Germination of semi-arid plants is usually cued to rainfall which is unpredictably distributed across years. Plants must cue their germination to rainfall events that are heavy enough to ensure sufficient moisture for seedling emergence and survival; hence, it is thought that this cue is moisture availability. In variable environments (i.e. drier systems with a large coefficient of variation in mean annual rainfall), it is predicted that a smaller fraction of seeds from a population will germinate at any one timeafter rainfall compared to species whose populations have evolved in more mesic or predictable environments (i.e. systems where the mean is higher and the coefficient of variation is lower). Despite this fairly simple hypothesis, it has not been well studied in Australia, reflecting the gap in fundamental autecological research that exists.
I am starting a project to determine whether seeds of semi-arid shrubs (in the genera Atriplex and Maireana) sampled from populations in xeric environments have lower germination fractions (at all moisture regimes) than seeds sampled from more mesic populations.
To do this, we will study the lab germination of several populations of semi-arid shrubs that have been collected across a strong climate gradient from southern NSW. This climate gradient represents a temperature, rainfall and evaporation gradient.Seeds will be counted into replicate seed lots and exposed to a range of experimental conditions to see if (a) xeric populations exhibit lower germination fractions than more mesic populations and (b) whether variability in germination is most affected by population differences or is outweighted by environmental conditions.
If you are interested in bet-hedging and seed germination, these paper give a nice entry into the field: