UC Santa Cruz

Plant-pollinator communities vary over landscapes, seasons, and years. This inherently high variation can be linked at least in part to seasonal climates. Early spring temperature and rainfall are important cues for plant germination (Ackerly 2004, Carta et al. 2013, Nonogaki and Nonogaki 2016), flowering (Glover 2008, Tooke and Battey 2010) and other life-cycle events (Chuine et al. 2013). Rearing temperature also influences the growth, development, and consequently emergence time of pollinating insects (Kemp and Bosch 2005, Kingsolver and Huey 2008). While these plant-pollinator cues often align, physiological responses to climate stress in plants or pollinators can induce mismatches that could negatively impact the larger community of interacting species.

Recent climate change alters ecologically important mutualistic interactions that may have far-reaching consequences for ecosystems (Tylianakis et al. 2008, Hegland et al. 2009). For instance, increasingly warmer temperatures cue some plant species to flower at unusual times (Wolkovich et al. 2012) that their pollinators do not always track (Inouye 2008, Lambert et al. 2010, Kudo and Ida 2013, Caradonna et al. 2014). Resource-based mismatches may also develop as plants physiologically respond to environmental stress by allocating resources away from reproduction and toward survival (Memmott et al. 2007, Scaven and Rafferty 2013, De la Luz 2018a). On a geographical scale, species’ ranges may shift as pollinators expand their foraging ranges to higher elevations or latitudes over time to escape temperature stress (Kelly and Goulden 2008, Parmesan and Hanley 2015).

Resource exchange between plants and pollinators ultimately influences population-level reproductive success for both trophic levels (Wang and Smith 2002, Bascompte and Jordano 2014a). Despite the ecological importance of pollination for both biodiversity and ecosystem function (Potts et al. 2006, Bascompte and Jordano 2007, Albrecht et al. 2012, Hanley et al. 2015, IPBES 2016) there are only a few known plant-pollinator datasets that cover a long enough time frame to specifically address the potential negative impacts of recent climate change or other stressors on those communities in natural ecosystems (reviewed in Burkle and Alarcón 2011, Burkle et al. 2013).

My dissertation explores how individual physiological responses to temperature and water stress, scale up to impact plant-pollinator network structure, ecosystem function, and biodiversity. I examine how plant-pollinator networks in two habitats are structured in relation to intrinsic abiotic stressors, and how those habitats have changed over the past twenty years. In the lab and greenhouse, I test how temperature and water availability influence nectar output in three native plant species. Then I follow bumblebee foraging patterns on those plant species in controlled choice trials. In a separate set of experiments, I test how nectar diet influences a bumblebee’s ability to metabolically cool their bodies while under acute heat stress.

Chapter one builds upon an existing dataset to examine how plant-pollinator network structure and stability vary over time in two habitats. Plant-pollinator communities are made up of pairs of mutualistic species. These communities are sometimes represented as bipartite interaction networks, where pollinator and plant species are ‘nodes’ and the interactions between them are ‘links’ (Bascompte and Jordano 2014b). Mutualistic ecological networks are often nested, a structure that develops out of an asymmetric arrangement of species interactions. Specifically, nestedness is a tendency for rare, usually specialist species to interact with generalist, usually common partners and for generalist species to most often interact with other generalists. This creates properly nested subsets of interacting species (Bascompte et al. 2003, Guimarães et al. 2006, Pires et al. 2011) that should buffer a community against ‘extinction cascades’, which are secondary extinctions due to the loss of an interaction partner(s) (Bascompte and Stouffer 2009). In a perfectly nested network, the inner-most species subset are ‘hub’ species, highly linked species that disproportionately contribute to ecosystem function (Olesen et al. 2007, Bascompte 2009, Guimarães et al. 2011). Identifying ‘hub’ species in a network and understanding how those species might directly respond to stress is important given the outsized role they likely play in species and network maintenance.

Mutualistic network theory predicts that a diverse and highly nested network should be more robust to climate stress than one that is less diverse or with a less defined structure (Bascompte and Jordano 2014a). To test this, I created plant-bee networks from historical survey data (years: 1991-1993) for two California habitats with different baseline temperatures and water availabilities: coastal grasslands and endemic Santa Cruz sandhills. I evaluated nestedness as a metric of network stability and identified changes in composition and structure in each network. Next, I resurveyed (years: 2013-2015) a subset (n=9) of the original survey sites (n=14) and evaluated how plant-pollinator networks changed in the ~20 years between surveys. During that time, both habitats experienced combined temperature stress, drought, and species invasions.

