Ecological risk assessment of genetically modified strawberries : the hybridization potential between cultivated and wild strawberries

Natural hybridization is a widespread phenomenon in many plant and animal species complexes and has been acknowledged as an important evolutionary process that can lead to new evolutionary lineages.
The introduction of genetically modified (GM) economic plants has raised questions about the potential for transgene escape from GM plants into populations of wild or weedy relatives via hybridization. To date, numerous studies have shown the potential of GM economic plants to hybridize with wild species. In the genus Fragaria (Rosaceae), the potential for hybridization between different species has been demonstrated repeatedly. Although future commercialisation of GM garden strawberries (F. x ananassa Duch.) is very likely, there is limited knowledge about the potential for hybridization between garden strawberries and wild relatives under natural conditions. The goal of my thesis was to assess the hybridization potential between cultivated garden strawberries and wild relatives in Switzerland, and thus to provide a basis for estimating the risks of a potential future cultivation of GM garden strawberries. In Switzerland, the most likely wild candidate species for hybridization with cultivated F. x ananassa seems to be the common woodland strawberry (F. vesca L.). The main research aims of this thesis were:
(1)Assessment of the hybridization potential between F. x ananassa and F. vesca
(2)Assessment of fitness of hybrid plants and the potential effects of hybridization on natural F. vesca populations
To detect past and present natural hybridization between F. x ananassa and F. vesca, a hybrid survey was conducted in the surroundings of farms in Switzerland and southern Germany, where F. x ananassa has been cultivated for at least ten years and wild F. vesca plants occur in the close vicinity. Samples of wild F. vesca plants were analysed with microsatellite markers and ploidy levels of plants were estimated by flow cytometry to identify putative hybrids. Furthermore, F. x ananassa and F. vesca plants were hybridized experimentally to assess the hybridization potential under controlled conditions. No hybrid plants were detected in the field. Experimental hand-crosses yielded some vigorous F. vesca x F. x ananassa hybrid plants but germination and survival rates of hybrids were generally very low.
Solitary bees are important pollinators that visit both F. x ananassa and F. vesca plants in the field. To assess whether natural hybridization between F. x ananassa and F. vesca is promoted by the behaviour of pollinators I studied the flower choice behaviour of the red mason bee (Osmia rufa L.) in a greenhouse experiment. Solitary bees did not show a preference for either F. x ananassa or F. vesca. The results suggest that the behaviour of solitary bees does not obstruct gene flow from cultivated to wild strawberries.
As a basis for the assessment of fitness of hybrid plants and the potential effects of hybridization on natural F. vesca populations, the demography of wild F. vesca populations was studied. Demographic data were used to parameterise matrix population models, and the importance of different growth parameters for population growth was assessed using prospective (elasticity analyses) and retrospective (life table response experiments) matrix analysis methods. It could be shown that clonal reproduction is of great importance for growth and maintenance of F. vesca populations, whereas sexual reproduction seems to be insignificant for population growth within established populations.
Furthermore, growth characteristics of F. vesca x F. x ananassa hybrids and F. vesca plants were directly compared. Different hybrid and F. vesca clones were grown under a competition and a control treatment in a greenhouse. I found that hybrids can exceed F. vesca plants in total biomass production. However, the number of clonal offspring and the biomass proportion that was allocated to clonal reproduction was significantly lower in all hybrid clones under competition and significantly lower for most hybrids in the control treatment. Furthermore, there was a large difference in the timing of clonal reproduction between F. vesca and hybrid plants with clonal reproduction starting later in hybrids. In summary, the results indicate that hybrids have a competitive disadvantage against co-occurring F. vesca plants due to inferior and delayed clonal reproduction.
In conclusion, the lack of hybrids at farm survey sites, the low experimental hybrid germination and survival rates and the differences in clonal reproduction parameters between F. vesca and hybrid plants indicate that there is low potential for hybrid establishment under natural conditions. Therefore, it seems that chances for transgene escape from transgenic F. x ananassa cultivars via F. vesca x F. x ananassa hybrids are also low as long as transgene effects can not compensate for any disadvantage of lower and delayed clonal reproduction rates of hybrid plants.