Wesleyan University The Honors College

Plant-pollinator interactions across California grassland and coastal scrub vegetation types on San Bruno Mountain, San Mateo County

by

Miles Gordon Brooks Class of 2020

A thesis submitted to the faculty of Wesleyan University in partial fulfillment of the requirements for the Degree of Bachelor of Arts with Departmental Honors from the College of the Environment

Middletown, Connecticut April, 2020

1

2 Abstract

Animal pollination of plants is a crucial ecosystem service for maintaining biodiversity and ecosystem function, worldwide. High pollinator abundance and diversity can likewise improve the reproductive success of the plant community.

Plant-pollinator interaction networks have the potential to identify dominant, specialist, and generalist pollinator species within a system, and their host plant counterparts. Understanding these relationships is paramount for buffering natural systems from biodiversity loss in a world where pollinator abundance continues to decline rapidly. San Bruno Mountain (SBM) in San Mateo County, California, is one of the last natural, open spaces in the urban landscape in the northern San

Francisco Peninsula. I conducted a series of timed meanders and vegetation surveys at eight sample sites within SBM (four grassland and four coastal scrub sites) to identify plant species prevalence and pollinator species visitation of flowering plants. I employed a multivariate approach for investigating plant and pollinator species richness, plant and pollinator community composition, and trophic-level interactions across the SBM landscape, and I evaluated differences in these relationships between grassland and coastal scrub habitats. A total of 59 pollinator species and 135 plant species were inventoried over the course of the study. While species richness did not vary significantly between vegetation types, the nonmetric multidimensional scaling results revealed significant differences in species composition and key indicator species between vegetation types. The bipartite analyses identified Bombus vosnesenskii, Eriophyllum stachaedifolium,

Grindelia hirsutula and others as generalist pollinator and plant species that are

3 important for the long-term biodiversity conservation of SBM due to their interactions with a diverse array of other plant and pollinator taxa. In the future, adaptive restoration activities could be used at SBM and other similar habitats to bolster the abundance of herbaceous flowering plants for pollinators to conserve biodiversity and promote ecosystem health in a world that continues to experience declines in pollinator abundance.

4 Introduction

Bees and other wild pollinators provide an important ecosystem service to agriculture crops and natural environments (Kearns et al. 1998, Aguilar et al.

2006, Klein et al. 2007, Kremen et al. 2007, Tuell et al. 2008, Potts et al. 2010).

Pollination is the act of transferring pollen grains from a plant’s male reproductive parts, the anthers, to a plant’s female parts, the stigma or pistil (Meeuse 2018).

Pollination is often facilitated by animals, who feed on the nectar and pollen of flowering plants. Approximately 35% of crops and 80% of native plants rely on animal pollination for reproduction (Klein et al. 2007, Potts et al. 2010). As such, pollination maintains biodiversity, facilitates the plant community’s reproductive health, and provides ecosystem services to ecosystems, worldwide (Waser et al.

1996, Kearns et al. 1998, Aguilar et al. 2006, Greenleaf and Kremen 2006, Klein et al. 2007, Kremen et al. 2007, Black et al. 2009, Potts et al. 2010, Elle et al.

2012, Smith DiCarlo et al. 2019).

High pollinator diversity facilitates plant community diversity and ecosystem function (Aguilar et al. 2006, Fontaine et al. 2006, Kremen et al. 2007,

Tuell et al. 2008, Blüthgen and Klein 2011). Many types of pollinators exist, but bees make up a significant proportion of pollinators around the world (Potts et al.

2003). They are one of the most diverse groups of insects, with an estimated

20,000 species, globally (Michener 2007, Smith DiCarlo et al. 2019). Other types of pollinators include butterflies, wasps, birds, and flies. These pollinator species have different plant preferences that allow them to fill complementary niches to bees (Fontaine et al. 2006, Blüthgen and Klein 2011, Venjakob et al. 2016,

5 Iwasaki et al. 2018). Greater pollinator diversity benefits plant community productivity because high pollinator diversity increases chance of plant pollination (Fontaine et al. 2006, Greenleaf and Kremen 2006, Tuell et al. 2008,

Potts et al. 2010). Higher plant species richness provides a feedback for pollinator species richness because it boosts resource availability, which attracts a wider range of pollinators taxa to an area (Potts et al. 2003, Ebeling et al. 2008, Carman and Jenkins 2016, Maia et al. 2019). Greater pollinator diversity increases plant visitation frequency and the pollinator population’s resilience to perturbation, such as extinction of a plant species (Memmot et al. 2004, Fontaine et al. 2006,

Ebeling et al. 2008, Tylianakis et al. 2010, Elle et al. 2012). Plant-pollinator interactions and richness are therefore crucial ecosystem health and integrity metrics.

Recent global declines in wild pollinator abundances threaten community health, on scales spanning from local to global (Kearns et al. 1998, Aguilar et al.

2006, Klein et al. 2007, Kremen et al. 2007, Williams et al. 2009, Winfree et al.

2009, Potts et al. 2010, vanEngelsdorp et al. 2017). The loss of pollinator diversity has stimulated new interest in understanding the diversity and distribution of pollinator species across a range of habitats and land cover types

(Grundel et al. 2010, Colteaux et al. 2013, Meiners 2016, Campell et al. 2018,

Luong et al. 2019, Smith DiCarlo et al. 2019). Climate change, pesticide use, and the increased prevalence of pests are responsible for bee population decline

(Kearns et al. 1998, Black et al. 2009, Potts et al. 2010, vanEngelsdorp et al.

2017). Habitat loss and fragmentation comprise other anthropogenic stressors on

6 wild pollinator populations (Aguilar et al. 2006, Fortuna and Bascompte 2006,

McFrederick and LeBuhn 2006, Kremen et al. 2007, Goulson et al. 2008, Black et al. 2009, Winfree et al. 2009, Carman and Jenkins 2016). Bee communities depend on the ecosystem integrity (Kremen et al. 2007, Grundel et al. 2010,

Ullmann et al. 2010), and the plant community depends on the presence of pollinators (Aguilar et al. 2006, Fontaine et al. 2006, Greenleaf and Kremen 2006,

Klein et al. 2007). Further pollinator population decline could lead to further decline in biodiversity in the vegetation community (Memmot et al. 2004, Klein et al. 2007, Lundgren et al. 2016, Venjakob et al. 2016). While a myriad of negative effects from pollinator decline are predicted for the future, understanding contemporary plant-pollinator relationships in natural areas that span different habitats and plant cover types represents a key step for predicting the effects of future global change on pollinator population dynamics.

Bipartite networks, networks that consist of interactions between two trophic levels, are useful ways to assess and conserve the mutualistic relationship between plants and pollinators (Blüthgen et al. 2008, Dormann et al. 2009,

Bascompte 2010, Tylianakis et al. 2010, Elle et al. 2012, Carman and Jenkins

2016, Petanidou et al. 2018, Maia et al. 2019). Unlike traditional measures of conservation that only measure the abundance of a key species or community species richness, plant-pollinator networks capture interactions and their frequency between the two trophic levels (Tylianakis et al. 2010, Elle et al. 2012).

In plant-pollinator networks there tend to be a few well-connected, generalist species that promote diversity by interacting lots of species (Waser et al. 1996,

7 Lopezaraiza-Mikel et al. 2007, Maia et al. 2019). Identifying key generalist plants and pollinators through bipartite networks allows conservation managers to determine important species for conservation to maintain ecosystem function and promote resilience to extinction (Memmot et al. 2004, Maia et al. 2019).

Understanding the mutualistic interactions between plant and pollinators is paramount for buffering natural systems from biodiversity loss in a world where pollinator abundance continues to decline rapidly.

San Bruno Mountain (SBM) in San Mateo County, California is an important site for conservation because it is an ecological ‘island’ surrounded by an urban matrix (Figure 1). The mountain has been protected by the San Bruno

Mountain Habitat Conservation Plan since 1982 (Ormshaw 2018), and has been called one of the most important and threatened biodiversity sites in the world

(Wilson 1999). SBM hosts both northern coastal scrub and grassland plant cover types, and a rich pollinator community. A lack of disturbance from wildfires, which were common on the mountain before European colonization, has resulted in widespread scrub encroachment and a subsequent decline of grassland community cover in the absence of fire (from 793 hectares of grassland in 1932 to

478 in 2014) (Weiss et al. 2015). Fire frequency is a well-known influence on the abundance and distribution of both woody and herbaceous vegetation in coastal habitats, where a lack of fire often results in shrub dominance at the expense of graminoid-dominated communities (Moyes et al. 2005, Zavaleta and Kettley

2006, Knapp et al. 2007, NPS 2007, Eviner 2016, Carlsen et al. 2017, Poulos et al. 2020). Disturbances can likewise affect pollinator community composition and

8 dynamics through mortality, injury, or displacement (Black et al. 2009, Campell et al. 2018, Smith DiCarlo et al. 2019). Yet, little is known about how different fire-maintained vegetation types support pollinator communities across landscapes.

SBM is home to several endangered pollinator species including the

Mission Blue Butterfly (Icaricia icarioides missionensis Hovanitz), the Callippe

Silverspot (Speyeria callippe callippe Boisduval), and San Bruno Elfin (Incisalia mossii bayensis Brown), which signals the importance of the site as a key biodiversity conservation area in an otherwise human-dominated, urban landscape. The abundance of these three endangered pollinators has been recorded across different transects over time (Reid et al. 1980, Ormshaw 2018), and results from this work highlight butterfly populations over different habitats; however, these butterflies only fly between March and June (Ormshaw 2018).

The prevalence of pollinators on SBM outside those months has yet to be examined. We know little about the population abundance or feeding behavior of other pollinators taxa at SBM, or other similar coastal habitats in northern

California. Such information is valuable for evaluating plant and pollinator community resilience in a large conservation area of San Francisco Bay that continues to be at risk of anthropogenic environmental stress due to habitat fragmentation and development.

In this study I examined plant and pollinator community dynamics and plant-pollinator interactions in two dominant habitat types in SBM. Study objectives included: 1) quantifying landscape, plant and pollinator species

9 richness, 2) comparing these patterns between grassland and coastal scrub vegetation types, 3) identifying indicator species for each vegetation type, and 4) modeling plant-pollinator interactions to elucidate key relationships across trophic levels at the landscape-scale, and between grassland and coastal scrub habitats. I hypothesized that grassland environments would host significantly higher pollinator abundance and richness based on the importance of this land cover type to the Mission Blue and Callippe Silverspot butterflies and the prevalence of forbs

(i.e. non-woody, herbaceous flowering plants) in graminoid-dominated cover types, worldwide.

Study Area

This research took place on San Bruno Mountain, a State and County Park located 14 km directly south of downtown San Francisco. SBM stands 402 meters tall, and consists of 1,214 hectares of open space (McClintock et al. 1990) (Figure

1, Figure 2). The mountain’s rocks formed over 100 million years ago. Marine sediment deposited during the Cretaceous Period created the greenish-grey sedimentary rock, called graywacke, found across the mountain (McClintock et al. 1990). The SBM range took shape around a million years ago when the earth’s crust fractured and caused the mountains to rise relative to other fault blocks

(McClintock et al. 1990). The range of hills extends from San Bruno Mountain through San Francisco to Land’s End. Since the mountain has uniform Franciscan graywacke bedrock material throughout, there is little variation in soil type

(McClintock et al. 1990), although soil depth varies depending on slope steepness,

10 with steep slopes having thin soil while thicker soil profiles occur on flatter slopes

(McClintock et al. 1990).

SBM’s topography and the regular presence of fog creates various microclimates. The western slopes of the mountain can experience cool wind and moisture from fog, while the eastern slopes will be warm and still. During the summer months (June-August), when this research was conducted, the weather on the mountain could be warm with clear skies until the fog rolled in the afternoon, or it could be completely covered by thick fog with high winds, cold temperatures, and heavy condensation throughout the day. The presence or absence of fog is influenced by the presence or absence of a high pressure system over the Pacific Ocean (McClintock et al. 1990, Gilliam 2002). The average summer temperatures range from a high of 18.8° C to a low of 10.7° C

(McClintock et al. 1990).

San Bruno Mountain is one of the few remaining natural environments in the Bay Area for many plants and animal species, which gives the mountain high conservation importance. San Bruno Mountain Watch, a local non-profit started in the 1970s, fights to protect the natural ecosystem on San Bruno Mountain from development and encroaching non-native species. Teams of volunteers, State and

County park staff, and Mountain Watch staff work together to manage the native habitats on the mountain.

The mountain is home to several vegetation types including grassland, coastal scrub, chaparral, riparian wetland, and woodland (McClintock et al. 1990).

Grassland and coastal scrub are the two predominant habitat types on the

11 mountain, and they are therefore the focus of this study. The Northern

(Franciscan) Coastal Scrub cover type dominates the north and west-facing slopes and ravines, which experience frequent fog, moisture, and strong winds

(McClintock et al. 1990). Common coastal scrub plant taxa include coyote brush

(Baccharis pilularis DC.), California sagebrush (Artemisia californica Less.), sticky monkey flower (Diplacus aurantiacus Curtis), and poison oak

(Toxicodendron diversilobum Greene) (McClintock et al. 1990). Sagebrush dominates on poor, shallow soils and inhibits the growth of adjacent plants. When coyote brush is also present the plant community becomes denser, taller, and reflects better soil conditions (McClintock et al. 1990). Grassland vegetation types like the Valley Needle Grassland and Coastal Terrace Prairie are found on drier, south, south-west, and east facing slopes and valleys. Prominent grassland plants include grasses like Purple Needle Grass (Stipa pulchra Hitchc.), California Oat

Grass (Danthonia californica Bol.), California Fescue (Festuca californica

Vasey), and forbs like California Poppy (Eschscholzia californica Cham.) and

Farewell-to-Spring (Clarkia rubicunda H. Lewis & M. Lewis). Both habitats contain noxious exotic plants such as Fennel (Foeniculum vulgare Mill.), Italian

Thistle (Cirsium vulgare Ten.), and French Broom (Genista monspessulana L. A.

