Human-managed and occupied urban green spaces may mimic the ecosystem functionally of naturally occurring habitats, either spontaneously or by design (Lundholm, 2006; Oberndorfer et al., 2007). Understanding how communities of organisms assemble and use these novel spaces provides a key opportunity to understand, and potentially shape, the ecosystem functions and services delivered in human-dominated landscapes (Groffman et al., 2017; Mallinger et al., 2016). The structural elements that make a habitat unique is termed “habitat template” (Southwood, 1977). The habitat template shapes which organisms can colonize by filtering out many species that are not suited for that habitat (Lundholm, 2006). Using this theory, Lundholm (2006) developed the habitat template approach to urban biodesign, in which built elements, such as extensive green roofs, can be designed using elements from various ground-level thin-soil rock barren environments, due to the similar physical attributes they share. It has been suggested that a structural diversity of plants and abiotic factors in green roofs influences insect diversity (Brenneisen, 2003).
Green roofs may, or may not, be designed to mimic similarly structured habitats in their region. The services that green roofs provide, including stormwater management, reduced energy consumption, accessible or visible green space, and habitat for organisms, among others, vary according to how the green roof is designed and maintained (Dunnett and Kingsbury, 2004). As part of the urban built environment, green roofs may experience high wind and solar radiation, as well as periods of flooding or drought brought on by the thin substrate on top of a hard surface. These conditions limit primary producers' survival (Lundholm, 2006). Thin soil (substrate) roofs are known as extensive green roofs, in which plants are intentionally grown on top of a human-built structure in shallow (typically 15–20 cm or less) growing medium (Getter and Rowe, 2006; Oberndorfer et al., 2007). As opposed to intensive green roofs, which have at least 15 cm of substrate and may host a wide range of vegetation types, extensive roofs put less stress on buildings and can be less expensive (Dunnett and Kingsbury, 2004).
In many locations, roofs designed without specific biodiversity goals in mind commonly use exotic Sedum (Phedimus) species because as succulents they have been shown to withstand the challenging growing conditions, especially drought, on roofs (Dunnett and Kingsbury, 2004; VanWoert et al., 2005). These Sedum based roofs initially became widely used in western Europe, especially Germany (Köhler, 2006; Ngan, 2004; Oberndorfer et al., 2007; Thuring and Grant, 2015), and are now popular in many places of the world including North America (Dunnett and Kingsbury, 2004; Dvorak and Volder, 2010; Snodgrass and McIntyre, 2010). When biodiversity service provision is a priority, designers may choose to use plants native to their region, creating a habitat analog. Prairie ecosystems are widely distributed in North America and commonly experience drought conditions, so these plants are well accustomed to the challenges often encountered on green roofs (Sutton et al., 2012). Although prairie plants, especially in tallgrass prairies, often have deep root systems (Nippert et al., 2012) many species root less deeply or will adapt to shallow growing mediums by growing roots horizontally (Sutton et al., 2012). The diversity of plant taxa found in prairies is also beneficial to their success as the richness supports ecosystem functioning (Cardinale et al., 2011; Tilman et al., 1996). Prairie analog roofs can be found in the Great Lakes Region of the United States of America (Dvorak, 2015; Hawke, 2015). Other ground-level thin-soil ecosystems that are structurally analogous to extensive green roofs, such as alvars, cliff edges, and barrens are found in the Great Lakes Basin. These natural environments experience similar environmental conditions to green roofs and have thin soils on top of bedrock, usually sandstone, limestone, or dolostone (Lundholm, 2006).
Insects inhabit practically all terrestrial and freshwater ecosystems, including urban environments, and play a variety of critical roles in ecosystem function and service (Rosenberg et al., 1986), making them the ideal ‘barometers’ to measure biodiversity functions of green roofs. Insects that provide services such as pollination, pest control, and decomposition are commonly referred to as beneficial insects, and these groups provide billions of dollars’ worth of ecosystem services to agricultural ecosystems in the United States each year (Losey and Vaughan, 2006). Numerous studies have reported a wide variety of invertebrate taxa occurring in green roof habitats (Brenneisen, 2003; Coffman and Davis, 2005; Coffman and Waite, 2011; Colla et al., 2009; Diethelm and Masta, 2022; Fabián et al., 2021; Jacobs et al., 2023; Kadas, 2006; Kratschmer et al., 2018; Ksiazek et al., 2012; Kyrö et al., 2020; MacIvor et al., 2015; MacIvor and Lundholm, 2011; MacIvor, 2016; Passaseo et al., 2021, 2020; Pétremand et al., 2018; Sánchez Domínguez et al., 2020; Starry et al., 2018; Tonietto et al., 2011; among others). Green roofs designed with insect biodiversity and native resources in mind can provide habitat in these urban landscapes that have lost some of this space on the ground level (Brenneisen, 2006, 2003; Lundholm, 2006). However, green roofs not necessarily designed for biodiversity may still provide habitat or foraging resources (Coffman and Davis, 2005; Coffman and Waite, 2011; MacIvor et al., 2015). In general, green spaces in human dominated landscapes can be important to support biodiversity and conservation (Tonietto et al., 2011), but the connectivity of these habitats may influence insect communities (Barr et al., 2021; Braaker et al., 2017, 2014).
To understand community composition several previous authors have compared green roofs and ground-level sites (Braaker et al., 2017; Colla et al., 2009; Ksiazek et al., 2012; MacIvor and Lundholm, 2011; Tonietto et al., 2011), however less is known about ground-level habitats under conservation protection, except Tonietto et al. (2011) examined prairies and green roofs, finding bee communities to be distinct. Regarding beneficial insects, pollinator studies on green roofs are more abundant (Colla et al., 2009; Jacobs et al., 2023; Ksiazek et al., 2012; MacIvor et al., 2015; Passaseo et al., 2021, 2020; Tonietto et al., 2011) than studies examining natural enemies (Diethelm and Masta, 2022; Fabián et al., 2021; Sánchez Domínguez et al., 2020). The majority of green roof insect studies use only one sampling method (Brenneisen, 2003; Coffman and Davis, 2005; Coffman and Waite, 2011; Colla et al., 2009; Fabián et al., 2021; Kratschmer et al., 2018; Ksiazek et al., 2012; Kyrö et al., 2020; MacIvor and Lundholm, 2011; Passaseo et al., 2021, 2020; Sánchez Domínguez et al., 2020; Starry et al., 2018), but using multiple sampling methods can improve understanding of the insect community (Aguiar and Santos, 2010; Campbell et al., 2023; McNamara Manning et al., 2022; Missa et al., 2009; Russo et al., 2011).
In this study, we examine insect communities and biodiversity characteristics in green roofs and ground-level thin soil habitats. While green roofs and ground-level thin soil habitats may share common structuring characteristics, due to greater connectivity with other habitats, we predict that insect richness and diversity will be greater in ground-level than green roof habitats. Between green roofs, we predict that insect richness and diversity will be greater among green roofs managed for biodiversity services versus those designed for other services like stormwater management.