Site-scale assessment of urban riparian ecological condition
Abstract
Rapid urbanisation is resulting in ecosystem degradation, including loss of ecosystem services and biodiversity at an alarming rate, with habitat conversion and fragmentation the key threats to ecological condition. Riparian ecosystems in cities are critical refugia from threatening processes for many obligate and non-obligate riparian taxa, and these are novel ecosystems that cannot easily be compared to historical baselines. Despite much research on the condition assessment of riparian ecosystems generally, there is currently no method designed to rapidly assess the ecological condition of urban riparian areas, particularly in Australia. This thesis aims to address this gap by developing a method to assess urban riparian condition at the site scale. I compiled from the literature an initial list of 44 compositional, structural and functional indicators that may predict urban riparian ecological condition at the site scale. I sampled these 54 sites which independent ecologists previously stratified qualitatively as either "good", "change (i.e., undergoing restoration)" or "poor." I then constructed an ordinal scoring index by identifying indicators that discriminate between sites, removing correlated indicators, and developing scores based on expected distributions of remaining indicators. I evaluated this index by analysing associations with native floristic and bird richness, macroinvertebrate diversity, the occurrence of several frog species and the occurrence of two riparian mammal species using generalised linear models (GLMs). Finally, I ran the same models again, but instead used Rapid Appraisal of Riparian Condition (RARC) scores to compare my findings to this widely used method of non urban riparian condition assessment. I found 22 core indicators that discriminated between the ecological condition of my 54 sites, and were not correlated. I constructed the Multi-taxa Urban Riparian Condition Index (MURCI), which was lowest at "poor" sites, intermediate at "change" sites and highest at "good" sites, with no overlap of 95% confidence intervals. I found the MURCI had a significant positively association with three taxa, (native plants, riparian mammals, and Uperoleia laevigata), with a further two (native birds and aquatic macroinvertebrates non-significantly associated). There was a significant positive association of RARC with three taxa (native plants, Litoria peronii and U. laevigata. In all five cases, models comprised of my scoring system performed better than models comprised of RARC based on both Pseudo R2 and _AICc. This approach could enhance urban riparian condition management, since the MURCI is comprised of indicators that are relatively efficient to sample in the field, discriminates between sites, and appears to be a reasonable measure of urban riparian ecological condition. It also performs better at discriminating between sites and predicting the richness or occurrence of different riparian taxa than RARC, highlighting the need for situationally relevant methods. It is suitable for use by a range of managers for assessing the condition of sites, prioritising sites for restoration or as a complement to other research, though may not be suitable for indicating the value of a site for biodiversity with specific habitat requirements, such as threatened species.