial” to impermanent and nonreproducing gate. Such properties might be seen as beneficial for life in the communities. These might come to show traits of interest to or very long run (2), but have never been adaptations as traditionally valued by us as researchers or citizens, such as functional stability, defined by a history of differential reproduction within pop- or resilience, or “eubiosis” (in the case of the “holobiont” that is ulations. Standard notions of ENS by differential reproduction, us and our gut ), these being the indirect byproducts and associated concepts of adaptation of things, cannot apply. of “lower-level” processes. But without some form of differential Nevertheless, because each step of the nitrogen cycle in iso- reproduction or differential and continuous physical persistence, lation can benefit a taxon that implements it, energetically or in these traits cannot be considered adaptations for communities if the provision of metabolites, and because the cycle has long been ’ anything resembling Lewontin s recipe is to be applied. in operation, many taxa (sometimes from all three domains and forming ecological “guilds”) have evolved that can and now do, ITSNTS Theory collectively, perform it (Fig. 1). Lateral gene transfer, whose vari- “ ” ITSNTS theory seeks to rationalize (or Darwinize ) such indirect ous mechanisms are adaptations for the subcellular agents of “ ” beneficial outcomes, casting these not as adaptations for the transfer, facilitates the expansion of guilds, enabling operation of individuals or collectives that implement a process, but for the the cycle under many environmental conditions. Because there’sa process itself. Indeed, it was initially motivated by the now fre- song, there are singers. “ ” quent claim that microbial community activities ( functions ) are Reciprocally, because there are many taxa capable of imple- more stable or ecologically resilient than are the taxonomic menting each step, and intermediates are variously diffusible, the compositions of the assemblages carrying them out, a phenom- cycle would spontaneously resume operation if disrupted. The enon demanding evolutionary explanation. Early in the develop- responsible taxa would re-produce it. Such disruption is no doubt ment of microbiomics as a , Turnbaugh et al. (20) for unimaginable in the case of the nitrogen cycle globally, but example, used an appealing ecological metaphor to describe this analogous re-production occurs locally (and daily for sunlight- phenomenon as it was observed in gut microbiota: dependent processes) and periodically for others assembled by recruitment, such as the squid–vibrio pair, decaying whale carcass This conservation suggests a high degree of redundancy in microbiota, and our own gut microbiomes. Because there are the gut microbiome and supports an ecological view of each singers, there’s a song. “ ’ individual as an island inhabited by unique collections of Together, evolutionary recruitment of capable taxa and their microbial phylotypes: as in actual islands, different species periodic reassembly, re-producing the nitrogen cycle or any of assemblages converge on shared core functions provided by these less global processes, promote their persistence and evo- distinctive components. lution as processes. All else being equal, we might imagine that How fine-scaled the mapping of genomic diversity to eco- the persistence of a process would depend on the taxa that have logical diversity is remains an active question (21, 22). We must evolved to perform its various steps, specifically on their number, ask how often taxa redundant with respect to participation in a population sizes, diversity (ecological and biochemical), and given process are specifically adapted to other, unaccounted-for, geographic dispersal. And the properties of a process will affect environmental constraints. In the context of ITSNTS, this will these variables and thus the persistence of that process, in a

4008 | www.pnas.org/cgi/doi/10.1073/pnas.1722232115 Doolittle and Inkpen Downloaded by guest on September 25, 2021 variety of ways. Among relevant properties would be: (i) the ad- Endosymbiosis and the Black Queen Hypothesis. The Black vantage in energetic or metabolic gain offered to taxa that can Queen Hypothesis was developed to accommodate the in- evolve to perform a given step; (ii) the genetic underpinning of creasingly common finding of natural microbial communities in such performance, for example close genetic linkage of the re- which genes necessary for life are distributed among cellular quired genes, facilitating simultaneous lateral gene transfer and lineages, such that some depend on the leaked metabolites of the formation of taxonomically and ecologically diverse “guilds”; others, seen as the result of a self-limiting process driven by se- (iii) sustainability, in not requiring abiotic components available lection for reduced genome size (34). This seems more surprising only in limiting supply; (iv) the range of physical conditions under when it involves separate cellular lineages rather than endosym- which the process can be implemented; and (v) ability to integrate biont and nuclear lineages in a single , but the evolutionary with other processes, either through interchangeable products or strategy is the same: a biochemical pathway remains in operation the availability or evolvability of taxa participating in multiple while the genes underwriting it are distributed among genomic processes. Comprising a cycle (rather than a linear pathway lineages that are henceforth bound to evolve together. The leading to the accumulation of an unmetabolizable end product) singers divide up the parts, but the song goes on (35). should also promote process persistence, and Life’s existence and With independent cellular lineages, one might speak of a continuity might depend on the prevalence and interconnected- “population pan-genome,” comprising all genes present in all ness of many biogeochemical cycles (24). Processes will change lineages at a given site and, as Fullmer et al. (36) suggest: over time as a consequence of changes (driven by ENS as usually understood) among the taxa that collectively implement