S. Johnson), which compete with native plants (Ormshaw 2018).

SBM has a rich history and unique ecology. The mountain supports a diverse community of plants and animals, including many native, endemic, endangered, and range-limited species (Ormshaw 2018). At the end of the

Pleistocene Epoch when sea levels were much higher, the northern San Francisco

12 Peninsula, including SBM, was an island, and the Pacific Ocean lapped at its western border (~80,000 to 120,000 years ago) (Molanphy 2016). Today, SBM is essentially an island of biodiversity. It is surrounded by urban landscape on three sides (San Francisco to the North, Colma and Daly City to the West, and South

San Francisco to the South) and the San Francisco Bay on the eastern side (Figure

1). This presents a barrier for plant and animal migration outside the state park.

This similarly puts severe population pressure on San Bruno Mountain’s smallest inhabitants, its pollinators.

The mountain’s three endangered butterflies (the Mission Blue, the

Callippe Silverspot, and San Bruno Elfin) rely on specific host plants to lay their eggs, to feed their larvae, and to provide nectar, including perennial lupines

(Lupinus albifrons var. Collinus Greene, L. formosus Greene, L. variicolor

Steud.) for the Mission Blues, viola (Viola pedunculata Torr. & A. Gray) for the

Callippe Silverspots, and Pacific Stonecrop (Sedum spathulifolium Hook.) for the

San Bruno Elfins (McClintock et al. 1990, Ormshaw 2018). The Mission Blue’s and Callippe Silverspot’s host plants are found in grasslands across the mountain, whereas the San Bruno Elfins’ host plant establishes predominantly on North- facing slopes (McClintock et al. 1990, Ormshaw 2018). SBM is protected from further development by the Habitat Conservation Plan because of the endangered status of these butterflies in the area (Longcore et al. 2010, Weiss et al. 2015,

Ormshaw 2018). Since two of these endangered species rely on grasslands for survival, grassland conservation has been a major focus of active vegetation management by conservation planners.

13 For thousands of years, until the Spaniards colonized the area in 1769, the indigenous Ohlone people inhabited SBM and the surrounding San Francisco Bay

Area. The Ohlone are the first evidence of anthropogenic disturbance on the mountain. They used fire to maintain grassland vegetation types, which attracted the grazers that they hunted. When the Spanish settled the area, they brought cattle that grazed on the mountain. Grazing is another form of disturbance that prevents scrub encroachment (Reid et al. 1980, Ormshaw 2018). The lack of a disturbance on SBM has led to the loss of grassland habitat and a corresponding shift towards woody shrub dominance on the mountain. Since 1932, an average of

3.8 hectares per year of grassland has been lost to coastal scrub habitat (793 hectares in 1932 to 525 hectares in 2004 and an estimate of 478 hectares in 2014)

(Weiss et al. 2015). Today, grassland habitats are maintained through manual and chemical scrub control because of the constraints of applying prescribed fire in an urban-wildland interface and due to the cost of controlled burns and grazing

(Beattie et al. 2017, Ormshaw 2018). Managers of San Bruno Mountain target shrubs like coyote brush or fennel using hand tools, brush cutters, or herbicide

(Ormshaw 2018).

Fires still naturally occur on the mountain (Allshouse 2017), but fire suppression and vegetation management has limited the frequency and intensity of fires. The most recent fires on San Bruno Mountain were a 4.5 hectare fire along the ridgeline in October of 2019 (CBS 2019), a ~16 acre fire in Juncus

Ravine in 2013 (Allshouse 2016), and a 121 hectare fire in Buckeye and Owl

Canyons in 2008 (Allshouse 2017). Fires in Coastal Scrub environments tend to

14 be more intense than grassland fires because of increased woody biomass in these sites, and they trigger lush new growth due to the increased space, nutrients, and high soil moisture (McClintock et al. 1990), while grassland fires move faster due to the lighter fuel loads and lack of woody vegetation (Allshouse 2016).

Study Sites

I established eight plots across San Bruno Mountain–four grassland sites and four coastal scrub sites during the summer of 2019 (Figure 2). To acquire a representative sample of the range of plant and pollinator communities on the mountain, the sites were randomly scattered across the mountain at different elevations and facing different directions. Site selection was limited to areas accessible by trail because, as one of the last large, contiguous conservation areas in the San Francisco Bay Area, I had to minimize impacts on the vegetation from sampling as requested by the San Mateo County Parks. I delineated the sample sites using Google Earth Pro and the advice of San Bruno Mountain Watch’s director Ariel Cherbowsky Corkidi. Sample sites were distributed among grassland and coastal scrubland plant communities based on the cover of different plant-types. Grassland sites had high grass cover and scrub cover was less than

10%, whereas coastal scrub sites, similar to a closed-canopy coastal chaparral, were defined as a continuous and often overlapping shrub canopy with shrub cover >60%. A hand-held GPS (Garmin eTrex 20) was used to record the outline of each study area, which was later imported into ArcGIS (version 10.7.1 ESRI

2019, Redlands, CA). Sampling took place in July and August of 2019.

15 Grassland site one (G1) on the south-east ridge is at the end of the ridge trail above Brisbane acres. The area is the last to get fog on a summer afternoon, and thus often has more sun exposure and less moisture than other parts of the mountain. This area intersects some of the mountain’s best lupine habitat

(Ormshaw 2018). It has high habitat value for Mission Blue and Callippe

Silverspot butterflies and moderate habitat value for plant diversity and dominance (Ormshaw 2018). In this area the County Parks department has used herbicides to treat fennel and coyote brush invasions. This site has high priority for scrub control to maintain grassland and butterfly habitat.

Grassland site two (G2) borders the ridge trail above Owl Canyon and follows down the north-facing slope of the mountain. Since it is along the top of the ridge, it receives a lot of wind and cloud cover. The convection of moist air to the higher altitudes causes fog to form regularly over the top of the mountain.

There were high observations of Mission Blue and Callippe Silverspot in this area in 2017 (Ormshaw 2018). This area has high priority for scrub control, but it has not received much maintenance because of relatively slow rate of scrub encroachment and the difficulty to reach it (Ormshaw 2018).

Grassland site three (G3) surrounds the Hillside ridgeline and faces the south-west direction (Figure 3). The area has high habitat value for Mission

Blues, and moderate habitat value for Callippe Silverspot and native plant diversity (Ormshaw 2018). This area is treated for fennel, prickly lettuce, Italian thistle, and coyote brush using herbicide and hand removal (Ormshaw 2018). San

Bruno Mountain Watch and volunteers have planted native grasses and nectar

16 plants within this area for the past couple years (Cherbowsky Corkidi 2018). In

2013 there was a fire near the Hillside ridge, which burned around 16 hectares of grassland and Eucalyptus (Allshouse 2016). The following spring, the area had lots of wildflowers.

Grassland site four (G4) is on the ridgeline east of Buckeye Canyon, which is on the north side of the mountain. This area is frequently sunny, but experiences high winds when the fog rolls in. There are a series of power lines along the gravel PG&E road on the ridge. There is high habitat value for endangered species and native plant community diversity (Ormshaw 2018). The site is covered by native grassland and prairie communities. There is a low extent of invasive, non-natives herbs and scrub encroachment due to active and frequent vegetation management through scrub removal and native plantings (Ormshaw

2018, Cherbowsky Corkidi 2018). There is an abundance of nonnative Scabiosas, whose removal is ongoing. In 2008 there was a massive, 121 hectare fire that burned most of the area (Allhouse 2017).

Coastal scrub site one (CS1) borders the main ridge trail and the hillside ridge trail on a south-west facing slope. When fog comes in, this part of the mountain experiences high winds and moisture as air rises up the slopes of the mountain. This area is composed of dense coastal scrub and rocky outcrops. No

Mission Blue or Callippe Silverspot were observed in this location in 2017, but nearby many San Bruno Elfin larvae were identified (Ormshaw 2018). This site has no known disturbance or management history.

17 Coastal scrub site two (CS2) is on the western side of the mountain on a north-facing slope off the Summit Loop Trail (Figure 4). The area is characterized by coastal scrub and rock outcrops that are slightly sheltered from the westward winds. The site has low habitat value for Mission Blue and Callippe Silverspot habitat, but high habitat value for native plant community diversity and moderate value for San Bruno Elfin habitat (Ormshaw 2018). CS2 is not prioritized for scrub control or treatment.

Coastal scrub site three (CS3) is located in the Dairy Ravine, which ranges from north of the summit to the parking lot. The site consists mostly of coastal scrub vegetation and rock outcrops. It receives lots of moisture because its north- facing slope does not easily evaporate moisture. There is moderate native plant community diversity and high habitat value for the San Bruno Elfin (Ormshaw

2018). There is no known recent disturbance or management.

Coastal scrub four (CS4) on the northern saddle neighbors a grassland community with moderate habitat value for Mission Blue and Callippe Silverspot.

The site consists of encroaching coastal and non-native scrub. It has moderate native plant diversity. The perimeter of the site was targeted for maintenance to support Mission Blue and Callippe Silverspot populations (Ormshaw 2018).

18 Methods

Vegetation Survey

I estimated plant species richness using a timed meander approach following Goff et al. (1982). I employed the timed meander by walking through each study area for at least 45 minutes (up to 90 minutes) and recording every new plant species I encountered. I captured the variation in vegetation type within each sample site by deliberately choosing a path to capture the most variation in plant species, which is consistent with the original methodology of Goff et al.

(1982). I continued the meander until no new species were encountered for at least 10 minutes. All plants were identified to the species-level (Beidleman and

Kozloff 2003). Unknown plants in the field were photographed and later keyed to species using the help of San Bruno Mountain Watch program director Ariel

Cherbowsky Corkidi or via iNaturalist (https://www.inaturalist.org).

Pollinator Survey

I conducted a separate pollinator-focused, one-hour timed meander to calculate the abundance of pollinators, pollinator species richness, and plant- pollinator relationships in the different habitat types following Westphal et al.’s

(2008) variable transect method. Before and after each pollinator meander, I also recorded the survey date, the temperature, and wind speed. Temperature and wind speed were important potential metrics influencing pollinator abundance because pollinators at San Bruno Mountain require a minimum temperature of 20°C and wind speeds less than 16 kmph, with a few exceptions (Ullmann et al. 2010,

Iwasaki et al. 2018). Temperature and wind speed were measured with a digital

19 anemometer (HoldPeak, Zhuhai, China) at one meter above the ground to mimic the approximate flight height range of pollinators within the study area following

Miller et al. (2018). All pollinator observations were conducted between 9am and

4pm. The pollinator timed meander involved moving through the plot and recording every pollinator that I observed visiting a flower (Westphal et al. 2008,

Ullmann et al. 2010). “Visiting,” in terms of pollination, means that an organism comes into contact with the reproductive parts of the plant (Figure 5) (Ullmann et al. 2010). I identified both the pollinators and the visited plant to the lowest taxonomic level possible in situ (Tuell et al. 2008, Ullmann et al. 2010), or after taking a picture and consulting iNaturalist.

Pollinator Habitat Characteristics

To understand the plant community composition in the different habitats, I sampled vegetation structure within each timed meander matrix in 5 x 5 m plots.

Using a random number generator, I walked off-trail in a random direction for a random distance from the middle of the study area to place the plot. I was limited in my off-trail access due to instructions from San Mateo County Parks

Department to avoid areas of endangered butterfly host plants. In each plot, I recorded the percent cover and species richness of all grasses, forbs, shrubs. I estimated percent cover for each plant group in one of six cover classes (< 1%, 1-

4%, 5-24%, 25-49%, 50-74%, 75-100%). Other factors measured at each site included percent rock and canopy cover (at 1 m above ground), visual obstruction, and the spatial location of each plot center using a hand-held Garmin eTrex20

20 GPS. Visual obstruction was measured by placing an upright meter stick four meters away from the observer following Robel et al. (1970). I recorded the lowest visible height (cm) on the meter stick in order to understand the height and density of vegetation at each site. For every plant species inside the subplot, I recorded the abundance of individual plants, maximum and mean plant height

(cm), mean distance to the nearest plant (cm), and current flowering status

(flowering or not flowering).

Data Analysis

T-tests

I tested for differences in the number of plant species and flowering plants species between grassland and coastal scrub via unpaired t-tests. I evaluated the differences in the number of total pollinator observations and pollinator species between grassland and coastal scrub via unpaired t-tests. I also used unpaired t- tests to evaluate the differences in the vegetation characteristics between grassland and coastal scrub sites.

Nonmetric multidimensional scaling

Variation in plant and pollinator species composition between grassland and coastal scrub sites was analyzed using nonmetric multidimensional scaling

(nMDS) in the vegan package (Oksanen et al. 2019) in version 3.6.1 of R

Statistical Language (R Core Team, 2019). Separate nMDS analyses were run on

1) the plant presence/absence matrix and 2) the pollinator abundance matrix from the timed meander surveys. nMDS was performed by first converting these

21 matrices into Bray-Curtis dissimilarity matrices and then the nMDS was run using random starting configurations. nMDS does not assume linear relationships between species and environmental gradients, and it creates ordinations based on dissimilarity between species (Poulos et al. 2007, Omand et al. 2018, Oksanen

2019, Smith DiCarlo et al. 2019). Plots were then plotted in nMDS species space with 95th percentile confidence ellipses to display species compositional differences between grassland and coastal scrub vegetation types.

The plant and pollinator species that significantly contributed to the dissimilarity in the species composition between grassland and coastal scrub vegetation types were evaluated using similarity percentage analysis and Bray-

Curtis dissimilarities. This test computes the percentage contribution of each species to the dissimilarities, standard deviation, average abundance in each vegetation type, and the cumulative sum of dissimilarity. Species with the largest contribution to dissimilarity underwent a randomization test with 999 permutations to test for statistically significant correlations to the nMDS species scores, separately for the vegetation and pollinator nMDS solutions. Species with significant P-values (alpha <0.05) were considered significant indicator species of either grassland or coastal scrub. These analyses were conducted separately for the plant species matrix and the pollinator species matrix.

