| If genetically engineered biocontrol organisms are to have utility in the agroecosystem, they must establish populations in niches occupied by indigenous mircroflora. It may become necessary for purposes of risk management to model the population dynamics of these organisms. I examined the growth of a wild-type strain of Trichoderma virens from alginate prills in agricultural soil in response to increasing levels of inoculum. The per capita growth did not decline with increased inoculum levels, in contradiction to predictions of the Verhulst-Pearl model. To explore these mechanisms of population regulation, I investigated the population response to inoculum density in vitro. I observed that on nutrient-rich media, the growth was density dependent, as predicted by the classical ecology models. In contrast, the per capita growth on minimal media, as in soil, did not respond to inoculum levels. In the case of one minimal medium, density dependent per capita growth could be produced by the addition of sugar, supporting the hypothesis that the observations in soil were due to the oligotrophic nature of the medium. Because indirect, resource-mediated competition is only practically studied in environments with density dependent per capita growth, I developed two, independent in vitro approaches for assessing the fitness and competitiveness of transgenic strains of T. virens. Results from the first method, an adaptation of the de Wit replacement series, suggested that one transgenic strain was less fit than the wild-type, parental strain, but more competitive. The system was validated by experiments between a wild-type strain and a near-isogenic, auxotropic mutant which were found to be similarly fit and competitive. Another experimental approach confirmed the conclusions that the transgenic strain is less fit, but more competitive than the parental strain. Through a statistical approach it was demonstrated that the fitness of a transgenic strain is not predictable on the basis of transgene type. |
