Complex Ecological Interactions and Ecosystem Services in Urban Agroecosystems

Stacy M Philpott¹-^§, Shalene Jha², Azucena Lucatero¹, Monika Egerer¹, and Heidi Liere'

• 1 ^{Department of Ecology and Ecosystem Management, School of Life Sciences - Weihenstephan, Technische Universitat Miinchen}
• 2 ^{Integrative Biology Department, University of Texas at Austin, Texas, USA}
• 3 ^{Environmental Studies Department, Seattle University, Washington, USA}

^§ Corresponding author: Email - This email address is being protected from spam bots, you need Javascript enabled to view it , 1156 High Street, Santa Cruz,

CA 95064, USA

KEY WORDS: ecological network, garden, pest control, pollination, herbivore, natural enemy, functional traits

Introduction

The local and landscape conditions that support biodiversity are fundamental features of a functioning ecosystem and its ecosystem services (Loreau et al. 2001, Hooper et al. 2005, Cardinale et al. 2012), including urban agroecosystems. Urban agroecosystems such as community gardens and urban farms are increasingly recognized for their contribution to biodiversity conservation and ecosystem service provision (Clinton et al. 2018. Clucas et al. 2018). Local features of urban agroecosystems include patch size, soil quality, ground cover, and vegetation diversity and structure, among other factors, while landscape characteristics include quantifications of natural habitat cover, development intensity, and landscape composition and complexity at larger spatial scales. Ecosystem services, processes provided by ecosystems that contribute to human well-being (Daily 1997, Millennium Ecosystem Assessment 2005), include pollination, pest control, and food production, among others, and are worth >$18 trillion globally (Costanza 1997. IPBES 2016). In rural agroecosystems, ecosystem services boost US crop production value by $57 billion per year (Losey and Vaughan 2006) providing strong financial motivation for optimizing service provision. While increasing biodiversity (e.g., of pollinators and natural enemies) is often associated with increases in ecosystem services (e.g.. Tscharntke et al. 2012), this is not always the case (e.g.. Cardinale et al. 2006). One potential explanation is that ecosystem services depend on species traits and representation in ecological networks (Lavorel and Gamier 2002, $chleuning et al. 2015, Perovic et al. 2018), not biodiversity alone. While many factors have been documented to affect biodiversity, ecological interactions, and networks in rural agroecosystems, less is known about urban agroecosystems; this system presents a new research context ripe for developing general urban agroecological principles. In this chapter, we will examine how a mechanistic understanding of the relationship between local and landscape factors, functional traits, and ecological networks may provide key information for reducing species or ecosystem function loss, better informing management, and optimizing overall ecosystem service acquisition in urban agroecosystems. In doing so, we contribute a new perspective on the future of urban agroecology studies, highlighting key research questions and methods.

Functional ecology and network ecology provide an ideal framework for understanding the impact of species loss and its implications for crop production and related ecosystem services. The functional ecology literature assesses which traits make species more sensitive to environmental change or impact provisioning of ecosystem services (Lavorel and Gamier 2002, $uding and Goldstein 2008, Lavorel et al. 2013). For ecosystem service providers, like pollinators and natural enemies, whose presence and functional effectiveness depends on ability to colonize, forage, and reproduce (Kremen et al. 2007), a broader investigation of species traits could yield valuable insight into function and resilience in agroecosystems. Networks depict and quantify interactions between species across trophic levels (Bascompte et al. 2003, Ings et al. 2009) and thus are a powerful tool to assess how species within a community respond to environmental change and how this potentially impacts the delivery of ecosystem services. Given that the loss of species interactions potentially weakens ecosystem services (Bohan et al. 2013), interaction metrics can serve as key indicators of ecosystem service provision and agroecosystem response to perturbation (Bliithgen 2010).