Plant-pollinator networks

I used the bipartite package (Dormann et al. 2008) in R to evaluate plant- pollinator interaction networks for grasslands and coastal scrub. Network

22 structure metrics were calculated by creating a weighted plant-pollinator network.

Bipartite networks identify connections (weighted by frequency) between species of one trophic level (i.e. pollinators) and another (i.e. plants). All plant-pollinator interactions recorded throughout the sampling period were combined by vegetation type for the bipartite analysis following Carman and Jenkins (2016).

Network metrics including connectance, nestedness, interaction strength asymmetry, H2’, niche overlap, and partner diversity, were calculated for grasslands and coastal scrub in an effort to characterize vegetation type-specific interactions among plants and pollinators. Connectance is the proportion of observed connections to possible connections (Dormann et al. 2009).

Connectance in plant-pollinator networks tends to be low, but higher values can signify greater network stability (Elle et al. 2012). Nestedness quantifies the degree to which the network organizes around a subset of generalist interactions

(Elle et al. 2012, Carman and Jenkins 2016). It indicates network stability because nestedness means that specialist plants interact with the same pollinators as generalist plants or the vice-versa with specialist and generalist pollinators

(Memmot et al. 2004, Fortuna and Bascompte 2006). Interaction strength asymmetry quantifies the imbalance in relationship between species in the two trophic levels, where one trophic level might be more reliant on the other than vice versa. For this measurement, I had to eliminate all ‘singletons’ or species that

I only encountered once, because that would skew the value (Dormann et al.

2009). Higher positive asymmetry values indicate higher specialization of the pollinators (Dormann et al. 2009). H2’, the specialization index, characterizes the

23 network-level measure of specialization (Dormann et al. 2009, Elle et al. 2012). It ranges from 0 (no specialization) to 1 (perfect specialization for given interactions). Niche overlap measures the mean similarity in interactions between species of the same trophic level and varies from 0 (no common use of niches) to

1 (perfect niche overlap) (Dormann et al. 2009).

I also used bipartite to create a plant-pollinator network map to examine the strength and abundance of species-specific plant-pollinator interactions. Plant- pollinator bipartite networks display a truncated power law degree distribution

(Dormann et al. 2009), meaning there are many interactions that rarely occur and a small amount of high occurring interactions. So, I filtered the observations to only include pollinators that were observed greater than 1% of the total observations in that habitat type following Iwasaki et al. (2018). I then superimposed the results of the indicator species analysis to identify plant- pollinator relationships for significant indicator species in grasslands and coastal scrub.

24 Results

Across all sites, I recorded 135 different plant species. I identified ninety- one plant species in four coastal scrub sites and ninety-seven in the grassland sites

(Figure 6; not significant, t-test = 0.899; P-value = 0.435) (Supplementary Table

1). During the pollinator visitation observations, twenty-five flowering coastal scrub plant species were visited by pollinators and twenty-two flowering plant species were visited by pollinators in grasslands (Figure 7; not significant, t-test =

0.895 ; P-value= 0.405).

During the timed observations, I identified 724 pollinators from 59 pollinator species that I observed visiting flowers across all plots over 16 observation hours (Supplementary Table 2). Bumble bees (35.2% of all observations) were the most abundant pollinator group on the mountain, followed by bees (30.7%), flies (17.5%), wasps (11.7%), butterflies (3.5%), and other miscellaneous species (1.4%). The top four observed species included Bombus vosnesenskii Radoszkowski (31.2%), Apis mellifera Linnaeus (11.5%), genus

Vespula Thomson (10.5%), and subgenus Dialictus Robertson (8.4%), respectively.

The number of pollinators visiting flowers in the coastal scrub plots (372 individuals; 51.4% of total observations) was comparable to those encountered in the grassland plots (352 individuals; 48.6%) (Figure 8; not significant, t-test =

0.252; P-value = 0.805). In the coastal scrub sites, the most abundant species was

Bombus vosnesenskii (32.5%) along with Vespula (16.7%) and Dialictus (10.2%).

In grassland sites, Bombus vosnesenskii (29.8%) remained dominant, but Apis

25 mellifera (19.3%) and genus Lepidanthrax Osten-Sacken (8.5%) were also abundant. A total of 41 pollinator species were observed in coastal scrub sites; 48 pollinator species were observed in grassland (Figure 9; not significant, t-test =

0.086 ; P-value = 0.932).

Seven of twelve vegetation characteristics significantly differed between grassland and coastal scrub sites (Table 1). Coastal scrub sites had higher shrub species richness, mean plant height, maximum plant height, percent shrub cover, and mean distance to nearest plant. Grassland sites had more grass species richness and grass cover.

Differences in Species Composition Across Vegetation Types

Plant species composition differed significantly between coastal scrub and grassland study sites according to the plant community nMDS solution (Figure

10). Significant grassland indicator species included Eschscholzia californica

Cham., Clarkia amoena A. Nelson & J. F. Macbr., Grindelia hirsutula Hook. &

Arn., Briza maxima L., Taraxacum californicum Munz & I. M. Johnst., Rumex acetosella L., Acaena pinnatifida Ruiz & Pav., Festuca californica Vasey,

Lupinus albifrons var. collinus, Plantago lanceolate L., Chlorogalum pomeridianum Kunth, and Avena barbata Pott ex Link (Table 2). Coastal scrub indicator species were Cirsium vulgare, Sambucus racemosa L., and Frangula californica A. Gray (Table 2). Cirsium vulgare, Eschscholzia californica, Clarkia amoena, and Grindelia hirsutula contributed the most to habitat dissimilarity

26 (~2% each ) (Table 3). The nMDS reached final stability after 20 tries and the stress was 0.0155.

In pollinator species space, the nMDS solution revealed differences between coastal scrub and grassland pollinator community, but the 95% confidence ellipses for the two vegetation types overlapped slightly. Four pollinators species contributed significantly to the dissimilarity between vegetation types (Figure 11). The nMDS reached final stability after 20 tries and a stress of 0.0918. Bombus vosnesenskii contributed most to dissimilarity (24%), but they did not favor either specific plant cover-type (Table 4). Apis mellifera

(13%) and Lepidanthrax (6%) were significant indicator species for grasslands

(Table 5). Vespula (10%) was a significant indicator for coastal scrub sites. These three species explained a large portion of the variation in the pollinator community across the two habitat types. Three environmental factors differed significantly in the nMDS pollinator species space: mean distance to nearest plant, maximum plant height (cm), and grass cover (% class) (Table 5). Mean distance to nearest plant and maximum plant height positively correlated to coastal scrub, which means coastal scrub sites had more space between taller plants compared to grassland sites. Grassland sites had greater grass cover percent.

Plant-pollinator interactions

The plant-pollinator networks displayed high complexity and many shared links in both the coastal scrub and grassland vegetation types. Connectance in both vegetation types was similarly low (Table 6). Bipartite network metric

27 analysis results indicated high nestedness in both networks, but the coastal scrub vegetation type had slightly higher nestedness (Table 6). Interaction strength asymmetry was near zero for both vegetation types, but grassland had a negative value and coastal scrub was positive. Based on the H2’ values, pollinators in grassland networks were more specialized than their coastal scrub counterparts.

Such specialist interactions in grassland sites included Rumex acetosella, Salvia spathacea (Greene), Scroophularia californica (Cham. & Schltdl.) only interacting with Bombus vosnesenskii, and Latuca virosa (L.) were only pollinated by Dialictus. Megachile perihirta (Cockerell) only visited Grindelia hirsultula, and Trimerotropis occidentalis (Bruner) specialized its visitation on Hypochaeris glabra (L.) (Figure 12). Carduus pycnocephalus (L.), Solidago spathulata (DC.),

Lonicera hispidula (Douglas ex Torr. & A. Gray), Prunella vulgaris (L.), and

Solanum physalifolium (Rusby) were only visited by one species, Bombus vosnesenskii, in coastal scrub cover types (Figure 13). Niche overlap and partner diversity for plant and pollinator populations were higher in coastal scrub than grassland (Table 6).

Pollinator-plant networks in each vegetation type were influenced most by a few indicator species. In the coastal scrub network, Bombus vosnesenskii and

Vespula were the most frequent pollinator interactors. Notably, Bombus vosnesenskii displayed a strong interaction with Eriophyllum staechadifolium

(Lag.), while Vespula was strongly linked to Frangula californica (Figure 13).

Vespula and Frangula californica were both significant coastal scrub pollinator and plant indicator taxa (Table 5). In our observations, Frangula californica was

28 only pollinated by Vespula. Cirsium vulgare, which was a significant plant indicator species for the coastal scrub vegetation type, was pollinated by multiple bee and bumble bee species.

In the grassland community, plant-pollinator interactions also had a few frequent interactions and many infrequent interactions. In grasslands, two key interactions were observed between pollinators and plants: Bombus vosnesenskii with Eschscholzia californica and Apis mellifera with Scabiosa japonica (L.)

(Figure 12). While there are many other network interactions, these two linkages were the strongest among trophic levels. Both Eschscholzia californica and Apis mellifera were significant grassland indicator species and these interactions also occurred more frequently in grassland than in coastal scrub (Table 5). Most other indicator species appear in the grassland network at varying frequencies.

Lepidanthrax was a generalist, and it visited many different plant species including these indicator plant species: Eschscholzia californica, Taraxacum californicum, and Grindelia hirsutula. Bombus vosnesenskii interacted with every indicator plant species in the grasslands, including Briza maxima and Rumex acetosella, but these two plant species experienced few total pollinator visits.

In summary, a subset of species and their interactions created significant differences between coastal scrub and grassland environments. The variation in species composition changed the dominant plant-pollinator interactions between vegetation types. However, there are also many species that are represented across both vegetation types, which led to insignificant differences in species richness and overlap in the pollinator nMDS species space.

29

CS4 ´

CS3 Figure 1: Aerial view of CS2 San Bruno Mountain surrounded by an urban landscape with Brisbane G4 to the North East, South San Francisco to the South, G2 CS1 and Daly City to the West. Legend Coastal scrub vegetation is Plot Outlines G1 green on the Veg Type G3 mountain, and Coastal Scrub 0 0.15 0.3 0.6 0.9 1.2 grassland Grassland Miles vegetation is Source: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AeroGRID, IGN, and the GIS User Community brown.

30

CS4 ´

CS3

CS2 San Bruno Mountain San Francisco Bay Figure 2: A topographic map of

G4 San Bruno Mountain in San Mateo County, California in the northern San G2 CS1 Francisco Peninsula. The 8 study areas are demarked by filled in G1 polygons. Stars Legend G3 indicate the location of the 5 x 5 m plots Plot 5O xu t5li nme Plotss Veg Type for vegetation Coastal Scrub structure analysis at Grassland each site. Contour Contour Features 0 0.125 0.25 0.5 0.75 1 lines represent 25m. Source: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AeroGRID, IMGNile, asnd the GIS User Community

31

Figure 3: Grassland site 3 on August 17th, 2019. Brown, dry grasses were typical of grassland environments at this time of year. The organge/yellow flowers of Eschscholzia californica (California poppy) can be seen in the bottom of the image, and a flowering Foeniculum vulgare (fennel) is present in the middle of the image.

Figure 4: Coastal scrub site 2 looking East along the Summit Loop Trail on August 13th, 2019. Baccharis pilularis (coyote brush) dominates the foreground, and the white flowers of Anaphalis margaritacea (pearly everlasting) groups to the left of the picture.

32

Figure 5: Examples of pollinators visiting flowering plants. On the left, Megachile perihirta (Western leaf cutter bee) on Cirsium vulgare (bull thistle). On the right, Apis mellifera (Western honey bee) visiting Scabiosa japonica (Pincushion flower).

33

60

50

40

Coastal Scrub 30 Grassland 20 # of Plant Species

10

0 Average of Total Plant Species Observed Figure 6: The average number of plant species observed in each vegetation type (four sites per vegetation type). Error bars represent standard error. T-test indicated that there was no significant difference between coastal scrub and grassland.

25

20

15 Coastal Scrub

10 Grassland # of plant species

5

0 Average of Total Flowering Species

Figure 7: The average number of flowering plant species in each vegetation type. Error bars represent standard error. T-test indicated that there was no significant difference between coastal scrub and grassland.

34 60

50

40

30 Coastal Scrub Grassland

#of Observations 20

10

0 Average of total observations per site

Figure 8: The average number of pollinators observed at each study site by vegetation type. Error bars represent standard error. T-test indicated that there was no significant difference between coastal scrub and grassland.

14.5

14

13.5

13

12.5 Coastal Scrub Grassland # of Species 12

11.5

11

10.5 Average of total species per site

Figure 9: The average number of pollinator species observed at each study site by vegetation type. Error bars represent standard error. T-test indicated that there was no significant difference between coastal scrub and grassland.

35 Table 1: Average vegetation characteristics by vegetation type, and the t-test results with significant p-values (<0.05) in bold.

Environmental Factor Grassland average Coastal scrub average T-test P-value Forb richness 8.5 6.5 1.3093 0.2383 Shrub richness 0.25 3.75 -6.4807 0.0013 Grass richness 5 0.25 4.6082 0.0192 Mean plant height (cm) 31.1397 81.2286 -6.9254 0.0009 Max. plant height (cm) 66.5 168.25 0.0029 0.0057 Rock cover (% class) 1.75 2.75 -1.1239 0.3239 Grass cover (% class) 5.75 1.25 12.7279 <0.001 Shrub cover (% class) 1 5.75 -19 0.0003 Forb cover (% class) 3 2.75 0.2425 0.8240 Canopy cover (%) 0 0.955 -1 0.3910 Visual obstruction (cm) 11 59 -1.9236 0.1943 Mean distance to nearest 2.7947 13.575 -5.5872 0.0050 plant (cm)

36 Plant Matrix RUAR* SARA* FRCA* BRMA** 0.5

TACA* CLAM* LUAL* ESCA* PLLA* GRHI* CHPO* Coastal Scrub 0.0 AVBA* Grassland

NMDS2 ACPI* FECA*

RUAC* CIVU* 0.5 −

CAOC* 1.0 − −1.0 −0.5 0.0 0.5 1.0 NMDS1 Figure 10: nMDS of plant species compositional differences in plot space for grassland and coastal scrub sites on San Bruno Mountain. Ellipses indicate 95% confidence of grassland sites (red) and coastal scrub sites (black). Asterisks (*) show significant indicator species contributing to differences in the two axis from nMDS analysis (* = p<0.05, ** = p<0.01, and *** = p<0.001). Species acronyms are described in Supplementary Table 1.

