Seed dormancy is considered to present an adaptation to unpredictable environmental conditions in many higher plants. However, many other factors besides environmental variability are known to influence germination fractions. One important factor is the maternal environment, which only in recent years has been accounted for.

Environmental maternal effects were reported, with a copious year for the mother plant increasing the dormancy of its seeds. In 2005 Tielbörger and Valleriany presented a bet-hedging model, where germination fractions were allowed to vary according to the yield in the year of seed production. The model predicts an evolutionary stable strategy with increased seed dormancy after favorable years as a means to reduce generalized competition in the following season. If this is true these maternal effects should be stronger in populations from environments with high competitive pressure.

To test this hypothesis we raised a total of 720 plants of two target species (Biscutella didyma (Brassicaceae), Bromus fasciculatus (Poaceae)), that descended from 4 and 2 populations, respectively. They were subjected to 12 levels of increasing water stress and the germination of their seeds, under common garden conditions, was evaluated in the following year (2007). The populations were situated along a natural climate gradient in Israel.

Results/Conclusions

Our results for Biscutella didyma support the model prediction but germination fractions from arid and semi arid source populations decreased more steeply with maternal water availability than from seeds with Mediterranean origin. Regression slopes of several of the 4 populations were significantly different from each other. On the other hand seeds from the two Bromus fasciculatus populations did not show any maternal effect. The difference between the two species in the effect of maternal environment on seed germination corresponds to the existence of adaptive germination fractions of fully- after-ripened seeds in Biscutella didyma and absence of these in Bromus fasciculatus.

We conclude that maternal environmental effects play an important role in the evolution of seed dormancy strategies. They may alter average germination fractions in the same order of magnitude as strong climatic differences between our source populations. However an adaptive dormancy fraction is a requirement for maternal effects to be able to operate.