School of Agriculture, Food and Ecosystem Sciences - Theses
Permanent URI for this collection
Search Results
1 - 2 of 2
-
ItemThe role of green roofs as native bee habitat and their potential to provide three-dimensional habitat connectivity in urban landscapes( 2023-09)With ongoing urbanisation, cities around the world experience a continuing loss of biodiversity habitat. Green roofs have the potential to add crucial habitat on vertical built structures in neighbourhoods where ground-level habitat is lacking. Green roofs are also known to provide additional ecosystem services related to urban heat island mitigation, stormwater capture, and human health and wellbeing benefits. For these reasons, green roof uptake is encouraged in many urban areas and this thesis sets out to understand to what extent green roofs currently function to support biodiversity in the context of Australian cities, such as Melbourne. Despite more than a decade of incentives for installing green roofs in Australian cities, their current extent is largely unknown. However, if cities aim to include green roofs into their biodiversity targets, their abundance, distribution, and typology needs to be understood. Since native bees are essential ecosystem service providers and bees in general are known to be experiencing a global decline, it is of high importance to understand how native bees perceive green roofs as habitat, and to what extent green roofs can help improve functional connectivity for bees in urban landscapes. The three research questions addressed in this thesis are: 1) what is the current distribution, abundance and typology of green roofs in Melbourne and how has this changed over last 20 years; 2) which bee species utilise green roofs in Melbourne, and what are the characteristics of the green roof and surrounding landscape that influence their abundance and richness; and 3) how can we model functional connectivity in a way that captures the 3-D arrangement of green roofs, and what do these models reveal about the role that green roofs play in supporting native bees in dense urban landscapes in Australia? To address the first question, I applied a remote sensing approach to identify and map the location and characteristics of green roofs across 17 of the 30 municipalities in the Melbourne metropolitan area. By 2020, there were 224 green roofs in Melbourne, 56% of which were constructed after 2014, and 65% were located within 5 km of the Central Business District. While these green roofs were recorded across most of the municipalities and included large public green roofs on commercial buildings, or smaller, non-accessible biodiversity green roofs at family houses, 65% formed a distinct green roof typology that is typical of many green roofs in temperate Australian cities. These typical green roofs in Melbourne are located on mid- to high-rise apartment buildings, offer passive recreation for the building residents, and contain approximately 30% vegetation cover mostly comprising of lawn. This is a distinctly different typology of green roof to those that have historically been constructed in cities across Europe and North America and reinforces previous studies which have highlighted that the climatic conditions in south-eastern Australia have an important influence on the types of green roof plantings that can be successfully maintained. Given that the typical green roof in Melbourne is quite distinctive and differs from the types of green roofs that have been studied internationally, my second question sought to understand whether green roofs in Melbourne were utilised by native bees, and how the design and spatial location influenced their composition and abundance. I used the database developed in the previous question to identify 20 green roofs which were surveyed for native bees using a combination of passive and active sampling techniques. During the survey, I recorded a total of 114 individuals belonging to 18 native bee species and 186 individuals belonging to 2 exotic bee species. Using Generalised Linear Models (GLM) and Hierarchical Modelling of Species Communities (HMSC) revealed that the abundance and species richness of bees were positively affected by increasing species richness of flowering plants, the proportion of native flowering plants and the overall floral availability on the green roof. The proportion of impervious groundcover in the surrounding 500 m and the increasing height of a green roof, on the other hand, had a negative effect on bee abundance and richness. Bee surveyed on green roofs were all polylectic, mostly ground-nesting but varied in their body size. Although, native bees were all generalist foragers, most were observed foraging from native plants only. These findings highlight that green roofs do offer valuable habitat, indeed may even play an important role in supporting ground-nesting bees in urban areas as the green roof substrates may remain more open and less compacted than exposed soil at ground level. Having determined that native bees do utilise green roofs as habitat, the final component of my thesis investigated the role that green roofs play in the functional connectivity of 3-D urban landscapes. To address this question, the first step was to develop a 3-D functional connectivity