Citation
Effective pollination in a rare Colorado endemic, penstemon degeneri

Material Information

Title:
Effective pollination in a rare Colorado endemic, penstemon degeneri
Creator:
English, Carol Theresa
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
x, 55leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Penstemons -- Pollination -- Colorado ( lcsh )
Degener's beardtongue -- Pollination -- Colorado ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (M.S.)--University of Colorado Denver, 2008. Biology
Bibliography:
Includes bibliographical references (leaves 53-55 ).
General Note:
Department of Integrative Biology
Statement of Responsibility:
by Carol Theresa English.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
318449830 ( OCLC )
ocn318449830

Downloads

This item has the following downloads:


Full Text

PAGE 1

EFFECTIVE POLLINATION IN A RARE COLORADO ENDEMIC, PENSTEMON DEGENERI By Carol Theresa English B.S., University of California, Santa Cruz 1985 A thesis submitted to the University of Colorado Denver in partial fulfillment of the requirements for the degree of Masters of Science Biology 2008 H

PAGE 2

This thesis for the Master of Science Degree by Carol Theresa English has been approved by Leo P. Bruederle Michele Engel Date

PAGE 3

English, Carol Theresa (M.S. Biology) Effective Pollinators in a rare Colorado endemic P enst e mon degen e ri Thesis directed by Associate Professor Leo P Bruederle ABSTRACT Penst e mon d e generi Crosswhite (Plantaginaceae) is a rare Colorado endemic about which little is known ; it is considered threatened and globally imperiled with few federal or state level conservation regulations in place Because P. deg eneri depends on local pollinators for successful reproduction determining effective pollinators is an essential step in conservation management for P. degeneri This research determines visitors and effective pollinators for Penstemon degeneri Effective pollination was assessed by correlating visitor frequency with visitor behavior, amount and placement of P d e g e n e ri pollen on the visitor s body and the visitor s fidelity toward P d e g e n e ri In 2007, three populations of P. degen e ri were studied that span an elevation range of7, 000-9200 in Fremont County Colorado. As with other rare endemic P en s temon species in the west P. degeneri attracts a diversity of visitors These visitors represent four orders and ten families including butterflies (Lepidoptera) flies (Diptera), bees and wasps (Hymenoptera) and one hummingbird (Trochillidae) based upon observations conducted during the 2007 field season. Only a few of the many visitors including Osmia (Megachilidae) and Bombu s (Apidae) were verified as effective pollinators. Unexpectedly the oligolectic pollen wasp, Ps e udoma s ari s v espoid e s was nearly absent in 2007 although preliminary results indicate that this

PAGE 4

preliminary results indicate that this pollen wasp was a frequent visitor in 2008. The most effective pollinators on P. degeneri were Osmia bees yet these bees did not occur with consistency at all three study sites. However as with the pollen wasp preliminary results indicate Osmia bees were frequent visitors at each ofthe sites in 2008. A high percentage of these individuals carried copious amounts of P. degeneri pollen on their bodies, and many had 100% fidelity toward P. degeneri pollen in 2007. Apids including Bombus centra/is and B huntii were also effective pollinators although less effective than Osmia bees This abstract accurately represents the content of the candidate s thesis. I recommend its publication. Leo P. Bruederle

PAGE 5

DEDICATION I wish to dedicate this thesis to my husband, Dave Elin, who without his loving support patience, and many long hours of assistance with photography and fieldwork, I would not have finished this degree :

PAGE 6

ACKNOWLEDGEMENTS I give thanks to my graduate advisor Dr. Leo Bruederle who has guided me with great patience and valuable knowledge to reach this goal. I would also like to thank Dr Vince Tepedino who answered numerous questions regarding pollination biology. Without his help this research would not have happened. Additionally I want to give out great thanks to both Virginia Scott and Terry Griswold who keyed out all the bee species, and offered field materials in support of collecting and storilig the insects.

PAGE 7

TABL E OF CONTENTS Figures ..... ................... ...................... .... ...... . ........... ..... .. ... ............ ... ........ ..... ......... v i i i Tables .. .............. ........... .......... ............... ........ .... . ............. ........ ............... . ... .. ....... ...... x Chapter 1 0 Int r oduction ........... ...... ......... ........................... ............ ........................ ............. 1 1 Genus P en ste m o n ..................... ... ........................ .... ....... ... ......... ..... ... ............ 1 1 2 P en ste m o n d ege n eri ......... ... ................... ............. ......... .. ...... ............ ...... ... . ..... 4 1 3 R e search Objecti ves .... ....... .... ............. ........ ........... .... ................. ..... . ........ 6 2.0 Materials and Method s .............. ..... ......... .... ........... .......................................... 8 2 1 Research Si tes .... ....... ............ ................................... ... ... .. ... ....... ......... .......... 8 2.2 Phenology ............ ......... ................... ........ ......... ................. ..... .... . ........ . ..... ..... 8 2.3 Effecti v e Pollinator s ........ ........................................... ............... .......... ....... ... ... 9 2.4 Visitor s ... ...... ...... ..... . ... .. ... ......................... . ................ .. .. ..... .... ..................... lO 2 5 Vi s itor Beha v ior . .... ............... ........ ... ...... ......... ........... ... ............. .... . ............ ll 2.6 Body Poll e n ... ............................ .... .... ............ . ............................ ............... 1 2 3 0 Res ults ................................. .................... ........................ ... ................... ..... ... 15 3 1 Phenolo gy ............ ....... .......... ..... ....... ........ ........... .. ........... ..... ................ ..... .... 15 3 2 Vi s itor s . ....... ....... .... ........................... .... ....... ... ........ ........ ..... ........ ........ ........ 16 3.3 Vi s itor Beha v ior .............. . . .......................................... ......... . .......... ............ ? 3 4 BodyPollen .. .......... ...................... ........ ........ ............... ....... ......... ....... ......... 19 4 0 Di s cussion ...................... ................... .... ........... .......... .......................... ....... ... 2 3 4 1 Vi s itor Guild ...................... .......... ..... ........ ............ ... ...... .......... ............ ..... ..... 23 4 2 Effecti v e Pollinator s .... ..... ............... ........... .... ...... ........ ..... ............... ......... . 25 4.3 Future Studie s ....................... .......... .......... ...... ...... .. ... .... .......... ....... ... ....... ..... 3 0 Appendix A Fuch s in G e la tin Recip e a nd Proc e dur e ........ ............. . ........... ....... ............. ... 51 B. Glyce rin J elly Slides Pro ce dur e .. ... ............ ........ ........................ .................... 5 2 References ......... . ................... ....... .... .................... .... ...................... ............. . .... .... 5 3 Vll

