Background/Question/Methods The next generation of transgenic crops will be engineered with increased abiotic stress tolerance. By altering cellular, metabolic, or regulatory processes, these transgenes are likely to cause more secondary effects than seen in previous transgenic crops. This study examines secondary effects of genetic engineering for increased salinity tolerance in the model species Arabidopsis thaliana under intra- and intergenotypic (50:50 with wild-type) competition under multiple conditions. Three transgenes, previously demonstrated to increase salinity tolerance in Arabidopsis, with different modes of action were selected: Salt Overly-Sensitive 1 (SOS1) encodes a plasma membrane Na+/H+ antiporter, mannose-6-phosphate reductase (M6PR) encodes a mannitol biosynthesis enzyme, and C-repeat binding factor 3/drought responsive element binding factor 1a (CBF3/DREB1a) encodes a transcription factor for abiotic stress gene regulation. Independently transformed lines (2-3) were used for each transgene. Transgene fitness was calculated by selectable marker screening of progeny seed from 14 intergenotypic competition populations per line and fecundity comparisons between wild-type and transgenic intragenotypic competition populations (5 populations/line). By performing multigenerational assessments under different conditions, both in the field and the greenhouse in the presence and absence of salinity stress, we are able to create a fitness profile of these transgenes detailing short and long-term fitness impacts of transgene expression.

Results/Conclusions Lifecycle analyses of field-grown populations demonstrated that, in the absence of salt stress, M6PR and SOS1 overexpression caused no significant developmental differences relative to wild-type (P≥0.59 for all transgenic lines). CBF3 overexpression in lines A28 and A30 led to delayed development and dwarfism as seen in previous studies (P<0.0001) while line A40, with lower CBF3 expression levels, showed non-significant differences from wild-type (P>0.12). A 2007 preliminary intergenotypic competition field trial found that CBF3 lines A28 and A40 had fitness values of 23.3% ± 7.2 and 23.9% ± 6.6 while the fitness of M6PR line M2-1 was 145.1% ± 12.0 relative to wild-type (mean ± SE). Preliminary 2008 field fitness data, of populations under intergenotypic competition, found a fitness of 18.4% ± 3.9 for CBF3 line A30 while M6PR line M2-1 and SOS1 line 7-6 had fitnesses of 121.3% ± 5.0 and 56.6% ± 6.2 relative to wild-type respectively. Under intragenotypic competition, average fitness relative to wild-type was 41.8% ± 7.8, 153.5% ± 21.9, and 82.9% ± 13.5 for CBF3, M6PR, and SOS1 lines respectively. Expression of abiotic stress resistance enhancing transgenes can therefore have variable effects on fitness not predicted by developmental or previous growth chamber analyses.