Units of evolution are those things that evolve; a unit of evolution is “a lineage (an ancestral-descendant sequence of populations) evolving separately from others and with its own unitary evolutionary role and tendencies” (Simpson 1960; in Hull 1976, 184). On this received view, a species is a unit of evolution composed of a temporal sequence of populations, in which each member of the population is a spatiotemporal part of the species and a unit of selection (Hull 1976). Evolution by natural selection occurs in a species-level unit of evolution when the members of the populations are characterized by variation (V), heredity (H), and differential fitness (F) (Lewontin 1970).
Stegenga (2014) shows that evolution by natural selection can occur in populations whose members belong to different species, as heterogeneous populations can also be characterized by V, H, and F. In this paper we extend Stegenga’s arguments for population pluralism by considering evolution by natural selection in populations constituted by symbiotic collectives. Over the last several decades, a number of biologists and philosophers have argued that multi-species symbiotic collectives function as units of selection (Margulis and Fester 1991), (Mindel 1992), (Zilber‐Rosenberg and Rosenberg 2008), (Dupré and O’Malley 2012), (Booth 2014), though little has been said about the populations from which these multispecies collectives are selected or the lineages such heterogeneous populations form. Others have argued that multispecies collectives cannot be units if selection, because they are not lineage-generating (Godfrey-Smith 2009, 2012), (Doolittle and Booth, forthcoming), (Skillings, forthcoming). In this paper we focus on a specific kind of multispecies collective, a tree coupled with a symbiotic fungus attached to its roots, and we argue that such collectives are in fact lineage-generating, at least to the extent that male animals are, so if the lineage-generating criterion does not prohibit men from functioning as units of selection, then it likewise does not prohibit a tree/fungus symbiotic collective from functioning as a unit of selection.
Proponents of the lineage-generating criterion on units of selection note that symbiotic collectives often fail to reproduce as a single unit, such that the offspring of coupled symbionts can find themselves uncoupled in future generations. In the case of a mushroom/tree collective, the lineages of both symbionts frequently do remain coupled for multiple generations. An acorn does not fall very far from the tree, so the ground where an acorn lands and sprouts is likely occupied by the same mycelium already colonizing the mother tree. In fact, mother trees are known to transfer nutrients to daughter trees via a mycelium connecting the roots of both daughter and parent. Such nutrient transfer is crucial in mature forests where young trees must spend their early years in the shade of their parents (Gorzelak 2015).
However, the offspring of mushroom/tree pairs can find themselves decoupled. Wind can carry mushroom spores many miles, and squirrels are known to move acorns around the forest. Even though the offspring of a mushroom must find some arboreal partner to continue its own lineage, and even though if a tree is to thrive it must find some mycorrhizal partner, the next generation of trees and mushrooms need not be paired with direct descendants of the parent’s symbiotic partner – symbiotic lineages often diverge at reproduction. When Godfrey-Smith and others deny that a symbiotic individual such as a mushroom/tree collective is lineage-generating, they mean that the lineages of the symbiotic partners diverge at reproduction; lineages of symbiotic partners cohere while they are engaged in symbiosis, but at reproduction, the two lineages cease to be cohesive.
Lack of cohesion in transgenerational symbiotic lineages is what prompts Godfrey-Smith and others to deny that symbiotic collectives can be units of selection, but cohesion is not a binary condition; transgenerational lineage cohesion comes in degrees, such that some transgenerational lineages are more cohesive than others. We argue that the lineages generated by populations of symbiotic collectives can be cohesive to the same degree as sexually reproducing monospecific population lineages, and so these multispecies population lineages can function as units of evolution.
A human male is a paradigm example of a unit of selection. A man who is a successful hunter, who slays his enemies in battle, and who wins the affection of women, is likely to leave more offspring than a man who fails at these tasks, so a man is a unit of selection, but a man does not pass on to his children all of his DNA. Analogous to the way lineages diverge when a mushroom/tree symbiotic collective reproduces, a man’s nuclear DNA and his mitochondrial DNA are separated at reproduction, such that only the nuclear DNA forms a lineage extending into the next generation. The man’s nuclear and mitochondrial DNA lineages are cohesive during his lifetime, but they diverge when the man reproduces, just as the DNA lineages of a tree and its mycorrhiza are cohesive for a time, but diverge when the tree and the mushroom reproduce. Lack of cohesion of constituent DNA lineages through reproduction does not prohibit a man from being a unit of selection, nor does it prohibit the human species from being a unit of evolution, as sexual recombination eventual brings divergent nuclear and mitochondrial lineages back together. Sexual reproduction requires the splitting of organism-level lineages, such that constituent DNA lineages cease to be cohesive at reproduction, but at the same time it affects cohesion of population-level lineages through sexual recombination in subsequent generations.
Reproduction in populations of symbiotic collectives is analogous to reproduction in sexual species, in that while reproduction splits lineages, causing a lack of cohesion when particular symbiotic collectives reproduce, the formation of new symbiotic collectives affects cohesion in the multispecies population lineage, as constituent symbiont lineages come back together in future generations. Symbiotic recombination is thus analogous to sexual recombination in the way it splits and merges organism-level lineages in a population. It follows that both symbiotic collectives and monospecific organisms can function as units of selection, and both monospecific population lineages and multispecies population lineages can function as units of evolution.