PAGE 8

LIST OF FIGURES 1. Locations for three Penstemon degeneri Crosswhite research ........................ 32 sites: Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), in Fremont County, Colorado. 2. Floral morphology and receptivity in Penstemon degeneri Crosswhite: ........ 34 (a) straight style, (b) slightly curved style still situated within groove on upper inside surface of flower, and (c) fully curved style almost touching staminode hairs. 3. Pollen grains viewed at 400X from: (a) Penstemon virens Pennel ex Rydb.,. 35 (b) P. degeneri Crosswhite, and (c) P. virgatus A. Gray. 4. Visitors captured on Penstemon degeneri Crosswhite at three ....................... 38 Colorado sites in 2007; organized by family or order. 5. Bees captured on Penstemon degeneri Crosswhite at three ............................ 39 Colorado sites in 2007; organized by family or order, with Bombus centra/is and B. huntii (Apidae) grouped to indicate their relative abundance. 6. Bee behavior on Penstemon degeneri Crosswhite flowers: (a) halictid .... ..... .41 (Halictidae) bee entering a flower inverted, remaining up to 60 seconds, and actively collecting pollen and possibly nectar; (b) Osmia (Megachilidae) bee about to enter a flower; (c) Bombus (Apidae) bee with large scopa (pollen) sac collecting nectar 7. Progression over 30-45 minutes depicting a female worker bee, Born bus ..... .42 huntii, on Penstemon degeneri Crosswhite collecting and accumulating pollen and nectar in her scopae: (a) early (5-10 mins) during the visit, with relatively small amounts of pollen in the scopal sac; (b) following approximately 15 mins. of pollen and nectar collecting, with larger scopal sac; (c) late (30-45 mins.), with very large scopal sacs making it difficult for the bee to hang onto the flower. 8. Percent megachilid (Megachilidae) and apid (Apidae) bees that carried ..... . .43 greater than 80% Penstemon degeneri Crosswhite pollen in their scopae at three Colorado sites in 2007. 9. Percent Penstemon degeneri Crosswhite scopal pollen found on 31 ............. .44 species of female worker bees captured at three Colorado sites in 2007. Vlll

PAGE 9

10 Percent apid (Apidae) bees caught at three Colorado sites in 2007 with ....... .47 moderate to copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA =ventral abdomen. Includes number of individuals captured (N) and average size. 11. Percent apid (Apidae) bees caught at three Colorado sites in 2007 with ....... .48 copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA=ventral abdomen Includes number of individuals captured (N) and average size. 12. Percent megachilid (Megachilidae) bees caught at three Colorado sites ....... .49 in 2007 with moderate to copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA=ventral abdomen. Includes number of individuals captured (N) and average size 13. Percent megachilid (Megachilidae) bees caught at three Colorado sites ........ 50 in 2007 with copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA =ventral abdomen Includes number of individuals captured (N) and average s1ze. lX

PAGE 10

LIST OF TABLES 1. Penstemon degeneri Crosswhite phenology at three Colorado sites in ........... 33 2007, including Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), Fremont County. Elevation is specified in meters 2. All visitors captured on Penstemon degeneri Crosswhite in 2007 at the ........ 36 Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), Colorado sites in 2007 Numbers with a slash indicate females / males. All visitors are identified to the lowest possible taxonomic level. 3. Behavior of visitors of Penstemon degeneri Crosswhite at three Colorado ... .40 Sites in 2007 4. Hymenoptera species collected from Penstemon degeneri Crosswhite ......... .45 at three Colorado sites in 2007, including: number collected at the Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP) sites, and in total (NC); number collected with pollen loads (NP); and mean percentage (SD) of P. degeneri pollen in the pollen loads(% PD). Includes total number of species captured at each site and time spent capturing. X