37 Table 2: Correlations of plant species presence/absence and environmental variables with plot scores for Axis 1 and 2 for nMDS of 8 plots sampled in 2019. Significant correlations with nMDS axes are indicated with asterisks (*) where * = p<0.05, ** = p<0.01, and *** = p<0.001. Pollinator Species Axis 1 Axis 2 R2 P-value Cirsium vulgare 0.95597 -0.29346 0.8662 0.018 * Eschscholzia californica -0.99968 0.02545 0.8624 0.036 * Clarkia amoena -0.99968 0.02545 0.8624 0.036 * Grindelia hirsutula -0.99968 0.02545 0.8624 0.036 * Raphanus sativus -0.98322 0.18241 0.5136 0.185 Scabiosa atropurpurea -0.7607 0.64911 0.7392 0.061 Rumex crispus 0.97897 0.20402 0.6958 0.086 Symphoricarpos albus 0.9732 -0.22994 0.4488 0.226 Briza maxima -0.83392 0.55188 0.9164 0.004 ** Rubus armeniacus 0.39718 0.91774 0.7656 0.049 * Heterotheca sessiliflora -0.66934 0.74296 0.6896 0.095 Sambucus racemosa 0.59312 0.80512 0.8093 0.021 * Hypochaeris glabra -0.98967 -0.14336 0.3742 0.283 Sonchus arvensis -0.53419 -0.84536 0.3355 0.337 Taraxacum californicum -0.96225 0.27218 0.8175 0.011 * Heteromeles arbutifolia -0.34546 0.93843 0.698 0.083 Clarkia davyi -0.49652 0.86802 0.6249 0.104 Eriophyllum staechadifolium 0.49652 -0.86802 0.6249 0.104 Rumex acetosella -0.92931 -0.3693 0.7924 0.014 * Lactuca virosa 0.45759 -0.88916 0.1138 0.693 Scrophularia californica 0.79001 0.61309 0.416 0.253 Eriogonum latifolium -0.60994 -0.79245 0.6335 0.092 Acaena pinnatifida -0.98989 -0.14184 0.8802 0.034 * Festuca californica -0.98989 -0.14184 0.8802 0.034 * Dudleya farinosa 0.03093 -0.99952 0.7197 0.057 Horkelia californica -0.22474 -0.97442 0.0849 0.766 Taraxacum officinale -0.99559 0.09384 0.3769 0.298 Lactuca serriola -0.99559 0.09384 0.3769 0.298 Silene scouleri ssp. grandis -0.99559 0.09384 0.3769 0.298 Frangula californica 0.63522 0.77233 0.8144 0.017 * Vicia sativa ssp. sativa 0.28526 0.95845 0.1535 0.624 Solanum physalifolium 0.8669 -0.49847 0.0507 0.877 Sonchus asper 0.82323 0.5677 0.0685 0.84 Cirsium quercetorum -0.84091 -0.54117 0.4151 0.271 Lonicera hispidula 0.48747 0.87314 0.3231 0.398 Castilleja wightii 0.48747 0.87314 0.3231 0.398

38 Salvia spathacea 0.06399 -0.99795 0.1667 0.621 Calystegia occidentalis -0.24292 -0.97005 0.8283 0.016 * Foeniculum vulgare -0.5505 0.83483 0.3596 0.33 Lupinus albifrons var. -0.99968 0.02545 0.8624 0.036 * collinus Plantago lanceolata -0.99968 0.02545 0.8624 0.036 * Chlorogalum pomeridianum -0.99968 0.02545 0.8624 0.036 * Avena barbata -0.99968 0.02545 0.8624 0.036 * Genista monspessulana -0.44837 0.89385 0.576 0.071 Diplacus aurantiacus 0.83567 -0.54923 0.2358 0.572 Clinopodium douglasii 0.17751 -0.98412 0.6732 0.071 Trifolium -0.98322 0.18241 0.5136 0.185 Helminthotheca echioides -0.37369 -0.92756 0.0924 0.861 Solidago spathulata -0.99352 0.11363 0.2758 0.507 Marah oregana 0.57103 0.82093 0.6946 0.076 Rubus ursinus 0.60806 0.79389 0.2293 0.551 Conium maculatum 0.51977 0.85431 0.1109 0.819 Leucanthemum vulgare -0.89199 -0.45205 0.061 0.89 Pseudognaphalium 0.34565 0.93836 0.3693 0.343 californicum Cotoneaster pannosus 0.10933 0.99401 0.1548 0.667 Carduus pycnocephalus 0.46593 0.88482 0.0276 0.953 Quercus agrifolia 0.8435 0.53713 0.1398 0.675 Artemisia californica 0.45299 -0.89152 0.4319 0.289 Hirschfeldia incana 0.99773 0.06739 0.0119 1 Dittrichia graveolens 0.31117 -0.95036 0.5165 0.124 Anaphalis margaritacea 0.36743 -0.93005 0.3229 0.51 Environmental Factors Forb Richness -0.41554 0.90958 0.3159 0.391 Shrub Richness 0.91862 -0.39515 0.7335 0.054 Grass Richness -0.92098 -0.3896 0.628 0.099 Mean plant height (cm) 0.99151 -0.13002 0.9336 0.003 ** Maximum plant height (cm) 0.97387 -0.22713 0.9 0.008 ** Rock cover (% class) 0.50252 0.86456 0.5957 0.084 Grass cover (% class) -0.99957 -0.02939 0.8179 0.025 * Shrub cover (% class) 0.99935 0.03594 0.8223 0.032 * Forb cover (% class) -0.25762 -0.96625 0.0943 0.838 Canopy Cover (%) 0.54901 0.83581 0.3976 0.254 Visual Obstruction (cm) 0.63162 -0.77528 0.5268 0.169 Mean distance to nearest plant 0.88429 -0.46693 0.7182 0.07 (cm)

39 Table 3: Similarity percentage (SIMPER) analysis differences between grassland and coastal scrub for plant species listed in order of strongest to weakest contribution to dissimilarity. In total, all of these species contribute 70% to site dissimilarity. Grassland and coastal scrub values represent whether the species was presence, absent, or flowering on average in each vegetation type, where 0 = absent, 1 = present at all sites, and 2 = flowering at all sites. Values between 0 and 1 mean that the plant species was present at some, but not all study sites, and values between 1 and 2 mean the plant species was flowering at some, but not all study sties.

Mean Grassland Coastal Cumulative Plant Species Dissimilarity sd ratio value scrub value sum Cirsium vulgare 0.013254 0.004219 3.1413 0.25 2 0.02084 Eschscholzia californica 0.012132 0.007386 1.6425 2 0.5 0.03991 Clarkia amoena 0.012132 0.007386 1.6425 2 0.5 0.05898 Grindelia hirsutula 0.012132 0.007386 1.6425 2 0.5 0.07805 Raphanus sativus 0.011536 0.007219 1.5981 1.5 0 0.09619 Scabiosa atropurpurea 0.011431 0.007165 1.5955 1.5 0 0.11416 Rumex crispus 0.011128 0.004899 2.2713 0.25 1.75 0.13165 Symphoricarpos albus 0.010752 0.006723 1.5992 0 1.5 0.14856 Briza maxima 0.010011 0.006178 1.6204 1.5 0.25 0.16429 Rubus armeniacus 0.009947 0.005843 1.7023 0.25 1.5 0.17993 Heterotheca sessiliflora 0.009922 0.008138 1.2191 1.5 0.5 0.19553 Sambucus racemosa 0.009353 0.007079 1.3212 0 1.25 0.21023 Hypochaeris glabra 0.008422 0.00629 1.339 1.75 0.75 0.22347 Sonchus arvensis 0.008158 0.005652 1.4434 1.25 0.25 0.2363 Taraxacum californicum 0.007974 0.005293 1.5065 1.25 0.25 0.24884 Heteromeles arbutifolia 0.007938 0.007048 1.1263 1.25 0.75 0.26132 Clarkia davyi 0.00789 0.008351 0.9447 1 0 0.27372

40 Eriophyllum staechadifolium 0.00789 0.008351 0.9447 1 2 0.28612 Rumex acetosella 0.007843 0.005272 1.4876 2 1 0.29845 Lactuca virosa 0.007807 0.00678 1.1514 0.75 1.25 0.31073 Scrophularia californica 0.007591 0.005522 1.3749 1 2 0.32266 Eriogonum latifolium 0.007576 0.006456 1.1735 2 1 0.33457 Acaena pinnatifida 0.007539 0.001239 6.0823 1 0 0.34642 Festuca californica 0.007539 0.001239 6.0823 1 0 0.35827 Dudleya farinosa 0.007535 0.006385 1.1802 1 1 0.37012 Horkelia californica 0.007525 0.005507 1.3663 1 1 0.38195 Taraxacum officinale 0.007359 0.007771 0.9469 1 0 0.39352 Lactuca serriola 0.007359 0.007771 0.9469 1 0 0.40509 Silene scouleri ssp grandis 0.007359 0.007771 0.9469 1 0 0.41665 Frangula californica 0.00728 0.0045 1.6179 0.25 1.25 0.4281 Vicia sativa ssp sativa 0.007044 0.004835 1.457 0.5 1 0.43917 Solanum physalifolium 0.006986 0.007349 0.9506 0.5 1 0.45016 Sonchus asper 0.00668 0.00601 1.1114 0.75 1 0.46066 Cirsium quercetorum 0.006672 0.005563 1.1992 1 0.25 0.47115 Lonicera hispidula 0.0065 0.006786 0.9579 0 1 0.48136 Castilleja wightii 0.0065 0.006786 0.9579 0 1 0.49158 Salvia spathacea 0.00645 0.005392 1.1962 1 0.75 0.50172 Calystegia occidentalis 0.006358 0.005242 1.2129 1 0.25 0.51172 Foeniculum vulgare 0.006208 0.006734 0.9219 1.75 1.25 0.52148 Lupinus albifrons var collinus 0.006066 0.003693 1.6425 1 0.25 0.53101 Plantago lanceolata 0.006066 0.003693 1.6425 1 0.25 0.54055

41 Chlorogalum pomeridianum 0.006066 0.003693 1.6425 1 0.25 0.55009 Avena barbata 0.006066 0.003693 1.6425 1 0.25 0.55962 Genista monspessulana 0.005981 0.007092 0.8433 0.75 0 0.56902 Diplacus aurantiacus 0.005895 0.007068 0.8341 1.25 2 0.57829 Clinopodium douglasii 0.005805 0.006548 0.8865 0.25 0.75 0.58742 Trifolium 0.005768 0.003609 1.5981 0.75 0 0.59649 Helminthotheca echioides 0.005625 0.005654 0.9949 0.75 0.25 0.60533 Solidago spathulata 0.005577 0.006287 0.8871 0.75 0 0.6141 Marah oregana 0.005413 0.003399 1.5923 0 0.75 0.62261 Rubus ursinus 0.005205 0.004459 1.1674 0.5 1.25 0.63079 Conium maculatum 0.005109 0.005411 0.9441 0 0.75 0.63882 Leucanthemum vulgare 0.005083 0.006875 0.7393 0.5 0.5 0.64681 Pseudognaphalium 0.005067 0.004315 1.1742 0.5 1.25 0.65478 californicum Cotoneaster pannosus 0.004989 0.004858 1.0269 0.25 0.75 0.66262 Carduus pycnocephalus 0.00478 0.00494 0.9677 1 1.25 0.67014 Quercus agrifolia 0.00468 0.003863 1.2115 0.25 0.75 0.67749 Artemisia californica 0.004594 0.003794 1.211 0.25 0.75 0.68471 Hirschfeldia incana 0.00448 0.005887 0.7609 1.5 1.75 0.69176 Dittrichia graveolens 0.004252 0.007623 0.5577 0 0.5 0.69844 Anaphalis margaritacea 0.004072 0.007407 0.5498 1.5 2 0.70484

42 Pollinator Matrix

1.0 Vespula * disttoplant shrubrich 0.5

Coastal Scrub 0.0 NMDS2

Grassland Bombus vosnesenskii *** 0.5 −

grass Apis mellifera** Lepidanthrax ** 1.0 −

−1.0 −0.5 0.0 0.5 1.0

NMDS1 Figure 11: nMDS of pollinator species compositional differences in plot space for grassland and coastal scrub sites on San Bruno Mountain. Ellipses indicate 95% confidence of grassland sites (red) and coastal scrub sites (black). Asterisks (*) show significant indicator species contributing to differences in the two axis from nMDS analysis (* = p<0.05, ** = p<0.01, and *** = p<0.001). Blue vectors indicate significant vegetation characteristics that differ between grassland and coastal scrub.

43

Table 4: Similarity percentage (SIMPER) analysis differences between grassland and coastal scrub for pollinator species, which are listed in order of strongest to weakest contribution to dissimilarity. Grassland and coastal scrub values represent the average number of observations in those vegetation types.