Urban agroecosystems are an ideal system for exploring the local and landscape management impacts on functional traits, ecological networks, and ecosystem services and may also alleviate local food security challenges. In this chapter, we consider urban agroecosystems as community, allotment, backyard, and rooftop gardens, urban farms, and other spaces within and at the edges of cities dedicated to cultivation of vegetables, medicinal plants, fruit trees, ornamental plants, and associated products (Lovell 2010, Zezza and Tasciotti 2010. Lin et al. 2015). Specifically, we focus on the ecological dimensions of these systems. By 2030, 80-90% of the global population will live in cities (United Nations 2005, Seto et al. 2012) and many urban residents lack access to fresh produce (e.g. Larson et al. 2009). Urban agroecosystems provide 15-20% of the global food supply (Smit et al. 1996, Hodgson et al. 2011), are an important source of vitamin-rich vegetables and fruits (Wakefield et al. 2007, Gregory et al. 2016), and promote gardener health and well-being (Brown and Jameton 2000, Classens 2015). Despite documented negative impacts of urban sprawl and increased urban developed cover, urban agroecosystems often host high biodiversity (Lin et al. 2015); however, the species assemblages in these disturbed habitats may be different from natural habitats, likely due to the different traits and species interactions that allow for their persistence in urban areas (e.g.. Cane et al. 2006. Faeth et al. 2005). In addition, gardeners often express challenges in food production in these urban agriculture environments, such managing pests (Oberholtzer et al. 2014). In order to provide needed technical assistance with greater certainty, we need to understand how garden management practices are best implemented to optimize pest control, pollination, and crop production (Lin et al. 2015). This is vital given the increased importance of urban agriculture for food security, especially in underserved communities (Alig et al. 2004. Pothukuchi and Thomas 2004, Ver Ploeg et al. 2009. Chappell and LaValle 2011).

In this chapter, we summarize current knowledge of local and landscape drivers of herbivore, natural enemy, and pollinator abundance, richness, and community composition, and resulting impacts on functional traits of service-providing organisms, the structure of natural enemy-herbivore and pollinator-plant networks, ecological interactions, and pest control and pollination services in urban agroecosystems. We highlight different observational and experimental approaches that have been used in ecological studies in urban agroecosystems, propose a unification of functional and network ecology, and end with an analysis of some challenges and future opportunities in ecological research in urban agroecosystems.

Local and Landscape Drivers of Herbivore, Natural Enemy, and Pollinator Communities and Traits

Local and landscape management strongly impacts the abundance, richness and composition of herbivores, natural enemies, and pollinators in agroecosystems. Local-scale vegetation diversity and complexity and higher floral abundance and richness boost abundance and richness of natural enemies (e.g., Andow 1991. Langellotto and Denno 2004) and pollinators (e.g., Baldock et al. 2015, Ballare et al. 2019). Landscapes with more natural habitat cover offer more resources (Landis et al. 2000. Ricketts et al. 2008. Schellhorn et al. 2014) and support a higher density and diversity of arthropods, even in sites with low vegetation diversity (Bianchi et al. 2006, Chaplin-Kramer et al. 2011). Although these relationships are well established for rural agroecosystems, less is known, especially in terms of landscape context, for urban agroecosystems, which may be different given heating effects of concrete, and perhaps different conceptualization of landscape diversity when multiple habitat types are highly degraded rather than a mix of natural and semi-natural habitats.

Knowledge of the functional traits of organisms, such as herbivores, natural enemies, and pollinators, occurring within urban agroecosystems is vital to understanding community structure and composition (Fountain-Jones et al. 2015) and can increase the predictive power of community studies beyond taxonomic analyses alone (Barton et al. 2011). A functional trait approach also reduces context dependency and allows for generalization across communities and ecosystems (Moretti et al. 2017). For natural enemies and pollinators, life history traits like voltinism or nesting habits are linked to fitness but are also sensitive to environmental stress, making them useful traits for assessing how species and their interactions might respond to global change (Moretti et al. 2017). Morphological and consumption-related traits, such as body size, diet, and hunting mode may influence colonization and likelihood of extinction, and effective delivery of ecosystem services (Peters and Wassenberg 1983, Woodward et al. 2005, Stang et al. 2009, Ibanez 2012, Ibanez et al. 2013). Increases in urban cover (e.g., concrete and other impermeable surfaces) change the physical and biotic properties of habitats, resulting in changes in soil and vegetation characteristics that act as environmental filters for arthropod communities via their functional traits (Aronson et al. 2016, Maisto et al. 2017). In particular, species (and their interactions) that exist in urban areas are present after passing through extra environmental ‘filters’ distinct from those in rural and especially natural habitats, such as influences of urban form (e.g.. buildings, density, sprawl), history of development and land use (e.g., toxic wastes, soil compaction), and human facilitation both at the regional and local site scales (Aronson et al. 2016). Here, we review the local and landscape drivers of herbivore, natural enemy, and pollinator abundance, richness, community composition and traits in urban agroecosystems.

