Seed Bank Dynamics of Four Co-Occurring Banksia Species

Publisher Website
Google Scholar
Abstract
(1) Age structure, phenology, infructescence (cone) and fruit (follicle) production, predispersal seed predation, and size and viability of the seed banks of four co-occurring Banksia species were studied in a fifteen-year-old stand of fire-prone scrub-heath in southwestern Australia. (2) Seeds released spontanously during autumn germinated in the winter but nearly all seedlings died during the following summer. Recruitment is therefore largely confined to the immediate post-fire period. Seed-bank analysis indicated that, in order to maintain populations of the four Banksia species, the scrub-heath should be burnt at a frequency of no less than ten years. (3) Apart from the summer-flowering B. attenuata, the species flowered in winter. There were large variations in cone and follicle production between plants and between years within plants of all species. There was no support for the hypothesis that strongly serotinous species would show the highest variation in reproductive output. Follicle set was low for all species (0.1-4.7% of florets). (4) All species stored some seeds in the canopy for at least five years. There was no viable soil-stored seed bank. The mean number of canopy-stored viable seeds per plant ranged from 1344 (B. leptophylla) to 2 (B. menziesii). Extreme serotiny was recorded for B. leptophylla and B. attenuata in which seed older than the current year''s crop comprised 92% and 80% respectively of the total viable seed bank. Corresponding values were 34% for B. prionotes and 25% for B. menziesii. The rate of decline in seed viability and pre-dispersal seed predation were lowest in the most serotinous species. (5) With the exception of lower follicle set in the two species which re-sprout after fires, there was no support for the hypothesis that such species would allocate less energy to sexual reproduction than the two non-sprouting species. Patterns of seed storage had a greater influence on reproductive allocation than mode of post-fire regeneration.