Natural selection acts upon the entire range of hereditary variation in a given population. There are, however, certain requirements to the nature of this variation. If these requirements are met, we can consider natural selection as an important mechanism of changing the modality of the genotypic composition of the population (to be entirely sure about its effectiveness, one should also analyse the importance of “system filters” – see section 3).

There are two major requirements. Firstly, the variation should be to more or less “homogenous” (or isotropic 53). This means that it should be characterized as unlimited, potentially representing a broad range of variants. The manifestations of hereditary deviations should be distributed more or less evenly across the axes of characters’ variation. The second requirement is as follows: the primary act of hereditary variation should not depend on the need in the act and the nature of this need^*.

*Note: In essence, these requirements have been formulated by Meyen and Chaikovsky 80, p. 12, paragraphs 5, 6 and 9 in their review of the ideas of A.A. Lyubishchev concerning selectogenesis.

If both these requirements are met, we may be fairly sure that the subsequent changes in the frequencies of characters indeed can be, in principle, be formed by natural selection. On the other hand, if these requirements are not met, we should rethink the leading role of natural selection in microevolutionary events.

For instance, the “non-homogeneity” of the character space (the first requirement) — in other words, canalized nature of hereditary variation, considerable differences in the frequency of manifestation of the hereditary variants, the discreteness of forms in the hereditary variants — testifies to the importance of the “construction-determined” evolutionary mechanisms and to the prevalence of orthogenesis (Box 1). The greater “non-homogeneity” is observed, the more important orthogenesis is (Figure 1). The role of natural selection is reduced to eliminating non-viable variants and the work within the range of the origin of forms, which islimited and predetermined by the variation 8121753707985.

Failure to fulfil the second requirement spells the acknowledgement of direct adaptogenesis (see Box 1. Three types of evolutionary concepts). Direct adaptogenesis occurs when the impact of the environment on the individual is accompanied by the formation of the hereditary variants of the progeny that are, in general, consistent with the impact (inheritance of acquired characters). This situation is essentially different from that envisaged by the natural selection model, where adaptations are formed indirectly, by selection from the available variants (indirect adaptogenesis).

What are, in fact, the manifestations of the hereditary variation? Are we in the position to say whether the two requirements are fulfilled and to provide the rationale of the effectiveness of natural selection?

The most general approach to the assessment of the “homogeneity” of the character space is the construction of potentially possible (mathematically calculated) spaces of certain characters. Within this approach, the calculated “morphological space” (“morphospace”) of the potential variation of a character is compared with that actually implemented in a certain species of organisms, group of species or group of organisms united into taxonomical categories of a higher rank. In all these cases it has been shown that only a minor proportion of the potential diversity is actually implemented in the nature 187578100101102113118119121. This discrepancy between the potential and the actual diversity may be an argument in favour of essential structural constraints of the origin of forms(that is, the primary “non-homogeneity” of variation). At the same time, this assessment is not necessarily incompatible with a different viewpoint: the initial variation is reasonably “homogenous” and the subsequent non-homogeneity is the result of the mechanism of differential mortality, that is, natural selection.

The analysis of hereditary variation of characters in closely related species allows one to make more definite conclusions. The range of variation and particular characters turn out to be rather conservative in species of the same genus and even the same family. Manifestation of the same range of forms in the variation of different species may concern both particular morphological characters and complexes of correlated characters representing certain “morphotypes”. This phenomenon is best described by Vavilov’s law: the law of homologous series of hereditary variation 123. It emphasizes, firstly, the similarity of series of variation in closely related species and genera (the similarity being directly proportional to the relatedness of the taxa under comparison) and, secondly, the similarity of cycles of variation in subordinate taxa within the category of a higher rank (such as a family). Noteworthy, discussing the formation of such parallelisms of variation, Vavilov admitted the possibility of a similar direction of the mutation process in closely related species (see 123). Homologous series reflect the nomothetic nature of variation: it is not stochastic but subject to general rules. Such ideas are incompatible with the initial hypothesis about the “homogeneity” of variation. Nomothetic nature of variation is discussed in detail by Vasil’yev and Vasil’yeva 122 and widely discussed in modern literature (see, for example 85385).

The nomothetic nature of variation at the level of macrotaxa is strongly supported by paleontological data on the evolutionary formation of large taxonomic groups of organisms: archaeocyathids 104, arthropods 95,96, mammals 117, echinoderms 105 and birds 71. Limitations applying to the origin of forms and numerous parallelisms indicate that Vavilov’s law of homologous series is expressed at the scale of evolution of marcotaxa 106. Therefore, the “non-homogeneity” of variation may play a pivotal role in directed character of the evolution.

