Seed Banks, Fire Season, Safe Sites and Seedling Recruitment in Five Co-Occurring Banksia Species

Abstract

Quantity of seed stored in the canopy of 5 co-occurring Banksia species varied by nearly 2 orders of magnitude. The 3 species which resprout vegetatively after fire, produced less seeds and retained smaller seed banks than 2 non-sprouting species. Contiguous patches of scrub-heath were burned in spring and autumn. The spring fire was cooler, and both sets of seed released subsequently did not germinate until the following (common) winter. Rate of seed release was higher after the autumn fire. The non-sprouting species released more seeds after each fire and yielded more seedlings per parent than the resprouting species. Percentage (field) germination of the non-sprouters was not consistently different from that of the resprouters. Seeds exposed on the soil surface during summer soon lost viability compared with buried seeds. The large number of seedlings established up to 8 months after the spring fire resulted from many seeds escaping exposure by dispersal into litter-covered safe' sites. Although litter microsites covered only 30% of the ground surface after the spring burn, they accounted for 80% of seedlings both before and after the summer drought. Litter microsites covered only 14% of the autumn-burned site but accounted for 60% of seedlings before summer and 40% after summer, suggesting density-dependent thinning of seedlings. The autumn-burned site supported more than twice the density of seedlings than the spring-burned site by the end of the 1st winter. Summer mortality of seedlings was 32% in the spring-burned site and 65% in the autumn-burned site, equalizing the seedling:parent ratio at both sites. Net recruitment of 1-yr-old seedlings varied from 0 for 2 resprouters to >100 per parent for a non-sprouter.