Orchids as Indicators of Ecosystem Health in Urban Bushland Fragments

In this thesis I investigate the utility of orchids as indicators of ecosystem health. The study areas were urban bushland fragments on the Swan Coastal Plain, a global biodiversity hotspot. The study focuses on the abundance, reproductive success, mycorrhizal abundance and seedling biomass accumulation of a suite of native terrestrial orchids common to Perth’s urban bushland fragments. A critical factor in exploring the ecological responses of these orchids to site condition and their application as indicators of ecosystem health is the assessment of the ecosystem health of each of the study sites. I studied the vegetation condition gradient across eleven urban bushland fragments using three known ecosystem health assessment methods. Correlations were found between the perimeter to area ratio, native vegetation cover, weed cover and canopy cover in relation to site condition gradients. Floristic complexity at sites was found to mask relationships with environmental variables that were apparent following classification into plant functional groups. Of the plant functional traits only facultative sprouter, sub-shrub, barochory and perennial trait frequencies correlated with the vegetation condition gradient, all traits showing a decline with decreasing vegetation condition. Multivariate analysis of orchid abundances and environmental parameters revealed three orchid species that could potentially be used as indicators of ecosystem health. Diuris magnifica and Microtis media correlated strongly with poor condition sites. Pterostylis sanguinea correlated strongly with very good condition sites. However, environmental parameters, floristic composition and plant functional groups provided weak correlation to orchid species presence and abundance. Reproductive response, mycorrhizal abundance and biomass accumulation across the vegetation condition gradient were then measured to determine the extent to which orchids can be used as indicators of ecosystem health. The effects of site condition on fruit set success were not found to be significant for any of the orchid species in this study. Widespread pollen limitation across sites revealed that fruiting success was likely to be too insensitive a measure for examining ecosystem health. Mycorrhizal distribution across the cline of condition was found to be patchy within fragments and revealed unoccupied niches capable of supporting orchid germination. Further evidence of the use of Microtis media as an indicator of poor condition sites was found in an increased abundance of the associated mycorrhizal symbiont. The abundance of mycorrhizal symbionts for Caladenia arenicola and Elythranthera brunonis at sites of very good condition indicated their potential use as indicators. An inverse relationship was found to exist between biomass allocation to leaf or tuber in sites of good and poor condition. In sites of poor condition, Diuris magnifica and Caladenia arenicola increased allocation of biomass to shoots presumably in order to obtain photosynthates. In sites of very good condition these two species increased their allocation of biomass to the tuber. Initial findings suggest biomass allocation in Caladenia arenicola and Diuris magnifica may be a useful tool in measuring ecosystem health. The lack of currently undisturbed urban remnants and a poor historical record of past disturbance events in the study sites make understanding the role of past disturbances on the current condition gradient difficult. The results of this study suggest that orchid presence and abundance, orchid growth and orchid symbionts can be used as indicators of ecosystem health, although work needs to be undertaken to refine the understanding of their response to specific disturbances. This study provides a baseline for investigating the utility of orchids as indicators of ecosystem health in highly fragmented systems