them, ... seriously consider as the operative units in and changes that promote process persistence will preferentially which genes are selected ... rather than exclusively individ- accumulate among the survivors. To the extent that changes ual organisms. Similar to how Richard Dawkins advocated promoting persistence buy time for further persistence-promoting thinking of an organism as a collection of generally agree- changes, “complex adaptations” will be progressively elaborated able, but selfish, genes perhaps we should be thinking of lineages and populations as the collections of genes, i.e., (25–27). pan-genomes, rather than the individual cells. Interestingly, some investigators now treat the history and development of biogeochemical cycles in the way one would The problem here, again, is that so long as lineages and treat those of organismal lineages (28, 29), showing both stability populations behave as in our model with large p and n, Lewontin’s and change over time. Both are “complex adaptive ” recipe does not apply. ITSNTS does. exhibiting “aggregation” (hierarchical organization), “nonlinearity” (path-dependant evolvability), “diversity” (redundant singers), and Types, Tokens, and Genes. In an essay on types and tokens, “flows” (implementations of process) as described, for example, by Wetzel (37) uses Gertrude Stein’s “Rose is a rose is a rose is a rose” Levin (30). Their evolutionary trajectories are constrained by the as an example. If “word” is taken as a type, there are three words laws of physics and chemistry but also by their own prior histories: in this sentence, while as tokens there are 10. Whatever biological we do not know if Life’s tape were rewound how similar its pro- processes are made of, it’s not the same kind of stuff as the things cesses would be (31). (genes, cells, organisms, species) that implement them. The differ- ence seems like the philosophical one between types and tokens. Parallels to ITSNTS Dawkins (38), in his popular explication of emerging theory ITSNTS thus offers an alternative to luck and the anthropic prin- (The Selfish Gene), also had something like this in mind. His genes ciple (2) for rationalizing Life’s long tenure on this planet, recasting as “immortal coils” are informational types and their embodi- “fortuitous benefits” as persistence-promoting adaptations of ments in DNA are tokens. In the same vein and explicitly, Haig (39) processes, not of collectives of things. It links process evolution by writes of this way of thinking that: differential re-production to ENS through the differential repro- duction of process-implementing individual organisms, address- ... material genes were physical objects but informational ing deficiencies of claims about holobionts and attributions of genes were the abstract sequences of which material genes function to nonmicrobial ecosystems (see below). The theory were temporary vehicles. Material genes were identified with ... nevertheless has similarities, some only apparent but some quite gene tokens and informational genes with gene types Continuity resides in the recursive representation of immortal relevant, to other areas of biological theory and philosophy. pattern by ephemeral avatars (emphasis ours). Group Selection. Group selection—as in popular models of the For ITSNTS, too, the continuity of process “resides in the re- evolution of altruism within species (for example, see ref. 32)— cursive representation of immortal pattern by ephemeral avatars might be seen as relevant to the evolution of microbial collectives. (collectively implementing taxa)” (39). Arguably unlike genes, the For example, a minor modification of our model still consistent pattern as a type need never be represented in one place at one with standard ENS would allow that groups with more of species X time: the taxa implementing it (the singers) know only their own produce more individuals overall (of any species). The frequency parts and can, in the case of something like the nitrogen cycle, be of X individuals in pools would thus rise, as would the frequency of on opposite parts of the globe in different centuries (29). communities dominated by species X formed from them. But this Although the focus here is on community-level processes, it is would be a direct effect of group composition on reproduction of also possible to apply such reasoning to within-lineage de- individuals within groups, not on the differential reproduction of velopmental evolution at the organismal, cellular, and molecular groups as groups (33). By Lewontin’s recipe, or “William’s Princi- levels. Wagner (40) argues, for example, that “it is the historical ple,” only the latter can generate group-level adaptations (19). continuity of gene regulatory networks rather than the expression Interspecies , whether or not taken to be a kind of of individual homologous genes that underlies the homology of altruism, counts as merely a “fortuitous benefit” of group selec- morphological characters.” In a sense, preexisting developmental tion of this kind. processes (regulatory networks) are the types maintained by

Doolittle and Inkpen PNAS | April 17, 2018 | vol. 115 | no. 16 | 4009 Downloaded by guest on September 25, 2021 (persisting through) ENS. These processes recruit genes, and not differential reproduction of things (1, 2). Whether or not such se- always the same ones, needed for their implementation (as to- lection, function, and adaptation will prove to be important in kens), just as biogeochemical cycles recruit microbial taxa. explaining what we see in the world is a separate, if not un- interesting, matter. Similarly one might uncouple Darwin’s for- Memes. There are equally obvious parallels between ITSNTS and mulation of the principle that reiterated acts of selection and the notion of meme selection, and objections to it might take a amplification of variants will produce complexity [which Sober (43) similar form (4, 41). Dennett (42) recently reviewed and rebutted calls an a priori] from the empirically testable claim that this pro- some of these objections in the case of meme theory. He dis- cess explains biological adaptedness and diversity overall, or even misses out of hand the