PAGE 11

1.0 Introduction 1.1 Genus Penstemon Penstemon Mitchell (Plantaginaceae) is the largest endemic genus in North America (ca. 270 spp.), with the majority of species located west ofthe Continental Divide (Wolfe 2006). This diversity is believed to represent a rare example of continental radiation characterized by high neoendemism and rapid speciation (Wolfe 2006, Lesica et al. 2006). Continental evolutionary radiations have been documented worldwide including: southwestern Australia, South Africa, Mediterranean Basin and western North America (Lesica et al. 2006). Western North America genera believed to exemplify recent radiations include Penstemon (Plantaginaceae), Erigeron L., (Asteraceae), Astragalus L. (Fabaceae), and Eriogonum Mich. (Polygonaceae). These genera, which are thought to have evolved relatively recently, are widespread and have large numbers of narrow endemics (Lesica et al 2006). Penstemon is hypothesized to have originated in the central Rocky Mountains in the late Tertiary with subsequent diversification due to evolutionary adaptations, including pollinator and niche specialization (Wolfe et al. 2006) P enstemon is a member ofPiantaginaceae tribe Cheloneae. Species within this tribe are characterized as having a cyrnose inflorescence and a sterile starninode (Wolfe et al. 2006). The genus P enstemon is divided into six 1

PAGE 12

subgenera including Cryptostemon and Dissecti, both of which are monotypic, Penstemon (ca.l82 spp.), Habroanthus (ca. 50 spp.), Saccanthara (ca. 28 spp ), and Dasanth e ra (ca. 9 spp ); these subgenera are further divided into sections and subsections. P en s t e mon is distributed coast to coast from Guatemala to Alaska, with over 100 species restricted in distribution to a single state; for example, approximately nine P en s t e mon species are endemic to Colorado (Wolfe eta!. 2006). Most Penst e mon species attract many different pollinators including bees wasps moths, and flies while the remainder (ca. 40 spp.) are adapted to attract hummingbirds specifically (Wolfe eta!. 2006). These two syndromes differ with respect to: anther dehiscence patterns corolla color and shape, and starninode morphology For example, differences in anther dehiscence have been correlated with pollination efficiency (Castellanos eta!. 2006). Bird pollinated flowers release pollen in large quantities over a short amount of time, whereas bee pollinated flowers release pollen in smaller quantities over longer periods of time (Castellanos eta!. 2006). Furthermore bird-pollinated P ens l e mons tend to have red narrowly pendant shaped corollas with reduced, glabrous starninodia. In contrast insect-pollinated P en s t e mons tend to have blue pink, or purple corollas, with wide ventral lobes that are used as landing pads and variously shaped pubescent staminodia that facilitate pollination (Wolfe eta!. 2006) The staminode which is located on the corolla throat directly under the anthers and 2

PAGE 13

stigma, is larger than the fertile stamens and has bristles on the adaxial distal surface (Dieringer et al. 2002). Recent studies on P. digitalis Nutt. have demonstrated that staminode bristles and pollinator size correlate with pollination efficiency (Dieringer et al. 2002). When staminodia are removed, pollen deposition efficiency has been shown to be reduced for small bees in comparison to larger bees (Dieringer et al. 2002) The staminode acts as a step stool" and allows the dorsal side of small bees to contact the stigma more consistently and efficiently as the bees enter the corolla orifice (Dieringer et al. 2002) The staminodia in other P e nstemon species such asP. palmeri Gray with their larger corollas and thus larger pollinators act as levers Heavy bees activate the lever and cause the style to drop onto the dorsal side of their body (Dieringer et al. 2002). Recent pollination biology studies on congeners (including P. p enlandii Weber, P d e bilis O'Kane and Anderson P. glaber Pursh, P. ha y d enii Gray P harringtonii Wats., and P strictus Benth) have shown a wide variety of bees, wasps, flies butterflies and beetles visiting flowers ; effective pollinators for these Penstemon species include bees in the Megachilidae (e.g., Osmia) and Apidae (e g., Bombu s). Ps e udomasari s v espoides a wasp in the Masaridae exhibits a specialized preference ( oligolecty) for Penst e mon pollen (Beaty et al. 2004). A pollination study performed on P e nst e mon serie s Gracil es demonstrated that small bees in the genus O s mia are the primary pollinators for most species in 3

PAGE 14

this alliance. Penstemon hirsutus (L.) Willd., P. tenuiflorus Penn. and P. oklC!Jhomensis Penn. are exceptions and are pollinated principally by small bees in the genera Hoplitis and C e ratina (Crosswhlte and Crosswhite 1966). This same study found evidence supporting that entire subgenera of O s mia are mostly dependent on P e nst e mon pollen and nectar for nutrition. Furthermore, a small group of species within Osmia subgenus Nothosmia are oligolectic for Pen s lemon series Gracil e s alliance Graciles. More recent studies have also shown that Osmia bees are the most effective pollinators for P p e ndlandii, P. ha y denii P harringtonii, and P. lemhi e nsis specifically (Tepedino et al. 1999 2006). 1.2 Penstemon degeneri Penst e mon degeneri Crosswhite, commonly referred to as Degener's Beardtongue is a rare species endemic to south central Colorado (Crosswhite 1967). The Colorado Natural Heritage Program considers P. d e g e n e ri to be a species of special concern, and ranks it S2 (vulnerable to extirpation endangered or threatened in the state) and G2 (globally imperiled due to its regional rare and endemic status) Due to the fact that very little is known about the biology of this species, there are few mechanisms at the federal or state level to regulate its conservation. These regulations may not be adequate to conserve P d e g e n e ri over the long term. P e nst e mon d e g e neri belongs to subgenus Habroanthus section Habroanthus subsection Humil e s series Gra c ilis alliance Oliganthi (Beaty et. al 2004) The 4

