An Overview of Dr Chloe Molineux Green Roof Research

January 21, 2015

Green roofs are simply rooftops (usually flat) that have been vegetated with plant species. There are three types of green roof: intensive, semi-intensive and extensive. They are categorized as such because generally, they serve different purposes in the urban environment and are constructed in very different ways. Green roofs were originally designed for environmental and economical reasons. They enhance thermal insulation to, increase the roof life, absorb air pollutants, greenhouse gases and dust, reduce noise (Grant et al. 2003) and even to reduce water run-off and improve the quality (Mentens et al. 2006; Berndtsson et al. 2006), having a positive effect on sewage systems that are prone to flooding. Although it was not originally in the design concept it appears that these roofs have another, incidental benefit – to provide a habitat for rare birds and invertebrates

Plate 1.1 Thriving green roofs in Basel, Switzerland

1.1.1 Intensive green roofs
Intensive green roofs are those that are found in easily accessible private or public places; often these are parks or attractive gardens, a good example being the Jubilee Gardens in Canary Wharf, London (Plate 1.2), situated on top of the Jubilee tube station (Grant et al. 2003). Although many intensive types are not obvious as being on a roof, many have the underground train system running beneath them or are over underground car parks (commonly found in Basel, Switzerland). Intensive roofs are constructed using 20 – 50 % soil or organic matter plus a lightweight aggregate; together, aggregate plus organic matter, are described as a ‘substrate’. These green roofs have substrates that are extremely deep (above 20cm), enough in many cases for trees and large shrubs to grow well in (Dunnett & Kingsbury 2004). These types of roofs are aesthetically very pleasing but they require heavy maintenance and a complex irrigation system to keep healthy. Intensive green roofs are therefore very expensive to install and to upkeep (livingroofs 2009) and their purpose is probably best described as that of an attractive roof garden.

Plate 1.2 Intensive green roof, Jubilee Gardens Canary Wharf

1.1.2 Semi-intensive green roofs
Semi-intensive green roofs are similar to the purely intensive variety but their substrate depths are significantly reduced, usually to between 10 – 20 cm (Dunnett & Kingsbury 2004). These types of green roofs are more obvious as being situated on rooftops, although they are also better described as roof gardens due to ornamental vegetation and their ability to sustain large shrubs (Plate 1.3). Lighter substrates are usually required for semi-intensives as the roof structure has a finite loading capacity; therefore the soil/organic components are often reduced to 20 – 30 % (as this is where the main body of weight is coming from).

1.1.3 Extensive green roofs
Extensive green roofs are the most common and can be categorized into two further sub-types. Firstly ‘Sedum matted’, are those placed directly onto a moisture blanket or onto a thin substrate layer (often less than 2 cm deep). These ‘mats’ are planted with the stonecrop plant genus Sedum (up to 7 species) and like rolls of carpet, are simply laid out over the building’s rooftop. They are extremely lightweight and so are often perfect for sheds, garages and buildings with a very low roof loading capacity (Plate 1.4). Secondly, substrate-based green roofs can be plug-planted with Sedum species (which are well known for being drought resistant) or can be seeded with wild flower mixes for more natural, species rich meadows. These roofs are constructed with substrates that contain little soil components typically between 5 – 20 % organic matter, and contain lightweight aggregates that make up the bulk of the substrate to between 5 – 10 cm in depth (Snodgrass & Snodgrass 2006).

Plate 1.3 Semi-intensive green roof Plate 1.4 Extensive Sedum matted green roof

Both types of extensive green roof require little or no maintenance. They are far less costly than intensive roofs as they only require thin substrate levels, making them very light weight – perfect for city rooftops. It is estimated that around 20,000 hectares of existing roof spaces in London (equivalent to an area 28 times size of Richmond Park) are currently unused and could be extensively vegetated with little or no structural modification to the buildings below (Grant et al. 2003).

