The effect of resource limitation on floral visitors of Monardella odoratissima

Isabelle Everhart1, Ashish Kothari2, Chris McCarron3, and Tricia Nguyen4

1University of California, Santa Cruz; 2University of California, Riverside; 3University of California, Berkeley; 4University of California, Davis

ABSTRACT

Fluctuating flowering patterns affect an environments floral abundance, impacting the spatial structure of nectar and pollen resources, ultimately influencing floral visitor foraging behavior. We examined the effect of reducing floral resources on floral visitors of Monardella odoratissima, commonly known as the mountain coyote mint. Isolated patches of M. odoratissima were created by removing all surrounding resources, in order to test its effect on floral visitor communities. Exploring how the resource- richness of a habitat affects this relationship, we conducted our study at two ecologically contrasting sites the Sierra Nevada Aquatic Research Laboratory (SNARL) dominated by sagebrush steppe and Valentine Camp Reserve made up of montane chaparral. For both sites we anticipated changes in floral visitor composition, time spent per inflorescence, and patch visitation time, but expected results to be stronger at Valentine. A change in species composition of flower visitors occurred in our Valentine but was not found at SNARL. Only a weak interaction was found between treatment and flower visitation that accounted for an insignificant reduction of time spent per inflorescence at SNARL, Valentine had no difference. Overall floral visitor time spent before and after treatment showed difference only among site. Our findings offer insight into ways species composition shifts when resource availability changes and our methods provide a platform that should be expanded upon to a larger area with a longer research period.

Keywords: Monardella odoratissima, floral visitor, resource limitation, Valentine Eastern Sierra Reserves, foraging behavior

INTRODUCTION that invertebrates rely on. Past research has shown that resource availability within Abiotic variables such as weather patterns landscapes can alter the foraging behavior and resource availability change from year of organisms that rely on particular floral to year and greatly impact plant species resources (Mustajarvi et al. 2001). For flower production and flowering phenology example, floral visitors have been shown to (Flo et al. 2018). These fluctuations can visit plants at different rates by spending greatly affect the vegetative landscapes

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 1/8

longer or shorter times at flowers in commonly known as mountain coyote mint, response to resource availability (Sih and is a flowering species endemic to Western Baltus 1987). For instance, when a plant North America and found in the Eastern species does not flower extensively or Sierra Nevada range. Due to its success in bloom at a different time of year than varying environmental conditions and usual, flower visitors foraging on this plant habitats, M. odoratissima is the most may be forced to look elsewhere for non- widespread Monardella species (Epling preferred resources (Flo et al. 2018). 1925). M. odoratissima has two subspecies Understanding community-level effects of and while they do overlap in ranges, M. variation in flower phenology provides odoratissima ssp. pallida is found only in insight into floral visitor success because montane areas while ssp. glauca extends its these organisms rely on floral abundance, range into Great Basin sagebrush habitat. composition and co-flowering species (Flo Through observation and experimental et al. 2018). A habitat’s floral abundance manipulation, our study examined the influences the spatial structure of nectar interaction between two subspecies of M. and pollen resources, which has an odoratissima and their surrounding visiting influence on floral visitor foraging behavior. insect communities. More specifically, we Examining how insect visitation to a single explored how limiting the surrounding species changes as plant and plant visitor resources of M. odoratissima would affect community structures shift provides insight its floral visitor species composition and into the ways plant visiting insects alter amount of time floral visitor species would their behavior under different environmental spend at flower patches. After treatment conditions to maximize energy gained we expected to find a slight shift in species versus energy lost (Westphal et al. 2006, composition, an increase in overall time Steffan-Dewenter and Tscharntke 1999). spent by floral visitors at patches, and a Widely distributed native plants that have decrease in the amount of time spent per adapted a generalist pollination strategy flower by individual visitors. Montane and occur in numerous habitats present chaparral contains a more diverse amount useful study species. They may provide of flowering plants as well as floral visitors, insight on how different native floral visitors therefore, after limiting the resources are affected by the availability and spatial surrounding ssp. pallida we anticipated to structure of resources. These same species see a greater effect there than in the can also shed light on how particular floral sagebrush chaparral with spp. glauca. visitors change behavior throughout different habitats within their range. METHODS The Eastern Sierra Nevada provides a range of species and ecological 2.1 Natural History of Study System communities due to its radical shift over Data was collected at the Valentine relatively short distances due to steep Eastern Sierra Reserves (VESR) from July 30 elevational gradients, presenting a study to August 4, 2019. VESR is made up of two system to examine dramatic plant-visitor reserves, the Sierra Nevada Aquatic interaction shifts between habitats within a Research Laboratory (SNARL) and Valentine short distance. Monardella odoratissima,

