Optimal ground preparation treatments for restoring lowland Sand Fynbos vegetation on old fields

A restoration field trial was designed to investigate herbaceous weed control and Sand Fynbos plant establishment on a fallow field at Riverlands Provincial Nature Reserve located on the West Coast forelands. The objectives were to investigate optimal treatments for reducing weed cover and the interaction between weed cover and shrub establishment. Herbaceous weeds were controlled using tilling, herbicide and carbon-rich mulch in a split-plot replicated design. Seedlings of two bioassay fynbos species were planted and survival monitored. Tilling in April reduced weed cover for four months, but thereafter had no overall effect. Herbicide applied twice, in May and early July, reduced weed cover significantly for a longer period, whereas mulch had no effect. Soil nitrogen levels were slightly higher than those of untransformed Sand Fynbos soils but no effects of the treatments on nitrogen mineralization rates or total available nitrogen were found. Bioassay plant establishment was highest in the herbicide plus mulch treatment and was negatively related to herbaceous weed cover. Molerat activity, which was positively related to weed cover, was noted to reduce the establishment success of shrub seedlings. There was no overall effect of tilling and mulch on plant establishment in the field, although mulch influenced growth rates in the pot experiment. In terms of cost-effectiveness, it is recommended that herbicide treatment be used to control herbaceous weeds prior to the re-introduction of fynbos species. Other factors to consider in planning a Sand Fynbos restoration, that may impact on native seedling establishment, include whether to: burn the site to remove standing alien biomass and litter; reduce soil N levels; control molerat and gerbil colonies; control browsers; and control of regenerating invasive alien plants. (C) 2007 SAAB. Published by Elsevier B.V. All rights reserved.