complaint that memes are not “real.” We any particular instance thereof. But ITSNTS does provide a spe- venture that this complaint is ontological, having to do with the cific mechanism (sketched in Fig. 1) for what might otherwise differing kinds of stuff of which types and tokens are made. It seem a mere logical necessity. seems incontrovertible that something like ITSNS happens in the For the cultural domain, Godfrey-Smith (4) insists that the exemplary case of words. Words as tokens do not reproduce, but means by which memes are reproduced must mimic organismal through the intermediary of speaking or writing (indeed re- reproduction (not entailing too many extraneous influences or, markably, through many media) they are re-produced, and as a especially, too many parents) for ENS to be the applicable model. consequence individual words persist as represented types. We Direct imitation will do it for him [as for the critic Sperber (41)], suggest that ITSNTS can similarly dismiss the reality concern: while more diffuse mixtures of group learning and idiosyncratic processes implemented by organisms are as real (or unreal) as creativity will not. For the evolution of microbial communities, words spoken by people. ITSNTS relaxes our understanding of what ENS requires, allowing Godfrey-Smith (4) articulates a second concern about mem- re-production to stand in for reproduction, and such a relaxation etics, writing: “As far as theory goes, recent work does establish may also make sense in the cultural context, on the condition that substantial links between cultural and biological [ENS] mecha- there remains a causal relationship between the properties of nisms. It gives us ‘how-possibly’ explanations for the maintenance populations over time, as noted below (44, 45). of various behaviours. Its empirical application—its ability to give us ‘how actually’ explanations—is another matter.” Major Transitions in Evolution. A longstanding problem with Indeed, even meme skeptics accept that a coherent selective Maynard Smith and Szathmàry’s “Major Transitions in Evolution” process can be imagined. The situation with ITSNTS seems simi- formalism, in for example its 1995 version (46), is that the transition lar. It is easier to argue that persistence is a legitimate target for from “Solitary individuals to colonies (non-reproductive classes)” ENS, in principle, than to show that this matters in practice. And is followed by “Primate societies to human societies (language),” the former was ITNTS’s initial purpose: to legitimize discussions of and these seem to be different in kind. The first (and several be- selection, adaptation, and function for community activities that fore it) are “transitions in individuality,” successive steps in the seem inexplicable to ENS, as long as this is focused on the hierarchy of life in which lower-level units are combined into

Fig. 1. A cartoon illustrating ITSNTS. Hand-drawn arrows are meant to represent the steps or components of a cyclical process with intermediate metabolites or stages A–D. Shaded symbols (rectangles, ellipses, and so forth) are microbial taxa comprising guilds that implement each step. The process or interaction, by existing, encourages the evolution of entities (taxa) that implement its steps, and these in turn reconstruct the process or interaction as a whole. The process thus persists, and changes in the mode of implementation that promote persistence are selected for. (As discussed briefly in the text, a similar scheme could be developed for developmental or regulatory interactions and the genes that carry them out.)

4010 | www.pnas.org/cgi/doi/10.1073/pnas.1722232115 Doolittle and Inkpen Downloaded by guest on September 25, 2021 higher-level collectives. Languages, however, are not comprised Rewriting Lewontin’s Recipe of the individuals that speak them, although each affects ENS of If ITSNTS is accepted as a form of ENS, then Lewontin’s three the other. ITSNTS thinking offers a way, analogous to gene- conditions, as commonly interpreted, are sufficient but in fact not culture coevolution (47), to understand interacting evolutionary necessary. We must allow that ENS can address differential per- process that are not nested, one within another, but comprise sistence as well as differential reproduction, and that re-production separate though interacting domains. can be how the former is underwritten. Arguments have been made for both elements of such a synthesis: ITSNTS integrates them. Process Ontology. Dupr ´e(48) argues that much of contemporary biology embraces a generally mechanistic and reductionist Differential Persistence and Differential Reproduction. Claims “thing” (or “substance”) ontology. According to such a meta- that ENS addresses “survival and reproduction” are common- physics, what the universe holds are material entities, while pro- place in textbooks and general folk understandings, but most cesses are secondary: just what happens to, or are performed by, evolutionists view differential reproduction (of alleles in particular) things. Conversely, process ontologists see processes as primary, as the fundamental feature of ENS, with survival (of individual things as their manifestations. For ITSNTS this would mean, in organisms) being of value only because it allows more time for effect, that taxa and the communities they form are adaptations of reproduction. Possibly, it is the contingent history of the modern the processes they implement, not the other way around. And of synthesis and population as its dominant formalism that course we, as multicellular individuals, are processes: few of our make us think like this. But as Bouchard (26) has pointed out: cells are “the same” as those with which we were born, and our ... identity is sustained throughout our lives by the continuity of persistence (viability, survival, etc.) is always in the picture. developmental and regenerative processes, not the or cells To put this a different way, differential persistence seems to be necessary to have evolution by natural selection while in our bodies at birth. Dupr ´emakes a similar case for sexual (not differential reproductive success seems to be a contingent asexual) species, linking