1.1.4 Extensive green roof construction
Extensive green roofs are constructed using standard procedures (as outlined by the German FLL guidelines, 2002). Firstly, there is a doubled layer of water proofing membrane placed directly onto the roof of the building. This is on top of the buildings original waterproofing, which should be in good condition in order to guarantee no leaks and ensure the maximum life span of the roof (between 45 – 60 years with a green roof, Langley Waterproofing pers. comm.). A root barrier (300 μm) is then laid over the water proofing along with a filter sheet (150 gsm) to prevent particulate matter from entering the water run-off, followed by an engineered egg-box-like drainage layer.

Plate 1.5 Layers of construction for an extensive green roof system

For a Sedum mat system, there may or may not be a substrate layer used, as mats can sit directly on top of a moisture retention blanket for an ultra-lightweight design (Plate 1.5). For Sedum plug-planted systems, there is a deeper substrate layer (5 – 10 cm) that is planted with several species of Sedum seedlings and in the case of a biodiverse roof, the substrate will be either seeded or left to colonise and vegetate naturally (Plate 1.6).

Plate 1.6 Layers of construction for a Sedum plug-planted extensive green roof system or for a biodiverse ‘brown’ roof.

1.2 Non-Sedum extensive roofs

Substrate-based extensive roofs that are seeded with a wild flower mix or left to colonise naturally have recently been termed ‘brown roofs or biodiverse roofs’ as they are comparable to natural Brownfield land. These rapidly decreasing wastelands are mainly found abandoned in urban areas but are increasingly being lost to planners and developers as the pressure for new housing for expanding cities mounts (DETR 2000; ODPM 2003). These wasteland sites generally have a substrate that has the consistency of crushed brick aggregate with little soil nutrition or retention of water, however re-colonisation can lead from bare ground to grassland, scrub and woodland, allowing a wide range of wildlife to become established (Gibson 1998; Angold et al. 2005). Brownfield areas are home to such animals as the brown hare, skylark and lapwing (these birds nest on open areas producing eggs that camouflage with the brick substrates), rare invertebrates (particularly spiders as found by Kadas 2002) and a variety of butterflies, reptiles and amphibians. These habitats also provide mitigation to the now very endangered Black Redstart (Frith & Gedge 2000; Grant et al. 2003). Red listed species could therefore be conserved by the green roof alternative that mimics these ever-decreasing natural habitats. The most species-rich site so far identified in the UK is a 27.5 hectares (55 acres) Brownfield site in Essex that has more recorded biodiversity per square foot than anywhere else in the country (Anon. 2005).

The idea of green roofs for enhancing biodiversity is simple; to create a similar environment to Brownfield sites but in places where they cannot be disturbed (Gedge 2001).

1.3 Previous green roof research

Although this is a fairly new research area in the U.K, green roofs have been a building requirement in Switzerland and Germany for a number of years. By law in Switzerland, 25% of all new commercial buildings must be ‘greened’ to preserve microclimates. In Germany 43% of cities provide a financial incentive for installing green roofs (Grant et al. 2003). There has been very little biodiversity research conducted on green roofs in the U.K, and less so on the plant species that are colonising the substrates found there. At the moment architects and developers install green roofs for non-ecological reasons, such as aesthetical appeal and economical value (i.e. thermal insulation). For this reason they tend to place down the commercially available ready-made Sedum matting; that generally does not allow natural plant colonisation nor offers the varied, species diverse environment that is desirable for most animals (Gedge & Kadas 2005).