Ideas about limitations of variation find an even more general expression in the concept of repeating polymorphic sets 78,113. The emphasis here is on the distributions of the modalities of separate characters (rather than character sets as in Vavilov’s law) and their similarity (“isomorphism” according to 78) in different taxa. Moreover, the notion of repeating polymorphic sets falls outside the scope of phylogenetic homology. Their repetition “…can be observed in obviously non-homologous parts …” (78, p.165). This is equivalent to an outright acceptance of the existence of some laws of variation and origin of forms that are not subject to “phyletic” relationships. From this viewpoint, the potential role of natural selection in origin of forms is reduced to secondary corrections.

An important argument in favour of a considerable “non-homogeneity” of the space of hereditary variation is the wide distribution of agamic, and parthenogenetic species among protists and multicellular organisms 3944107. Their morpho-functional distinctness is not associated with any special mechanisms uniting their gene pool and at the same time isolating it from other genetic entities. The very existence of distinct species in such organisms is an evidence of the discreteness of the origin of forms and its canalized nature.

Another argument jeopardizing the hypothesis about the “homogeneity” of the space of hereditary variation is provided by the existence of numerous so-called cryptic species. These are genetically separate groups endowed with all the prerequisites for morpho-functional divergence on the basis of their genetic separateness. The scale of genetic differences in such groups may vary; in fact, they are often more than sufficient for these species to be considered as reliable “biological species” 16. However, despite the genetic divergence, the representatives of these groups are characterized by a conservative structure and an extreme paucity of morpho-anatomical differences. Cryptic species occur in a variety of taxonomic groups of eukaryotes, from protists to multicellular animals 44, for a review of multicellular animals see 120. The importance of “non-visual signals” for the establishment of reproductive isolation and the stabilizing selection for a certain morphotype have been used to explain the evolutionary formation of such complexes of species 16. However, the problem of their morphological conservatism remains. Actually, under conditions of an acquired (no matter how) reproductive isolation there should be an extremely effective mechanism of maintaining the conservativeness of morphological characters (stabilizing selection?), also explaining an impressive intraspecific plasticity in some species (directional selection at the same time ?) 4897108 (“anti-cryptic selection”?) 16. The explanations of this phenomenon from the point of view of limitations in the origin of forms, the “non-homogeneity” of character space, are more parsimonious.

Several attempts have been made to assess directly variation in organisms at the early stages of the ontogenesis (that is, before the time when variants may be eliminated by natural selection). An example of such an analysis is the study of variation of postcranial skeleton (spinal column, sacrum, pectoral and pelvic girdle, limbs) of tailless amphibians 585960. The results indicate that а) the range of actual variation is considerably more narrow than that of the theoretically calculated potential variation; it is represented by series of discrete variants; b) there exist “prohibited” variants, which are never implemented, and the set of such variants is different in different species; c) under extreme developmental conditions the proportion of anomalous individuals increases but the general set of variants remains the same. Therefore, both the “non-homogeneity” of the character space and the influence of environmental conditions on the frequency of actual anomalous variants is observed.

In many instances hereditary variation is characterized by a considerable discreteness of the manifestation of the entire complexes of characters in an organism 2, 17. A theoretical basis of these observations can be found in the ideas about correlative connections in an organism and the canalization of the ontogenetic ways 109,125.

Selectogenetic model of evolution is based on the idea that potential hereditary changes are in principle unlimited and multidirectional. However, ideas about systemic organization of genomes 3451112128 leave little space for interpreting molecular mechanisms underlying hereditary variation as stochastic processes. Failures of matrix processing in prokaryotes and eukaryotes are mitigated by an array of reparation systems 265098. These systems, in turn, can be fine-tuned and regulated, resulting in regulated changes in the level of mutational variation 31,94. The activity of horizontal transfer in prokaryotes is associated with the systems of DNA reparation. Activation of stress-induced mutagenesis leads to mobilization of conjugative elements, which are used as vectors of genes responsible for antibiotic resistance 7,43.

Much of hereditary variation is associated with the transfer of mobile genetic elements in the genome of different species 32832545673. Their transfer is often directional, sometimes involves large numbers of elements and may result in quite definite hereditary changes 103.