that to the widely accepted notion that strategy for a lineage to survive. The irony is that this account such species are spatio-temporally continuous biological individ- was always in our face: Darwinism is usually understood, not uals, not essence-defined classes. Processes can be individuals, as “reproduction of the fittest” but as “survival of the fittest.” although not all phenomena we call processes (erosion, inflation, Indeed, and in line with a suggestion by Dennett (55) in a re- evolution itself) have the necessary coherence. ITSNTS envisions view of Godfrey-Smith’s 2009 book Darwinian Populations and that some community-level processes (which Dupr ´ewould call Natural Selection (56), ITSNTS theory might integrate persistence- “stabilized”) do have the necessary coherence. Some individual based and reproduction-based ENS as extremes on a seamless processes (the squid–vibrio interaction) will be relatively easy to spectrum. It is not so much that Nature holds two distinct kinds of circumscribe, while others (many biogeochemical cycles or the evolving entities in either the domain of things or the domain of metabolic interactions in our gut) will be highly interconnected. processes: there is a spectrum in both. But there are two distinct ways of thinking about variation-and-selection, one emphasizing ITSNTS and . There are especially obvious reproduction and the other persistence. “The nitrogen cycle,” for parallels to formulations of niche construction theory (49, 50), example, might be individuated as a single persistent process taken as an evolutionary process in which “organisms, through evolving by the ITSNTS mechanism, being re-produced rather their , their activities, and their choices, modify their than reproduced. But equally, one might regard nitrogen cycles own and/or each other’s niches” (50) in such a way as to affect performed at different temperatures or values of pH, or by dif- further ENS. By prioritizing processes as units of selection, ferent microbial guilds, as multiple different progeny of some accepting as adaptations characteristics that have promoted dif- original cycle whose reproductive success is thereby demon- ferential process persistence, and allowing re-production to stand strated. Reproduction in this case might be a gradual divergence, in for reproduction, ITSNTS may offer additional resistance to the as in some cases of . Note that for ITSNTS it would be skeptical view that niche construction is about nothing more than types (processes) and not tokens (implementations of processes “ ” extended phenotypes (51). Indeed, Haig (52) sees these as by things) whose persistence or reproduction is inferred. “ ‘ ’ usefully distinguished, writing that [t]he extended phenotype As Godfrey-Smith (4) stresses in his essay on cultural evolution, ‘ ’ and niche construction are sometimes presented as alternative the “grain” of the analysis matters. With fine-grained individua- labels for a single concept, but I prefer a division of labor in which tion (many types of nitrogen cycles, individual organisms in a the former term refers to differences under selection and the latter species) differential reproduction makes sense, but as the grain to evolved parts of the environment. From this perspective, gets coarser (nitrogen versus sulfur cycles, for example, or entire bodies and genomes are constructed niches that select among species or higher clades in the domain of things), differential genetic differences.” ITSNTS similarly admits that “evolution by persistence of processes becomes more satisfying as an expla- natural selection converts that which is selected (phenotype) into nation, and the role of reproduction as just one mechanism for that which selects (environment)” (52), but sees this conversion, promoting it becomes more obvious (25–27, 57–59). Having more too, as part of a larger process, selectable among alternatives for (or more diverse) progeny benefits individuals within a species persistence. As well, processes as patterns of interaction are not through differential reproduction but can benefit the species committed to any single type of material implementation: the through differential persistence (relative immunity to ). same “immortal pattern” can be realized in fundamentally distinct Similarly, having nitrogen cycles implemented in many places “ephemeral avatars.” Thus, we might allow bridging of awkward under many conditions favor the persistence of the cycle, transitions in retelling the , such as that between broadly construed. unicells and multicellular organisms, the RNA and DNA- As reproduction becomes persistence, the need for identifying worlds, and indeed prebiotic -independent chemical a relevant “Darwinian population” becomes less obvious: there is processes and the RNA world (53, 54). already a literature that makes this point (26, 27, 57–59). What

Doolittle and Inkpen PNAS | April 17, 2018 | vol. 115 | no. 16 | 4011 Downloaded by guest on September 25, 2021 ITNSTS has additionally to offer, we think, is a mechanism (evo- concepts also tied to recent discussions of the “health” of such lutionary recruitment) by which the processes that are re-produced systems, and about which there is arguably “an urgent need for a evolve, as do persistent clonal organisms (like aspen groves) or general framework” (64). Particularly germane is the long history clades above species (57). ENS indeed is about “survival and re- of attempts in community and ecosystem to discover and production,” but the relationship is more interesting than that justify the conditions for stable persistence or resilience of as- organisms that don’t survive won’treproduce. semblages through time (indeed, of complex adaptive systems as collectives more generally) (30, 65–67). Recent authors have ar- Re-Production and Reproduction. Persistence of a process of gued that these conditions involve at least functional redundancy, necessity entails continual or continuous re-production, and vice response diversity, spatial heterogeneity, and modularity (68). versa. Sperber (41) and other meme-skeptics complain of memes Such a move away from traditional explanations that treat com- that they don’t reproduce except when