PAGE 15

five species within alliance Oliganthi occur in mountainous areas of three western states These include: P d e g e n e ri and P. griffinii Nels. in southern Colorado P injlatus Crosswhite and P. p se udoparvus Crosswhite in northern and central New Mexico, and P oliganthus Cro s swhite in eastern Arizona (Fig. 1) (Cro s swhite 1967) Penst e mon d e g e neri has pale lavender, zygomorphic corollas that are 14-19 mm long. The corolla throat is obscurely ridged with sparse yellow non glandular hairs; the mouth has a few white hairs; and the staminode is sparsely yellow bearded. The four free stamens are didynamous terminating in glabrous anthers (Crosswhite 1965). Penst e mon degeneri is endemic to Fremont, Teller and Custer Counties, Colorado where it occupies habitats varying from pinyon-juniper woodlands (1830 to 2042 m elevation) to montane meadows and ponderosa pine parklands (2073 to 2743 m) (Beaty et al. 2004). In pinyon-juniper wood l ands associated species include open canopies of Pinus edulis Engleman (Pinaceae) with a mixed understory of shrubs, including Juniperus c ommuni s (Smith) Celakovsky (Cupressaceae) grasses and forbs. In montane meadows and ponderosa parklands, associated species include Danthonia parry i Scribner (Poaceae) Pinus ponderosa Douglas (Pinaceae) Populu s tremuloides Mich (Salicaceae) and P se udot s uga m en s i es ii (Mirbel ) Franco (Pinaceae). Other associated species include A c hill e a lanulo s a Nutt. (Asteraceae). A nt e nnaria spp Gaertner 5

PAGE 16

(Asteraceae) Art e misiafrigida Willd (Asteraceae), Car ex L. spp. (Cyperaceae) Lupinus L. spp. (Fabaceae) Muh/ e nb ergia montana Schreber (Poaceae) The rmop s i s montana Nutt. (Fabaceae) and P en s t e mon v ir en s Pennell (Beaty et al. 2004) The Colorado Natural Heritage Program prepared status reports for P d e g e n e ri in 1981, based on five known populations and again in 1991 based on nine known populations (Beatty et al. 2004). In 1998 several more populations were found during a floristic survey bringing the total number of known populations to 14. Since that time ten of these populations have been re affirmed. Population size varies greatly from 25 to thousands of individuals (Beatty et al2004). Population abundance trends have not been documented yet s ome populations appear to have declined severely with some evidence of herbivory (Beaty et al. 2004). A short term insect visitation study completed in 2004 verified that a variety of insects including flies, bees and wasps visit P deg e n e ri (Spackman 2004) yet it remains unknown which insects are most effective pollinators 1.3 Research Objectives The objectives of this research are to develop a comprehensi v e list o f floral visitors for P d e g e neri and subsequently determine which visitors are most effective pollinators across the range in elevation for this species. I hypothesize that a variety of insects will v isit P d e g e n er i including butterflies (Lepidoptera) 6

PAGE 17

flies (Diptera) and bees and wasps (Hymenoptera) Furthermore I hypothesize that small bees within genus O s mia, a variety of bumblebee species within genus Bombus and the oligolectic wasp P se udoma s ari s vespoide s are the most effective pollinators for P deg e n e ri (Tepedino et al. 1999 2006, Wilson et al. 2004) Additionally effective pollinator species have been shown to vary between years and in different geographic locations (Tepedino et al. 1999 2006). This variation is suspected to be due to the fact that pollination effectiveness depends on the coordinated timing of floral reproductive maturity and insect activity (Tepedino et al. 1999 2006). Therefore I expect effective pollinators for P d e g e n e ri to vary both spatially and temporally Determining effective pollinators for a rare endemic and entomophilous plant is important since pollinators are essential for reproductive success Furthermore because previous studies have revealed that a co-evolutionary relationship exists between certain hymenopterans and P e nst e mon the former also depend on this species for successful reproduction (Tepedino et al. 1999 2006 Wilson et al. 2004) Therefore pollinators and their habitat must be conserved along with the plant. Preserving these effective pollinators also preserves a breeding system that maintains the appropriate and healthy levels of genetic diversity for P d e g e neri (Keams 1993) Finally, discovering the effective pollinators for another P e n s temon species may help scientists better understand the evolutionary mechanisms leading to speciation in genus P en s t e mon 7

