Differences in species composition of urban bushland sites that had been subjected to increasing degradation and progressive invasion of Pittosporum undulatum Vent. were assessed using nonparametric multivariate and other statistical techniques. Increasing suburb age was found to significantly affect community pattern as a whole, specifically through increased proportions of exotic species, decreased native species richness and sustained shifts in the relative importance of individual species. Older suburbs were typified by species that were mesic, fire-sensitive, shade-tolerant and adapted to relatively moist, nutrient-rich edaphic conditions. These species included P. undulatum and many invasive exotics. Many native species decreased substantially in rank importance with increasing suburb age, to the point of elimination in older suburbs. These included one vulnerable taxon (Tetratheca glandulosa Sm.). Overall community pattern was correlated with abundance of P. undulatum, fire and human disturbance. Relative cover of P. undulatum was found to be significantly correlated with increased proportions of exotic species and reduced native species richness and diversity. While most exotic species were concentrated within 30 m of the suburban edge, it is suggested that most management effort should be directed at those exotic species that commonly establish throughout bushland remnants. The study also provided an opportunity to test the application of the multivariate software package PRIMER in assessing environmental impact on vegetation communities.

Suburban edges are shown to exert a major infiuence on invasion ofthe native Pittosporum undulatum Vent, in dry sclerophyll busbland of northern Sydney. Transect data from fifteen urban busbland sites spanning approximately 90 years of development indicate significant increases oi P. undulatum witb time. Basal area, density and frequencies of plants in all size-classes increased significantly witb age of adjoining development. Mean basal area at edges of old sites was 5700 times greater tban in comparable busbland in larger reserve interiors. Tbe effect of age was compounded by the greater impact of edge effects in narrower reserves of older suburbs. Basal areas and proportion of reproductive plants decreased significantly witb distance into tbe reserve. Older sites contained larger but fewer individuals at tbe edge, compared witb high densities of smaller plants further into the reserve, suggesting an advanced successional stage. The overall relationship of density with distance was not significant. The observed clumping of seedlings beneath canopies is consistent witb previous work on seed dispersal by vectors, and enhanced seedling recruitment in safe sites. Pittosporunt basal area and density reached no maximum levels witb age, implying ongoing invasion. The study indicated some factors which appeared to promote tbe establisbment of P. undulatum in tbe study area. Signs of fire were negatively correlated across all sites witb basal area and density of i? undulatum. Human disturbance was positively correlated witb the same variables. On the basis of findings in tbis study, some broad priorities for management oi Pittosporum in urban busbland are suggested.

Natural bushland is a dwindling resource in the Sydney Metropolitan Area, not only because of direct clearing, but also because uncontrolled impacts from surrounding development have caused increasing degradation over time. Community change has been characterized by invasions of environmental weeds, and consequent displacement of many indigenous species. A few locally native species, Pittosporum undulatum in particular, have emerged as environmental weeds. Invasion of new habitats within a species' natural geographic range must be recognized, since their impact on natural communities can be as serious as that of the worst exotic invaders. Factors implicated as the immediate cause of environmental weed invasions include increased human-induced dispersal by birds, anthropogenic disturbance, suppression of fire and increased moisture and nutrients. Management of plant invasions often includes mitigation to reduce impacts of the immediate causes. Strategies to maximize the success of mitigation by identifying and ameliorating the factors which impact bushland from remote sources are discussed. Practical recommendations for restoration of degraded sites are given. Native weeds such as P. undulatum require specialized treatment within an integrated environmental weed management strategy. It is stressed that application of the best management strategies available will be wasted if more fundamental issues related to current planning practices are ignored. Uncoordinated and ad hoc decisions are continuing to reduce urban bushland to mere remnants with high edge-to-interior ratios. Only when these practices are recognized and addressed can long-term success in the treatment of environmental weed invasions be achieved by the many excellent restoration techniques being developed.

Both land use intensification and abandonment within grasslands lead to a homogenisation of vegetation structure. Therefore, specially structured microsites such as vegetation gaps with bare ground play an important role for species conservation within grasslands. Vegetation gaps are crucial for the establishment of low-competitive plant species and offer special microclimatic conditions essential for the development of the immature stages of many invertebrate species. The influence of small-scale soil disturbance in the form of mounds created by ecosystem engineers such as ants or moles on biodiversity is therefore of special scientific concern. The effects of mound-building species on plant species diversity have been extensively studied. However, knowledge on the significance of these species for the conservation of other animals is rare. In this study we analyse the importance of mounds created by the European mole (Talpa europaea) as an oviposition habitat for the small copper (Lycaena phlaeas) within Central European mesotrophic grasslands. Our study showed that host plants occurring at molehills were preferred for oviposition. Oviposition sites were characterised by an open vegetation structure with a high proportion of bare ground (with a mean coverage of about 50 %), a low cover of herbs and low-growing vegetation (mean height: 4.5 cm). Our study clearly illustrates the importance of small-scale soil disturbance for immature stages of L. phlaeas and the conservation of this species within mesotrophic grasslands. Mound-building ecosystem engineers, such as T. europaea, act as important substitutes for missing dynamics within mesotrophic grasslands by diversifying vegetation structure and creating small patches of bare soil.