Mean Cumulative Pollinator species dissimilarity sd ratio Grassland Coastal scrub % Bombus vosnesenskii 0.174524 0.151775 1.1499 13.125 15.125 24% Apis mellifera 0.092001 0.132421 0.6948 8.5 1.875 37% Vespula 0.072922 0.042223 1.7271 1.75 7.75 47% Lepidanthrax 0.044737 0.052836 0.8467 3.75 0 53% Dialictus 0.039258 0.034248 1.1463 2.875 4.75 59% Halictus tripartitus 0.026679 0.026876 0.9927 0.625 2.25 63% Eristalis tenax 0.021483 0.019998 1.0742 0.5 1.875 66% Geron 0.018058 0.030692 0.5884 1.375 0.375 68% Bombus melanopygus 0.017545 0.024101 0.728 0.625 1.125 70%

Table 6: Network-level metrics from all plant-pollinator interactions in grassland and coastal scrub, except for interaction strength asymmetry, which used all non-singleton interactions.

Interaction Niche Niche Partner Partner Vegetation Strength overlap overlap Diversity Diversity Type Connectance Nestedness Asymmetry H2’ (pollinators) (plants) (Pollinators) (plants) Grassland 0.1152 8.328 -0.0085 0.5349 0.1643 0.1498 0.9742 1.391 Coastal 0.1076 4.176 0.0026 0.3175 0.2711 0.236 1.453 1.687 Scrub

44

Table 5: Correlations of pollinator species abundance and environmental variables with plot scores for Axis 1 and 2 for nMDS of 8 plots sampled in 2019. Significant correlations with nMDS axes are indicated with asterisks (*) where * = p<0.05, ** = p<0.01, and *** = p<0.001. Pollinator Species Axis 1 Axis 2 R2 P-value Bombus vosnesenskii 0.92633 -0.37671 0.6941 0.001 *** Apis mellifera -0.51669 -0.85617 0.5553 0.006 ** Vespula 0.43626 0.89982 0.386 0.04 * Lepidanthrax -0.2945 -0.95565 0.5356 0.006 ** Dialictus 0.8893 0.45733 0.1755 0.275 Halictus tripartitus 0.29515 0.95545 0.1791 0.268 Eristalis tenax 0.52682 0.84997 0.3646 0.057 Geron -0.95696 0.2902 0.2225 0.203 Bombus melanopygus -0.5014 0.86522 0.1292 0.392 Environmental Factors Forb Richness -0.99705 -0.0767 0.1819 0.262 Shrub Richness 0.52223 0.8528 0.3758 0.042 * Grass Richness -0.35492 -0.9349 0.2717 0.122 Mean plant height 0.67999 0.73322 0.3277 0.062 (cm) Maximum plant height 0.6657 0.74622 0.3231 0.065 (cm) Rock cover (% class) 0.2189 0.97575 0.0381 0.782 Grass cover (% class) -0.55737 -0.83026 0.389 0.037 * Shrub cover (% class) 0.53076 0.84752 0.3527 0.053 Forb cover (% class) -0.36732 0.93009 0.0827 0.584 Canopy Cover (%) 0.91453 0.40451 0.0837 0.602 Visual Obstruction 0.53387 0.84557 0.1983 0.224 Mean distance to 0.45526 0.89036 0.3753 0.043 * nearest plant (cm)

45 HorkeliaHOCA californica V. Vanessa.Carduicardui

B. Bombus.melanopygusmelanopygus

ScabiosaSCJA japonica

A. Apis.Melliferamellifera

Cirsium CIQUquercetorum Coenonympha.tullia HeteromelesHEAR arbutifolia C. tullia RaphanusRASA sativus ZadontomerusSubgenus..Zadontomerus BrizaBRMA Maxima RumexRUAC acetosella SalviaSASP spathacea Scrophularia californica SCCA B. Bombus.vosnesenkiivosnesenskii

EschscholziaESCA californica

B. Bombus.Californicuscalifornicus H. Halictus.tripartitustripartitus EriogonumERLA Latifolium ColletesGenus..colletes HirschfeldiaHIIN incana DialictusSubgenus..Dialictus Taraxacum TACAcalifornicum LatucaLAVI Virosa LepidanthraxGenus..Lepidanthrax GrindeliaGRHI Hirsutula M. Megachile.Perihirtaperihirta

HypochaerisHYGL glabra GeronGenus..Geron T. Trimerotropis.occidentalisoccidentalis HeterothecaHESE Sessilliflora E. Eristalis.tenaxtenax T. Taxomerus.marginatusmarginatus SolidagoSOSP spathulata T. Toxomerus.Occidentalisoccidentalis HelminthothecaHEEC echioides FoeniculumFOVU vulgare VespulaGenus..Vespula

Figure 12: Grassland bipartite network representing cumulative interactions (grey lines) between flowering plants (left) and pollinators (right). The thickness of lines and boxes represent the frequency of observations. Stars represent indicator species. This network only contains observations of pollinators that occurred >1% of the total observations in grassland communities to reduce congestion due to infrequent pollinator observations. Plant species acronyms are described in Supplementary Table 1, and pollinator full names are described in Supplementary Table 2.

46 ClarkiaCLAM amoena ColletesGenus..colletes GrindeliaGRHI Hirsutula leucanthemum vulgare LEVU ZadontomerusSubgenus..Zadontomerus EschscholziaESCA californica Mimulus MIAUAurantiacus A. Apis.Melliferamellifera HeteromelesHEAR arbutifolia H. Halictus.tripartitustripartitus HirschfeldiaHIIN incana

BaccharisBAPI Pilularis DialictusSubgenus..Dialictus

Anaphalis MargaritaceaANMA

E. Eristalis.tenaxtenax CirsiumCIVU Vulgare Rubus RUARarmeniacus CalliphoraGenus..Calliphora

ScaevaGenus..Scaeva

DipteraOrder..Diptera ERST Eriophyllum staechadifolium T. Toxomerus.Occidentalisoccidentalis

B. Bombus.melanopygusmelanopygus ErigeronERGL glaucus HorkeliaHOCA californica HeracleumHEMA maximum Cardus pycnocephalusCAPY B. vosnesenskii SolidagoSOSP spathulata LoniceraLOHI Hispidula Bombus.vosnesenkii PrunellaPRVU Vulgaris Solanum PhysalifoliumSOPH

SymphoricarposSYAL Albus

S. Scaeva.Pyrastripyrastri

B. Bombus.Californicuscalifornicus ScrophulariaSCCA californica

FoeniculumFOVU vulgare VespulaGenus..Vespula CotoneasterCOPA pannosus FrangulaFRCA californica

Figure 13: Coastal scrub bipartite network representing cumulative interactions (grey lines) between flowering plants (left) and pollinators (right). The thickness of lines and boxes represent the frequency of observations. Stars represent indicator species. This network only contains observations of pollinators that occurred >1% of the total observations in coastal scrub communities to reduce congestion due to infrequent pollinator observations. Plant species acronyms are described in Supplementary Table 1, and pollinator full names are described in Supplementary Table 2.

47

48 Discussion

My results demonstrate the importance of plant-pollinator interactions in maintaining landscape-scale biodiversity of both the pollinator and plant communities of SBM. Interactions across plant and pollinator trophic levels are paramount for maintaining ecosystem function in coastal northern California plant communities. This is especially important on SBM because it is one of the few remaining natural, open spaces in an otherwise human-dominated, urban landscape.

Species Richness and Pollinator Abundance

Pollinator species richness was high at SBM compared to other similar coastal habitats of California (McFrederick and LeBuhn 2006, Colteaux et al.

2013, Luong et al. 2019), which demonstrates the conservation value of SBM as the only large, contiguous natural area of the northern San Francisco Peninsula.

The lack of significant differences in pollinator species richness and the overlap in pollinator species composition between grassland and coastal scrub habitats indicates that the pollinator community moves through, and utilizes the entire

SBM vegetation matrix, regardless of plant species composition. Prior research in other similar habitats demonstrates that high flowering plant diversity, and forb

(i.e. non-woody, non-graminoid herbaceous flowering plants) species richness in particular, is positively correlated with both pollinator abundance and species richness (Potts et al. 2003, Ebeling et al. 2008, Grundel et al. 2010). Increasing the floral resources through habitat restoration on the mountain in response to

49 recent woody plant encroachment in the absence of fire could further promote pollinator abundance and diversity, and especially generalist pollinators that visit a wide variety of plant species. A pollinator species’ ability to generalize, that is to visit different plant taxa, allows them to forage in different vegetation types.

Generalization benefits the pollinator because it increases the amount of available resources for the pollinator at any given place or time (Waser et al. 1996), as opposed to specialization, which restricts the pollinator’s viable resource sources to a few target plant species. Generalists plants are likewise pollinated by a wide range of pollinator taxa, which increases a plant’s chance of pollination and reproductive success (Fontaine et al. 2006, Greenleaf and Kremen 2006). Plant and pollinator generalists can buffer the loss of low abundance species because they can provide food resources or pollination for many species (Waser et al.

1996, Memmot et al. 2004, Maia et al. 2019). Generalist pollinators are important because of their ability to move through the vegetation matrix and visit different types of plant species. Generalists on SBM can thereby maintain the survival of key native species in this large, contiguous natural area.

Given the importance of SBM as a biodiversity hotspot in an otherwise urban landscape, maintaining a habitat for the pollinator community is a key component of biodiversity conservation. This is especially true in light of the many negative impacts that human disturbance and destruction in the surrounding matrix can have on both plants and pollinator abundance and distributions (Potts et al. 2003, Aguilar et al. 2006, Fortuna and Bascompte 2006, Kremen et al. 2007,

Goulson et al. 2008, Winfree et al. 2009, Potts et al. 2010, Carman and Jenkins

50 2016). Habitat restoration that promotes the recruitment of forbs to the site could reverse the effects of shrub encroachment from fire exclusion in SBM in recent decades. Grassland cover has decreased over time at SBM, and in other similar coastal habitats in the absence of fire (Zavaleta and Kettley 2006, Knapp et al.

2007, NPS 2007, Eviner 2016, Campell et al. 2018, Poulos et al. 2020). While forb cover and dominance was not significantly lower in coastal scrub cover at

SBM, habitat homogenization through scrub encroachment in the absence of fire has resulted in declines in forb cover and abundance elsewhere (Anderson et al.

2000, MacDougall and Turkington 2007). Habitat heterogeneity, floral diversity, and the maintenance of grassland habitats may be important factors promoting pollinator diversity (Potts et al. 2003, Ebeling et al. 2008, Grundel et al. 2010,

Luong et al. 2019); adaptive management activities including prescribed fire, mowing, disc harrowing, or tilling in conjunction with native seed additions of herbaceous species may maintain habitat heterogeneity at SBM into the future

(Dunwiddie et al. 1997, Moyes et al. 2005, Wheeler et al. 2015, Carlsen et al.

2017, Omand et al. 2018, Poulos et al. 2020).

Another factor that could increase pollinator richness is the availability of nesting resources like dead plant ground cover (Grundel et al. 2010). Nesting resources can affect pollinator richness, especially wild bee and bumble bee populations, because wild pollinators tend to nest in the ground within flying distance to floral resources (Thorp et al. 2002, Greenleaf and Kremen 2006, Black et al. 2009). While I did not measure pollinator nesting resource availability or

51 nest site abundance, such information could also be valuable for guiding pollinator management in SBM.

Although I conducted my surveys during just two months of the growing season in July and August, 2019, pollinator observations over the entire flowering season could be valuable since flowering plants and pollinator abundance can vary throughout the growing season (Moldenke 1976, Wojcik et al. 2008, Iwasaki et al. 2018, Smith DiCarlo et al. 2019). In my surveys, I likely observed just a subset of the total flowering plant and pollinator species and interactions. For instance, I did not encounter any of the endangered native butterflies (Mission

Blue, Callippe Silverspot, and San Bruno Elfin) because their flight season occurs earlier in the year between March and June (Ormshaw 2018). To compile a complete record of plant and pollinator species richness, abundance, and interactions on SBM, future pollinator surveys should be conducted every few weeks from the beginning of the flowering season until the end.

Plant-pollinator interactions

SBM plant-pollinator networks have many generalist species and overlapping interactions, which reduces the chance of large-scale biodiversity loss. This is very important because SBM hosts higher plant and pollinator diversity relative to surrounding urban landscape. The bipartite network metrics, and particularly their high nestedness and low specialization, indicate that there is a low extinction risk of plant and pollinator species across SBM, as long as the site remains under conservation. Plant-pollinator interaction webs are structured

52 in a nested way, where less abundant, or specialist species, interact with a subset of the species with which the most abundant generalist species interact (Elle et al.

2012). Such a pattern creates networks that are resilient to extinction because of the many overlapping interactions between taxa and across trophic levels that facilitate plant pollination and ecosystem function (Memmot et al. 2004, Fontaine et al. 2006). Low specialization index, H2’, indicates more species are generalists rather than specialists. This adds further redundancy to the network because there are more species to fill the role left by a potentially extinct species (Waser et al.

1996, Memmot et al. 2004, Elle et al. 2012). The network metrics for SBM corroborate other studies that identify large plant-pollinator networks as being resilient to extinctions (Waser et al. 1996, Memmot et al. 2004, Elle et al. 2012,

Venjakob et al. 2016), which is important because SBM is the last biodiversity hotspot in the largely anthropogenic landscape of the northern San Francisco

Peninsula. The high nestedness and generalization of the plant-pollinator network means that the loss of one species will not cause a collapse of the entire plant- pollinator network because of the many connections among species within the network. Coastal scrub cover type has higher nestedness and lower specialization than grassland, which may be due to more generalized plant species that attract all types of pollinators. Since grassland networks display slightly higher plant and pollinator specialization, they have less niche overlap and partner diversity across both plant-pollinator trophic levels in comparison with the coastal scrub communities. Grasslands may have more specialized plant and pollinator species, which would benefit from an increase in generalized pollinator or plant

53 abundance. In turn, more generalists would increase the nestedness and niche overlap of the grassland plant-pollinator network (Maia et al. 2019).

Connectance, i.e. the proportion of plant-pollinator links to possible links between all present plant and pollinator species (Dormann et al. 2009, Elle et al.