PAGE 18

2.0 Materials and Methods 2.1 Research Sites Pollinator studies were carried out at three research sites in Fremont County Colorado: Sunset City Phantom Canyon and Locke Park (Figure 1). These sites represent more or less the ecological range of P d e g e n e ri Sunset City Colorado is pinyon-juniper woodland at 2,300 m elevation and occurs on land administered by the Bureau of Land Management (BLM) It is located approximately 11 miles east of Canyon City and six miles south of the Royal Gorge Bridge that spans the Arkansas River at N38 24.780 W105 24.894. Phantom Canyon is a montane meadow at 2 700 m elevation and is located on BLM land approximately 20 miles north-east of Canon City and the Arkansas River at N38 36.821 W105 08 089 Locke Park is a ponderosa parkland at 3 070 m elevation and is located in the San Isabel National Forest approximately 20 miles south of Canyon City and the Arkansas River at N3815.867 W105 15.091. All field and lab-work was completed with the help of one to several student field assistants in 2007 and 2008 2.2 Phenology Between May and September 2007 the following phenological stages were recorded at each of the aforementioned P degen e ri sites: rosette emergence; floral budding ; early middle and late anthesis ; and mature fruit set. Anther dehiscence 8

PAGE 19

and style movement (curving down) was observed while performing the pollination observation experiments. 2.3 Effective Pollinators Relati v e pollinator effecti v eness for all insect v isitors was determined following the techniques ofTepedino et al. (1999 2006). In so doing I asked four questions regarding P d e g e n e ri visitors Which insects are the most frequent visitors to P. d e g e n e ri? Do frequent visitors carry relatively large amounts of P. deg e n e ri pollen on their bodies? Does the behavior of the visitors promote the transfer of pollen from their bodies to the stigma? Which female bees have high fidelity toward P. d e gen e ri pollen? To answer these questions I observed the behavior of all insects visiting P d e g e neri. I quantified P d e g e n e ri body pollen on all insect visitors that I captured. Subsequently I correlated the body pollen results with the behavior observation results For example if a particular bee carried large amounts of P d e g e neri pollen on its dorsal thorax I correlated this result with the part of the body that had been observed contacting the stigma Finally I quantified scopa pollen to determine a female bee s degree of fidelity (faithfulness) toward P d e g e n e ri pollen By integrating and interpreting these results I subsequently determined the relative effectiveness for all P d e g e n e ri insect visitors in 2007. For example highly effective pollinators frequently v isited P d e g e n e ri flowers behaved in such a way to promote pollen transfer to the stigma carried copious 9

PAGE 20

amounts of P d e g e n e ri pollen on their bodies and carried a high percentage ( > 80 % ) of P d e g e n e ri pollen in their scopae. 2.4 Visitors In order to quantify the relati v e v isitation frequency and di v ersity of all visitors toP. d ege n e ri insects were captured by passing slowly through a P d e g e n e ri population over a period of two hours and capturin_g any insects seen on or inside P d e g e n e ri flowers. Insects were caught by me or a field assistant using a Bioquip insect net with an 18" handle and a 12" ring, and were then placed in a killing jar. Insects were subsequently pinned using stainless steel Bohemia Insect Pins labeled and stored in a Styrofoam insect box. Some insect species including butterflies and bombyliid flies, were excluded from frequent capture, as they have long proboscises and can obtain nectar without contacting the reproductive organs ; field observations confirmed v ery little pollen transfer occurred by these insects during v isits to P d ege n e ri At Sunset City I captured insects over seven days between June 18 and July 5 for a total of21 hours Nine of the catch hours were in the morning and 12 were in the afternoon. At Phantom Canyon I captured insects between July 1 and July 17 for a total of three da y s and 12 hours including two morning hours and ten afternoon hours At Locke Park, insects were captured over three days between July 6 and July 18 for a total of 12 hours with six morning hours and six afternoon hours 10

PAGE 21

Insects were pinned, labeled using acid-free 25# we ight paper and deposited at the University of Colorado Museum ofNatural History. Virginia Scott, Collections Manager for the Entomology Section, completed most of the identifications A few Osmia and halictid bees were identified to species by Dr. Terry Griswold, Research Entomologist with the USDA-ARS Bee Biology and Systematics Laboratory in Logan, Utah. The data on visitors to P deg e neri were summarized by taxon In most cases, visitors were identified to species although some insects were only identified to genus 2.5 Visitor Behavior Insect behavior on P. degeneri, both on plants and inside flowers, as well as their movement between plants was recorded These observations help to discern the extent to which different insect species may be transferring and depositing pollen toP. degeneri stigmas. As mentioned previously, correlating pollinator behavior and body pollen deposition aids in determining effective pollinators for P. degeneri. For example, a frequent visitor may or may not behave in such a way that promotes the accumulation and transfer of P degeneri pollen to other P. degeneri flowers and plants. Pollinator observations at the Sunset City site occurred June 18 to July 5 for a total of six days and 21.25 hours, with 15.5 observation hours in the morning and 13.5 observation hours in the afternoon. At Phantom Canyon, observations 11

PAGE 22

occurred on July 1, 10, and 17 for a total of eight afternoon hours. At Locke Park observations occurred July 7-19 for a total of2.5 hours in an afternoon. With the help of a field assistant, I initially noted the number of plants and flowers being observed, and described the habitat and weather conditions. As an insect approached a P. degeneri flower, the insect was identified to the lowest taxonomic level possible using photographic keys. Additionally, the insect was described noting various physical features, including approximate size, color, and hairiness. The number of plants and the number of flowers visited on each plant were noted. Additionally, I observed insects as they entered P. degeneri flowers and noted whether the insect collected pollen and/or nectar, if the insect deposited pollen on the flower's stigma, and how the pollen was deposited (by which body section). For each taxon (described above), means were calculated for number of plants visited, number of flowers per plant, and length of visit. 2.6 Body Pollen In order to accurately identify P. degeneri pollen on bee bodies I examined pollen from 24 other species that were observed flowering at the three sites. Anthers from these species were collected, preserved in 70% ethanol, and subsequently dried, during which they dehisced. Basic fuchsin gelatin was prepared following Beattie (1971) lab procedures in preparation for pollen staining (Appendix A). The gelatin was cut into small squares and, using forceps, dabbed onto the pollen samples following Erdtnab (1960, 1969) (Appendix B). 12

