My lab has a long-standing interest in the evolution of seed dormancy and the formation of natural soil seed banks. Models of the evolution of seed banks predict that the probability of a dormant seed dying in the soil should be positively correlated with the percent emerging and negatively correlated with the percent dormant. We have conducted a complex set of field and controlled environment studies using subsets of 24 populations to quantify correlations among these three seed fates (emerging, dormant and dead), to examine the size of the dormant seed bank in the field, and to quantify the broad-sense heritabilities (h2b) of seed fates. Our experiments clearly indicate that phenotypic differences in seed fates that contribute to the variance in seed bank formation among populations and significant h2b for seed fates. Counter to expectations, correlations across populations showed that percent emergence was strongly and negatively correlated with percent dead while percent dormant was uncorrelated with percent dead. These correlations were highly consistent between the field and controlled environment experiments. These projects are collaborative with Mark McPeek and Denise Thiede.