Ecological restoration of dryland kānuka communities in an irrigated agricultural landscape

Abstract

The Canterbury Low Plains ecological district exists as a highly modified, productive landscape containing little of its original indigenous vegetation. The residual native plant communities comprise scattered fragments of dryland shrubland with varying degrees of legal protection. One such area, the Eyrewell Forest, comprised a matrix of small dryland kānuka (Kunzea serotina) communities subsumed within an exotic pine plantation (6,764 ha) and pastoral land, and was described as the largest kānuka shrubland stand on the Canterbury Plains. Since 2010, the plantation has been progressively removed for conversion to precision-irrigated dairy farming. Within this landscape matrix, the current project focused on the establishment of ecological restoration areas and corridors to re-construct, protect and enhance indigenous species and habitats. The community structure and ecology of plant assemblages and associated edaphic conditions within existing remnants were investigated to assist in design, implementation and management of 18 ecological restoration blocks encompassing approximately 150 ha. A particular focus was the importance of moss carpets within natural kānuka-dominant ecosystems. I investigated the responses of the habitat and its plant species to nutrient spillover from adjacent farmland and evaluated restoration techniques suitable for the task. Study of the remnant kānuka habitats revealed that the native vegetation is characterised by low-nutrient soils and the dry climate, supporting assemblages of plants with obvious xerophytic traits. Small-scale disturbance was found to be critical for regeneration of the single canopy species, kānuka, and an understory shrub, Pomaderris amoena. The dominant ground cover moss species, Hypnum cupressiforme, and two woody associates, Leptecophylla juniperina and Carmichaelia australis, regenerated in areas of low disturbance and high canopy cover. Germination trials had some difficulties; P. amoena seeds required heat treatment and naturally followed fires, but L. juniperina could not be germinated artificially. H. cupressiforme exerted significant influence over the functioning of the ecosystem with regards to hydrology, nutrient cycling and biotic interactions. Extremes of soil temperature and moisture were mediated by the moss carpet; soil nutrients, particularly mineral nitrogen, were reduced. The moss layer also had species-specific effects on flora but generally decreased germination and establishment of native species. Glasshouse trials showed that K. serotina, P. amoena and Kunzea robusta grew better with soil additions of nitrogen. C. australis and H. cupressiforme responded positively to lime additions, but L. juniperina showed no response. In the field, fertilisation had either negligible or detrimental effects on native species. Increased soil nutrients from organic amendments elicited exotic weed encroachment. Nutrient spillover from farmland was observed at least 30m into remnant kānuka ecosystems and was positively correlated with exotic weeds but negatively related to moss cover. Successful introduction of planted native species required protection from desiccation and competition from weeds. Tree guards successfully and cost-effectively sheltered young plants and moss. Weeds were best controlled by herbicide application and, moderately with moss and pine litter mulch. This research has contributed to the restoration of 150 ha of landscape restoration, generating practical applications for similar projects elsewhere. A major finding is the importance of integrating moss into dryland kānuka restoration projects.