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 2/8

Camp. The study sites at each reserve were odoratissima flower from six individual located at 37.6102 N, -118.8466 W and plants was surveyed (Table 2). To determine 37.6327 N, -118.9966 W, respectively. the floral visitor community at each plot, Despite being only 12 miles away from each we spent an hour at each flowering species. other, SNARL and Valentine Camp offered During each hour trial, every new floral very different study environments. These visitor species was noted and collected for two reserves differ in elevation (SNARL: identification. Stephen et al. (1969) was 1,250–4,012 m, Valentine: 2,437–2,605 m) used to key bee species to genera and and precipitation rates (SNARL: 25–38 inaturalist.org was used for other insects. cm/yr, Valentine: 51–61 cm/yr), but overlap Since flower visitation rates have been in the average summer temperature (28–29 known to peak at 12:30 PM for M. °C). SNARL hosts a Great Basin sagebrush odoratissima, our observation window was steppe vegetation. Glaciation shaped set from 10 AM to 1 PM plus/minus 30 Valentine Camp’s landscape, creating minutes (Akiba et al. 2018). Once flower surrounding lateral and terminal moraines, visitor species were collected and identified which host a different plant community by genera, they were sorted into functional assemblage of sagebrush steppe, including groups: Lepidoptera, Hymenoptera (solitary), M. odoratissima. Additionally, Valentine Hymenoptera (social), and other (Table 3). Camp lies in a glacier-carved basin, but is For our control, visitation rate and length dominated by Jeffrey pine and lodgepole of visits for flower visitors of Monardella pine forest. It also hosts montane was surveyed. At both sites, we observed meadows, montane chaparral, and patches four separate patches of Monardella, of sagebrush steppe. Plant diversity and ranging in number of inflorescences from composition among the reserves differs, 13 to 38. At SNARL our plot contained 88 and they host two different subspecies of inflorescences of ssp. glauca, whereas in Monardella odoratissima. M. odoratissima Valentine the patch contained 94 ssp. glauca is found at SNARL, whereas M. inflorescences of ssp. pallida. Surveys took odoratissima ssp. pallida is found at place over 1 hour, with four 15 minute trials Valentine. from 12 PM to 1 PM +/- 30 minutes. Each floral visitor species was listed, as well as 2.2 Research Design the total time spent at each patch in seconds. Data collection and manipulation began at SNARL from July 30 to August 1, and then at Valentine from August 2 to August 4. At both sites, we found a similar sized patch of M. odoratissima that co-occurred with several other flowering species (Table 1). Using the largest patch of M. odoratissima found as the center point, we created a 10 m x 10 m (100 m2) plot. Within this plot, the inflorescence width and pistil, stamen, tube, and petal length of one M.

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 3/8

Table 1. Plant community species list Table 3. Insect community species list

To create resource limitations, we Table 2. The two subspecies of M. odoratissima differed in inflorescence size (N=6, t=4.33 p=0.0015), expanded our original 10 x 10 m plot to 30 x 2 floral tube length (N= 6, t= 2.84, p= 0.017) and the 30 m (900 m ), and eliminated all flower exertion of the pistil (N= 6, t= 3.16, p= 0.010). resources surrounding our four M. odoratissima control patches. This was done by covering all flowering shrubs in the plot with plastic tarps and cutting any flowers that remained exposed. We allowed

for a 45 minute period to pass before treatment sampling in order to allow flower visitors to acclimate to the manipulated environment. The post-treatment floral visitor surveying was done with the same methods as the control.

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 4/8

2.3 Statistics three of which were in flower. At Valentine, 20 plant species were found, 10 of which All statistical analyses were conducted were in flower in our 900 m² plot. 10 plant using JMP statistical software v14.0. After species were present within the 100 m2 accounting for the different genera found plot, four of which were in flower (Table 1). between SNARL and Valentine, we made Floral measurements of ssp. glauca found direct comparisons between the number of larger inflorescence head width by 4.17cm floral visits per genera to determine the (N=6, t=4.33 p=0.0015). Measurements of differences between sites and the ssp. pallida had a longer floral tube by 1.66 experimental manipulation to observe any cm(N= 6, t= 2.84, p= 0.017) along with a overlap in species composition and more exerted pistil by 2 cm (N=6, t=3.16, p= visitation. T-tests were performed on flower 0.010) (Table 2). Comparisons of total size measurements between the two visitation to flower patches among sites and subspecies of M. odoratissima to determine treatments shows that SNARL had far fewer if there were any differences in floral visits but the number of visits didn’t change morphology. To understand the effects of dramatically with treatment (Figure 1). different habitats and the experimental manipulation on pollinator community Table 4. Floral visitor genera found on structures, we performed a chi-square test M. odoratissima subspecies glauca and pallida. to look at the differences in the total order proportions and the effects of the treatment at both sites. To test our prediction of an increase of treatment effect at SNARL versus Valentine we ran a 2- way ANOVA on log transformed proportion of time spent per inflorescence number in flower patches, with site and treatment as predictors including the predictor term.

