[Informative predation: Towards a new species concept].

We distinguish two types of predations: the predation of matter-energy equals the food chain, and the informative predation is the capture of the information brought by the sexual partners. The cell or parent consumes energy and matter to grow, multiply and produce offspring. A fixed amount of resources is divided by the number of organisms, so individual growth and numerical multiplication are limited by depletion resources of the environment. Inversely, fertilization does not destroy information, but instead produces news. The information is multiplied by the number of partners and children, since each fertilization gives rise to a new genome following a combinatorial process that continues without exhaustion. The egg does not swallow the sperm to feed, but exchange good food for quality information. With the discovery of sex, that is, 1.5 Ga ago, life added soft predation to hard predation, i.e. information production within each species to matter-energy flow between species. Replicative and informative structures are subject to two competing biological constraints: replicative fidelity promotes proliferation, but limits adaptive evolution. On the contrary, the offspring of a couple obviously cannot be a copy of both partners, they are a new production, a re-production. Sexual recombination allows the exponential enrichment of the genetic diversity, thus promoting indefinite adaptive and evolutionary capacities. Evolutionary history illustrates this: the bacteria proliferate but have remained at the first purely nutritive stage in which most of the sensory functions, mobility, defense, and feeding have experienced almost no significant novelty in three billion years. Another world appeared with the sexual management of information. Sexual reproduction actually combines two functions: multiplicative by "vertical transfer" and informative by "horizontal transfer". This distinction is very common: polypus - medusa alternations, parasite multiplication cycles, the lytochal and deuterotochal parthenogenesis of aphids, and the innumerable para- and pseudo-sexual strategies of plants opportunistically combine the two modes of asexual replication and sexual combination. However, for the majority of animals and multicellular plants that produce many gametes, numerical proliferation by descendants and informative diversity by sexuality are mutually implicated, for example in the seed. The true discovery of eukaryotes may not be the "true nucleus", as their name implies, but an orderly informative function. The field of recombinations circumscribes a class of partners genetically compatible with each other, each simultaneously prey and predator of the DNA of the other. The mythical Maxwell demon capable of tracing entropy by sorting molecules according to their state does exist: each mate is the other's Maxwell's demon. While a sexless bacterium is simply divided into two cells, two sexual parents work together to produce a single offspring a time. Added to this are the burdens involved in meiosis and crossing-over, cellular diploidy, and mating. Sex produces an information gain that is paid for by a cost of energy-material, and this barter must be fair to survive. The domains of sexual intercourse are very diverse: uniparental reproduction, alternation of asexual proliferation and sexual information, self-fertilization, endogamy, exogamy, panmixis, diffuse or structured polymorphism, fertile or sterile hybridization, horizontal transfers. Each species is a recombination field between two domains, cloning and hybridization. Multiplicative descent and informative fertilization are organically distinct, but selectively associated: the information produced by the parents' sexuality favors the predation of matter-energy and therefore the proliferation of offspring, and this proliferation in turn favors the sexed producers of information. The equation specific to each species is: enough energy to proliferate, enough information to diversify. Alternatively, two other reproductive modes obtain or transmit less information at lower cost: not enough recombinations=repetitive clonal proliferation, and too many recombinations=disordered hybridization. But these marginal modes have poor prospects, as the model of the species is successfully attractive. Better discriminate to better inform. In bacteria, the exchanged and incorporated DNA segments are directly identified by the parity of the complementary strands, which determines simultaneously the similarity, the offspring, and the pairing. In eukaryotes, on the contrary, somatic growth and germinal information are segregated. During speciation, adaptive information is compacted, delocalized, codified and published to inform the species about its own state: the prezygotic relationship governs viable mating. Under the effect of sexual selection, the runaway and the reinforcement of the characters related to courtship testifies to their identifying function, which explains the paradox of the singularity and luxuriance of the sexual hypertrophies. The speciation discretizes a balanced recombination field and validates the informative relations. The species is without degree. Mates of a species recognize each other quickly and well because the logic of coding disengages from the ecological game of adaptations. The system of mate recognition has a function of cohesion and its regularity allows the adaptations of the less regular being, it is neither elitist nor normative, it is subjected neither to a level of aptitudes, nor to sexual performances, but permissive; it protects the variability and polymorphism. Two mutually irreducible relationships triggered the debate between the taxonomists who support the phyletic definition of the species by the descendance, and the proponents of the definition by interfertility. Such a taxonomic disagreement is not insurmountable, but the issue is deeper than taxonomic concepts, because these concepts relate to two different modes of evolution. According to the phyletic model, each species is a lineage passively isolated by external circumstances; on the contrary, in the sexual model each species is actively produced by an internal process of adjustment between replicative costs and informative gains. Each species develops a solution of the equation that matches material-energy expenditures with informative gains. A species concept based on a lasting relationship between these two quantities or on the limits of certain values or their equilibrium is therefore legitimate. It is this equilibrium that all couples resolve, without our formulation being as clearly as biology desires and as physics demands. Energy expenditures and informative gains in sexuality are almost impossible to measure, yet observation and experience allow an approximate ranking of the energy/information ratio. For example, endogamy is more economical, but less diversifying than exogamy, polymorphism increases information, the reinforcement of sexual isolation limits the rate of unproductive fertilization, between neighboring species hybridization allows certain genetic contributions, etc. A closed species evolves naturally towards another just as closed. On the contrary, the artificial transfer of DNA opens the species. The natural boundaries that isolate the species are easily trespassed as energy costs and constraints of sexual recognition are easily controlled; and the perspectives of manipulations are visible, whereas natural selection never anticipates and thus works blindly. Informative, artificially directed predation stimulates the evolution of species.