Introduction
Shifts in phenology, i.e., the timing of seasonal biological events, are among the most noticeable impacts of human-caused global change (Cleland et al., 2007; Parmesan, 2007). Numerous studies have shown that recent climate warming advances the timing of spring events, e.g., budburst, breeding of amphibians, and arrival of migrating birds (Parmesan, 2007; While and Uller, 2014; Cohen et al., 2018). Beyond global climate change, large-scale landscape modification can also impact phenology. For example, urbanization has been shown to advance the timing of seasonal events such as plant flowering, in part through the urban heat island effect (reviewed in Neil and Wu, 2006). However, more recent work has revealed that urbanization can also delay phenological events, especially in warmer regional climates (Li et al., 2019), although the mechanisms driving these delays have yet to be determined. Overall, there is considerable variation in phenological responses to global change both within and among species (Edwards and Richardson, 2004; Thompson and Clark, 2006; Park et al., 2019). If interdependent species differentially respond to human-mediated changes, phenological mismatches may occur with potentially significant, negative demographic consequences (Miller-Rushing et al., 2010; Renner and Zohner, 2018).
Better understanding insect phenological responses is critical given their vast diversity, temperature-dependent developmental timing, and critical role in ecosystems and the services they provide. These services include dung burial, pest control, pollination, and wildlife nutrition and are valued at over $57 billion annually in the United States (Losey and Vaughan, 2006). Further, several recent papers have reported dramatic declines in insect populations (Hallmann et al., 2020; Wagner, 2020), potentially due to human-caused land-use change, climate change, introduced species, and pollution (Wagner et al., 2021). Phenological shifts may exacerbate losses due to mismatches, but might provide a means to adapt to warmer temperatures and could even lead to overall population growth rates, particularly in species that can successfully add a generation due to extended growing seasons (Kerr et al., 2020).
Most insect phenological studies focus on how climate drives theemergence of insects. In general, warmer-than-average years cause adult insects to emerge earlier (Bartomeus et al., 2011; Roy et al., 2015; Villalobos-Jiménez and Hassall, 2017). Much less is known about what determines termination or total duration of insect activity (Forrest, 2016). Given extended growing seasons for many plant species (Steltzer and Post, 2009), it might be expected that insects also delay termination of adult insect activities in warmer regions. Longer growing seasons are increasing the number of generations per year (voltinism) of some insects (Altermatt, 2010a; Pöyry et al., 2011), but many species are obligate univoltine across their entire range, including warm regions (Forrest, 2016). For these reasons, life history traits, known to be important in determining insect activity (Diamond et al., 2011; Zografou et al., 2021), may strongly determine adult insect termination. For example, Stemkovski et al. (2020) found that timing of bee emergence was most influenced by climatic variation, but termination of adult bee foraging was better explained by life history traits. Specifically, bee species that nest below ground ended foraging earlier than species nesting above, but bee species that overwinter as prepupae ended foraging later than those that overwinter as pupae (Stemkovski et al., 2020).
Interannual regional climate variation is not the only driver of insect phenology. Urbanization is in general leading to earlier flowering in many plant species due to the urban heat island effect, although these responses may be complex and context dependent (Jochner and Menzel, 2015). Less is known about insects’ phenological response to urbanization. The emergence of some insect species appears to be advancing in urbanized areas (Diamond et al., 2015; Chick et al., 2019), but other studies have found no change in phenology across urbanization gradients, despite phenological advances in co-occurring plant species (Seress et al., 2018; Fisogni et al., 2020). The interaction between urbanization and regional temperature can be an important driver of spatial phenology patterns, as urbanization appears to advance plant phenology in cold areas but causes delays in warm areas (Li et al., 2019). Diamond et al. (2014) also demonstrated that the urbanization effect on phenological responses of butterflies depends on regional temperature. This expected interaction between urbanization and temperature remains untested at larger spatial scales and across a broader range of insect groups.
Establishing generalities about determinants of emergence, termination, and duration of adult insect activity is challenging because most phenological research relies on surveys that provide much needed information on species’ population abundance but are limited spatially and taxonomically. Recent enormous growth in open and freely accessible and curated community science photographs, such as those available via the iNaturalist platform, are allowing researchers to ask novel phenological questions at greater spatial and taxonomic scales (Li et al., 2021). However, careful data curation and specialized analytical methods must be used to generate biologically meaningful results (Larsen and Shirey, 2021).
Here we use community science generated digital vouchers and digitized museum specimens to investigate how the emergence, termination, and total duration of adult insects varies spatially in response to climate and urbanization. We also examine how such responses differ across species-specific life history traits. We predict strong interactions between traits and key climate predictors. For example, we expect the seasonal activity of insects with thermally buffered larval stages to be less sensitive to variation in temperature than species without thermally buffered larval stages, aligning with a recent study on bees (Stemkovski et al., 2020). We further predict adult duration to vary based on voltinism, with multivoltine species showing stronger responses to temperature than univoltine species. Lastly, we predict adult insect termination to be later and adult duration longer in warmer and urbanized areas, consistent with recent studies in plants (Li et al., 2021).