model that represents the reality that green roofs may be isolated from other greenspaces vertically as well as horizontally. This was achieved by combining a habitat suitability model with a combination of least-cost path and graph theory modelling. Partial dependence plots from the habitat suitability model revealed a stepped negative relationship between height of green roof and the probability of occurrence for native bees. This is the first time a non-linear relationship has been recorded and it offers a useful framework to inform the placement of green roofs where biodiversity gain is an intended goal. The 3-D functional connectivity analysis revealed that ground-level greenspaces are highly isolated within the inner city of Melbourne (Scenario 1) but the presence of existing green roofs (Scenario 2) increases the number of patches functionally connected for native bees with short (101 m), medium (205 m) and long (586 m) dispersal ranges. Hypothetically maximising the green roof area across the landscape (Scenario 3) led to an increase in overall connectivity, however, it mostly improved the inter-connectedness between green roofs rather than improving the connectivity of ground-level greenspaces. However, for all the scenarios and native bee dispersal distances modelled, the most important role that green roofs played in connectivity was as additional patches of habitat (dPCintra values of 56 – 92%), followed by local connected networks (dPCflux values of 8 – 44%), but their role as critical stepping stones (dPCconnector) were consistently less than 0.1%. This suggests that while green roofs should continue to be added to support biodiversity of native bees, we need to reframe their role in urban landscapes as they are most likely to be used as a habitat patch, rather than as stepping stones. The findings of this thesis highlight the existence of a distinct typology of green roofs in Melbourne, Australia, provide evidence that green roofs can be important habitat for Australian native bees if offering sufficient floral resources and being located at an adequate height, and that green roofs have higher value in providing habitat than improving connectivity. This new knowledge and the novel 3-D functional connectivity approach can help inform local policies and stakeholders on how to design and where to prioritise new green roof installations if the goal is to create a more biodiversity-friendly urban landscapes.
-
ItemDirect seeding onto green roof substrate supports species rich, high cover novel grassland( 2016)The vegetation of green roofs is central to their functioning and ability to provide ecosystem services. When vegetation performs well, green roofs contribute to storm water mitigation, thermally buffer buildings, improve biodiversity and provide aesthetic and recreational relief in the grey city landscape. However, poor vegetation performance is common, with a decline in both species richness and cover over time. This decline can in part be attributed to design, failing to consider community assemblage mechanisms that lead to quality vegetation performance. Direct seeding of grassland species could offer a randomness in distribution and abundance of seedlings that supports early community self-organization and co-existence. Comparatively, adult plant establishment does not provide this early opportunity. This thesis determines, the ability of a scoria based green roof substrate to support the germination and establishment of a species rich, high cover, novel grassland community, and the direct seeding sowing methods to achieve this. Additionally, species richness, and abundance were investigated as potential drivers of cover. A grassland forb only species seed mix was applied in two experiments. Experiment One (n=7), in glasshouse conditions, investigated application of seed with and without a sand bed, and depth of sowing; six treatments. Experiment Two (n=10) in green roof module conditions outside under irrigation, investigated depth of sowing and rate; four treatments. Main results showed a species rich and abundant germination on scoria based green roof substrate. Results indicated that both depth of sowing at greater than 10 mm and application in a sand bed, reduced species richness and abundance. In green roof module conditions, surface sowing indicated a slight species richness advantage and an abundance disadvantage, in comparison to sowing between 0 to 10 mm depth. Sowing rate approaching that of on ground grassland restoration rates, were shown to be as effective as a doubled sowing rate in producing a species rich, high cover. This study found; no support for species richness as a key driver of cover, however abundance is indicated as an early key driver of cover, and may not act in isolation during rapid cover development. These findings are relevant to management practices. Quality vegetation performance, achieved at sowing rates approximating on ground restoration, suggest that further investigation into lowering rate and species richness and cover response, is warranted. Long term studies investigating community dynamics, would give insight into this novel community’s ability to continue co-existence as a functional resilient system, as a predictor of ecosystem service potential.