mindlessly imitated: munities as defined by a list of member species (as things) is at memes are otherwise re-produced by a diversity of influences and least implicit in trait-based (or “process-focused”) accounts of do not show parent–offspring lineages. This is true but not fatal to and functional , now a standard ap- ENS (for either memes or community ) as long as proach in literature on resilience (69, 70). Jax (71) writes that “In there is some causal connection between populations over time, a this type ... the processes are the very focus of the ’s relationship between implementations of a process now and in definition. In a conservation and management context, it the future such that the former can be seen as necessary for the is often the processes (and their bearers) selected as ecosystem latter. There are currently several competing formulations bearing services that are of concern; they guide the idea of what consti- on this possibility. Charbonneau (58) argues that population-level tutes a functioning ecosystem.” processes of ENS (“generation” and “memory”) do not require On the advantage of this approach, Gagic et al. (72) write that “local-level lineages.” Bourrat (59) shows that evolution for per- “[i]ndices solely based on the numbers and abundances of spe- sistence without reproduction is consistent with the Price equa- cies [i.e., taxonomic diversity measures] were consistently poor at tion. And De Monte and Rainey (53) claim that reproduction and predicting ecosystem functioning... Moreover, they performed “heritable variation in fitness” are unnecessary as long as gene- worse than indices using a trait-based approach, both in previous alogies of collectives are traceable. studies of plants ...and in our current analysis of animals.” One might also refer to re-production as recurrence (60), as is ITSNTS provides an evolutionary framework for rationalizing done for organs like the heart that are reconstructed during de- collective community properties—such as redundancy or diversity velopment but do not reproduce or generate parent–offspring of guilds—as pattern or process adaptations, rather than adap- lineages: I am my father’s son, but my heart is not the offspring of tations of communities themselves or mere fortuitous byproducts my father’s heart. But hearts are encoded in genomes that are of the “selfish” interactions and adaptations of their individual reproduced, while re-production in ITSNTS theory requires no members. Selection for the persistence (or stability) of these comparable single information repository nor any single lineage patterns as types is underwritten by their re-production as re- of such repositories unless, as very expanded forms of de- peatable tokens—by implementing species assemblages—such velopmental (61) or niche construction might hold, as the longleaf pine forest (73). ITSNTS thus rationalizes and re- one regards the larger and associated abiotic compo- defines ecological “community function,” potentially in one (or nents as that repository. both) of two ways (74, 75). Still, for ITSNTS there is a causal relationship between imple- The first is forward-looking, casting functions as contributions mentations of a process now and in the future such that the former to current or future persistence-promoting traits of processes, are in the long run necessary for (and determine the properties of) implemented by a current collection of taxa. Such an approach the latter: if a process dies, taxa that benefit from performing its was recently suggested by Dussault and Bouchard (76), who components likely will as well. That is, successive re-productions nevertheless see the community itself as adaptation-bearing. For are necessary for the persistence of a process in the same way that ITSNTS, if a property, such as being implementable by diverse continued reproduction of its members is necessary for the per- taxa under diverse conditions, contributes to the persistence of a sistence of a species or continual replacement of cells is necessary pattern, then it is ascribed a forward-looking function. Such for our survival as individual organisms. Godfrey-Smith (45) has functions are explanatory in the sense that they provide answers stressed that it is the complexity of any causal chain that deter- to “why?” questions about the contribution of pattern properties mines the appropriateness of “replicator” language in such situ- to observable pattern persistence. ations: “persistence language” may be more forgiving. A backward-looking account, equally consistent with ITSNTS thinking and perhaps more in line with something like William’s ITNSTS Beyond Microbial Communities Principle (19) can also be entertained. Such explanations answer a ITSNTS is, as Sober argues in the case of some other biological different explanatory question: they explain why the pattern has models, a priori true in the sense that no experiment could dis- the properties it has, and they do so by referring to the function of prove it (43, 62). The empirical question then is about range of those properties as past selected effects (77). The evolution of application and utility as explanation. One obvious extension patterns of interaction is premised on some patterns persisting beyond and is into traditional longer than others as a result of certain properties. The function of (animal and plant) community ecology and the field of a property, on this account, is that of its effects that contributed to biodiversity-ecosystem function research (63). Indeed, Turnbaugh the property’s maintenance under selection for persistence, re- et al. (20) (quoted above) likened the stability of microbial com- gardless of how the diversity of a pattern or process arose. Either munity functions across sites (guts) to the stability of functional or both of these ITSNTS understandings of function may prove types of nonmicrobial species across islands. helpful for current discussions of ecosystem functioning and ser- ITSNTS offers a sensible way of discussing adaptations and vices (71), or for helping to clearly delineate different explanatory functions relevant to communities and ecosystems generally, projects in different branches of ecology.

