Plant–animal interactions in suburban environments: implications for floral evolution

Abstract

Plant interactions with mutualists and antagonists vary remarkably across space, and have played key roles in the ecology and evolution of flowering plants. One dominant form of spatial variation is human modification of the landscape, including urbanization and suburbanization. Our goal was to assess how suburbanization affected plant-animal interactions in Gelsemium sempervirens in the southeastern United States, including interactions with mutualists (pollination) and antagonists (nectar robbing and florivory). Based on differences in plant-animal interactions measured in multiple replicate sites, we then developed predictions for how these differences would affect patterns of natural selection, and we explored the patterns using measurements of floral and defensive traits in the field and in a common garden. We found that Gelsemium growing in suburban sites experienced more robbing and florivory as well as more heterospecific but not conspecific pollen transfer. Floral traits, particularly corolla length and width, influenced the susceptibility of plants to particular interactors. Observational data of floral traits measured in the field and in a common garden provided some supporting but also some conflicting evidence for the hypothesis that floral traits evolved in response to differences in species interactions in suburban vs. wild sites. However, the degree to which plants can respond to any one interactor may be constrained by correlations among floral morphological traits. Taken together, consideration of the broader geographic context in which organisms interact, in both suburban and wild areas, is fundamental to our understanding of the forces that shape contemporary plant-animal interactions and selection pressures in native species.

Urbanization reshapes a food web

“…Beyond preserving food webs for their own sake, species interactions are often integral components of key ecosystem processes with implications for human well‐being (Tylianakis, Didham, Bascompte, & Wardle, ; Tylianakis et al, ). For instance, the degradation of food webs in urban areas may reduce the efficacy of urban agriculture (through pollinator loss; Irwin et al, ) or cause humans to use more pesticide to control insect outbreaks in cities (because of enemy loss; Christie, Cassis, & Hochuli, ). In short, better incorporating our understanding of interaction networks into applied issues of global anthropogenic change may yield an improved understanding of conservation issues, facilitating conservation planning (Tylianakis et al, ).…”

“…Beyond preserving food webs for their own sake, species interactions are often integral components of key ecosystem processes with implications for human well-being (Tylianakis, Didham, Bascompte, & Wardle, 2008;Tylianakis et al, 2010). For instance, the degradation of food webs in urban areas may reduce the efficacy of urban agriculture (through pollinator loss; Irwin et al, 2014) or cause humans to use more pesticide to control insect outbreaks in cities (because of enemy loss; Christie, Cassis, & Hochuli, 2010).…”

“…food web modules). For instance, plant-pollinator interactions may become more or less frequent (Irwin, Warren, Carper, & Adler, 2014), competitive interactions can be rearranged (Francis & Schindler, 2009), and patterns of parasitism can be altered (Theodorou et al, 2016) by urbanization. More generally, it has been repeatedly demonstrated that high trophic-level species tend to be absent from heavily urbanized environments ( Figure 1; Faeth, Warren, Shochat, & Marussich, 2005;Raupp, Shrewsbury, & Herms, 2010;Start et al, 2018), possibly because of an increased sensitivity of predators to fragmentation (Gravel, Massol, Canard, Mouillot, & Mouquet, 2011).…”

Hierarchical filters determine community assembly of urban species pools

“…For instance, a comparison of plant-pollinator networks across the United Kingdom habitats found that pollinator communities in urban areas had higher generality (i.e., they visited a greater number of plant species) but also higher specialization (i.e., they visited a lower proportion of available plant species) compared to pollinator communities in non-urban habitats, due to the higher species richness of non-native plants in cities (Baldock et al 2015). The higher generality of pollinators in urban habitats could reduce pollinator efficiency with negative consequences for plant fitness (Geslin et al 2013, Irwin et al 2014, but see Verboven et al 2014). Other trophic interactions can also break down in urban habitats, such as between hosts, parasitoids, and parasites Forbes 2014, Calegaro-Marques andAmato 2015), between songbirds and nest predators (Rodewald et al 2011, Fischer et al 2012, and between birds and their arthropod prey (Faeth et al 2005, but see Bang et al 2012).…”

Variation in composition of two bumble bee species across communities affects nectar robbing but maintains pollinator visitation rate to an alpine plant, Salvia przewalskii

“…E-mail: qfwang@wbgcas.cn or cfyang@wbgcas.cn pollinators (Inouye, 1980a). Nectar robbing may affect plant reproductive success directly by altering or damaging floral resources or structures and/or indirectly by changing the plant-pollinator interactions, resulting in important ecological and evolutionary consequences (Maloof & Inouye, 2000;Zhang et al, 2009;Irwin et al, 2010Irwin et al, , 2014Zhang et al, 2014;Irwin et al, 2015). Flowering plants with tubular flowers or nectar spurs are likely to be robbed of nectar (Irwin & Maloof, 2002), and animals that are capable of consuming nectar in a way that could potentially lead to pollination may also act as nectar robbers, including species of insects, birds, and even mammals (Irwin et al, 2010).…”