The endpoint of a successful sexual act, particularly as considered from a microbial evolution perspective, can be viewed as the movement of genetic material from the cytoplasm, nucleus, or virion capsid from one individual such that it becomes stably incorporation into the genetic material of a recipient individual. At this point the now hybrid sexual product or products can be subject to natural selection. This movement can be horizontal (between otherwise independent individuals) as well as vertical (in the course of the transfer of genetic material from parent to offspring). Note that other processes of movement of genetic material from one individual to another exist which are less strictly sexual since genetic material sourced from different parents does not (necessarily) end up into the same cell. These other processes include especially the movement of whole organisms, such as symbionts between hosts, to form an organism's microbiome, though there too the movement can be distinguished into horizontal versus vertical. Overall this gene exchange impacts genetic variation, the efficiency of natural selection, the potential for the coevolution of specific alleles, the nature of genetic drift, etc. In the table that follows I provide an overview of various ways that sex can impact biology such as of microorganisms.

Table: Phenomena Impacted by Gene Exchange.

Species cohesion In truly clonal organisms, natural selection acting on individual organisms in combination with mutation prevalence is what constrains variability, with each lineage otherwise free to fully explore sequence space. This results in at least a potential for profound divergence between lineages because reproductive barriers between those lineages exist by default. To the degree that gene exchange both occurs and is potentially beneficial to organisms, however, then the maintenance of sequence similarity can foster such gene exchange. That is, with increasing sequence similarity, at least to a point, there can be an increasing potential to realize benefits that may arise from especially homologous recombination. At the same time, gene exchange can help to increase sequence similarity, and this is particularly so given selection for products of homologous recombination and thereby greater potential for survival and fixation of those products should they occur. Thus, species cohesion, that is, genotypic similarity among members making up the same population, can both foster gene exchange (i.e., increase the potential for homologous recombination to occur) and be a consequence of having homologous-recombination-associated gene exchange within species be beneficial to recipient organisms (i.e., increase the likelihood of survival of more similar versus less similar genotypes). These ideas are taken to an extreme by organisms that are obligately sexual reproducers, where the primary benefit of sex is reproduction itself, but with this benefit most readily realized only given high degrees of sequence homology between mating individuals.
Biological species concept The biological species concept is the idea that an ability to successfully engage in sexual interactions implies the placement of individuals within the same species. This comes with the caveat that successful participation in sex is defined in terms of outcomes, i.e., reproductive success of the resulting sexual products (such as progeny), rather than in terms of the sexual act itself. We can define these outcomes along a continuum, however, ranging from a high potential for sexual interactions to result in reproductively successful offspring and/or hybrid products at one end (as seen within species of obligately sexually reproducing organisms), a lower potential to result in reproductively successful hybrid products in the middle (as seen with non-obligately sexual but nonetheless products of homologous recombination, particularly so-called orthologous replacements), and very low potential (i.e., as potentially contributing to what can be described as a "zone of paralogy") or indeed no potential at all at the other extreme. (Orthologous replacement is the swapping of one homologous gene for another where the position of a gene on a chromosome is not otherwise changed whereas "zone of paralogy" refers to acquisition by genomes of alleles in novel locations, i.e., as particularly equivalent to insertion events.) It would be at some point of potential for success that would define two organisms as members of the same biological species, e.g., >90% of the time, >50% of time, etc., with the former the case for obligately sexual organisms and the latter in some cases perhaps for products of orthologous replacement.
Breaking up of epistatic interactions between coevolved genes This can be viewed as a negative consequence of sex. Two genes, or more precisely their alleles, are free to become highly epistatically integrated (coevolved) so long as reasonable assurance exists that the genes/alleles will not be separated via sexual processes. Highly sexual organisms, such as those that are obligately sexual, avoid this concern by severely constraining the "gene pool" within which they engage in regular sexual interactions. Particularly, hybridizations that results in reduced offspring fitness, such as due to separation of coevolved genes and/or the interaction of non-coevolved genes, are selective against, resulting in increasing levels of selection against such matings the greater the fraction of a parent's offspring that are so affected. In obligately sexual, particularly monogamous, low-fecundity pairings, such low fitness offspring can be disastrous. Alternatively, in clonal organisms sex can be sufficiently infrequent that fitness costs due to separation of coevolved genes are for the most part irrelevant to the parent, though presumably not irrelevant to the resulting low-fitness hybrid progeny.
Loss of genotypes that have stood the test of time Though not necessarily epistatically integrated, nonetheless multiple beneficial alleles can be found within a single individual but with the survival of these allele combinations threatened by sexual interactions. An example of such genotype losses can be seen with hybrid seeds such as you can purchase for gardening. These are generated in a manner that results in favorable allelic combinations. By definition, however, the resulting hybrid plants are unable to "breed true", which is another way of saying that sex eliminates existing genotypes. As with the breakup of coevolved genes, this is much less of an issue the lower the prevalence or importance of sex to a lineage, with mostly clonal organisms displaying a low likelihood that genotypes that have stood the test of time will be lost from a population, even if recombinant offspring are severely selected against, since relatively few offspring in fact will be recombinant. With obligately sexual organisms, on the other hand, all genotypes no matter how successful are lost from the population every generation.
Avoidance of Muller's ratchet Without sex, wild-type – that is, deleterious-mutation-free genotypes – can be lost from populations with little potential for recovery. This is more true given smaller populations as well as populations possessing per-individual mutations rates that are relatively high since with this combination of mutation rates and low population sizes then mutation-free genotypes should be less likely and therefore more likely to be lost to genetic drift. The idea in particular is that wild-type genotypes can be lost from populations without corresponding loss of wild-type alleles from the same population, with those alleles instead distributed among many individuals rather than found within a single, that is, wild-type genotype. Because all not-mutated alleles within a population that is being subject to Muller’s ratchet should remain within the population even if not within a single individual, with sex non-wild-type individuals that possess not-mutated alleles at complementary loci can mate and thereby potentially restore the wild-type genotype to a population. Note that these issues are less applicable to populations that possess substantial amounts of neutral variation in part because a population possessing large amounts neutral variation is possible only given low levels of genetic drift but also because the key issue with Muller’s ratchet is the existence of genotypes that lack (or relatively lack) detrimental mutations rather than simply that possess an otherwise invariant wild-type genotype.