PAGE 23

Each pollen reference slide was labeled, examined at 400x using a compound microscope and photographed with a digital camera. Only pollen from other P e nstemon species appeared similar enough to cause difficulty with identification ; however it was possible to discriminate among the three species with P d e g eneri pollen intermediate in size between P. v ir en s and P vi r gatus. (Figure 3a-c) Furthermore peak anthesis varied among the three species blooming at my research sites. Insect body pollen was analyzed following Tepedino et al. (1999 2006). All bees wasps and flies that had been captured were placed in an insect relaxing chamber containing chlorocresol to relax connective tissue Subsequently the length of the insects was measured in millimeters using a Starrett 120Am-150 metric dial caliper. Each insect s head thorax and abdomen were measured separately and these three measurements were summed to calculate the total insect body length. Body pollen was removed from insects by dabbing each of five body parts (including the head ventral and dorsal sides of the thorax and ventral and dorsal abdomen) individually using fuchsin gelatin. Due to copious amounts of pollen it was often necessary to use the forceps or a probe to gently scrape loose the pollen from insect body parts especially on the face and caudal thorax before dabbing the gelatin on the area. After pollen was removed from each section the forceps and probes were cleaned by flaming before gathering pollen from the next insect. The gelatin from each of the fiv e sections was placed 13

PAGE 24

on a glass microscope slide, labeled, and slowly melted using a Bunsen burner. The melted gelatin was covered with a cover-slip allowed to cool, labeled, and stored at 4 C for future reference. Pollen was removed from only a few flies and wasps since very little pollen could be seen on their bodies using a dissecting microscope. Pollen collected from each of the five body sections was identified as either P. d e gen e ri or that of another taxon using the aforementioned reference photographs. For each insect the total amount of P.de g e n e ri pollen on each body section was then nominally categorized as 0 (0-50 pollen grains) 1 (50-150 pollen grains), 2 (150-500) or 3 ( > 500) and recorded. For each insect visitor that I examined the percentage of individuals was calculated with respect to the relative proportion of body pollen (e g. copious) on each body section (e .g., head) Using forceps scopal pollen (when present) was removed from female worker bees and similarly prepared for microscopic examination. The percentage (mean and standard deviation) of P. d e g e n e ri pollen in bee scopae was obtained by averaging three counts of 500 pollen grains For each count the pollen was identified as P d e g e neri or when possible to one of the other species of flowering plants observed at the site. A bee that carried > 80 % P d e g e n e ri scopal pollen was considered to exhibit high fidelity. 14

PAGE 25

3.0 Results 3.1 Phenology In 2007, young plants were fir s t observed on May 7, 17 and 22 at Sunset City, Phantom Canyon, and Locke Park sites respectively. Flower buds appeared approximately two weeks later at Sunset City and approximately four weeks later at both Phantom Canyon and Locke Park (Table 1). Flowering defined as early middle, and late anthesis occurred over a period of two weeks from June 16-30 at Sunset City from July 1-17 at Phantom Canyon, and from July 6-20 at Locke Park (Table 1 ) Fruit development including the time of fruit set (late an thesis) to the time at which fruit matured, took approximately five weeks from June 30 to August 7 at Sunset City, and six weeks from Jul y 17 to August 30, and July 20 to September 4 at Phantom Canyon and Locke Park, respectively (Table). During the first day of anthesis, the longer pair of anthers dehisced first while the shorter anther pair dehisced subsequently over a period of one to four hours Once anthers dehisced there appeared to be very little pollen remaining. As the anthers dehisced the style remained straight and hidden within a groove along the upper inside surface of the corolla tube (Figure 2a) Within 24 hours the style began to curve downward toward the inside, lower surface of the flower toward the starninodiurn (Figure 2b ). Within 48 to 72 hours the sty le was completely bent, with the stigma nearly touching the dehisced anthers and starninodiurn ( Figure 2c). When deprived of pollinators approximately 38% of P. d egeneri flowers set 15

PAGE 26

fruit. These observations were the result of a breeding system experiment performed in 2007 (Carol English pers. obs.). 3.2 Visitors A total of 44 species comprised the guild of visitors for P. degeneri including one hummingbird (Table 2, Figure 4). Visiting insects included bees representing the Megachilidae (megachilids), Apidae (apids), Halictidae (halictids), and Colletidae (colletids); flies representing the Syrphidae (syrphids) and Bombyliidae (bombyliids); wasps representing the Masarinae (masarines) and Eumeninae (eumenines); and butterflies including western swallowtails (Papilio, Papilionidae) and skippers (Hesperiidae ) Bees were the most frequent visitors to P. degeneri and made up approximately 73% of all taxa visiting these flowers. Flies made up 23% of all visitors, while wasps and butterflies made up the remaining 4% (Figure 5). Of the bee taxa, 76% were apids, 15% were megachilids, and 7% were halictids. Two Bombus (bombid) species, Bombus centra/is and B. huntii, represented approximately 70% of the apids and 53% of all bees captured on P. degeneri (Figure 5). Variability among sites existed, both for insect abundance and diversity. For example, even though B centra/is was the most abundant visitor captured on P. degeneri, this species was not found at Sunset City, and nearly twice as many individuals were captured at Phantom Canyon as compared to Locke Park. 16

