Do soil microbes and abrasion by soil particles influence persistence and loss of physical dormancy in seeds of tropical pioneers?

• Main Authors: Paul-Camilo eZalamea, Carolina eSarmiento, A. Elizabeth eArnold, Adam eDavis, James W. eDalling
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2015-01-01
• Series: Frontiers in Plant Science
• Subjects: pioneer plants; Barro Colorado Island; germination cue; physical dormancy; seed dormancy loss; seed persistence
• Online Access: http://journal.frontiersin.org/Journal/10.3389/fpls.2014.00799/full

Germination from the soil seed bank (SSB) is an important determinant of species composition in tropical forest gaps, with seed persistence in the SSB allowing trees to recruit even decades after dispersal. The capacity to form a persistent SSB is often associated with physical dormancy, where seed coats are impermeable at the time of dispersal. Germination literature often speculates, without empirical evidence, that dormancy-break in physically dormant seeds is the result of microbial action and/or abrasion by soil particles. We tested the microbial/soil abrasion hypothesis in four widely distributed neotropical pioneer tree species (Apeiba membranacea, Luehea seemannii, Ochroma pyramidale, and Cochlospermum vitifolium). Seeds were buried in five common gardens in a lowland tropical forest in Panama, and recovered at one, three, six and twelve months after burial. Seed permeability, microbial infection, seed coat thickness and germination were measured. Parallel experiments compared the germination fraction of fresh and aged seeds without soil contact, and in seeds as a function of seed permeability. Contrary to the microbial/soil abrasion hypothesis the proportion of permeable seeds, and of seeds infected by cultivable microbes, decreased as a function of burial duration. Furthermore, seeds stored in dark and dry conditions for two years showed a higher proportion of seed germination than fresh seeds in identical germination conditions. We determined that permeable seeds of A. membranacea and O. pyramidale had cracks in the chalazal area or lacked the chalazal plug, whereas all surfaces of impermeable seeds were intact. Our results are inconsistent with the microbial/soil abrasion hypothesis of dormancy loss and instead suggest the existence of multiple dormancy phenotypes, where a fraction of each seed cohort is dispersed in a permeable state and germinates immediately, while the impermeable seed fraction accounts for the persistent SSB.