Selection de deux caracteres chez une espece de Trichogrammes, efficacite parasitaire des souches obtenues (Hym trichogrammatidae)

Английский. Two characteristics, fecundity and "host-finding capacity", were selected in T brassicae Bezdenko over 5 generations. Three experiments were carried out. The first experiment concerned only fecundity, selected according to 2 methods (mass selection and "by strain" selection) with a pressure of 50%. It allowed fecundity to be improved by up to 52 or 71%. The second experiment concerned only the "host-finding capacity", selected under the same conditions. It did not allow improvement of character. This failure can be explained by an overestimation of the heritability or by a lack of standardization of the measurements. The third experiment concerned the 2 characteristics, separately or simultaneously (by means of an index) selected according to the "by strains" method with a pressure of 66%. The characters were not improved and this failure can be explained by a loss of fecundity variability during the rearing period between the 1st and 3rd experiment. The parasitic efficacy of the strains both selected or not was tested in greenhouses planted with corn where corn borers were released. It was concluded that the increase in fecundity had no clear influence on the biological control. The advantages and disadvantages of different biological control strategies have been discussed: maintaining, as is the current practice, the greatest possible genetic variability; selection of some characteristics in the laboratory; selection of the parasitic efficacy under natural or seminatural conditions. Methodologies have been suggested and compared.

Французский. Deux caractères ont été séparément ou simultanément sélectionnés chez T brassicae Bezdenko, la fécondité et la «capacité de recherche de l’hôte». Le 1er a connu une amélioration allant jusqu’à 71 % contrairement au 2e. Cet échec peut s’expliquer par une surestimation de l’héritabilité ou par un manque de standardisation de la mesure du caractère. L’augmentation de la fécondité des souches n’a pas eu d’influence nette sur l’efficacité parasitaire appréciée dans des serres. Nous discutons les avantages et inconvénients de diverses stratégies envisageables pour la lutte biologique : maintien, comme actuellement, de la plus grande variabilité génétique possible; sélection de certains caractères au laboratoire; sélection de l’efficacité parasitaire dans des conditions naturelles ou semi-naturelles.