1. Introduction
This article critically examines whether integrative pluralism can successfully explain holobiont individuality. Integrative pluralism is epistemically motivated: it posits multiple theoretical approaches to explain complex entities (Mitchell Reference Mitchell2003; Brigandt Reference Brigandt2010).Footnote 1 I will assess its applicability to holobionts, which are complex associations of various organisms belonging to different species. I argue that integrative pluralism has problematic implications, which complicate its ability to capture holobiont cases. I will focus on puzzles concerning the evolutionary individuality status of holobionts in order to demonstrate that claim. But first I briefly introduce the notion of holobionts and the problem one faces when considering integrative pluralism in that case.
‘Holobiont’ denotes an obligatory multispecies symbiotic relationship observed between an individual host and its microbial community. A holobiont is not merely a collection of organisms. Rather, holobionts are composed of various biological entities, which together satisfy conditions for being a distinct biological entity—a biological individual (Roughgarden et al. Reference Roughgarden, Gilbert, Rosenberg, Zilber-Rosenberg and Lloyd2018). A pluralist attitude toward biological individuality has already been both epistemologically and metaphysically motivated (Love and Brigandt Reference Love, Brigandt, Lidgard and Nyhart2017). The case of holobionts contributes an additional dimension to pluralism: if multispecies holobionts satisfy individuality conditions, then there are multispecies individuals (Gilbert, Sapp, and Tauber Reference Gilbert, Sapp and Tauber2012). I suggest that single-species and multispecies individuality accounts both generate successful explanations for holobionts, depending on the context. Indeed, I argue that partners within holobiont associations are single-species individuals; however, the holobiont complex including both partners is itself a multispecies individual. This means that depending on the focus (i.e., partners or the entire complex) different individuality explanations are required (i.e., single species or multispecies).
As we will see, integrative pluralism is unable to capture the above scenario. Integrative pluralism does authorize the use of different theoretical accounts. It does not allow the application of multiple models to concrete cases without integration (Mitchell Reference Mitchell2003, 189). I briefly introduce integrative pluralism in section 2. In section 3, I argue that integrative pluralism’s viability is questionable because holobiont cases challenge the integration of single-species and multispecies individuality accounts. Thereafter, I discuss two outcomes of my critique in section 4.
2. Integrative Pluralism in a Nutshell
Integrative pluralism employs three assumptions. The first explains why scientific explanations require a pluralistic framework. Since various causal factors are responsible for complex phenomena, pluralism is needed to address those different factors. The second assumption suggests that a successful explanation for a complex case requires generating an integrative explanation. Because different theoretical approaches focus on distinct causal factors, they need to be integrated. The third assumption states that true integration, which promotes a successful explanation, is attained by means of pragmatic aims. There are two implications that can be drawn from integrative pluralism. They show that integrative pluralism is too restrictive: it cannot capture the case of holobionts. After discussing those implications, I demonstrate the restrictiveness of integrative pluralism through a particular holobiont case.
2.1. The First Implication: The Most Successful Explanations Are Integrative Ones
Mitchell (Reference Mitchell2003, Reference Mitchell2009) indicates that a limited number of causal factors contribute to the existence of a particular biological phenomenon. Accordingly, she explains that different sorts of factors should be taken seriously during a scientific investigation. In addition, Mitchell endorses that multiple research questions remain unconnected to one another in the absence of integration. As we learn that these factors are interconnected, she suggests that those various models need to be integrated. For instance, for depression, the genetic model explains the relationship between depression and the genome, whereas the environmental model of depression explains the relationship between the environment and depression. Both the genome and the environment are causal factors in depression. Here one sees that the most successful explanation of depression is an integrative one. Thus, a satisfactory explanation of depression requires integrating these different explanations (Mitchell Reference Mitchell2009, 105–7). Mitchell states that pragmatic considerations help us focus on a problem. Since different problems are the products of different pragmatic concerns, and since different problems require developing different epistemic techniques, pragmatic concerns and cognitive capacities must collaborate with each other (114–15).
