Several complementary methods of plant establishment are required for effective restoration of banksia woodland

Abstract

AbstractNew banksia woodland vegetation was established at two sites totalling 50 ha in the Perth region of Western Australia as part of an offset-funded project. Restoration methods included topsoil transfer (16 ha), planting of nursery-raised local provenance seedlings (46,000 seedlings over 39.5 ha) and direct seeding with machinery or by hand (16.5 ha) with treatments overlapped. Six years of rigorous monitoring revealed trends in plant diversity, density and cover and allowed comparison of vegetation structure and composition to reference sites. Of the 162 native plants recorded, 115 originated primarily from the topsoil seed bank and the remainder from planting and seeding. Native plant germination from topsoil peaked at 700,000 stems per ha in year 2, but there was very high attrition during extreme summer drought. By year 5, native perennials averaged 20,000 stems per ha, well above the target of 7,000, but there was high spatial variability in plant density with 1/3 of quadrats below target. Saplings of Banksia spp. (the dominant local trees) plateaued at 150-220 stems per ha due to high summer mortality. Native plant cover reached 20% and perennial weed cover stabilised at under 5% within 5 years. Several complimentary methods were required for successful restoration, since transferred topsoil established most of the plant diversity. However, trees required planting or seeding, due to their canopy stored seed. Direct seeding and planting without respread topsoil led to lower overall diversity and density, but higher tree density. Most completion criteria targets were reached after 5 years. Areas with respread topsoil are trending towards recovery as a banksia woodland, but areas with only planting and seeding are likely to remain a separate vegetation type. Evidence for resilience of restored areas was provided by abundant pollination and seed set and second-generation seedlings. We suggest it may be possible to restore banksia woodland despite major challenges due to unpredictable offset funding, climate, weeds, grazing, recalcitrant species, and seed availability, but long-term monitoring is required to confirm this.