Characterization of soil seed banks in California vernal pools: Links to restoration success

Background/Question/Methods

California’s vernal pools are ephemeral wetlands that host numerous endemic species, many of which are now rare and endangered due to extensive habitat loss from agriculture and development. Conservation of existing wetlands, restoration projects, and long-term studies of population and community dynamics are critical aspects of vernal pool recovery. One little-studied aspect of vernal pool plant ecology is the role of the soil seed bank in the persistence of restored populations. We used a long-term experiment involving 256 constructed pools to determine whether constructed pools maintain a sufficient seed bank to perpetuate native species. We collected soil samples from 60 constructed pools to assess the identity and quantity of seeds present. Each soil sample was split into two equal parts, with one half sieved for manual seed sorting and the other planted in a greenhouse germination trial. Here, we compare these two methods for characterizing the seed bank in vernal pools.

Results/Conclusions

Paired comparisons (paired t-tests) of seed and seedling counts indicated that the two methods yielded significantly different abundance estimates. Seed counts greatly exceeded germinant numbers in nearly all samples. However, our analyses indicated that tallied monocot seed abundance and monocot germinant counts were significantly correlated, suggesting that germination rates can be predictive of monocot abundance in the seed bank. In contrast, we found no correlation between the abundance of dicot seeds and of dicot germinants. Dicot seeds may be more difficult to detect in soil samples than typical monocot seeds, or dicot germination rates may be more variably affected by the environment. High variance both within and among the paired dicot counts seems to support species life-history influences. Variable germination of dicots, coupled with the high seed abundance in field collected soils used in this study suggest that restoration success is not limited by seed availability, but more likely by factors that influence species-specific germination rates such as pool environment or soil characteristics.