According to Mitchell (Reference Mitchell2003), the existence of a plurality of idealized models does not entail a plurality of applications of those models (189). Indeed, her argument depends on the distinction between the idealized models and their specifications to concrete cases. While scientific models generate idealized explanations of some causal factors, she argues that the applications of those models require compatible claims (207) because “however complex, and however many contributing causes participated, there is only one causal history” (216). Therefore, Mitchell suggests that an integration of applications provides “one true integrated explanation” for each specific concrete and complex case (217).
Mitchell does not assert that all explanations need to be integrative. She mentions that reductive explanations can be helpful in some situations (Mitchell Reference Mitchell2009, 116–17). Yet, Mitchell’s analysis implicitly entails that the most successful explanation for a complex case is either mostly integrative or at least reductionist in the sense that nonintegrative explanations are less desirable. For instance, she states that “pluralism with respect to models can and should coexist with integration in the generation of explanations of complex and varied biological phenomena” (Mitchell Reference Mitchell2003, 218; emphasis added). Hence, the first implication is that a successful explanation of complex phenomena demands an integration of multiple models. At first glance this might seem important for covering complex cases like holobionts; however, as we will see, that is not the case.
2.2. The Second Implication: Explanations of Evolutionary Novelties Require Integration
Ingo Brigandt adopts integrative pluralism to investigate the evolutionary origin of novelties. He uses the distinction between a problem and a problem agenda to stress the importance of integration among epistemic units (e.g., methods, concepts, and models). A problem agenda refers to a set of interrelated questions that involve various epistemic units. He suggests that all problem agendas are associated with some criterion of adequacy. The criterion of adequacy provides an epistemological evaluation to decide whether appropriate epistemic units are equipped to solve the problem at hand.
In Brigandt’s analysis, satisfactory explanations for problem agendas call for the integration of multiple epistemic units. For example, Brigandt thinks that explanations concerning evolutionary novelties form a problem agenda. Hence, he argues that the “complete explanation” of an evolutionary novelty requires integration of epistemic units collected from multiple areas, such as phylogeny, paleontology, ecology, biogeography, and functional morphology (Brigandt Reference Brigandt2010, 299–302).
Like for Mitchell, Brigandt’s explanatory integration is not universal but case by case. For instance, Brigandt rejects pluralism as a general or regulative ideal because he suggests that increasing the number of methods, models, theories, and so on, would not help solve complex problems (Reference Brigandt2010, 304). Nevertheless, his pluralistic approach implies that integration is necessary for some problem agendas, such as his case of the evolutionary origin of novelties. For example, he indicates that “the nature of the biological problem considered here [the explanation of evolutionary novelties] requires the integration of different approaches, involving biological fields such as developmental biology and morphology that traditionally study entities above the cellular level” (302).
In summary, according to Mitchell, integrative pluralism implies that successful explanations of complex phenomena require integration. According to Brigandt, integration is necessary for cases that include sets of interrelated questions—the origins of evolutionary novelty being one of them. Since holobiont individuality is a complex phenomenon, which often includes the emergence of evolutionary novelty at the level of the holobiont complex, prima facie it appears to need an integrative approach. However, in what follows, I show how integrative pluralism is too restrictive for the case of holobiont individuality.
3. Shades of Holobiont Individuality
Holobionts are complex organisms consisting of multiple partners from different species. Although organismality is not a sufficient condition for biological individuality, some holobionts are examples of biological individuals (Godfrey-Smith Reference Godfrey-Smith, Bouchard and Huneman2013). In addition, holobiont cases point out the problems of traditional evaluations of individuality (Gilbert et al. Reference Gilbert, Sapp and Tauber2012). I suggest that explanations of holobiont individuality appeal to two theoretical approaches. These approaches can be introduced by two models of individuality. There is the single-species individual model and the multispecies individual model. However, their success as explanatory approaches is not achieved through integration but rather through contextual application. Consider the following definitions:
Single-Species Individuality Model (SSI): An individual is formed by an organism that belongs to one species.
