摘要：
Aims Many resistance genes against fungal pathogens show costs of resistance. Genetically modified (GM) plants that differ in only one or a few resistance genes from control plants present ideal systems for measuring these costs in the absence of pathogens.
Methods To assess the ecological relevance of costs of pathogen resistance, we grew individual plants of four transgenic spring wheat lines in a field trial with three pathogen levels and varied the genetic diversity of the crop.
Important findings We found that two lines with a Pm3b transgene were more resistant to powdery mildew than their sister lines of the variety Bobwhite, whereas lines with chitinase (A9) or chitinase and glucanase (A13) transgenes were not more resistant than their mother variety Frisal. Nevertheless, in the absence of the pathogen, both the GM lines of Bobwhite as well as those of Frisal performed significantly worse than their controls, i.e. Pm3b #1 and Pm3b #2 had 39% or 53% and A9 and A13 had 14% or 23% lower yields. In the presence of the pathogen, all GM lines except Pm3b #2 could increase their yields and other fitness-related traits, reaching the performance levels of the control lines. Line Pm3b #2 seemed to have lost its phenotypic plasticity and had low performance in all environments. This may have been caused by very high transgene expression. No synergistic effects of mixing different GM lines with each other were detected. This might have been due to high transgene expression or the similarity between the lines regarding their resistance genes. We conclude that costs of resistance can be high for transgenic plants with constitutive transgene expression and that this can occur even in cases where the non-transgenic control lines are already relatively resistant, such as in our variety Frisal. Transgenic plants could only compete with conventional varieties in environments with high pathogen pressure. Furthermore, the large variability among the GM lines, which may be due to unpredictable transgene expression, suggests that case-by-case assessments are necessary to evaluate costs of resistance.
