Potentilla Fruticosa (rosaceae) as a Nectar Plant for Butterflies

Despite its wide distribution, little documentation exists to confirm that butterflies use the woody shrub Potentilla fruticosa (Linnaeus) (Rosaceae) as a nectar plant. During the summer of 2004, we observed 36 individual butterflies belonging to 10 species nectaring on P. fruticosa in the Jemez Mountains, New Mexico. Butterflies spent 56% of total observed nectaring time on P. fruticosa, where it composed 26% of total blooming forb availability (out of 17 plant species). We used the anthrone method for carbohydrate analysis of P. fruticosa nectar samples and found significantly more (χ±SE μg/2ml) carbohydrates (i.e., nectar) in flowers (n=68) excluded from nectivores (26.83±1.35 mg/2ml) than available (n=63) to nectivores (6.71+1.40 mg/2ml). Carbohydrate levels were also significantly higher in nectar later in the sampling season (Two-way ANOVA with repeated measures, p<0.05). Although anecdotal observations suggest that P. fruticosa is not a preferred nectar source for butterflies in the northern Rocky Mountains and in others areas of its range, our results indicate that P. fruticosa is an important nectar resource for adult butterflies in the Jemez Mountains, New Mexico. Additional key words: Pentaphylloides floribunda, flower visitation, Lepidoptera, nectar, nectaring The wide distribution and blooming phenology of shrubby cinquefoil Potentilla fruticosa (Linneaus) (Rosaceae) [syn: Pentaphylloides floribunda (Pursh) A. Löve] make it a potential nectar source for many insect species. The life history and ecological characteristics of this shrub are well documented (Elkington & Woodell 1963, NRCS 2006, USGS 2005). In North America, ornamental and wild cultivars bloom from May to September with two main flowering periods in May and August. The latter is more vigorous and produces larger flowers (Elkington & Woodell 1963). P. fruticosa is intolerant of shade, and the wild North American form produces yellow flowers, while some horticultural varieties originating from Asia produce white, pink, orange, or red flowers (Elkington & Woodell 1963, NRCS 2006). Inflorescences are terminal and appear solitary or in small clusters. Flowers have five petals, triangular ovate sepals, and open nectaries (Elkington & Woodell 1963, USGS 2005). In North America, P. fruticosa ranges from the Arctic slope of northern Alaska to Newfoundland, south to the Sierra Nevada and Rocky Mountains, and east through the Great Lakes to New England and Labrador (Elkington & Woodell 1963, NRCS 2006, USGS 2005). North American individuals appear to be uniformly diploid, but populations are tetraploid in Europe and diploid or hexaploid in Asia (Elkington & Woodell 1963). Little documentation exists on the use of P. fruticosa as a nectar source by insects, other than a limited number of species belonging to the orders Diptera, Coleoptera, Hymenoptera (Elkington & Woodell 1963) and Lepidoptera (Voss 1954, Emmel 1964, Emmel et al. 1992, Opler & Krizek 1984). The Brooklyn Botanic Garden (2007) lists P. fruticosa as a bee-pollinated species, and several gardening and horticultural websites recommend P. fruticosa for butterfly gardens. Voss (1954) observed five species of butterflies nectaring on P. fruticosa in Michigan. These were Danaus plexippus (Linnaeus), Nymphalis milberti (Godart), Satyrium titus, (Fabricius) (Nymphalidae), Lycaena dorcas (Kirby) (Lycaenidae) and Erynnis lucilius (Scudder & Burgess) (Hesperiidae). In the central Rocky Mountains of Colorado, Emmel (1964) and Emmel et al. (1992) observed five species of butterflies nectaring on P. fruticosa; Euphydryas anicia eurytion (Mead), Polygonia zephyrus (Edwards) (Also known as Polygonia gracilis (Grote & Robinson), Satyrium titus titus, Euphydryas anicia capella (Barnes) (Nymphalidae), and Lycaena rubidus (Behr) (Lycaenidae). The Wisconsin Department of Natural Resources (2006) listed P. fruticosa as a nectar source for the endangered Calephelis mutica (McAlpine) 223223 JOURNAL OF THE LEPIDOPTERISTS’ SOCIETY (Riodinidae) and Opler and Krizek (1984) listed P. fruticosa as the larval host plant for L. dorcas. Webster & deMaynadier (2005) listed P. fruticosa as the host and principle nectar plant for L. dorcas claytoni (Brower) in Maine. We have been studying butterfly abundance and species richness in Bandelier National Monument and the Valles Caldera National Preserve, New Mexico, 1999–2004 (Kleintjes Neff et al. 2007, USGS 2005). In an experiment evaluating the impact of ungulate browsing upon butterflies and their host plants, we found butterfly richness and abundance to be greatest in areas containing P. fruticosa (unpubl. data). We additionally found butterflies nectaring on P. fruticosa, yet in the literature we found little documentation to confirm its use as a nectar source