Does the reduction of seed dormancy during ex situ cultivation affect the germination and establishment of plants reintroduced into the wild?

Plants or seeds produced in botanic gardens or nurseries have become an important source of plant material for reintroductions or population reinforcements. However, recent research has shown that these living collections bear the risk of being genetically impoverished and adapted to the artificial habitat. In particular, many studies have reported a decline of seed dormancy during ex situ cultivation, which may compromise their suitability for reintroduction programmes. However, the impact of those ex situ derived changes on the germination and establishment of reintroduced plant populations is still unclear. We studied the germination behaviour, population establishment and plant fitness over 3 years of reintroduced plants of the short-lived perennial Digitalis lutea, comparing plants grown from (1) a 30-year botanic garden population, (2) seeds from a seed bank representing the initial starting point of the botanic garden culture and (3) a re-sampled corresponding wild population. Under laboratory conditions, wild-collected seeds required cold stratification to germinate, whereas seeds from the garden population germinated without stratification. This pattern was strongly reduced in an outdoor pot experiment, where only few garden seeds germinated before winter, and all seeds remained dormant when seeded in the natural area of origin. In a transplant experiment, reintroduced plants from the wild population outperformed both, the garden and the seed bank plants, in their fitness in the first 3 years after reintroduction suggesting adaptation to current climatic conditions. Synthesis and applications. Our study demonstrates that trait changes that occurred during ex situ cultivation can negatively impact the establishment of reintroduced plants. We conclude that wild plant material collected from contemporary populations is best suited for reintroduction and should be preferred over ex situ cultivated and seed bank stored material, especially when the cultivation spanned multiple generations. However, our study also shows that germination requirements change in complex ways, and the loss of dormancy observed under laboratory conditions may not always be directly transferable to natural conditions. When established standards are respected, ex situ propagated material may thus still be a valuable resource, especially when wild material is not available in sufficient quantities.