RESULTS

Upon surveying all insects in the 900 m2 plots, 35 genera of floral visitors were recorded between the two sites, 20 genera at SNARL and 23 genera at Valentine. Twenty-five of the total genera across sites were recorded visiting M. odoratissima over Figure 1. Total amount of visits to flower patches among sites as well as experimental trials. the course of surveying, 15 of those genera were at SNARL and 17 genera were at Statistical analysis of data showed that Valentine (Table 4). Our 900 m2 plot at SNARL and Valentine both had a different SNARL had a total of 21 plant species, six of proportion of species when they were which were in flower. Our smaller 100 m2 binned into categories reflective of their plot had seven of the overall plant species, orders, except for Hymenoptera (social)

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 5/8

(N=4, Chi=22.6 p=0.0001). Treatment had extremely weak interaction (N=512, F=2.27, an effect on species proportions at p=0.11). However, the main effect of site Valentine, driven by shifts in Hymenoptera showed a lower time spent of floral visitors (Solitary) and Lepidoptera (N=4, Chi=14.29, at SNARL (N=512, F=2.27, p=0.032). p=0.0025) (Figure 2). Examination of Treatment was not significant in this model individual species driving these patterns (N=512, F=2.27, p=0.18) (Figure 3), but we showed that Polites sp. (Lepidoptera) has a expect that including the interaction term 166.67% increase in visitation with effects this result. treatment, while Steniolia sp., Andrena sp., and Megachile sp. (Hymenoptera (Solitary)) all have reductions of 43.64%, 65.22%, and 58.33% respectively (Table 5).

Figure 3. Inflorescence visitation. Visitation of inflorescence showed an extremely weak interaction Figure 2. Functional groups. When genera were (N=512, F=2.27, p=0.11) with main effect of site sorted into functional groups and analyzed, the showing a lower time spent at SNARL (N=512, proportional composition was different among sites F=2.27, p=0.032) , treatment was not significant (N=4, Chi=22.6, p=0.0001). Treatment also had an (N=512, F=2.27, p=0.18). effect on the proportions of genera (N=4, Chi=14.29 p=0.0025). DISCUSSION Table 5. Four genera were responsible for changes in community composition Polites sp, Steniolia sp, Despite initial predictions that resource Andrena sp, Megachile sp at Valentine Reserve post limitation would cause floral visitors to stay treatment (Figure 2). Social Hymenoptera, a category at M. odoratissima for longer, we saw no of single genus, Bombus sp., did not change across effect of treatment on visitation time at all sites. either site (Figure 1). Instead, we found that visitation rate differed overall, with SNARL having far fewer species visitations. The floral visitor community had large differences among habitats (Table 3) as did plant species (Table 1). It was observed at the time of surveying that Tetradymia The test of site and treatment on the canescens was beginning to bloom; bushes normalized time spent per inflorescence had numerous species visiting at all times. with the interaction term showed an This could indicate that, due to phenological