Chapter two tests resource-based mismatches as one possible mechanism that disrupts plant-pollinator interactions. Plants often allocate resources toward survival in favor of lifetime fitness when they are stressed (e.g., life history trade-off; Stearns 1989, Ashman et al. 1994). For instance, perennial plants might postpone reproduction within a season, or across years until conditions are more favorable (Willmer 2011a); annual plants have less flexibility and may produce fewer, smaller flowers that are less attractive to bees. For plants, these strategies conserve resources to support reproduction through to seed (Pleasants and Chaplin 1983, Galen 2000, Carroll et al. 2001, Liu et al. 2012). However, small, resource-poor flowers often receive shorter and less frequent visits from bees (McCallum et al. 2013), which reduces seed production in many plant species (Wright and Schiestl 2009, Burger et al. 2010, Willmer 2011b, Essenberg et al. 2015, Milet-Pinheiro et al. 2015).

I initially grew 8 California native plant species (all visited by bumblebees in the field) in five temperature and humidity–controlled growth chambers and under three levels of water stress. Only three of those species survived under the most stressful temperature and water conditions with enough replication to measure plant and floral traits. Then, for each plant species, I offered one plant from each of the 15 temperature/water combinations to single bumblebee workers in foraging choice trials, where I recorded each bee’s behaviors over a full array of experimental plants for 20 min. To determine the downstream impact of temperature and water availability on plant fecundity through foraging, I measured seed production in choice trial plants that survived to the end of the experiment.

Bumblebee diet and heat stress: In Chapter three I test how nectar diet could limit a bumblebee’s ability to cool down when heat stressed. The interface of nectar diet and temperature regulation in bumblebees, or other bees that can thermoregulate, has been studied though largely in the context of warm-up to prepare the flight muscles for foraging at cooler temperatures (Heinrich 1976). The recent large-scale declines in bumblebee diversity and abundance (National Academy of Sciences 2007, Goulson et al. 2008, Potts et al. 2010, Cameron et al. 2011, Koch 2011, Hatfield et al. 2015a, Thomson 2016) have been attributed in part to disease, a decline in floral resources, habitat fragmentation, and climate stress, but, the specific mechanisms driving species losses remain unclear.

Bumblebees are among a group of large-bodied insects capable of active thermoregulation (Heinrich 1979, May 1979). This physiological adaptation allows bumblebees to persist in cold climates as far north as Alaska (Heinrich and Vogt 1993). In temperate climates, bumblebees “shiver” to warm flight muscles so they can forage longer and at cooler temperatures than smaller bees that remain in torpor when cold (Esch et al. 1991, McCallum et al. 2013). Bumblebees use carbohydrate-rich nectar to offset the energetic costs of metabolic warm-up and flight (Esch et al. 1991, Heinrich and Vogt 1993, McCallum et al. 2013). High temperatures induce increased activity and foraging rates in bumblebees; however, bumblebees can also easily overheat due to their high flight metabolism (Heinrich 1977). Bumblebees could also use metabolic energy to quickly offload excess body heat (via convection) in response to temperature stress. One study addresses bumblebee thermoregulation in the context of climate stress, but at the colony level (Holland and Bourke 2015), and I have not found any papers that address the impact of nectar diet on the ability of individual bumblebees to thermoregulate in response to heat stress.

To address the question of thermoregulation and nectar diet in bumblebees, I tested three bumblebee species’ abilities to offload excess heat after being fed a pre-determined nectar diet in the lab. I restrained individual bumblebees on a Styrofoam platform with insect pins around the waist (petiole) and partitioned the abdomen and thorax with an aluminum heat–shield. I used a heat lamp to heat the head and thorax (but not to the shielded abdomen) and measured the

temperature change between body segments every 30 seconds for five minutes. Understanding how pollinators tolerate high-temperature stress is important, particularly as average spring temperatures exceed record highs in a trend that is likely to continue moving forward (Parmesan 2006)