Multispecies Individuality Model (MSI): An individual is formed by at least two organisms of different species.
The partners of a holobiont are considered separate single-species individuals; however, the entire holobiont complex is a united multispecies individual. From an evolutionary point of view, for example, the host and its symbionts are one multispecies individual when a beneficial phenotype for the whole system is produced by the contribution of both partners (Gilbert et al. Reference Gilbert, Sapp and Tauber2012; Lloyd Reference Lloyd, Gissis, Lamm and Shavit2018). But they are separate single-species individuals when there is an evolutionary conflict between them (Douglas and Werren Reference Douglas and Werren2016).
In order to understand holobiont individuality, both models are required but not in an integrative fashion. Their application will rather depend on the context. I explore the aphid-Buchnera holobiont as a case study in order to show how integrative pluralism does not provide a proper guide to address aphid-Buchnera’s individuality. Specifically, I focus on the question: Are aphid-Buchnera holobionts evolutionary individuals? Evolutionary individuals are biological individuals that serve as units of natural selection. I argue that in order to establish evolutionary individuality status, both SSI and MSI are required but not in an integrated way.
3.1. Motivating Pluralism as Contextual Dependence with Aphid-Buchnera Holobionts
As previously mentioned, the partners that make up a holobiont are single-species individuals, and holobionts themselves are multispecies individuals. While holobionts are good candidates for evolutionary individuality, their evolutionary status is revealed context dependently.
One way to show that holobionts are evolutionary individuals is to consider how phenotypic variation at the level of a holobiont complex can only be acquired by a genetic variation in its partner symbionts (Bosch and Miller Reference Bosch and Miller2016, 70). Since natural selection acts on phenotypic variations, there are new “holobiont traits” caused by interactions among partners in the holobiont association, but the partners lack those traits individually (Császár et al. Reference Császár, Ralph, Frankham, Berkelmans and van Oppen2010). Some holobionts include a containment: the host physically contains its symbiont partner that is vertically transmitted the host’s offspring. As a result, the host and symbiont have intertwined reproductive histories, and together the host and the symbiont partner form the holobiont complex. Sometimes the host needs the inherited symbionts to survive, and the symbionts need the host for its survival, such that the relationship is obligatory. The aphid-Buchnera holobiont illustrates that kind of survival relation among its partners.
Acyrthosiphon pisum is a sap-sucking pea aphid and Buchnera aphidicola is a bacterium that lives inside those insects. In this symbiotic relationship, the pea aphid provides essential nutrients to the bacteria, and the bacteria provide essential amino acids to the pea aphid. An essential point is that these bacteria are maternally inherited. Thus, from generation to generation, the symbiotic relationship between pea aphid and Buchnera aphidicola is established (Skaljac Reference Skaljac and Vilcinskas2016). In addition, neither pea aphids nor Buchnera bacteria can survive without the help of each other. It follows that the symbiosis between a pea aphid and its Buchnera bacteria forms a single aphid-Buchnera holobiont that includes members belonging to different species (Bosch and Miller Reference Bosch and Miller2016).
According to Peter Godfrey-Smith, aphid-Buchnera holobionts are good candidates for evolutionary individuality. Importantly, evolutionary individuals must transmit traits with fidelity to offspring. Those heritable traits must make a difference to the individual’s fitness in order for populations of individuals to evolve. In other words, evolutionary individuals—or Darwinian individuals in Godfrey-Smith’s model—must reproduce in a particular way. He thinks that the aphid-Buchnera holobiont satisfies the condition of reproduction. The obligatory relationship means both partners share the same evolutionary fate: if one fails, then the other does. Holobiont associations can be created vertically or horizontally—through a parent-offspring fashion or from the environment. When the hosts vertically transmit their symbiont partners to offspring, just as aphids do with their Buchnera bacteria, they both share the same evolutionary fate. Given that the symbiotic relationship among pea aphids and Buchnera aphidicola is established, generations of aphid and generations of bacteria depend on each other. Therefore, according to Godfrey-Smith’s model, aphid-Buchnera holobionts are evolutionary individuals (Godfrey-Smith Reference Godfrey-Smith, Bouchard and Huneman2013, 30–32).