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 6/8

shifts in plant species, other resources were ssp. glauca potentially dissuaded the floral preferred by floral visitors at SNARL. visitors from traveling to it (Blaauw & Isaacs Due to M. odoratissima’s generalist 2014). These species have less floral characteristics, a large proportion of insects resources in their environment overall, and were found visiting the plant; 75% of total thus, are less likely to venture to a patch floral visitors were found visiting ssp. glauca with only one flowering species. In contrast, at SNARL and 74% were found on ssp. floral visitors at Valentine are less resource pallida at Valentine (Table 3). Treatment did limited in their environment, making them not affect the composition of floral visitors more willing to travel to a single species at SNARL but did cause a shift in those at patch. Future studies could strengthen this Valentine (Figure 2). These findings were potential pattern by studying the driven by four genera: Polites sp. relationship throughout an entire flowering (Lepidoptera) had 166.67% more visitation season, as well as by testing competition. with treatment, while Steniolia sp., Andrena When looking to the future, the sp., and Megachile sp. (Hymenoptera manipulation used in our experiment may (Solitary)) all had reductions in visitation by provide an interesting platform to expand 43.64%, 65.22%, and 58.33% respectively upon. M. odoratissima has a large range (Table 5). Previous research has shown that and can be found in a variety of different after a habitat is altered in the spatial habitats, therefore this species can be used density of its floral resources, it may to answer other hypotheses about floral become so unfavorable for certain species visitor communities across its range. Using that they avoid it entirely (Mustajarvi et al. our manipulation method and expanding it 2001). The reduction in Hymenoptera to a larger area could also yield interesting visitation specifically may be because findings. Establishing a site for a longer species in this order are limited in foraging amount of time throughout the season and distance (Gathmann et al. 2002). examining how species alter their behavior Conversely, visitation by species within the through phenological transitions would order Lepidoptera increased dramatically, allow for numerous questions to be asked. which may be explained by strong flying One could expand on literature addressing abilities. competition, foraging strategies, and plant An extremely weak interaction was found fitness when floral visitors across a between treatment and time spent per community have limited resources in their flower, although not the one we expected. natural environment. Yearly variability of Our hypothesis predicted that there would abiotic factors suggests that running a be a treatment effect of decreased time manipulation plot over multiple years spent per flower, but that it would be larger would assist in answering questions related at Valentine. However, floral visitors spent to resource fluctuations at the landscape less time spent per flower at SNARL and scale, allowing for additional comparisons there was no difference at Valentine. This to previous studies on the subject (Flo et al. may be caused by SNARL having fewer 2018). flowering species. Floral visitors increased While our research provides only a with flower richness so eliminating all glimpse into patterns occurring when this flowering resources at SNARL aside from experimental manipulation is done, we

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 7/8

hope this will inspire future research. Blaauw, B. R., & Isaacs, R. 2014. Larger patches of Conducting this rewarding study was our diverse floral resources increase insect pollinator density, diversity, and their pollination of native first attempt at research as budding wildflowers. Basic and Applied Ecology 15:701– ecologists and we learned valuable skills by 711. developing our experimental design, collecting and analyzing data, and finally Epling, Clawson. 1925. C. Monograph of the Genus writing our publications and presenting our Monardella. Ann. Missouri Bot. Gard. 12:1–106. findings. We look forward to becoming Flo, V., Bosch, J., Arnan, X., Primante, C., Martín, A. more deeply involved in the greater M., Barril-graells, H., & Rodrigo, A. 2018. Yearly scientific community. fluctuations of floral landscape in a Mediterranean scrubland: consequences on pollen-nectar ACKNOWLEDGMENTS availability. PloS One 13. e0191268. https://doi.org/10.1371/journal.pone.0191268 This work was performed at the Gathmann, A., and T. Tscharntke. 2002. Foraging University of California Natural Reserve ranges of solitary bees. Journal of Animal Ecology System’s Valentine Camp 71:757–764. doi:10.21973/N3JQ0H and Sierra Nevada Aquatic Resource Laboratory Goulson D. 2010. Bumblebees; their behaviour, doi:10.21973/N3966F. We would like to ecology and conservation. Oxford University Press, Oxford. thank the University of California Natural Reserve System for allowing us to conduct Mustajarvi, K., P. Siikamaki, S. Rytkonen, and A. experiments at different reserves over the Lammi. 2001. Consequences of plant population course of this class. We would also like to size and density for plant-pollinator interactions thank our wonderful teachers, Krikor and plant performance. Journal of Ecology 89:80– 87. Andonian, Ph.D., and Tim Miller, Ph.D. along with the course T.A. Prahlada Papper Sih, A., and Baltus, M. 1987. Patch size, pollinator for sharing their knowledge with us and behavior, and pollinator limitation in catnip. assisting in our growth as budding Ecology 68:1679–1690. ecologists. Perhaps the most important Steffan-Dewenter, I., and T. Tscharntke. 1999. people to thank are our dear friends, the Effects of habitat isolation on pollinator course assistants Amelia Maurer and Selena communities and seed set. Oecologia 121:432– Vengco, who tackled the challenge of 440. coordinating meals for 31 people with an amazing attitude. Lastly, we would like to Stephen, W. P., G. E. Bohart, and P. F. Torchio. 1969. The biology and external morphology of bees with thank our loving parents for their guidance a synopsis of the genera of North-western and support. America. 144 pp. Oregon State University, Corvallis. REFERENCES Westphal, C., I. Steffan-Dewenter and T. Tscharntke. Akiba,S., Lougee, E., & Thompson III, R. 2018. 2006. Foraging trip duration of bumblebees in Pollinator community structure and interactions relation to landscape-wide resource availability. on Monardella odoratissima spp. glauca. CEC Ecological Entomology 31:389–394. Research Summer:1–7.

CEC Research | https://doi.org/10.21973/N3QS95 Summer 2019 8/8