4012 | www.pnas.org/cgi/doi/10.1073/pnas.1722232115 Doolittle and Inkpen Downloaded by guest on September 25, 2021 Another point of connection between ITSNTS and ecological natural selection, was a significant (for him the principal) cause of theory are notions of niche-based environmental filtering and such diversification. What biologists have seen over the last cen- species sorting, which have been well-developed as part of the tury is an expansion of levels in the biological hierarchy, from theory of community assembly, both in microbial and non- stretches of “selfish DNA” to species, at which this law-like pro- microbial ecology (66, 78). These theories add rigor and sub- cess is thought to operate. We might also embrace ENS in the stance to our claim that “because there’s a song, there are cultural domain: surely dairying practices both explain and are singers” by explaining the process through which species are explained by the frequencies of lactose intolerance in dairying “selected” from larger pools to join as members of smaller com- and nondairying populations (81). But cultures are not just another munities, as in Fig. 1. level in the hierarchy, so this might be seen as bidirectional Finally, ITSNTS side-steps an important problem that has been “sideways,” rather than “upward” and “downward” causation (82). raised for ecosystem evolution and function: the inclusion of abiotic ITSNTS is similar and could play a similar role in explaining the components. Odenbaugh (79) has argued that a backward-looking prevalence of taxa and their activities. What we need is identifi- (selected effects) definition of function cannot apply to abiotic cation and understanding of factors that make some biologically components of ecosystems because those components are not mediated processes more persistent than others, and how these “ ” reproduced with the rest of the ecosystem. Heritability seems to affect and are affected by differential reproduction of taxa that be a stumbling block. More recently, Bouchard (80) has argued that implement them. This understanding could answer “why?” ques- the inclusion of abiotic components in ecosystems causes prob- tions that are left unanswered by Lewontin’s recipe, at whatever lems for ecosystem evolution in general, because, level it is focused, questions such as: “Why has nitrogen metabo- lism evolved towards completion as a cycle?” (29), “What drives the inclusion of abiotic material as parts of an evolving system genome reduction in cross-feeding microbial communities?” (34), makes the evolution of whole ecosystems a nonstarter for and “Why has life on this planet lasted long enough for us to be most evolutionary approaches: ecosystems do not have a ” unified genome, and even a meta-genome will leave out asking anything about it? (2). These are questions about biological essential parts of ecosystems, such as the abiotic factors that systems, generously defined, and search for commonalities across maintain much of the emergent properties of said ecosys- systems as addressed by systems biologists seems to require tems. Evolution as change in allelic frequencies does not some definition of the subject matter (83, 84). ITNTS provides seem to apply to systems that have a motley crew of alleles a framework in which biological systems are those “stabilized” and abiotic material interacting in a systematic way. (48) processes that can be individuated as units of selection ITSNTS provides a way of eliding these issues by focusing for persistence. Indeed, in a spirit of explanatory pluralism (85), attention instead on the processes that evolve, not on the ITSNTS encourages thinking about process/thing coevolution things (biotic or abiotic) that implement them. Reproduction is quite broadly. not required. Acknowledgments ITSNTS and Evolutionary Explanation We thank Austin Booth, Carlos Mariscal, Tyler Brunet, Scott McCain, Letitia Meynell, Richmond Campbell, Elliott Sober, and Andrew Hendry for comments Darwin argued, first, that the diversity of living organisms can be on earlier drafts. This work was supported by Grant GLDSU447989 from the understood historically and, second, that a law-like principle, Natural and Engineering Research Council of Canada.