1.3.1 Biodiversity benefits
Studies on Swiss green roofs have investigated how these artificial systems could be utilized to provide benefits for biodiversity. Brenneisen (2001) conducted one such study, looking at invertebrate populations on extensive green roofs. He showed that several roof design variables influenced green roof biodiversity, but that the most important factor seemed to be substrate depth. This is logical as deeper areas would retain more moisture and therefore provide a habitat for species requiring thicker and more diverse vegetation cover. Thinner areas of substrate would provide habitat that is bare and less vegetated, preferable to a number of drought tolerant invertebrates. By varying aggregate depth, a series of microhabitats are created on a green roof increasing biodiversity of the artificial habitat (Brenneisen 2001). Horticultural researchers have also been investigating substrate depth. They have suggested that thinner mediums should have a supplementary watering regime for plants, due to reduced water storage capacities (Dunnett & Nolan 2004; VanWoert et al. 2005) and conclude that media < 4 cm should ideally be Sedum matted to provide sufficient vegetation cover. Brenneisen’s research in Switzerland has recently been supported by PhD research in London by Gyongyver Kadas (2007). She also looked at invertebrate populations on the different types of extensive green roof and showed that although Sedum matted roofs supported larger invertebrate population numbers, substrate-based ‘brown’ roofs were more species rich – communities were actually 70 % similar to those seen in nearby brownfield sites. She suggested that this was due not only to the increased plant species richness on ‘brown’ roofs compared to the Sedum mats but also because of the diverse plant architecture that this provided. Importantly, 20 % of spiders and 15 % of beetles recorded on London roofs were either Notable or Red Data Book species (Kadas 2007). Furthermore, 10 % of the whole UK national (Harvey et al. 2002) and almost 20 % of the Greater London (Milner 1999) spider fauna were recorded from green roof sites in her study, again highlighting the importance and potential of green roofs in the urban environment. 1.3.2 Water run-off quality benefits Most studies by green roof researchers seem to centre on water run-off quality and thermal properties provided by vegetated roofs. Water run-off quality is measured by the quantities of leachate contaminates, e.g. high phosphorous levels from too much organic fertilisation. Young roofs can act as a source of contamination, with metals and nutrients being leached into storm water systems. This means that the use of dissolvable fertilisers should be avoided and careful planning of green roof substrates is essential (Berndtsson et al. 2006). Studies have also been conducted to find out what effects substrate depth and roof slope have on water absorption and therefore quantities of run-off (Nicholaus et al. 2005). Vegetated roofs are able to retain on average, 82 % of rainfall compared to 27 % on pure gravel rooftops; and this is even better on roofs with a 2 % slope (up to 87 % retention). Researchers in Sweden have also observed this hydrological function of extensive green roofs and have recorded peak flows (Bengtsson et al. 2005; Bengtsson 2005). The substrate itself is the most important factor in retention of water, however evapotranspiration of plants can increase this by 10 % (Nicholaus et al. 2005). This suggests that well-performing growing media will act not only to increase water-holding capacity (and so reduce storm water run-off) but may also enhance this benefit further if good vegetative cover is supported. Encouragingly, this storm water retention function of green roofs is beginning to be used as a management tool in many cities (Carter & Jackson 2007). Thus more and more green roofs are being installed in urban areas for their positive hydrological benefits, and not just for aesthetic appeal. 1.3.3 Thermal benefits Investigations into the thermal properties of green roofs have revealed that plants themselves significantly reduce air temperatures both inside the building and in the immediate environment (Niachou et al. 2001). Other studies have suggested that green roof plants (specifically their leaves) can reduce the thermal energy of solar radiation by 70 – 90 % and showed that vegetation cover and total leaf thickness is key to this reduction (Fang 2007). Experiments in Toronto, Canada, comparing common flat roofs, painted white roofs and green roofs found that, on average, roof temperatures on dry summer days were 65˚C, 42˚C and 35˚C respectively. Hence, green roofs contribute to the cooling of spaces below the roof during the summer and can increase heat in these spaces during the winter by controlling temperature fluctuations via increase thermal capacity (Niachou et al. 2001; Spala et al. 2008). Thus green roofs have the ability to significantly reduce the heat island effect in urban environments (Saiz et al. 2006; Alexandri & Jones 2007; Takebayashi & Moriyama 2007) because better insulation means less heat is escaping from the roof into the local vicinity. 