An impressive array of epigenetic mechanisms involved into the formation of variation provides an insight into the molecular basis of hereditary changes. At the same time, the work of epigenetic systems is essentially non-stochastic. It is a reflection of a system of negative feedbacks in the genome in the narrow sense of the word (that is, a set of species-specific sequences of DNA bases) and other molecular components of the cell associated with the regulation of the protein-synthesis apparatus, the dynamic chromatin structure, “protein” heredity and “small RNA” systems (see a series of reviews in Epigenetics 27).

Ideas about the systemic organization of the genetic-epigenetic mechanisms are reflected in the conclusion that the reconstructions of the hereditary apparatus comply to the external influences. It is not so much the matter of correlations between the impact of the environment and the frequency of hereditary changes (see above, regulation of reparation systems) but rather the compliance of molecular reconstructions of the genetic-epigenetic system to this impact. Examples of this compliance are a higher level of mutations in those bacterial sites whose products are involved in the potential compensation of the external influence 41, 42; amplification of functional areas of genome associated with the resistance to chemicals in protists 15; specific inserts into genome CRISPR cassettes in bacteria ensuring resistance to bacteriophages 555691 similar mechanisms in eukaryotes associated with the activity of piRNA and siRNA 3,52; paramutations, whose manifestation also seems to be associated with the mechanisms of RNA interference 47. All this prompts one to revisit the idea of the direct inheritance of adaptive molecular changes at the epigenetic and even the genome level in the light of molecular mechanisms 34555665.

It has been shown that at least in some cases variation is directional and its directionality is sometimes induced (by environmental impacts). Recalling the two major requirements to hereditary variation, we can make the following conclusion: observations on intra- and interspecies variation point to a considerable “non-homogeneity” of the variation implemented in the character space. The fact that this non-homogeneity, rather than being a post hoc result of differential mortality, is associated with the character of variation is proved by: a) the presence of general trends, variation series; b) direct assessments of variation at early stages of ontogenesis; c) the analysis of molecular mechanisms underlying hereditary variation; d) the possibility of direct observation (in case of laboratory models for the study of variation). Moreover, it can be considered as a fact that the frequency of hereditary reconstructions increases under the influence of external factors. There are also indications of correlations between the character of the impacting factor and the type of hereditary reconstructions. The structure of variation is “non-homogenous” in its manifestations at the level of organisms and considerably “non-homogenous” in manifestations of hereditary changes at the level of genome and the associated epigenetic processes.

To conclude, at the imaginary axis of the “homogeneity” of hereditary variation (from entirely random to strictly canalized) its modal values are evidently confined to the area of consistent expression of characters. This entails a conclusion that orthogenetic mechanisms of the formation of variation predominate while natural selection plays a secondary role. Its importance diminishes even further in the context of data indicating that hereditary variation may conform to certain environmental impacts.

Further analysis would focus on the possibility of natural selection “using” hereditary variation. Let us imagine a change in the genotype of an individual as a signal which should correlate with the ability of this individual to make a greater or a smaller contribution into the gene pool of the next generation. The character of variation is no longer of interest to us; the emphasis will be on the consequences of its indisputable presence.

Variation determines the range of the morpho-functional variants of individuals of a given population or species. We will consider each change in the genotype structure as a “signal”, which is to be reflected in the characteristics of the individual (regardless of the nature of the inherited reconstructions, be they gene mutations, chromosomal reconstructions, changes associated with ploidy, reconstructions in the course of crossing-over or recombination variation of combinations during sexual process). In the long run, these characteristics should be reflected in the differential contribution of individuals into the gene pool of the next generation. From now on, we will ignore the changes that occur in the genotype but are not manifested, by definition, in the morphological and functional characters of an individual (for instance, synonymous base substitutions).

The “hereditary signal” becomes involved in interactions representing various hierarchical levels of the systemic organization of living nature. Starting at the suborganismic level, the interactions ascend to the biocenotic one. Each level is characterized by specific mechanisms of maintaining systemic integrity; the systemic integrity of the next level comprises all subordinate mechanisms but cannot be reduced to their sum. These holistic notions are reflected in the concept of emergence and can be aphoristically expressed as “the whole is bigger than the sum of its parts“ (68; see 22 for historical context and development of the notions of emergence into the Synergism Hypothesis).

Hereafter the term “system filter” will be used to denote the system of feedbacks and compensatory regulations at each of the hierarchical systemic levels and the passage of the hereditary “signal” through a series of “system filters” will be considered. Special attention will be given to the stability of the “signal” since only stable passage through the filters may result in the implementation of the natural selection mechanism.