2012), tends to be low in plant-pollinator interactions because the amount of possible interaction increases exponentially with each additional species (Elle et al. 2012, Petanidou et al. 2018). So it is no surprise that connectance is low across grassland and coastal scrub habitats in the present study. Positive interaction strength asymmetry indicates higher specialization of the pollinator trophic level, but my values are close to zero, which means that neither trophic level in the environments is more specialized than the other (Dormann et al. 2009). Network size and species richness do have an effect on connectance, nestedness, interaction strength asymmetry, and specialization (Blüthgen et al. 2008).

Incomplete interaction sampling may result in low connectedness and high nestedness (Blüthgen et al. 2008). This suggests that measuring plant-pollinator interactions over multiple years and across the entire growing season may be important for elucidating the strength of the plant-pollinator network on SBM and its resilience to anthropogenic stressors including urban development, pollinator decline from habitat loss, pesticide use, and climate change (Memmot et al. 2004,

Fortuna and Bascompte 2006, Kremen et al. 2007, Goulson et al. 2008, Potts et al.

2010, Petanidou et al. 2018).

Through the plant-pollinator networks, I was able to identify key indicator plant and pollinator species and their trophic-level interactions across different

54 vegetation types on SBM. A non-native thistle, Cirsium vulgare, is an indicator species of coastal scrub vegetation types, which do not receive active vegetation management unlike grassland vegetation types. Cirsium vulgare may facilitate abundance of native pollinators to other plant species in coastal scrub environments (Lopezaraiza-Mikel et al. 2007), or it may outcompete native vegetation (Luong et al. 2019). Frangula californica, a native, woody, evergreen shrub that dominated the coastal scrub habitat type, was only pollinated by

Vespula in my observations. Vespula, the indicator pollinator in coastal scrub, are eusocial, ground nesting wasps (Reed and Landolt 2019). I observed a nest in

CS4, where many individuals were flying in and out of an old, small (<10 cm) rodent hole. Coastal scrub environments have more bare ground that would be accessible for their nesting needs (Reed and Landolt 2019). The indicator plant species in grassland habitats highlight the importance of both native and non- native graminoids and forbs in this habitat type. Apis mellifera, the European honey bee, is a domesticated pollinator and was predominately observed visiting the non-native Scabiosa japonica in G4, the grassland site closest to Brisbane, a nearby town that may have bee hives. In all other sites, Apis mellifera prevalence was low. Lepidanthrax, a type of small bee fly (family Bombyliidae), frequented aster flowers. Although its nesting habits are unknown, it likely does not fly far from its nest due to its small size (Greenleaf et al. 2007), thus I expect them to nest in grasslands. Bombus vosnesenskii is an indicator species for restored

California grassland habitats (Luong et al. 2019) and was abundant throughout

SBM, which is a good sign for the mountain. Bombus vosnesenskii nest in holes

55 underground often created by rodents (Thorp et al. 2002, Greenleaf and Kremen

2006, McFrederick and LeBuhn 2006), and they could benefit from vegetation management activities that focus on opening up the vegetation matrix through woody plant removal. Other Bombus species use dead plant cover as nesting material (Thorp et al. 2002, Kells and Goulson 2003), so maintaining a matrix of bare ground and dead plant material is important for diversity of bumble bees and other pollinators. One flaw of the plant-pollinator interaction analysis is that it does not account for the effectiveness of each floral visitor to transfer pollen between plants (Elle et al. 2012). For instance, bumble bees are hairier and carry more pollen than wasps. Future studies could observe the pollen transfer ability by identifying the pollen on various pollinators.

Pollinators, such as Bombus vosnesenskii or Lasioglossum (Dialictus), with high numbers of linkages to different species have high conservation value as generalist pollinators that maintain biodiversity (Carman and Jenkins 2016,

Maia et al. 2019). Generalist pollinator species can facilitate pollination across a wide range of plant species and functional types (Memmot et al. 2004, Fontaine et al. 2006), which increases the pollination and reproduction of the plant community. Their positive influence on the plant community also increases the abundance and richness of other pollinators species that rely on a subset of the generalists’ range of visited plants (Carman and Jenkins 2016, Maia et al. 2019).

Generalists, like Bombus vosnesenskii, represent an important keystone pollinator species in the SBM system. Bumble bee populations are in decline (Goulson et al.

2008, Williams et al. 2009), and, since Bombus vosnesenskii pollinates so many

56 different plant species, its loss at SBM could lead to extinctions of plants and pollinators that rely on its host plant matrix (Memmot et al. 2004). Therefore, conservation of this species by ensuring the abundance of a range of floral resources within the matrix of flowering plants at SBM would be beneficial for both grasslands and coastal scrub communities, alike (Goulson et al. 2008, Elle et al. 2012, Carman and Jenkins 2016, Luong et al. 2019, Maia et al. 2019).

Alternatively, Apis mellifera, the non-native European honey bee, may have negative impacts on grassland habitats by competing with bumble bees for floral resources (Thomson 2004) and it may also facilitate the invasion of non-native plants, such as Scabiosa japonica, through frequent flower visitation and plant fertilization (Lopezaraiza-Mikel et al. 2007, Colteaux et al. 2013).

Generalist plant species also attract a range of pollinator functional types at landscape-scale (Fontaine et al. 2006, Maia et al. 2019). In coastal scrub communities, Eriophyllum stachaedifolium, Anaphalis margaritacea (Benth.), and Scrophularia californica were visited by the widest range of different pollinator species visitors (Figure 13). Both Eriophyllum stachaedifolium and

Anaphalis margaritacea are asters (family: Asteraceae), which are highly attractive to a wide range of pollinators taxa due to their open flowers and high nectar content (Fontaine et al. 2006, Baude et al. 2016, Maia et al. 2019). While, their nectar and pollen may not be as beneficial to generalist pollinators (Thomson

2004, Sedivy et al. 2011), generalists still visit asters occasionally and therefore these plant generalists should be prioritized in habitat restoration. In grasslands,

Grindelia hirsutula, Heterotheca sessiliflora (Shinners), and Hypochaeris glabra

57 (all asteraceae) had the highest pollinator visitor richness, which means they represent plant generalist species that are accessible by pollinators with both small and large mouthparts (Fontaine et al. 2006). Future work that examines both flower and pollinator functional traits, and their effect on the relationship between specialist plants and pollinators could provide additional information about the intricate nature of plant-pollinator relationships in SBM and elsewhere. While this research provides a detailed census of pollinator visitation in an important conservation area, my ability to characterize specialist pollinators was limited by a single year of pollinator observations since pollinators may only visit one plant in one season and then visit a different species the next year (Petanidou et al.

2008). Additional multi-year monitoring over the entire growing season, from

March to September could further contribute to our understanding of such plant- pollinator relationships at SBM.

Plant-pollinator networks allow us to identify the most connected pollinators and plants that form the core of each vegetation cover type. Generalist species, such as pollinators Bombus vosnesenskii and Lasioglossum (Dialictus) and plants Eriophyllum stachaedifolium, Anaphalis margaritacea, Scrophularia californica, Grindelia hirsutula, Heterotheca sessiliflora, and Hypochaeris glabra, are important to protect because they play a key role in maintaining the biodiversity of San Bruno Mountain, which itself is an island of biodiversity in a sea of urbanization. While grassland and coastal scrub sites differ in their species composition, they both have high nestedness and high generalization, which indicates the plant-pollinators interactions may provide a buffer to extinction in

58 case of species loss. The protection of SBM as an important habitat for local species likely contributes to the complexity and health of the plant-pollinator community.

Conclusion and Management Applications

In the future, all plant-pollinator interactions should be considered in conservation and restoration tactics at SBM and in other similar coastal habitats to promote biodiversity conservation and ecosystem health, rather than focusing conservation and adaptive management on a select few indicator species (Elle et al. 2012, Carman and Jenkins 2016). Plant-pollinator networks could help managers identify keystone species that are critical for maintaining habitat diversity and the ecosystem service that these networks provide.

San Bruno Mountain has been protected from human disturbance based on its Habitat Conservation Plan (Ormshaw 2018). This conservation area should continue to be protected given the value of this large natural habitat within a landscape dominated by human development. Many volunteer and non-profit groups, including San Bruno Mountain Watch, Weed Warriors, and the San

Mateo Parks Department, have a long history of working on habitat restoration at

SBM. Past restoration activities have included manual, mechanical, and herbicidal scrub removal (Ormshaw 2018) and native grass and forb enrichment planting

(Cherbowsky Corkidi 2018). However, these management strategies target only small areas of mountain, and they have not been effective in reducing coastal scrub cover that has increased in recent decades at the expense of grasslands.

59 Additional restoration techniques that could increase the spread of native flowering species, and thus pollinator abundance, include: tilling with seeding of native plant to increase native species richness and percent cover (Wheeler et al.

2015), disc harrowing coastal scrub habitats with native seed additions (Omand et al. 2018), and repeated mowing or prescribed fire to reduce scrub encroachment and increase herbaceous plants (Dunwiddie et al. 1997, Poulos et al. 2020).

Mowing, prescribed fire, and tillage can have direct, short-term negative impacts on ground nesting wild pollinators, but applying management to only small portions of the habitat (less than a third) could provide refuge for pollinators

(Black et al. 2009). Fire may have an indirect positive affect on pollinator species due to higher forb abundance in subsequent years (Smith DiCarlo et al. 2019).

Adaptive management through experimentation will be necessary for achieving grassland restoration at SBM, with the goal of promoting and maintaining plant and pollinator diversity over the long-term. The effects of restoration on the pollinator community remain unknown, and continued monitoring of pollinators is important for evaluating the effects of restoration on the plant and pollinator community structure and function.

Expanding pollinator habitat into the urban landscape could also promote the diversity and abundance of pollinator species. This can be done by expanding green spaces with generalist plants and nesting resources into neighboring cities.

Neighboring residences could also promote pollinator diversity by reducing herbicide use, minimizing concrete cover, creating nesting areas for pollinators, and planting native plants, which are cultivated by San Bruno Mountain Watch at

60 their Mission Blue Nursery. These solutions will expand the available resources of wild pollinators and improve community biodiversity in the urban landscape of the San Francisco Peninsula.

Acknowledgments:

I would like to thank everyone who helped me throughout the process of this project including my parents, Ariel Cherbowski Corkidi, Helen Poulos,

Wesleyan’s College of the Environment, Kim Diver, and my fellow

Environmental Studies majors. I would especially like to thank Ariel for familiarizing me with the history, ecology, and geography of San Bruno

Mountain—without his support and guidance this project would not have been possible. I would also like to specifically thank my advisor, Helen Poulos, for helping me along the way with her knowledge, advice, and reassurance. Finally thank you to the San Mateo County Parks Department for granting me the scientific permit so that I could conduct this research.

References: Aguilar, R., L. Ashworth, L. Galetto, and M. A. Aizen. 2006. Plant reproductive susceptibility to habitat fragmentation: review and synthesis through a meta-analysis. Ecology Letters 9:968-980. Allshouse, D. 2016. Life has returned to Hillside Trail. Glimpse From the Field. San Bruno Mountain Watch. Allshouse, D. 2017. San Bruno Mountain Fire, June 2008. Yerba Buena Chapter of the California Native Plant Society. Anderson, R. C., J. E. Schwegman, and M. R. Anderson. 2000. Micro-scale restoration: a 25-year history of a Southern Illinois Barrens. Restoration Ecology 8:296-306. Bascompte, J. 2010. Structure and Dynamics of Ecological Networks. Science 329:765-766. Baude, M., W. E. Kunin, N. D. Boatman, S. Conyers, N. Davies, M. A. Gillespie, R. D. Morton, S. M. Smart, and J. Memmot. 2016. Historical nectar

61 assessment reveals the fall and rise of floral resources in Britain. Nature 530:85-88. Beattie, K. C., J. M. Karberg, K. A. Omand, and D. I. O’Dell. 2017. Sheep grazing as a grassland management tool: Lessons learned on Nantucket Island, Massachusetts. Northeastern Naturalist 24:45-66. Beidleman, L. H., and E. N. Kozloff. 2003. Plant of the San Francisco Bay Region: Mendocino to Monterey. 2 edition. University of California Press, Berkeley, California. Black, S. H., M. Shepherd, M. Vaughan, C. LaBar, and N. Hodges. 2009. Pollinator Conservation Strategy. The Xerces Society for Invertebrate Conservation, Sacramento, California. Blüthgen, N., J. Fründ, D. P. Vázquez, and F. Menzel. 2008. What do interaction network metrics tell us about specialization and biological traits? Ecology 89:3387-3399. Blüthgen, N., and A.-M. Klein. 2011. Functional complementarity and specialisation: The role of biodiversity in plant–pollinator interactions. Basic and Applied Ecology 12:282-291. Campell, J. W., P. A. Vigueira, C. C. Vigueira, and C. H. Greenberg. 2018. The Effects of Repeated Prescribed Fire and Thinning on Bees, Wasps, and Other Flower Visitors in the Understory and Midstory of a Temperate Forest in Nothern Carolina. Forest Science 64:299-306. Carlsen, T. M., E. K. Espeland, L. E. Paterson, and D. H. MacQueen. 2017. Optimal prescribed burn frequency to manage foundation California perennial grass species and enhance native flora. Biodiversity and Conservation 26:2627-2656. Carman, K., and D. G. Jenkins. 2016. Comparing diversity to flower-bee interaction networks reveals unsuccessful foraging of native bees in disturbed habitats. Biological Conservation 202:110-118. CBS. 2019. Fire Crews Gain Upper Hand on San Bruno Mountain Brush Fire. CBS SF Bay Area. Cherbowsky Corkidi, A. 2018. San Bruno Mountain Habitat Conservation Plan: Report on the 2017-2018 Planting Season. San Mateo County Parks Department. Colteaux, B. C., C. McDonald, M. Kolipinski, J. B. Cunningham, and S. Ghosh. 2013. A survey of pollinator and plant interactions in meadow and grassland habitats of Marin County, California. Bios 84:1-7. Dormann, C. F., J. Fruend, and B. Gruber. 2008. Introducing the bipartite Package: Analysing Ecological Networks. R News 8:8-11. Dormann, C. F., J. Fründ, N. Blüthgen, and B. Gruber. 2009. Indices, Graphs and Null Models: Analyzing Bipartite Ecological Networks. The Open Ecology Journal 2:7-24. Dunwiddie, P. W., W. A. Patterson III, J. L. Rudnicky, and R. E. Zaremba. 1997. Vegetation management in coastal grasslands on Nantucket Island, Massachusetts: effects of burning and mowing from 1982 to 1993. Pages 85-97 in P. D. Vickery and P. W. Dunwiddie, editors. Grasslands of