1.3.4 Economical benefits Life cycle assessments of buildings’ input/output costs have also been calculated and published in many building and environment journals. A study from Pittsburgh, USA, calculated that although annual energy savings – through reduced air conditioning and heating bills – were small, there was a significant decrease in environmental impact over the life cycle of the building (Kosareo & Ries 2007). Similar publications have indicated that the need for air conditioning in summer months can actually be reduced by 6 % and, combined with a 1 – 2 % reduction of the urban heat island effect; greened cities could see electrical demand come down by 5 % (Booth 2006). This would not only have positive impacts on the environment but would also save hundreds of millions of dollars (Peck 2006). At present, thin-layer green roofs are only about 10 % more expensive than traditional roofing systems on new urban builds and it is thought that this will be reduced (by as much as 20 %) as the green technology becomes more advanced and widespread in the coming years (Cater & Keeler 2007). These studies concentrate on economic benefits rather than biodiversity, but are nonetheless vital if green roofs are to become part of planning and developing in the U.K and other developed countries. 1.4 Green roof policy Policy in London has recently included a section on green roofs (Policy 4A.11 Living Roofs and Walls) and states the following: ‘The Mayor will and boroughs should expect major developments to incorporate living roofs and walls where feasible and reflect this principle in local development framework (LDF) policies. It is expected that this will include roof and wall planting that delivers as many of these objectives as possible: • Accessible roof space • Adapting to and mitigating climate change • Sustainable urban drainage • Enhancing biodiversity • Improved appearance Boroughs should also encourage the use of living roofs in smaller developments and extensions where the opportunity arises’ (Policy 4A.11 Living Roofs and Walls). This policy is not U.K wide, however there are calls for this to become national as the pressure for official guidelines mounts as green roofs become more and more popular. German green roof standards are available in English, from the German Landscape Development Research Society (or FLL). These are strict guidelines on every aspect of green roof construction from membrane layers to substrates to plant seed mixes (FLL 2002). However, these have been produced for the market in Germany and have little relevance to green roofs in the U.K. They are not flexible, as is needed for biodiverse green roof construction, and do not consider the possibility of using any other alternative recycled aggregates in substrates except for crushed brick. 1.5 Problems with existing green roofs Until now, construction of extensive green roofs has predominantly been by the use of Sedum matted and Sedum plug-planted systems. The substrate-based ‘brown or biodiversity’ roofs have become more popular in London, however these too have been made in a uniform fashion. Crushed brick, crushed concrete and general demolition waste has been the prime aggregate for extensive roof substrate; often with the result of poor plant growth and severe drought stress during the summer months (Plate 1.7). The low-nutrient and free-draining characteristics of these types of green roof are by far the largest problems. Plants need organic matter to remain healthy and water storage is vital in the summer months to sustain the vegetation cover. However, too much organic matter will have adverse effects on water run-off quality (Berndtsson et al. 2006) and water pooling would not only cause stagnation and root rotting but may also lead to leakages through the roof membranes (livingroofs 2009). One way of solving these problems could be through the use of biological applications, such as introducing a sustainable and viable microbial population to substrates in order to alleviate both nutrient – and to some extent – drought stress, on plants above. Plate 1.7 Poor vegetation cover on a roof in Canary Wharf Preliminary studies of microbial communities on green roofs have shown that there are relatively few microorganisms present within the substrates (Molineux 2005). This may mean that nutrient recycling is not as efficient as it could be, especially where there is no green roof maintenance, resulting in reduced plant growth and diversity. Therefore it will be important to look at the overall composition of microbial communities in green roof substrates, to see if manipulation experiments can firstly improve the health and then increase the growth of target plants and increase plant