Herbivores

Herbivorous insects are potentially detrimental to crop yield and quality, and managing pests is challenging for gardeners, who may lack pest management knowledge or avoid using chemical control due to health concerns and community garden regulations (Oberholtzer et al. 2014, Gregory et al. 2016). Herbivore populations can be regulated by intrinsic factors such as growth rates, survival, reproduction, dispersal capacity, as well as by bottom-up and top-down forces (Figure 2.1).

Major pest species differ in life history strategies and dispersal ranges which inform their responsiveness to habitat and management changes at different scales (Raupp et al. 2009, Mazzi and Dorn 2012). Thus, understanding variable pest responses to local and landscape features of urban systems requires consideration of pest traits and life histories in order to prevent pest outbreaks and minimize

FIGURE 2.1 Depiction of selected top-down (a. b) and bottom-up (c-e) impacts on herbivore insect pests and crop damage (f) in urban agroecosystems. Top-down impacts can result from planting flower strips (a) to support, or direct release of natural enemies (b). Bottom-up impacts result from changes to soil characteristics (c). water management (d) crop species identity and diversity (e). Drawings by Charlotte Grenier.

crop damage. For example, herbivorous species with life-histories that depend on specific host plants benefit when their host plants are cultivated and these insects may then become pests (Raupp et al. 2009). Further, pest traits such as tolerance to pollutants and urban heat may influence pest density in urban settings. For example, scale insects in urban trees surrounded by impervious landscapes experience higher temperatures resulting in higher fecundity and population growth (Dale and Frank 2014). Similarly, aphids exposed to polluted urban air have higher population growth rates and shorter development times due to greater availability of amino acids in plant phloem (Bolsinger and Fliickiger 1989). However, no studies have specifically examined the distribution of pest functional traits in urban agroecosystems.

Both local and landscape changes can alter bottom-up effects on herbivores. At the local scale, gardeners alter plant communities and add soil amendments and water; all of these activities control the quality of plant resources for herbivores (Faeth et al. 2011, Ceplova et al. 2017). For example, in California, garden soils with greater water holding capacity support larger plants, and thus higher cabbage aphid densities (Egerer et al. 2018b). Similarly, herbivorous bugs and aphids are more abundant in urban green spaces with higher vegetation volume and host plant density (Mata et al. 2017, Egerer et al. 2018b). Urban agroecosystems also support high crop diversity (compared with rural farms), and evidence from urban green spaces shows species-specific variation in insect responses to plant diversity, with some herbivorous species positively affected by plant species diversity and others negatively affected (Raupp et al. 2009. Mata et al. 2017). This suggests that gardeners hoping to manipulate the composition of the crops they cultivate as a pest management strategy will likely face trade-offs in controlling different herbivore species, and they may have to prioritize the most destructive or populous pests over more innocuous herbivores. Urban landscapes are characterized by a variety of land-use types, and the composition of landscapes surrounding urban agroecosystems may also influence pest populations. In rural agroecosystems. pest densities increase in more complex, well-connected landscapes (Kruess and Tscharntke 1994, Martin et al. 2015). In urban settings, few studies focus on the influence of landscape composition on herbivore pests, and the existing evidence on how urban impervious land cover (a component that strongly affects connectivity) affects pests is inconclusive. One study found a positive correlation between herbivore abundance and impervious land cover (Parsons and Frank 2019) while others have found no effect (Lowenstein and Minor 2018, Egerer et al. 2018b).

Similar to what rural farmers have done for many years, urban gardeners can also manage herbivore populations by manipulating top-down control (Figure 2.1). A common strategy used to enhance biological pest control in rural systems is planting flower strips or borders to prevent pest population build-up (Uyttenbroeck et al. 2016). This practice could be easily adopted by urban gardeners and farmers (Altieri and Nicholls 2018). Prey removal experiments in urban agroecosystems suggest that herbivore control by natural enemies can be very effective, with predators removing 15-100% of sentinel prey within 24 hours (Morales et al. 2018). But understanding more details about the interactions between herbivorous insects and their natural enemies in urban agriculture settings may help further enhance biological pest control and minimize crop losses to herbivory.