62 Northeastern North America: Ecology and Conservation of Native and Agricultural Landscapes. Massachusetts Audubon Society. Ebeling, A., A.-M. Klein, J. Schumacher, W. W. Weisser, and T. Tscharntke. 2008. How does plant richness affect pollinator richness and temporal stability of flower visits? Oikos 117:1808-1815. Elle, E., S. L. Elwell, and G. A. Gielens. 2012. The use of pollination networks in conservation. Botany 90:525-534. Eviner, V. T. 2016. Grasslands.in H. Mooney and E. Zavaleta, editors. Ecosystems of California. University of California Press, Berkeley, California. Fontaine, C., I. Dajoz, J. Meriguet, and M. Loreau. 2006. Functional diversity of plant-pollinator interaction webs enhances the persistence of plant communities. PLoS Biology 4:0129-0135. Fortuna, M., and J. Bascompte. 2006. Habitat loss and the structure of plant- animal mutualistic networks. Ecology Letters 9:281-286. Gilliam, H. 2002. Weather of the San Francisco Bay Region. University of California Press, Berkeley, California. Goulson, D., G. C. Lye, and B. Darvill. 2008. Decline and conservation of bumble bees. Annual Review of Entomology 53:191-208. Greenleaf, S. S., and C. Kremen. 2006. Wild bee species increase tomato production and respond differently to surrounding land use in Northern California. Biological Conservation 133:81-87. Greenleaf, S. S., N. M. Williams, R. Winfree, and C. Kremen. 2007. Bee foraging ranges and their relationship to body size. Oecologia 153:589-596. Grundel, R., R. P. Jean, K. J. Frohnapple, G. A. Glowacki, P. E. Scott, and N. B. Pavlovic. 2010. Floral and nesting resources, habitat structure, and fire influence bee distribution across an open-forest gradient. Ecological Applications 20:1678-1692. Iwasaki, J. M., K. J. M. Dickinson, B. I. P. Barratt, A. R. Mercer, T. W. D. Jowett, and J. M. Lord. 2018. Floral usage partitioning and competition between social (Apis mellifera, Bombus terrestris) and solitary bees in New Zealand: Niche partitioning via floral preferences? Austral Ecology 43:937-948. Kearns, C. A., D. W. Inouye, and N. M. Waser. 1998. Endangered Mutualisms: The Conservation of Plant-Pollinator Interactions. Annual Review of Ecology and Systematics 29:83-112. Kells, A. R., and D. Goulson. 2003. Preferred nesting sites of bumblee queens (Hymenoptera: Apidae) in agroecosystems in the UK. Biological Conservation 109:165-174. Klein, A.-M., B. E. Vaissiére, J. H. Cane, I. Steffan-Dewenter, S. A. Cunninghame, C. Kremen, and T. Tscharntke. 2007. Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society B 274:303-313. Knapp, A. K., J. M. Briggs, S. L. Collins, S. R. Archer, M. S. Bret-Harte, B. E. Ewers, D. P. Peters, D. R. Young, G. R. Shaver, E. Pendall, and M. B. Cleary. 2007. Shrub encroachment in North American grasslands: shifts in

63 growth form dominance rapidly alters control of ecosystem carbon inputs. Global Change Biology 14:615-623. Kremen, C., N. M. Williams, M. A. Aizen, B. Gemmill-Herren, G. LeBuhn, R. Minckley, L. Packer, S. G. Potts, T. a. Roulston, I. Steffan-Dewenter, D. P. Vázquez, R. Winfree, L. Adams, E. E. Crone, S. S. Greenleaf, T. H. Keitt, A.-M. Klein, J. Regetz, and T. H. Ricketts. 2007. Pollination and other ecosystem services produced by mobile organisms: a conceptual framework for the effects of land-use change. Ecology Letters 10:299-314. Longcore, T., C. S. Lam, P. Kobernus, E. Polk, and J. P. Wilson. 2010. Extracting useful data from imperfect monitoring schemes: endangered butterflies at San Bruno Mountain, San Mateo County, California (1982-2000) and implications for habitat management. Journal of Insect Conservation 14:335-346. Lopezaraiza-Mikel, M. E., R. B. Hayes, M. R. Whalley, and J. Memmot. 2007. The impact of an alien plant on a native plant-pollinator network: an experimental approach. Ecology Letters 10:539-550. Lundgren, R., Ø. Totland, and A. Lázaro. 2016. Experimental simulation of pollinator decline causes community-wide reductions in seedling diversity and abundance. Ecology 97:1420-1430. Luong, J. C., P. L. Turner, C. N. Phillipson, and K. C. Seltmann. 2019. Local grassland restoration affects insect communities. Ecological Entomology 44:471-479. MacDougall, A. S., and R. Turkington. 2007. Does the type of disturbance matter when restoring disturbance-dependent grasslands? Restoration Ecology 15:263-272. Maia, K. P., I. P. Vaughan, and J. Memmot. 2019. Plant species roles in pollination networks: an experimental approach. Oikos 128:1446-1457. McClintock, E., P. Reeberg, and W. Knight. 1990. A Flora of the San Bruno Mountains San Mateo County California. California Native Plant Society, Sacramento, CA. McFrederick, Q. S., and G. LeBuhn. 2006. Are urban parks refuges for bumble bees Bombus spp. (Hymenoptera: Apidae)? Biological Conservation 129:372-382. Meeuse, B. J. D. 2018. Pollination. Encyclopædia Britannica. Encyclopædia Britannica, inc. Meiners, J. M. 2016. Biodiversity, community dynamics, and novel foraging behaviors of a rich native bee fauna across habitats at Pinnacles National Park, California. Utah State University, ProQuest Dissertations Publishing. Memmot, J., N. M. Waser, and M. V. Price. 2004. Tolerance of polliantion networks to species extinctions. Proceedings of the Royal Society London 271:2605-2611. Michener, C. D. 2007. The Bees of the World. 2nd edition. The Johns Hopkins University Press, Baltimore, Maryland. Molanphy, T. 2016. Ancient Sand Dunes Under Threat from Development in California’s Bay Area. Earth Island Journal. Earth Island Institution.

64 Moldenke, A. R. 1976. California USA pollination ecology and vegetation types. Phytologia 34:305-361. Moyes, A. B., M. S. Witter, and J. A. Gamon. 2005. Restoration of Native Perennials in a California Annual Grassland after Prescribed Spring Burning and Solarization. Restoration Ecology 13:659-666. NPS. 2007. Point Reyes National Seashore Wildland Fire Resource Advisor Guide.in U. D. o. t. Interior, editor. National Park Service, Point Reyes Station, CA. Oksanen, J. 2019. Vegan: an introduction to ordination. Oksanen, J., F. G. Blanchet, M. Friendly, R. Kindt, P. Legendre, D. McGlinn, P. Minchin, R. B. O’Hara, G. L. Simpson, P. Solymos, M. H. H. Stevens, E. Szoecs, and H. Wagner. 2019. vegan: Community Ecology Package. R package version 2.5-6. Omand, K. A., J. M. Karberg, D. I. O’Dell, and K. C. Beattie. 2018. Harrowing and Seed Addition for Sandplain Grassland and Heathland Restoration. Natural Areas Journal 38:356-369. Ormshaw, H. 2018. San Bruno Mountain Habitat Conservation Plan: Year 2017- 18 Activities Report for Federally Listed Species. San Mateo County Parks Department, Redwood City, CA. Petanidou, T., A. S. Kallimanis, M. Lazarina, T. Tscheulin, J. Devalez, A. Stefanaki, E. Hanlidou, A. Vujic, A. Kaloveloni, and S. P. Sgardelis. 2018. Climate drives plant-pollinator interactions even along small-scale climate gradients: the case of the Aegean. Plant Biology 20:176-183. Petanidou, T., A. S. Kallimanis, J. Tzanopoulos, S. P. Sgardelis, and J. D. Pantis. 2008. Long-term observation of a pollination network: fluctuation in species and interactions, relative invariance of network structure and implications for estimates of specialization. Ecology Letters 11:564-575. Potts, S. G., J. C. Biesmeijer, C. Kremen, P. Neumann, O. Schweiger, and W. E. Kunin. 2010. Global pollinator declines: trends, impacts and drivers. Trends in Ecology & Evolution 25:345-353. Potts, S. G., B. Vulliamy, A. Dafni, G. Ne’eman, and P. Willmer. 2003. Linking Bees and Flowers: How do floral communities structure pollinator communities? Ecology 84:2628-2642. Poulos, H. M., R. C. Freeman, J. M. Karberg, K. C. Beattie, D. I. O’Dell, and K. A. Omand. 2020. Effects of Mowing and Prescribed Fire on Plant Community Structure and Function in Rare Coastal Sandplains, Nantucket Island, MA, USA. Environmental Management 65:111-121. Poulos, H. M., A. H. Taylor, and R. M. Beaty. 2007. Envionmental controls on dominance and diversity of woody plant species in a Madrean, Sky Island ecosystem, Arizona, USA. Plant Ecology 193:15-30. Reed, H. C., and P. J. Landolt. 2019. Ants, Wasps, and Bees (Hymenoptera).in G. R. Mullen, L. A. Durden, and J. G. King, editors. Medical and Veterinary Entomology. Elsevier. Reid, T. C., B. Castro, K. Weismann, and R. A. Arnold. 1980. Endangered Species Survey, San Bruno Mountain, Biological study - 1980-1981.in T. R. Associates, editor., Palo Alto, California.

65 Sedivy, C., A. Müller, and S. Dorn. 2011. Closely related pollen generalist bees differ in their ability to develop on the same pollen diet: evidence for physiological adaptations to digest pollen. Functional Ecology 25:718- 725. Smith DiCarlo, L. A., S. J. DeBano, and S. Burrows. 2019. Short-Term Response of Two Beneficial Invertebrate Groups to Wildfire in an Arid Grassland System, United States. Rangeland Ecology & Management 72:551-560. Thomson, D. 2004. Competitive Interactions Between the Invasive European Honey Bee and Native Bumble Bees. Ecology 85:458-470. Thorp, R. W., P. C. Schroder, and C. S. Ferguson. 2002. Bumble Bees: Boisterous pollinators of native California flowers. Fremontia 30:26-31. Tuell, J. K., A. K. Fiedler, D. Landis, and R. Issacs. 2008. Visitation by Wild and Managed Bees (Hymenoptera: Apoidea) to Eastern U.S. Native Plants for Use in Conservation Programs. Environmental Entomology 37:707-718. Tylianakis, J. M., E. Laliberté, A. Nielsen, and J. Bascompte. 2010. Conservation of species interaction networks. Biological Conservation 143:2270-2279. Ullmann, K., M. Vaughan, C. Kremen, T. Shih, and M. Shepherd. 2010. Califronia Pollinator Project: Citizen Scientist Pollinator Monitoring Guide. The Xerces Society for Invertebrate Conservation The University of California Berkeley, Portland, OR. vanEngelsdorp, D., K. S. Traynor, M. Andree, E. M. Lichtenberg, Y. Chen, C. Saegerman, and D. L. Cox-Foster. 2017. Colony Collapse Disorder (CCD) and bee age impact honey bee pathophysiology. PLoS ONE 12. Venjakob, C., A.-M. Klein, A. Ebeling, T. Tscharntke, and C. Scherber. 2016. Plant diversity increases spatio-temporal niche complementarity in plant- pollinator interactions. Ecology and Evolution 6:2249-2261. Waser, N. M., L. Chittka, M. V. Price, N. M. Williams, and J. Ollerton. 1996. Generalization in pollination systems, and why it matters. Ecology 77:1043-1060. Weiss, S. B., L. Naumovich, and C. Niederer. 2015. Assessment of the past 30 years of habitat management and covered species monitoring associated with the San Bruno Mountain habitat conservation plan., Creekside Science, Menlo Park. Westphal, C., R. Bommarco, G. Carré, E. Lamborn, N. Morison, T. Petanidou, S. G. Potts, S. P. M. Roberts, H. Szentgyörgyi, T. Tscheulin, B. E. Vaissiére, M. Woyciechowski, J. C. Biesmeijer, W. E. Kunin, J. Settele, and I. Steffan-Dewenter. 2008. Measuring Bee Diversity in Different European Habitats and Biogeographical Regions. Ecological Monographs 78:653- 671. Wheeler, M. M., C. Neill, E. Loucks, A. Weiler, B. Von Holle, M. Pelikan, and T. Chase. 2015. Vegetation removal and seed addition contribute to coastal sandplain grassland establishment on former agricultural fields. Restoration Ecology 23:539-547. Williams, P., S. Colla, and Z. H. Xie. 2009. Bumblebee vulnerability: common correlates of winners and losers across three continents. Conservation Biology 23:931-940.

66 Wilson, E. O. 1999. The Diversity of Life. W. W. Norton & Company, Inc., New York, N.Y. Winfree, R., R. Aguilar, D. P. Vázquez, G. LeBuhn, and M. A. Aizen. 2009. A meta-analysis of bees’ responses to anthropogenic disturbance. Ecology 90. Wojcik, V. A., G. W. Frankie, R. W. Thorp, and J. L. Hernandez. 2008. Seasonality in Bees and Thier Floral Resource Plants at a Constructed Urban Bee Habitat in Berkeley, California. Journal of the Kanses Entomological Society 81:15-28. Zavaleta, E. S., and L. S. Kettley. 2006. Ecosystem change along a woody invasion chronosequence in California grassland. Journal of Arid Environments 66:290-306.