diversity. The addition of arbuscular mycorrhizal fungi (see section 1.6) – which has been shown to be very limited in the substrates of green roofs (Molineux 2005) – may also help plants source scarce water in summer months by increasing the surface area of roots. Compared to the Sedum green roofs, brown roof systems have been shown to provide the most biodiversity (Kadas 2007) but also have so much potential to be improved. At present, substrate-based green roofs have predominantly been constructed with crushed brick and crushed concrete. Alternative materials to demolition waste are commercially available however these are often from primary sources and are expensive to install – such as lightweight pumice (Fentiman & Hallas pers. comm.). Other relatively cheap, secondary aggregates, local to the site of the new green roof, would be desirable and may be a way to recycle wastes that would otherwise be sent to landfill. Previous research has shown that varied plant communities, vegetation cover, architecture, topography and substrate depth all increase biodiversity of green roofs (Brenneisen 2001; Kadas 2007); thus diversity of the actual substrate should enhance and increase this further still. 1.6 Importance of soil microbial communities The soil microbial community is a vital ecosystem that supports successful colonization of a substrate by plants (Lavelle et al. 2006). These communities include many species of bacteria and fungi that work in an equilibrium to produce stable, coexisting viable populations. Arbuscular mycorrhizae (AM) fungi are an important group, they comprise of about 150 known fungal species and are said to be associated with around 70% of all plant species root systems (Hodge 2000). AM fungi do not cause any visible changes to a colonised plant, e.g. tumours as by some pathogenic fungi, and therefore can only be detected under the microscope. Plants become colonised by mycorrhizal fungi when hyphae, that can penetrate plant roots, emerge from AM fungal spores found within the soil. These form characteristic arbuscules (Plate 1.8) that can fill entire individual plant root cells. From five known genra, three (Glomus, Acaulospora and Entrophosphora) form vesicles inside the plant root cells on the hyphae of the fungus (Smith et al. 1994). These vesicles are storage vessels for nutrients and are also characteristic of these VA (vesicular arbuscular) mycorrhizal fungi (Plate 1.9). Plate 1.8 Characteristic arbuscules of a Plate 1.9 A vesicle on the hyphae of a mycorrhizal fungus within plant root cells mycorrhizal fungi, inside plant root cells Arbuscular mycorrhizal fungi are generally beneficial to herbaceous vegetation as a mutualistic association is formed. The plants are able to maximise uptake of limiting minerals and nutrients from their growing mediums because the fungus stimulates growth in the roots, allowing them to cover a much larger surface area – they are also thought to alleviate drought stress via the same mechanisms (Ruiz-Lozano et al. 1995a). This mediated access to limiting nutrients by the fungus to the plant is particularly important for phosphates, and to some extent nitrates (Bücking & Sharchar-Hill 2005). The uptake of P from the soil is improved by the fungus, not only via a larger root surface area but also due to a decreased distance for nutrient diffusion, a faster diffusion rate (up to six times) through the hyphae compared to root hairs alone (Bolan 1991) and through the production of acids that catalyse the release of P from organic complexes within the soil (Marschner & Dell 1994). As the fungus is connected to the host root cells, it is able to absorb carbon from the plant (Smith & Read 1997); this is vital for its growth, as it cannot do this from its environment unlike other mycorrhizal fungi. Arbuscular mycorrhizas also seem to have other positive affects such as: resistance to insect herbivory (Gange and West 1994) and reductions in foliar diseases (Borowicz 2001), but can also have less positive effects on plant productivity in some communities (Klironomos et al. 2000). It has been shown in many studies that well-established underground communities have huge benefits on the aboveground forbs and grasses (Gange & Brown 2002; Wardle et al. 2004). If there is a viable microbial community then plant growth – and to some extent diversity (Gange et al. 1990) – should be increased. Due to the nature of green roof substrates, there may be limited microbial activity (Molineux 2005), which could be a disaster for plant and insect populations. The presence of arbuscular mycorrhizal fungi in particular may be crucial in determining plant species composition, by strongly influencing the outcome of competition (Allen & Allen 1984). For this reason it will be very important to study presence and absence effects of AM fungi on green roof plant diversity, as has been studied in grassland communities by Hartnett & Wilson (2002). 