However, not all aphid-Buchnera holobionts are evolutionary individuals. To evaluate the fitness of the host organism (i.e., the aphid) and the fitness of its symbionts (i.e., the Buchnera bacteria) requires considering them separately in some contexts. The ibpA heat-shock protein increases the fitness of aphid-Buchnera holobiont in hot conditions, and it is harmful in cool conditions. So a single-base mutation in the Buchnera ibpA gene decreases the fecundity of aphids in extremely hot conditions but increases the fecundity of aphids in cool conditions (Dunbar et al. Reference Dunbar, Wilson, Ferguson and Moran2007). Hence, “the ibpA promoter mutation of Buchnera is expected to undergo either positive or negative selection at the level of aphid hosts, depending on the thermal conditions encountered by a population” (Burke et al. Reference Burke, McLaughlin, Simon and Moran2010, 367). Furthermore, there can be coevolutionary conflict between aphids and Buchnera. If a Buchnera mutant reproduces too many offspring, an abnormal abundance of bacteria can lead to the existence of selfish symbiont mutations (Chong and Moran Reference Chong and Moran2016). This means that the same evolutionary fate is not always shared in every context. Therefore, the partners of holobionts are different evolutionary individuals in those situations.
The aphid-Buchnera case points out that whether a holobiont is an evolutionary individual is context dependent. Next, I turn to how explanations about holobiont individuals appeal to two models separately, depending on the context: the SSI and the MSI. In order to assess whether holobiont individuality is evolutionary, these models need not be integrated but rather applied contextually.
3.2. Pluralism without Integration
Recall the first implication of integrative pluralism from section 2: a true integration of the applications of multiple models generates the most successful explanations. Holobiont individuality requires two models, namely, the SSI and the MSI. However, the integration of those models does not apply in the case of aphid-Buchnera holobionts.
In an aphid-Buchnera population, I discussed how some holobionts synthesize the heat-shock protein ibpA, but some others do not. Although the ibpA increases the fitness of aphid-Buchnera holobionts in hot conditions, it is lethal in cool conditions. In addition, although a single-base mutation in the Buchnera ibpA gene decreases the fecundity of aphids in extremely hot conditions, the same mutation increases the fecundity of aphids in cool conditions (Dunbar et al. Reference Dunbar, Wilson, Ferguson and Moran2007). Therefore, it is context dependent whether a particular holobiont is an evolutionary individual. For example, for the case of a single-base mutation, the contexts include either hot conditions or cold ones: Buchnera only contribute to the fitness of the whole aphid-Buchnera system in cool conditions. As such, the holobiont including both aphid and Buchnera partners is an evolutionary individual only in cool conditions. Therefore, MSI only applies to aphid-Buchnera holobionts in cool condition contexts.
However, in the context of hot conditions where a single-base mutation in the obligate symbiont Buchnera is detrimental to their aphid host’s fitness, it is better to consider the partners as separate individuals. That is, each of the aphid and the Buchnera bacteria are better considered as separate, single species individuals because of their dissimilar evolutionary fates. Therefore, SSI only applies to each partner in hot condition contexts. So, the application of SSI or MSI models depends on contextual factors, even for the very same holobiont relationship. These models are useful tools; however, intertwined interactions among a host, its symbionts, and their environment require the employment of those models on a contextual basis.Footnote 2
Recall that according to integrative pluralism, best explanations should integrate several models. Integrative pluralism is compatible with the claim that there can be different idealized models holding incompatible assumptions, but it rejects that applications of those models yield incompatible results. For instance, Mitchell writes that “while the idealized and abstract character of models allows compatibility at the theoretical level, the realistic and concrete nature of explanation entails integration and resolution” (Reference Mitchell2003, 89; emphasis added). However, in thermotolerance research of the aphid-Buchnera holobiont, the differences in fitness concerning hot and cool conditions suggest that the application of SSI and MSI introduces conflicting results.