67 Supplemental Table 1: List of plant species observed on San Bruno Mountain.

Grassland Plant or Coastal Name Latin Native Scrub or Acronyms Name Common Name Status Both? Acacia Australian Coastal ACME melanoxylon Blackwood nonnative Scrub Acaena ACPI pinnatifida California Acaena native Grassland Achillea ACMI millefolium Yarrow native Both Adenostoma Coastal ADFA fasciculatum Chamise native Scrub Aesculus Coastal AECA californica California Buckeye native Scrub Agoseris California AGGR grandiflora Dandelion native Grassland Aira AICA caryophyllea Silver Hair Grass nonnative Both Anaphalis ANMA margaritacea Pearly Everlasting native Both Angelica ANHE hendersonii Coast Angelica native Grassland Artemisia California ARCA californica Sagebrush native Both AVBA Avena barbata Wild Oat nonnative Both Baccharis BAPI pilularis Coyote Brush native Both BEPI Berberis pinnata California Barberry native Grassland BRMA Briza maxima Rattlesnake grass nonnative Both Calystegia Bush Morning CAOC occidentalis Glory native Both Carduus CAPY pycnocephalus Italian Thistle nonnative Both Coast Indian Coastal CAAF Castilleja affinis Paintbrush native Scrub CAEX Castilleja exserta Purple Owl Clover native Grassland Wright's Coastal CAWI Castilleja wightii Paintbrush native Scrub Ceanothus Blue Blossom Coastal CETH thyrsiflorus Ceanothus native Scrub Chlorogalum Small Flowered CHPA parviflorum Soaproot native Grassland

68 Chlorogalum CHPO pomeridianum Soaproot native Both Cirsium CIQU quercetorum Brownie Thistle native Both CIVU Cirsium vulgare Bull Thistle nonnative Both CLAM Clarkia amoena Farewell-to-spring native Both CLDA Clarkia davyi Davy's Clarkia native Grassland Clinopodium CLDO douglasii Yerba Buena native Both Conium Coastal COMA maculatum Poison Hemlock nonnative Scrub Coastal COSE Cornus sericea Creek Dogwood native Scrub Cotoneaster Silverleaf COPA pannosus Cotoneaster nonnative Both Cupressus Coastal CUMA macrocarpa Monterey Cypress native Scrub Cynosurus Bristly dogtail CYEC echinatus grass nonnative Both Danthonia DACA californica California Oatgrass native Grassland Deinandra DECO corymbosa Coastal tarweed native Grassland Coastal DEOD Delairea odorata Cape-ivy nonnative Scrub Deschampsia cespitosa ssp Coastal Tufted DECE holciformis Hair Grass native Both Diplacus Sticky Monkey DIAU aurantiacus Flower native Both Dittrichia Coastal DIGR graveolens Stinkwort nonnative Scrub DUFA Dudleya farinosa Sea Lettuce native Both ELGL Elymus glaucus Wild rye native Grassland Coastal ERGL Erigeron glaucus Seaside Daisy native Scrub Eriogonum California ERFA fasciculatum Buckwheat native Grassland Eriogonum ERLA latifolium Coast Buckwheat native Both Eriogonum luteolum var Tiburon Coastal ERLU caninum Buckwheat native Scrub Eriophyllum ERST staechadifolium Lizard Tail native Both

69 Erodium Common Stork's ERCI cicutarium Bill nonnative Both Eschscholzia ESCA californica California Poppy native Both Eucalyptus Eucalyptus/Tasman Coastal EUGL globulus ian Blue Gum nonnative Scrub Festuca FECA californica California Fescue native Grassland Rattail Sixweeks FEMY Festuca myuros Grass nonnative Grassland FERU Festuca rubra Red Fescue native Both Foeniculum FOVU vulgare Fennel nonnative Both Coastal FRVE Fragaria vesca Wild Strawberry native Scrub Frangula FRCA californica Coffeeberry native Both Genista GEMO monspessulana French Broom nonnative Grassland Grindelia GRHI hirsutula Gumplant native Grassland Helminthotheca HEEC echioides Bristly Ox tongue nonnative Both Heracleum HEMA maximum Cow parsnip native Both Heteromeles HEAR arbutifolia Toyon native Both Heterotheca HESE sessiliflora Golden Aster native Both Hirschfeldia HIIN incana Mustard nonnative Both Holodiscus Coastal HODI discolor Creambush native Scrub Horkelia HOCA californica California Horkelia native Both Hypericum HYPE perforatum Klamathweed nonnative Grassland Hypochaeris HYGL glabra Smooth Cat's Ear nonnative Both Coastal ILOP Ilex opaca American Holly nonnative Scrub IRDO Iris douglasiana Douglas Iris native Both Koeleria Coastal KOMA macrantha June Grass native Scrub LASE Lactuca serriola Prickly Lettuce nonnative Grassland

70 Poison Wild LAVI Lactuca virosa Lettuce nonnative Both Leucanthemum LEVU vulgare Oxeye Daisy nonnative Both Lonicera California Coastal LOHI hispidula Honeysuckle native Scrub Lonicera Twinberry Coastal LOIN involucrata Honeysuckle native Scrub Lotus LOCO corniculatus Bird's Foot Trefoil nonnative Grassland Lupinus albifrons Douglas Silver LUAL var. collinus Lupine native Both Lysimachia LYAR arvensis Scarlet Pimpernel nonnative Grassland Coastal MAOR Marah oregana Coast Man-Root native Scrub Monardella MOVI villosa Coyote Mint native Both Phacelia PHCA californica California phacelia native Grassland Phacelia Coastal PHNE nemoralis Shade Phacelia native Scrub Plantago PLEL elongata Coast Plantain native Grassland PLER Plantago erecta California Plantain native Grassland Plantago PLLA lanceolata English Plantain nonnative Both PLOV Plantago ovata Wooly Plantain native Grassland Coastal PRVU Prunella vulgaris Common Selfheal native Scrub Pseudognaphaliu California PSCA m californicum Everlasting native Both Pteridium Common PTAQ aquilinum (Western) Bracken native Both QUAG Quercus agrifolia Coast Live Oak native Both RASA Raphanus sativus Wild Radish nonnative Grassland Ribes Hillside Coastal RICA californicum Gooseberry native Scrub Ribes Coastal RIQU quercetorum Oak Gooseberry native Scrub Rubus Himalayan RUAR armeniacus Blackberry nonnative Both California RUUR Rubus ursinus Blackberry native Both RUAC Rumex acetosella Sheep Sorrel nonnative Both

71 RUCR Rumex crispus Curly Doc nonnative Both RUPU Rumex pulcher Fiddle Dock nonnative Grassland SASP Salvia spathacea Hummingbird Sage native Both Sambucus Coast Red Coastal SARA racemosa Elderberry native Scrub Scabiosa SCJA japonica Pincushion Flower nonnative Grassland Scrophularia SCCA californica Bee Plant native Both Silene scouleri SISC ssp. grandis Scouler Catchfly native Grassland Silybum Coastal SIMA marianum Milk Thistle nonnative Scrub Sisyrinchium SIBE bellum Blue Eyed Grass native Grassland Solanum SOPH physalifolium Hairy Nightshade nonnative Both Solidago Coast/dune SOSP spathulata goldenrod native Grassland SOAR Sonchus arvensis Field sowthistle nonnative Both SOAS Sonchus asper Spiny Sowthistle nonnative Both Sonchus SOOL oleraceus common sowthistle nonnative Grassland Spergularia Boccone's Sand SPBO bocconi Spurry nonnative Grassland Purple Needle STPU Stipa pulchra Grass native Grassland Symphoricarpos Common Coastal SYAL albus Snowberry native Scrub Symphoricarpos Western Coastal SYOC occidentalis Snowberry nonnative Scrub Taraxacum California TACA californicum Dandelion native Grassland Taraxacum Common TAOF officinale Dandelion nonnative Grassland Toxicodendron TODI diversilobum Poison Oak native Both TRIF Trifolium Clover Grassland Trifolium TRWI willdenovii Tomcat Clover native Grassland Coastal ULEU Ulex europaeus Gorse nonnative Scrub Verbascum VEBL blattaria Moth Mullein nonnative Both

72 Vicia sativa ssp. VISA sativa Spring Vetch nonnative Both Wyethia Narrowleaf mule- WYAN angustifolia ears native Grassland Zantedeschia Coastal ZAAE aethiopica Callalily nonnative Scrub

73

Supplemental Table 2: List of pollinator species found on San Bruno Mountain.

Generalist Pollinator Native or Common Name Name Family Order Group Status Specialist Ocean Spray Fairy Adela Adelidae Lepidoptera Butterflies Native Moth septentrionella and Moths Texas Striped-Sweat Agapostemon Halictidae Hymenoptera Bee Native Generalist Bee texanus European wool carder Anthidium Megachilidae Hymenoptera Bee Exotic Generalist bee manicatum California Digger Bee Anthophora Apidae Hymenoptera Bee Native Generalist californica Western Honey Bee Apis mellifera Apidae Hymenoptera Bee Exotic Generalist

Peridot Green-Sweat Augochlorella Halictidae Hymenoptera Bee Native Generalist Bee pomoniella Pipevine Swallowtail Battus Papilionidae Lepidoptera Butterflies Native Generalist philenor and Moths American Sand Wasp Bembix Crabronidae Hymenoptera Wasp Native Generalist americana California Bumble Bombus Apidae Hymenoptera Bumble Bee Native Generalist Bee californicus

74 Fernald's Cuckoo Bombus Apidae Hymenoptera Bumble Bee Native Generalist Bumble Bee fernaldae Black-tailed Bumble Bombus Apidae Hymenoptera Bumble Bee Native Generalist Bee melanopygus Red-belted Bumble Bombus Apidae Hymenoptera Bumble Bee Native Generalist Bee rufocinctus Yellow-faced Bumble Bombus Apidae Hymenoptera Bumble Bee Native Generalist Bee vosnesenskii Greater Bee Fly Bombylius Bombyliidae Diptera Fly Native Generalist major Blue Bottle Fly Calliphora Calliphoridae Diptera Fly Native Generalist vomitoria Moss's Elfin (San Callophrys Lycaenidae Lepidoptera Butterflies Native Generalist Bruno Elfin) mossii and Moths Anna's Hummingbird Calypte anna Trochilidae Caprimulgiformes Hummingbird Native Generalist

Echo Azure Celastrina Lycaenidae Lepidoptera Butterflies Native Generalist echo and Moths Common Ringlet Coenonympha Nymphalidae Lepidoptera Butterflies Native Generalist tullia and Moths Eristalis Syrphidae Diptera Fly Exotic dimidiata Drone Fly Eristalis tenax Syrphidae Diptera Fly Exotic Generalist Hoverflies Family: Syrphidae Diptera Fly Generalist Syrphidae

75 Potter Wasps Genus: Vespidae Hymenoptera Wasp Generalist Ancistrocerus Bottle Fly Genus: Calliphoridae Diptera Fly Generalist Calliphora Cellophane Bees Genus: Colletidae Hymenoptera Bee Colletes Carrot Wasps Genus: Gasteruptiidae Hymenoptera Wasp Gasteruption Bee Flies Genus: Geron Bombyliidae Diptera Fly Sweat Bee Genus: Halictidae Hymenoptera Bee Generalist Halictus Sweat Bee Genus: Halictidae Hymenoptera Bee Generalist Lasioglossum Bee Flies Genus: Bombyliidae Diptera Fly Lepidanthrax Blister Beetle Genus: Lytta Meloidae Coleoptera Beetle generalist Long-horned bee Genus: Apidae Hymenoptera Bee generalist Melissodes Mining Bees Genus: Andrenidae Hymenoptera Bee Panurginus Hoverflies Genus: Syrphidae Diptera Fly Paragus Hoverflies Genus: Scaeva Syrphidae Diptera Fly Hoverflies Genus: Syrphidae DIptera Fly Sphaerophoria Yellow Jacket Genus: Vespidae Hymenoptera Wasp Vespula

76 Tripartite Sweat Bee Halictus Halictidae Hymenoptera Bee Native generalist tripartitus Sinuous Bee Fly Hemipenthes Bombyliidae DIptera Fly Native generalist sinuosa Acmon Blue Icaricia Lycaenidae Lepidoptera Butterflies and Native generalist acmon Moths Western Leafcutter Megachile Megachilidae Hymenoptera Bee Native generalist perihirta Hoverflies Myathropa Syrphidae Diptera Fly Exotic generalist florea Woodland SKipper Ochlodes Hesperiidae Lepidoptera Butterflies and Native Generalist sylvanoides Moths Order: Diptera Diptera Fly Pale Tiger Swallowtail Papilio Papilionidae Lepidoptera Butterflies and Native Generalist eurymedon Moths Western Tiger Papilio Papilionidae Lepidoptera Butterflies and Native Generalist Swallowtail rutulus Moths Anise Swallowtail Papilio Papilionidae Lepidoptera Butterflies and Native Generalist zelicaon Moths FIeld Crescent Phyciodes Nymphalidae Lepidoptera Butterflies and Native Generalist pulchella Moths Cabbage White Pieris rapae Pieridae Lepidoptera Butterflies and Native Generalist Moths European Paper Wasp Polistes Vespidae Hymenoptera Wasp Exotic Generalist dominula

77 Scaeva Syrphidae Diptera Fly Native pyrastri Sweat Bee Sphecodes Halictidae Hymenoptera Bee Exotic Generalist davisii Metallic Sweat Bee Subgenus: Halictidae Hymenoptera Bee Dialictus Small Carpenter Bee Subgenus: Apidae Hymenoptera Bee Zadontomerus Toxomerus Syrphidae Diptera Fly Native Generalist marginatus Toxomerus Syrphidae DIptera Fly Native Generalist occidentalis Occidental Grasshopper Trimerotropis Acrididae Orthoptera Grasshopper occidentalis Boyote Brush Beetle Trirhabda Chrysomelidae Coleoptera Beetle Native flavolimbata Painted Lady Vanessa Nymphalidae Lepidoptera Butterflies and Native Generalist cardui Moths

78