Institute of Evolution and Ecology (EvE)

Role of seed dormancy in providing resistance for annual species in the Eastern Mediterranean

Decreased rainfall, increased temperature and increased temporal variability are proposed to have great impacts on the future persistence of species in Eastern Mediterranean regions, already imposed upon by extreme conditions. However, predictions of high vulnerability neglect to include current adaptive strategies already expressed by plants in these communities, which could buffer further change.

To understand potential buffering mechanisms, we performed a 2-year greenhouse seed germination experiment, testing germination responses to different levels of aridity. Overall we tested 58 annual species collected from Israel, from two sites which vary in their rainfall quantity but more importantly, their variability and unpredictability.

This experiment was designed to explicitly test seed dormancy strategies - a common adaptive bet-hedging strategy which spreads the risk of extinction for plants in more arid variable climates. We want to identify if these seed dormancy strategies are able to provide resistance for species to climate change.

In addition, we are working in collaboration with Mark Rees (University of Sheffield, UK) and Leonor Álvarez-Cansino (University of Bayreuth, Germany) to further test seed dormancy, and other rainfall and grazing related plant strategies, within a fully explicit simulation model closely linked to vast quantities of long-term data we have collected as part of the GLOWA project.

Overall Aims:

  • Determine intra- and inter- specific seed dormancy strategies for a large number of species across a natural rainfall gradient, using species established at two sites varying in rainfall quantity and variability.
  • Establish if patterns of germination fractions are consistent with bet hedging theory ie more dormancy in more unpredictably varying environments.
  • Determine under what conditions seed dormancy strategies are likely to be more variable within a community.
  • Use the results to validate and parameterize a mechanistic simulation modeling approach.

Main Investigators: Mark Bilton & Katja Tielbörger

Collaborators:

Mark Rees - University of Sheffield

Leonor Álvarez-Cansino - University of Bayreuth, Germany