While SSI holds the idea that the aphid and Buchnera are two separate evolutionary individuals, MSI holds the idea that the aphid and Buchnera form one multispecies evolutionary individual. Since SSI and MSI involve two incompatible applications concerning holobiont individuality, the integration of their applications is not a feasible option. Not all complex biological cases are best explained by integrating multiple approaches because models can yield conflicting results. It is worth mentioning that holobionts are pervasive biological organizations rather than exceptional entities. For instance, all animals need their symbionts to pursue their biological viability (McFall-Ngai Reference McFall-Ngai2014), and holobionts occupy the entire Eukarya domain (Bosch and Miller Reference Bosch and Miller2016). Therefore, holobionts are not exceptional biological organizations, and they have an important place in biological research.
So integrative pluralism is too restrictive because it entails the idea that the best explanations are generated by integration of the applications of models. However, a nonintegrative pluralism still informs us that in some cases aphid-Buchnera holobionts are evolutionary individuals, while in other cases some holobionts are not. Their evolutionary individuality status is context dependent.
3.3. Explanation without Integration
Recall the second implication of integrative pluralism from section 2: explanations of evolutionary novelties require explanatory integration. However, this is also too restrictive for the case of holobionts. The presence of novel traits at the level of holobiont complexes is important for their status as evolutionary individuals because individuals in selection require varying, heritable traits that make a difference to their fitness. And so, a novel trait can make that fitness difference and is something we should expect to see if indeed holobionts are evolutionary individuals. The thermotolerance trait may just be an evolutionary novelty, so how do we investigate its origin?
Brigandt suggests that explaining the evolutionary origin of novel structures necessarily demands an integration among several epistemic units, such as theories, concepts, and models. For instance, he says that “accounting for the evolutionary origin of novelties also makes it necessary to integrate different theoretical models and modes of explanations” (Brigandt Reference Brigandt2010, 304; emphasis added). In addition, according to Brigandt, the criteria of explanatory adequacy shape the proper integration for successful explanations (304). For example, he thinks the criteria of explanatory adequacy for evolutionary novelties concern multiple factors such as “addressing the causes that change the internal properties of an individual” and “focusing on how variation across individuals causes evolutionary changes” (304–5).
According to Brigandt, an evolutionary novelty “is a qualitatively new morphological structure or function feature in a group of organisms that did not exist in an ancestral species” (Reference Brigandt2010, 299). If the thermotolerance ability of aphid-Buchnera holobionts is an evolutionary novelty, it is a new trait at the level of the holobiont complex that neither partner possesses alone, yet it makes a fitness difference. There are other cases that include evolutionary novelty at the level of the holobiont complexes, in addition to the aphid-Buchnera case. For instance, chemosynthetic symbioses between bacteria and several animal phyla “create holobionts with novel metabolic capabilities” (Rosenberg and Zilber-Rosenberg Reference Rosenberg and Zilber-Rosenberg2014, 60). Furthermore, the digestion system of cows is generated by the help of microorganisms as a novel evolutionary trait (Margulis and Sagan Reference Margulis and Sagan2008) and the squid-Vibrio holobiont gains a new light organ as a result of symbiotic relations with Vibrio bacteria (Booth Reference Booth2014).
