The ecological impact and control of an invasive weed Tradescantia fluminensis in lowland forest remnants : a thesis submitted for the degree of Doctor of Philosophy at the Institute of Natural Resources (Ecology), Massey University, Palmerston North, New Zealand

Abstract

While there is a general awareness of the global march of invasive weeds, there are relatively few studies which measure the ecological impact of these species on the systems they invade. Tradescantia fluminensis Vell. is an invasive weed of canopy-depleted native forest remnants in New Zealand where it carpets the ground and prevents regeneration. In three lowland forest remnants in the lower North Island I measured the ecological impact of Tradescantia by comparison of affected and non-affected areas of forest. In addition, I evaluated three methods for control of Tradescantia in two heavily infested forest remnants in the lower North Island. The impact of Tradescantia on native forest regeneration is evident by the decreasing native forest seedling species richness and abundance with increasing Tradescantia biomass. Forest regeneration was prevented because of decreasing light levels beneath Tradescantia. The compositions of the extant vegetation, seed rain and seed bank are consistent with this interpretation. Seedlings of some native species were more tolerant of Tradescantia than others, though the growth to emergence of even the most tolerant species was compromised in dense Tradescantia. While shading of native plants by invasive weeds is a well-studied phenomenon, comparatively little is known about the effects of weeds on ecosystem processes. Tradescantia increases litter decomposition and alters nutrient cycling by modifying the litter quality and microclimate within these forest remnants. The annual uptake of nutrients by Tradescantia was a significant amount of the nutrient inputs via litterfall, which (with the exception of Ca) exceeded the amounts of these nutrients held within the forest litter layer, but was only a small amount of these nutrients held within the topsoil. It is likely that the microclimate within Tradescantia that promotes increased litter decomposition also affects invertebrate communities. Epigaeic invertebrates were sampled using pitfall traps. RTU richness was lower in Tradescantia plots compared with non-Tradescantia plots, though not statistically significant. Two-way indicator species and detrended correspondence analyses separated Tradescantia and non-Tradescantia plots within sites. Overall, impacts of Tradescantia were apparent despite large differences in invertebrate assemblages among sites. The possible biological consequences of the community and ecosystem impacts outlined in this and the preceding two paragraphs are discussed. A reduction of Tradescantia biomass to ~80 gm-2 (~40% cover) is compatible with native forest regeneration. Chemical and manual control methods had limited success in controlling Tradescantia, whereas artificial shading significantly reduced its biomass after 17 months. Native sub-canopy trees were planted into Tradescantia to achieve natural shading over large areas of forest, but were too young to assess whether or not they will overshade the Tradescantia. My research supports the theory that management needs to target the attributes of these forest systems that make them invasible rather than Tradescantia, otherwise they remain invasible to other weeds. Therefore, an integration of targeted control and tree planting to improve canopy cover is suggested as a means to arrest the decline of Tradescantia-affected forest remnants.