1.7 New Research To date there has been no published research on the microbial communities found in green roof substrates or on presence/absence effects on the above vegetation. Furthermore, there has been no published work on plant species diversity of green roofs in London or on the performance (in terms of supporting healthy plant communities) of different secondary aggregates used as extensive roof substrate. Instead floral research has mainly been lead by horticulturists who experiment with different herbaceous plants seeking out those able to withstand conditions up on a rooftop (Dunnett & Kingsbury 2004). These are nearly always non-native plants (Snodgrass & Snodgrass 2006), which look aesthetically pleasing but do not necessarily contribute to biodiversity or brownfield mitigation (livingroofs 2009). 1.7.1 Alternative green roof substrates Green roof research to date has predominantly focused on economical benefits for the urban environment and a small amount of work has been done on animal biodiversity in these unusual habitats. These are all highly important research topics, without which, green roofs may still be unknown in the U.K; however, now more and more popular, green roof research needs to expand on the most crucial element, and that is the substrate. As new applications for waste materials becomes ever more imperative, it seems that the production of lightweight secondary aggregates will become increasingly wide spread. These recycled materials should be commercially available and so seem perfect for use on a green roof. I do not expect that every material will be suitable for plant growth, but my study will characterise some selected aggregates (see chapter 2) and determine their potential as green roof growing media. 1.7.2 Plant and microbial diversity Another largely untouched area of research is that on the diversity of native plant species on existing green roofs. As many roofs in the U.K are Sedum based systems, there is little vegetative biodiversity to observe. There has also been no published research on the microbial communities in substrates on green roofs. Preliminary work suggests that these habitats are deficient in healthy microbial populations (Molineux 2005) therefore experiments to determine if microbial enhancements would improve substrate conditions, will be investigated in this study (see chapters 4 & 5). 1.7.3 Objectives of the study The research objectives in this thesis are divided into two main themes: physical improvements and biological improvements to extensive green roof substrates, in order to enhance plant growth and maximise plant species diversity. The first theme (physical improvements) investigates the potential for new, alternative and recycled aggregates to be used as substrate, by characterising the materials and testing their performance – in terms of plant growth and diversity – on a new experimental site (see chapter 3). Specifically, the study aims to address the following questions: • Can recycled secondary materials be produced into usable aggregates? • Are these aggregates suitable for green roof applications? • How do plants perform on a green roof containing the recycled aggregates? • How diverse are the plant communities that establish? • Are the alternative substrates equally as good or better than the standard crushed brick and concrete green roof substrate? The second theme (biological improvements) investigates the possibility of improving existing green roofs by the addition of a microbial community, in the form of bacteria and AM fungi, to the substrate. Specifically, this part of the study aims to answer the following questions: • Can a soil microbial community be established and sustained on a green roof? • Can plant performance on an existing green roof be improved biologically with soil microbes? • Does microbial diversity and plant species diversity change with soil manipulation experiments (see chapter 4)? • How do the microbial communities of green roofs compare to those of natural brownfield habitats? 1.8 Introduction to the chapters Chapter 2 is composed of material science techniques used for the rigorous testing of six potential new aggregates for extensive green roof substrates. Chapter 3 is dedicated to preliminary greenhouse experiments leading to the construction and monitoring of a new green roof – incorporating the potential new substrates. Chapter 4 describes microbial manipulation experiments on an existing green roof, in the hope of biologically improving plant growth and maximising biodiversity. Chapter 5 compares microbial communities found on two existing green roofs (including the one described in chapter 4) and from two brownfield sites to see if green roof communities are similar to those found in natural wasteland habitats; and if experiments in chapter 4 have changed or improved these communities. Finally chapter 6 is a general discussion of the results and suggestions for future green roof construction strategies.