PAGE 27

Bombus huntii, the second most common visitor to P degeneri was found in low numbers at Sunset City and Phantom Canyon, yet in relatively high numbers at Locke Park (Table 2). Halictid diversity and individual abundance decreased with elevation gain while megachilid and apid div ersity and individual abundance increased with increasing elevation gain (Table 2). Bombyliids were observed in approximately equal numbers at all three sites, yet syrphids only occurred at Sunset City (Table 2). In general, few wasp species were seen on P. degeneri flowers with masarines occurring in low numbers only at the Sunset City site and eumenines occurring only at Phantom Canyon and Locke Park (Table 2). 3.3 Visitor Behavior A total of 31 hours was spent observing visitor behavior at the three sites, including 21.5 7.0 and 2.5 hours at Sunset City Phantom Canyon and Locke Park respectively. Flower visitors were identified to genus, family or order. Apids were observed visiting P. d e generi flowers 174 times. On average during an observation period apids visited 160 plants and 5.3 flowers per plant, with sustained visits of22. 6 minutes (Table 3). Megachilids were observed visiting P degeneri flowers seven times. On average, megachilids visited three plants and 2.5 flowers per plant, with sustained visits of 43 seconds (Table 3). Halictids were observed visiting P degeneri flowers 19 times. On average halictids visited two plants and two flowers per plant with sustained visits of 64 seconds (Table 3). The behavior of dipterans was similar to megachilids and halictids but visit 17

PAGE 28

length was shorter a v eraging 27 seconds per plant (Table 3). Butterflies were observed six times averaging nine plants and two flowers per plant with sustained vi s its o f two minutes (Table 3) Only one hummingbird was observed v i s iting flowers; it returned for two 15 minute v isits a v eraging ten plants per visit and two flowers per plant (Table 3) As mentioned pre v iously insect behavior also consisted of observing the specific behavior of insects inside and near P d e g e n e ri flowers As bombids entered the flowers their heads and caudal thorax contacted both the stamens and stigma which was correlated with a noticeable change in buzzing. These bees were actively gathering nectar yet passively gathering pollen. After exiting flowers, they spent many seconds cleaning pollen off their face near the antennae and eyes They vi s ited and collected nectar from flowers in an area for up to 35 minutes at both Phantom Canyon and Locke Park. Scopal sacs could be observed increasing in size until finally the bee would fall toward the ground while attempting to land on a flower (Figure 7a-c). The bees would than fly quickly and directly in a straight path toward the southeast. Another bombid would arrive shortly and begin to work the flowers seemingly avoiding previously v isited flowers. At Phantom Canyon and Locke Park, numerous G e ranium caes pito sum and Oxytropi s lamb e rtii plants were blooming along with the P d ege n er i Many bombids were observed landing briefly on G. c a es pito s um flowers then quickly 18

PAGE 29

flying back to P. degeneri flowers. Bombids were observed foraging on both 0. lambertii and P degeneri flowers, yet usually on one or the other species without switching between the two species in a single foraging trip. Halictids would flip upside down after entering P d e g e n e ri flowers and stay inside the flowers for up to a minute at a time while collecting pollen and possibly nectar The ventral thorax and abdomen would contact the stigma (Figure 6a) Megachilids entered a flower with the dorsal abdomen contacting the stigma. After exiting the flower they sometimes entered 1-2 additional flowers on the same plants and then quickly flew off or landed on the ground to groom their bodies (Figure 6b ) 3.4 Bod y Poll e n The pollen wasp P se udoma s ari s ve spoid es was observed and captured very few times and carried little body pollen Additionally zero to very few pollen grains were detected on insects including eumenines collectids and dipterans; therefore only apids megachilids and halictids were examined in detail for P d e gen e ri body pollen. For most bombids a majority of the pollen was deposited on the head with decreasing amounts on the thorax and abdomen respectively; that was not the case for B ce ntra/is in which most of the pollen was deposited on the dorsal thorax (Figure 10). Approximately 73% of B c e ntra/is bees, the most abundant species captured on P. d e g e n e ri carried moderate to copious amounts of pollen on their heads while 78 % carried moderate to copious amounts 19

