Journal and Proceedings of The Royal Society of New South Wales Volume 120 Parts 1 and 2 [Issued September, 1987]




старонка4/7
Дата канвертавання21.04.2016
Памер105.81 Kb.
1   2   3   4   5   6   7

WHEELS WITHIN WHEELS: SELECTION AT MANY LEVELS


Selection is the only evolutionary force resulting in adaptation, because only selection discriminates between the mutations that constantly rush like a gale through the genome. But selection can take place at many levels, and adaptation at one level can take place at the expense of adaptation at another.

While the levels at which selection can occur are many (Table 1), we can certainly fix our ideas by noting that the levels of the individual, above the individual, and within the individual satisfactorily include them all!




TABLE 1
LEVEL OF SELECTION

Levels of selection during genetic evolution, determined on the basis of which entities are interactors (see text). With the possible exception of ecosystems, entities higher than the species are not interactors. Genera, for example, may evolve through selection within or between their constituent species, but are not themselves interactors. Other levels of selection may occur without affecting genetic evolution: for example, competition between cell types within a metazoan or metaphyte is unlikely to lead to an increase in the frequency of the winners’ genes next generation.

  1. Direct selection on DNA sequences

    1. Selection for the number of copies

      1. Sequences capable of infective transmission

      2. Sequences incapable of infective transmission (highly repreated sequences e.g. Alu)

    2. Selection between aleles (meiotic drive)

  2. Selection on genotypes

    1. Selection on haploid phases (bacteria, gametes of higher eukaryotes)

    2. Selection on diploid phases

      1. Cells within organisms
        (likely to be only rarely converted to evolutionary change)

      2. “Normal” complete higher organisms

  3. Supra-individual selection within species

    1. Intra-demic group selection

      1. Groups of non-relatives

      2. Groups of relatives (=kin-selection)

    2. Inter-demic group selection
      (this is the classical meaning of grouip selection)

  4. Selection at the species level

    1. Non-random speciation
      (bias in characteristics of new species)

    2. Non-random distribution of speciation rates (species with more optimal values of some characteristic speciate at a higher rate than others)

    3. Non-random distribution of extinction probabilities (species with more optimal values are less likely to become extinct)

  5. Selection at the ecosystem level (Wilson, 1976). Assemblages of species differ in their persistence, thus favoring “good mixers” among their constituent species
It is worthwhile pursuing the levels-of-selection approach a little further. Selection may take place at different levels in two ways: by acting at different levels of complexity in the physical hierarchy of life (as displayed in Table 1), or by acting at both the genetic and behavioral levels (of which more later). With regard to the hierarchical approach, we can turn to Hull (1980), who built on ideas enunciated by Dawkins (1976), and distinguish between replicators and interactors. To paraphrase Hull:

A replicator is an entity that passes on its structure directly in replication.

An interactor is an entity that interacts directly with its environment.

An interactor is made up of one or more replicators, usually very many. Selection acts by affecting the relative success of different interactors in passing on the replicators of which they are composed. The properties of interactors are of course formed by more than their replicator makeup, with environment and history playing their parts too. Selection is thus effective in causing evolution only to the extent that fitness differences between interactors mirror differences in their replicator makeup.

Levels of selection are thus determined by the level of interactor. But how can this level be recognised? It is not enough to look for average fitness differences between interactors because, for example, two populations may differ in average fitness only because one is made up of fitter individuals than the other.

I suggest that the best criterion for determining the level at which selection is acting in any one case is to discover which is the lowest level allowing a complete formal description of the situation. Thus, if population 1 is made up only of AA individuals, which are of higher fitness than the BB individuals of which population 2 is exclusively composed, then fitness differences between genotypes (level 2(b) in Table 1) is the appropriate level, and higher levels such as 3(a) or 3(b) are not usefully invoked.

Interactors above the gene level are usually not replicators as well, because their reproduction involves their dissolution into their constituent parts: when you reproduce, you do not pass on your genetic endowment intact but rattier shuffled subsets of your genes packaged as gametes.

Higher taxonomic categories such as genera and families cannot therefore be units of selection, because they are not interactors. It is also problematic whether they “evolve” as some authors consider they do (e.g., Arnold and Fristrup, 1982), because they lack objective reality, being simply assemblages of species set up to facilitate cataloging. But 11 would be justifiable to consider a lineage evolving, because that can be objectively defined, in principle, and genera and families that are defined strictly on lineage lines could be said to evolve. Such lineage-defined taxa would then be roughly analogous to populations, which evolve even where selection is strictly at the level of the individuals within them.

There are exceptions to the rule that only genes are replicators. Individuals reproducing asexually and without meiosis do replicate their genetic makeup exactly, and lichens, which are simple communities made up of particular combinations of fungal and algal species, reproduce by dispersing fragments that form new colonies elsewhere.

1   2   3   4   5   6   7


База данных защищена авторским правом ©shkola.of.by 2016
звярнуцца да адміністрацыі

    Галоўная старонка