Increased genetic variability This perhaps more correctly can be stated as an increased amount of genotypic variability across populations. The basic idea is that given sex then mutations that originate in different individuals can find their way into the same individual, ideally creating individuals containing greater numbers of beneficial alleles or, alternatively, bringing together alleles that by chance can have synergistically beneficial interactions. Alternatively, and essentially equivalently (as well as considered in more detail below), this increased genetic variability should create individuals possessing greater number of deleterious alleles as well as bring together alleles that by chance synergistically interact, but in ways that cause an excessive loss of fitness. Just as the former should allow for more efficient selection for those beneficial alleles that happen to be found within the same genotype, the latter should allow for more efficient selection against those deleterious alleles that also happen to be found in the same genotype (a concept which as noted will be addressed again immediately below). Which specific genotypes are subject to such selection, however, should change with each new recombination event.
Bringing together of alleles otherwise found in different lineages This is an expansion of the immediately previous entry though with the intention of bringing the concept of genetic migration into the conversation, that is, horizontal gene transfer and/or introgression. Another way of stating this is that alleles that are substantially evolutionarily separated can also be brought together by sexual processes, versus alleles that are found in separate individuals but nonetheless are found within what more precisely can be described as the same population.
Efficient removal of deleterious alleles from populations Just as beneficial alleles can be brought together into the same individual via sexual processes, so too as noted above can deleterious alleles be brought together into the same individual. The result is that deleterious alleles can be more efficiently removed from populations, i.e., by the removal of single individuals who by chance happen to have a larger than average concentration of these alleles. Furthermore, this process can operate more efficiently if negative synergistic interactions, essentially a kind of epistasis, occur between alleles such that multiple deleterious allele carriers are even less fit than one would expect based upon the fitness impacts of these alleles as measured individually.
Avoidance of clonal interference In clonal populations it is genotypes that are competing whereas individual alleles are able to compete among themselves with lower efficiency. The result is that different beneficial alleles found in different individuals can display similar fitnesses, making it difficult for any one of those alleles to rise to fixation. The pace of adaptation accumulation as a consequence can be hindered, i.e., the sequential acquisition of individual beneficial alleles by entire populations tends to not occur within populations that are both finite and purely clonal. Alternatively, these arguments can depend on how one defines "population" since these multiple competing genotypes are, of course, not sexually interacting so therefore arguably together do not make up a population so such as consist of multiple individual lineages. The genetic diversity that results from clonal interference on its face a good thing, however, since the population may subsequently be faced with environmental challenges that are better met by some genotypes versus others. Should clonal interference break down, however, such as following strong selection for a single allelic variant, then that diversity in fact can be wiped out through the fixation of a single genotype.
Avoidance of periodic selection Periodic selection represents the triumph of avoidance of disruption, via sexual processes, of well-adapted genotypes. I use the term "triumph" because the end point of successful periodic selection is fixation of that genotype, including all of its alleles, within a population (i.e., as associated with a selective sweep/clonal expansion). The cost of this process is extinction, or at least marginalization, of those alleles that don't happen to have been present in that now-fixed genotype. These other alleles can include ones which potentially could have provided the now-fixed genotype with additional benefits. Sex thus can be viewed as an important route towards retaining allelic variability and particularly beneficial alleles within populations.
Increased rates of adaptation Related to the idea of both retaining beneficial alleles within populations and being able to bring those alleles into the same individuals, sex can foster more efficient adaptive evolution. This is another way of saying that given sex then rates of adaptive evolution that are limited by mutation rates can take advantage of the mutation rates of entire populations, the products of which can then be brought together sexually. This is rather than each individual within populations generating each beneficial alleles independently. Furthermore, given highly promiscuous mating – that is, as resulting in horizontal gene transfer – then multiple, coevolved genes can move between lineages to allow for an even more dramatic building up of adaptations, e.g., such as acquisition of the mitochondrial endosymbiont which brought cellular respiration into an otherwise obligately fermentative lineage.
Disruption of coevolved selfish gene complexes Sexual processes, especially as occur with more obligately sexual organisms, can be problematic in terms of their potential to break up coevolved and thereby beneficial allele combinations. Alternatively, those same processes presumably can limit the evolution of coevolved but detrimental allele combinations. Thus, at a minimum, such alleles are forced to be linked together into relatively small patches of genomes, rather than be spread across genomes as otherwise unintegrated wholes which would be more easily disrupted sexually. Similarly, in less sexual lineages, the potential for these complexes to move between lineages is a function of the extent to which they are physically integrated, that is, with their component genes linked together. While this requirement for linkage can be seen as an impediment to movement of beneficial allele combinations (i.e., a "bad" thing), the same impediments also limit the movement of detrimental allele combinations (a "good" thing). Plasmids and perhaps especially phages, for example, though they are seemingly self-contained they nonetheless have host ranges that are limited as a consequence of reliance on genes to which they otherwise are not linked, such host polymerases. With viruses, in general, though, that reliance tends to be reduced the larger the viral genome, where the possession of a larger viral genome is another way of saying that more genes are linked together to form a genomically contiguous, potentially detrimental entity. In addition, the more genes that can move as coherent wholes then the greater the capacity to move between lineages without disrupting epistatic interactions.