All the mentioned examples require accepting the whole holobiont as an evolutionary individual. Explanations concerning holobiont individuality involve two approaches to individuality. The partners of a holobiont are considered separate single-species individuals; however, the entire holobiont complex is a united multispecies individual. A population can include both single-species and multispecies individuals. Hence, both SSI and MSI are required. If holobiont traits are evolutionary novelties, their origin cannot be explained by integrating single-species models and multispecies models. Because, in some contexts, an evolutionary novelty benefits the holobiont, and in other contexts it does not. To detect that, we need to employ SSI and MSI models separately.
In addition, Brigandt thinks that the criteria for the explanatory adequacy of evolutionary novelties show that successful explanations need to focus on relevant factors, such as variation across individuals. Yet, some factors can require separating epistemic models, such as SSI and MSI. For instance, focusing on variation across individuals in holobiont populations involves two theoretical approaches that are not integrated. For example, the thermotolerance case of aphids shows that the same trait can be a novelty or not according to context. It would be an evolutionary novelty for some holobionts but not some others. Therefore, the idea that explanations of evolutionary novelties require explanatory integration is too restrictive in light of intertwined interactions among a host, its symbionts, and a holobiont’s environment. Thus, integrative pluralism cannot explain the evolutionary origin of novelties when it comes to holobionts.
4. Concluding Reflections on Integrative Pluralism
I have argued that integrative pluralism is too restrictive in the case of holobionts. Integrative pluralism entails that successful explanations of complex cases and explanations of evolutionary novelties require the integration of models. On the contrary, I have discussed how SSI and MSI explain holobiont complexes and their evolutionary novelties in a nonintegrated, context-dependent account. There is no resolution between the applications of those models: they yield incompatible results by classifying holobionts as evolutionary individuals in some contexts and not others.
Overall, holobiont research reveals Mitchell’s and Brigandt’s versions of integrative pluralism to be unfruitful concerning the evolutionary individuality of holobionts. While integrative pluralism might be applicable in some cases, Mitchell’s analysis of integration implies that the best biological explanations are integrative ones for complex and concrete biological cases. However, I have shown that in complex cases like holobionts, the best explanations are context dependent, not integrative. And the application of SSI and MSI leads to incompatible results that prevent the compatibility needed for integration. Additionally, Brigandt’s analysis of novelty implies that explanations for the origin of evolutionary novelties require integration. However, some explanations for the origin of evolutionary novelties in holobionts require using SSI and MSI models separately. Therefore, establishing the evolutionary individuality status of holobionts requires a nonintegrative pluralism that can accommodate both single-species individuals and multispecies individuals.
The negative lesson of this discussion is that integrative pluralism provides an inadequate starting point to address holobiont evolutionary individuality. Nevertheless, there is one positive lesson. Integrative pluralism emphasizes an important problem: both Mitchell and Brigandt argue that multiple theoretical approaches should be used together and be connected to attain a satisfactory explanation for complex cases. Holobiont individuality shows that multiple theoretical approaches can be used together but without integration. Perhaps, then, we should be more pluralistic about the sort of pluralisms used to approach cases in biology. A liberal pluralism recognizes various pluralistic approaches and that some pluralisms may not apply to all cases.
A more liberal pluralistic view increases our understanding of holobiont research and evolutionary individuality. Liberal pluralism is pluralistic about pluralisms: one needs to be liberal concerning the relationship among theoretical approaches, scientific models, and epistemic units. That relationship need not always be integrative, even for the same scientific problem. The case of holobionts can help us develop a better account of scientific pluralism: although integrative pluralism may not apply to holobiont cases, a contextual pluralism—one in which explanatory models are applied or not according to conditions—does capture holobiont evolutionary individuality. The lesson for liberal pluralists is not to restrict approaches to only one kind of pluralism.
To sum up, I suggest two points. First, a pluralistic view helps us better understand holobiont research and biological individuality. Second, the case of holobionts can help us develop a better account of scientific pluralism.
