agresearch.co.nz). This project was part of an indigenous biodiversity restoration research project conducted in the context of a wider integrated catchment management program at the Whatawhata Research Centre, located near Hamilton, New Zealand.
SummaryOur study aimed to examine the applicability of a native plant restoration technique involving close grazing, blanket herbicide spraying and mob stocking (to trample the seed into the soil) for establishing New Zealand native tree and shrub species into exotic hill country pasture. This approach represents an alternative to the usual spaced-planting of seedlings in the context of an indigenous vegetation restoration programme. Field-collected seed of Cabbage Tree ( Cordyline australis ), Kahikatea ( Dacrycarpus dacrydioides ), Kanuka ( Kunzea ericoides ), Karamu ( Coprosma robusta ), Kohuhu ( Pittosporum tenuifolium ), Koromiko ( Hebe stricta ), and Manuka ( Leptospermum scoparium ) were germination tested in the laboratory, and broadcast sown in two field experiments in spring 2001 and autumn 2002. The spring-sown experiment tested the effect of mob stocking and the autumn-sown experiment tested the effect of mob stocking and sowing rate. Laboratory germination varied widely among species (0 -88%). In the field, the two shrubs, Koromiko and Karamu, established densities of up to 5 and 1.2 plants/m 2 , respectively, after 2.5 years. Autumn emergence for these two species strongly related to sowing rate. Three tree species, Cabbage Tree, Kahikatea and Manuka, emerged in the field but did not survive beyond 6 months. Mob stocking was effective in spring but not autumn sowing, implicating soil moisture content as an important factor in germination success. The study demonstrated that the direct seeding technique was successful for two shrub species, highlighting the potential for development of a cost-effective native plant restoration approach.