1 Doolittle WF, Booth A (2017) It’s the song, not the singer: An exploration of holobiosis and evolutionary theory. Biol Philos 32:5–24. 2 Doolittle WF (2017) Darwinizing Gaia. J Theor Biol 434:11–19. 3 Lewontin RC (1985) . Annu Rev Genet 19:81–102. 4 Godfrey-Smith P (2012) Darwinism and cultural change. Philos Trans R Soc Lond B Biol Sci 367:2160–2170. 5 Zilber-Rosenberg I, Rosenberg E (2008) Role of in the evolution of animals and plans: The hologenome theory of evolution. FEMS Microbiol Rev 32:723–735. 6 Gilbert SF, Sapp J, Tauber AI (2012) A symbiotic view of life: We have never been individuals. Q Rev Biol 87:325–341. 7 Gordon J, Knowlton N, Relman DA, Rohwer F, Youle M (2013) Superorganisms and holobionts. Microbe 8:152–153. 8 Bordenstein SR, Theis KR (2015) Host biology in the light of the microbiome: Ten principles of holobionts and hologenomes. PLoS Biol 13:e1002226. 9 Lemanceau P, Blouin M, Muller D, Moënne-Loccoz Y (2017) Let the core microbiota be functional. Trends Plant Sci 22:583–595. 10 Vellend M (2010) Conceptual synthesis in community ecology. Q Rev Biol 85:183–206. 11 Ereshefsky M, Pedroso M (2015) Rethinking evolutionary individuality. Proc Natl Acad Sci USA 112:10126–10132. 12 Roughgarden J, Gilbert JF, Rosenberg E, Zilber-Rosenberg I, Lloyd EA (2017) Holobionts as units of selection and a model of their and evolution. Biol Theory 13:44–65. 13 Moran NA, Sloan DB (2015) The hologenome concept: Helpful or hollow? PLoS Biol 13:e1002311. 14 Douglas AE, Werren JH (2016) Holes in the hologenome: Why host-microbe symbioses are not holobionts. MBio 7:e02099. 15 Skillings D (2016) Holobionts and the ecology of organisms: Multi-species communities or integrated individuals. Biol Philos 31:875–892. 16 Clarke E (2016) Levels of selection in biofilms: Multispecies biofilms are not evolutionary individuals. Biol Philos 31:191–212. 17 Song SJ, et al. (2013) Cohabiting family members share microbiota with one another and with their dogs. eLife 2:e00458. 18 Godfrey-Smith P (2015) The ant and the steam engine. Lond Rev Books 37:18–20. 19 Sober E, Wilson DS (2011) Adaptation and natural selection revisited. J Evol Biol 24:462–468. 20 Turnbaugh PJ, et al. (2009) A core gut microbiome in obese and lean twins. Nature 457:480–484. 21 Cordero OX, Polz MF (2014) Explaining microbial genomic diversity in light of . Nat Rev Microbiol 12:263–273. 22 Louca S, et al. (2016) High taxonomic variability despite stable functional structure across microbial communities. Nat Ecol Evol 1:15. 23 Nyholm SV, McFall-Ngai MJ (2004) The winnowing: Establishing the squid-vibrio symbiosis. Nat Rev Microbiol 2:632–642. 24 Falkowski PG, Fenchel T, Delong EF (2008) The microbial engines that drive Earth’s biogeochemical cycles. Science 320:1034–1039. 25 Doolittle WF (2013) Natural selection through survival alone, and the possibility of Gaia. Biol Philos 29:415–423. 26 Bouchard F (2011) Darwinism without populations: A more inclusive understanding of the “survival of the fittest”. Stud Hist Philos Biol Biomed Sci 42:106–114.

Doolittle and Inkpen PNAS | April 17, 2018 | vol. 115 | no. 16 | 4013 Downloaded by guest on September 25, 2021 27 Bouchard F (2008) Causal processes, fitness, and the differential persistence of lineages. Philos Sci 75:560–570. 28 Braakman R, Follows MJ, Chisholm SW (2017) Metabolic evolution and the self-organization of ecosystems. Proc Natl Acad Sci USA 114:E3091–E3100. 29 Klotz MG, Stein LY (2008) Nitrifier and evolution of the nitrogen cycle. FEMS Microbiol Lett 278:146–156. 30 Levin SA (1998) Ecosystems and the biosphere as complex adaptive systems. Ecosystems 1:431–436. 31 Falkowski P (2015) Life’s Engines: How Microbes Made Earth Habitable (Princeton Univ Press, Princeton). 32 Sober E, Wilson DS (1999) Unto Others: The Evolution and Psychology of Unselfish Behavior (Harvard Univ Press, Cambridge, MA). 33 Okasha S (2006) Evolution and the Levels of Selection (Oxford Univ Press, Oxford). 34 Morris JJ, Lenski RE, Zinser ER (2012) The Black Queen Hypothesis: Evolution of dependencies through adaptive gene loss. MBio 3:e00036-12. 35 Estrela S, Kerr B, Morris JJ (2016) Transitions in individuality through symbiosis. Curr Opin Microbiol 31:191–198. 36 Fullmer MS, Soucy SM, Gogarten JP (2015) The pan-genome as a shared genomic resource: Mutual cheating, cooperation and the black queen hypothesis. Front Microbiol 6:728. 37 Wetzel L (2014) Types and tokens. The Stanford Encyclopedia of Philosophy, ed Zalta EN, Spring 2014 Ed. Available at https://plato.stanford.edu/archives/ spr2014/entries/types-tokens/. Accessed December 22, 2017. 38 Dawkins R (1976) The Selfish Gene (Oxford Univ Press, Oxford). 39 Haig D (2014) Fighting the good cause: Meaning, purpose, difference, and choice. Biol Philos 29:675–697. 40 Wagner GP (2007) The developmental genetics of homology. Nat Rev Genet 8:473–479. 41 Sperber D (2000) An objection to the memetic approach to culture. Darwinizing Culture: The Status of Memetics as a Science, ed Aunger R (Oxford Univ Press, Oxford), pp 143–162. 42 Dennett D (2017) From Bacteria to Bach and Back: The Evolution of Minds (W. W. Norton, New York). 43 Sober E (2011) A priori causal models of natural selection. Australas J Philos 89:571–589. 44 Claidière N, Scott-Phillips TC, Sperber D (2014) How Darwinian is cultural evolution? Philos Trans R Soc Lond B Biol Sci 369:20130368. 45 Godfrey-Smith P (2000) The replicator in retrospect. Biol Philos 15:403–423. 