in the wild. Moreover, peer review of our initial qualitative observations was met with skepticism by reviewers. As a result, the objective of our study was to quantify the use (e.g., species visitation rates, nectar carbohydrate content) of P. fruticosa flowers by adult butterflies in the Jemez Mountains, New Mexico. MATERIALS AND METHODS Study Area. During the summer of 2004 (9 July–9 August), we worked within four study plots in the Jemez Mountains, New Mexico. Two were located in Bandelier National Monument (mixed conifer-MC4, meadowMD) and one each in the adjacent Santa Fe National Forest (ski basin-SB) and Valles Caldera National Preserve (Valles Caldera-VC). We chose sites that had ~25% total available blooming forb cover of P. fruticosa due to little or no elk browsing (inside exclosures and near human traffic). All sites were located within openings surrounded by mixed conifer-aspen forest between 2700m and 2830m in elevation. Sites were approximately 1200 m2 in size except for SB, which was approximately 576 m2 in size. Sites were considered independent (>2 km from each other). We collected butterfly foraging observations and estimated nectar plant availability in the MC4, MD, and SB sites. We sampled nectar availability for analysis at all four sites. Adult butterfly foraging behavior. We compiled butterfly foraging observations (1000–1500hr) from 10 July–4 August. Once an individual was sighted, we noted the species and then waited 5 sec. before initiating a 5 min. observation period. Butterflies were identified to species by sight and if necessary compared with voucher specimens or photographs (Glassberg 2001). For each butterfly, we collected detailed foraging time budgets which included recording the percent total observation time (sec) spent flying, basking, nectaring, mating, and in combat. We also recorded the percent time each individual spent nectaring per plant species and basking per substrate. Floral abundance and phenology. We estimated flower availability for eleven randomly selected plants in each site. We categorized the availability of open P. fruticosa flowers/plant by intervals of 50 flowers (1–50, 51–100, 101–150, 151–200 flowers). In three sites (MD, MC4, SB) we randomly selected five flowers for phenology studies (bud-to-bloom-to-closing) to determine the average time an individual flower was open (i.e., nectar was available). We marked individual flowers with flagging and noted the stage (bud, bloom, closing, closed) of development every day until they senesced. Nectar Plant Availability. We conducted a rapid assessment of blooming forb availability in each butterfly foraging observation site using Foxx and Hoard (1995) for plant species identification. We walked three (20-m) transect lines through each site. At every four meters along each transect we noted the closest blooming forb in a 2-m radius from the observer. We computed frequency of occurrence for available forb species by dividing the occurrence of each species closest to the observer from the total number of observations. Data were collected on 13 July and 6 August. Nectar availability and analysis. We measured the greatest height and width of each randomly selected P. fruticosa shrub per site (n=11). On each plant, we randomly selected 12 freshly opened flowers for nectar analysis. We bagged [treated] six flowers per plant and left six unbagged [untreated]. Bags excluded nectivores 24 hr prior to sampling. Bag material consisted of soft window screening, which excluded nectivores yet allowed air circulation. We extracted nectar from three flowers per treatment with micro-pipettes (10 ml; 1 ml increments) and paper wicks of ©Whatman’s filter paper (2x8 mm paper insect “points”) (Kearns & Inouye, 1993). We removed flower stamens with forceps prior to nectar sampling to decrease potential contamination by pollen and to make nectaries more accessible. We used the pipette samples to quantify nectar volume in the field and the wicks to analyze total carbohydrate content and type in the lab. We took samples between 0930–1440 hrs within two sampling periods (16–27 July, 29 July–5 August) to correspond with earlyand late-season nectar availability. Samples were stored at room temperature and analyzed in lab at the University of Wisconsin-Eau Claire during September–November. We used a Spectronic 20D+ spectrophotometer to estimate the absorbance of total carbohydrate in solution as done by McKenna and Thomson (1988). We pooled individual nectar samples VOLUME 61, NUMBER 4 224 per plant and averaged the absorbance to estimate mean nectar production per flower. We used a Two-way ANOVA with repeated measurements to test for significant differences (p<0.05) between treatments, periods and the interaction of treatment*period. Proportional data received a squareroot arcsin transformation.