46 Szathm ´aryE, Smith JM (1995) The major evolutionary transitions. Nature 374:227–232. 47 Richerson PJ, Boyd R, Henrich J (2010) Colloquium paper: Gene-culture coevolution in the age of genomics. Proc Natl Acad Sci USA 107:8985–8992. 48 Dupr ´eJ (2017) The metaphysics of evolution. Interface Focus 7:20160148. 49 Laland K, Matthews B, Feldman MW (2016) An introduction to niche construction theory. Evol Ecol 30:191–202. 50 Odling-Smee FJ, Laland KN, Feldman MW (2003) Niche Construction: The Neglected Process in Evolution (Princeton Univ Press, Princeton). 51 Wells DA (2015) The extended phenotype(s): A comparison with niche construction theory. Biol Philos 30:547–567. 52 Haig D (2017) The extended reach of the selfish gene. Evol Anthropol 26:95–97. 53 De Monte S, Rainey PB (2014) Nascent multicellular life and the of individuality. J Biosci 39:237–248. 54 Pross A (2013) The evolutionary origin of biological function and complexity. J Mol Evol 76:185–191. 55 Dennett DC (2011) Homunculi rule: Reflections on Darwinian populations and natural selection by Peter Godfrey Smith. Biol Philos 26:475–488. 56 Godfrey-Smith P (2009) Darwinian Populations and Natural Selection (Oxford Univ Press, Oxford). 57 Doolittle WF (2017) Making the most of clade selection. Philos Sci 84:275–295. 58 Charbonneau M (2014) Populations without reproduction. Philos Sci 81:727–740. 59 Bourrat P (2015) How to read “heritability” in the recipe approach to natural selection. Br J Philos Sci 66:883–903. 60 Godfrey-Smith P (2015) Reproduction, symbiosis, and the eukaryotic cell. Proc Natl Acad Sci USA 112:10120–10125. 61 Griffiths PE, Gray RD (1994) Developmental systems and evolutionary explanation. J Philos 91:277–304. 62 McCain K, Weslake B (2013) Evolutionary theory and the epistemology of science. The Philosophy of Biology, ed Kampourakis K (Springer, Dordrecht, The Netherlands), pp 101–119. 63 Naeem S, Duffy JE, Zavaleta E (2012) The functions of biological diversity in an age of extinction. Science 336:1401–1406. 64 Palmer MA, Febria CM (2012) Ecology. The heartbeat of ecosystems. Science 336:1393–1394. 65 Diamond JM (1975) Assembly of species communities. Ecology and Evolution of Communities, eds Cody ML, Diamond JM (Harvard Univ Press, Cambridge, MA), pp 342–444. 66 Weiher E, et al. (2011) Advances, challenges and a developing synthesis of ecological community assembly theory. Philos Trans R Soc Lond B Biol Sci 366:2403–2413. 67 Folke C, et al. (2004) Regime shifts, resilience, and biodiversity in . Annu Rev Ecol Evol Syst 35:557–581. 68 Desjardins E, Barker G, Lindo Z, Dieleman C, Dussault AC (2015) Promoting resilience. Q Rev Biol 90:147–165. 69 McGill BJ, Enquist BJ, Weiher E, Westoby M (2006) Rebuilding community ecology from functional traits. Trends Ecol Evol 21:178–185. 70 Violle C, Reich PB, Pacala SW, Enquist BJ, Kattge J (2014) The emergence and promise of functional biogeography. Proc Natl Acad Sci USA 111:13690–13696. 71 Jax K (2010) Ecosystem Functioning (Cambridge Univ Press, Cambridge, UK). 72 Gagic V, et al. (2015) Functional identity and diversity of animals predict ecosystem functioning better than species-based indices. Proc Biol Sci 282:20142620. 73 Steen DA, Barrett KE, Guyer C (2017) Conceptualizing communities as natural entities: A philosophical argument with basic and applied implications. Biol Philos 32:1019–1034. 74 Mitchell S (1993) Dispositions or etiologies? A comment on Bigelow and Pargetter. J Philos 90:249–259. 75 Godfrey-Smith P (1994) A modern history theory of functions. Nousˆ 28:344–362. 76 Dussault A, Bouchard F (2017) A persistence enhancing propensity account of ecological function to explain ecosystem evolution. Synthese 194:1115–1145. 77 Garson J (2017) A generalized selected effects theory of function. Philos Sci 84:523–543. 78 Robinson CJ, Bohannan BJ, Young VB (2010) From structure to function: The ecology of host-associated microbial communities. Microbiol Mol Biol Rev 74:453–476. 79 Odenbaugh J (2010) On the very idea of an ecosystem. New Waves in Metaphysics, ed Hazlett A (Palgrave, London). 80 Bouchard F (2014) Ecosystem evolution is about variation and persistence, not populations and reproduction. Biol Theory 9:382–391. 81 Laland KN, Odling-Smee J, Myles S (2010) How culture shaped the human genome: Bringing genetics and the human sciences together. Nat Rev Genet 11:137–148. 82 Jablonski D (2008) Species selection: Theory and data. Annu Rev Ecol Evol Syst 39:501–524. 83 Bizzarri M, Palombo A, Cucina A (2013) Theoretical aspects of . Prog Biophys Mol Biol 112:33–43. 84 O’Malley MA, Dupre ´ J (2005) Fundamental issues in systems biology. BioEssays 27:1270–1276. 85 Sterelny K (1996) Explanatory pluralism in . Biol Philos 11:193–214.

4014 | www.pnas.org/cgi/doi/10.1073/pnas.1722232115 Doolittle and Inkpen Downloaded by guest on September 25, 2021