PAGE 30

of pollen on their dorsal thorax (Figure 1 0). Forty-six percent maintained copious amounts of P.degeneri pollen on their dorsal thorax, while only 28% maintained copious amounts of pollen on their heads (Figure11). Fewer than 55% of B. huntii bees, the next most abundant species, carried moderate to copious amounts of pollen on their heads, while 27% individuals carried moderate to copious amounts of pollen on their dorsal thorax (Figure 10). Few of these individuals maintained copious amounts of pollen on their heads and dorsal thorax, with none in the moderate category (Figures 10, 11). Fifty-three and 36% of the 12 Bombus appositus individuals captured, carried moderate to copious amounts of P. degeneri pollen on their heads and dorsal thorax, respectively, with very few individuals carrying copious amounts of pollen on their bodies (Figures 10, 11). Other apids, including B. nevadensis nevadensis, Anthophora Montana, and A. terminalis, carried very little P. degeneri pollen on their bodies (Figures 10, 11 ). Often 100% of the megachilids the second most abundant family of bees captured on P. degeneri, carried moderate to copious amounts of pollen on their heads, dorsal thorax, and ventral abdomens, in contrast to Bombus (Figures 12, 13). Many of these individuals, often 100%, also carried copious amounts of P. degeneri pollen on their bodies when the moderate category was eliminated (Figures12, 13). Approximately 60-70% of Osmia brevis, the most abundant megachilid found on P. degeneri, carried moderate to copious P. degeneri pollen on their dorsal thorax and heads and ventral abdomens (Figure 12). Although 20

PAGE 31

this sample is small, all other megachilids, including Osmia juxta, 0. buchephala, 0. proxima, 0. bruneri, 0. paradisica, and Hoplitus algentifrons, carried moderate to copious amounts of P. degeneri pollen on various body parts (Figure 12, 13). One halictid Lasioglossum sisymbrii, carried moderate to copious amounts of P. degeneri pollen on their bodies, at least in half of the individuals from which data were collected. Fifty-seven percent of all female worker bees that carried greater than 80% P. degeneri scopal pollen were megachilids. The remaining 43% were apids (Figure 8). Bombus centra/is, the most abundant bee species on P. degeneri, carried 3100% P. degeneri scopal pollen (mean= 82.2 28.0, N = 75) (Table 4). The second most abundant species on P. degeneri, B. huntii, carried 12-99% P. degeneri pollen (mean= 78.5% 29.3, N = 46) (Table 4, Figure 9). Forty-two Osmia bees, representing 11 species, made up the next most abundant group of bees captured on P. degeneri. These individuals had scopalloads that ranged from 8-100% P. degeneri pollen (mean= 86.3% 26.8, N = 30) (Table 4, Figure 9). The three most abundant Osmia species (O.brevis, 0. penstemonis, and 0. juxta) carried a high percentage of P. degeneri scopal pollen at 99.8% 0.60, 99.6% 0.58, and 82.6% 35.1, respectively (Table 4, Figure 9). Infrequently captured species, such as 0. albo/ateralis, 0. proxima, 0. bruneri, 0. densa, and 0. paradisica also carried high percentages of scopa pollen (Table 4, Figure 9). In general, less than expected numbers of halictid bees were captured on P. 21

PAGE 32

degeneri, with Lassioglossum sisymbrii representing the most abundant halictid captured; they carried 4-93% P. degeneri scopal pollen (mean= 67% 42.4, N = 4)) (Table 4, Figure 9). 22

PAGE 33

4.0 Discussion 4.1 Visitor Guild As expected this research conducted during the 2007 field season revealed that a large variety of insects visited P. d e g e n eri. These visitors represented four orders and ten families including butterflies (Lepidoptera) flies (Diptera) bees and wasps (Hymenoptera) and one hummingbird (Trochilidae) (Table 2). These results corroborate the findings of other pollination studies which have repeatedly shown that P e n s t e mon species with blue to purple corollas and wide ventral lobes are primarily visited and pollinated by a wide variety of insects (Beaty et al. 2004 Kimball2008 Spackman 2004 Tepedino et al. 1999 2006 Wilson et al. 2004). Furthermore effective pollinator studies by Tepedino et al. (1999 2006) on rare and endemic congeners have shown that only a few bees among the many insect visitors are effective pollinators. Again the re s ults of my research corroborate the findings ofTepedino et al (1999 2006) and suggest that even though P d e g e neri has a large assortment of insect visitors only a few bee taxa are effectively pollinating flowers. Tepedino eta! (1999 2006) found that the frequency of visitation was directly correlated with effective pollination which is also supported herein. Bombus and O s mia bees were the most frequent visitors to P d e g e n e ri in 2007, yet their frequency varied greatly among s ites (Figure 5 Table 2). For example Bombu s centra/i s was only captured at Phantom Canyon and Locke 23

PAGE 34

Park. Since B. centra/is is known to occur typically above 2133m elevation, its absence from lower elevation sites is not surprising (Keams 1993, V. Scott, pers. Comm.). And while I expected to encounter Osmia bees at similar frequencies across the elevation range, they were observed and captured predominantly at the Locke Park site in 2007. As there was a relatively high frequency of both blooming plants and large bee visitors, especially in genus Bombus (Table 2), I may have simply overlooked the smaller Osmia bees during my fieldwork at the other sites, which was conducted earlier in the season and before my skills had improved. Interestingly, preliminary results from 2008 reveal a relatively high frequency and diversity of Osmia bees on P. degeneri at all three study sites. This variation is not surprising and, to some extent, matches results from other studies on effective pollination (Tepedino et al. 1999, 2006). Whereas our study found Bombus to be the most frequent visitor followed by Osmia, Tepedino et al. (1999, 2006) found Osmia to be the most frequent visitors to Penstemon pendlandii and P. haydenii. However Bombus were not frequently captured by Tepedino et al. (1999, 2006) and, therefore, relative frequency of Bombus to Osmia was not calculated. Less frequently encountered insect visitors to P. degeneri in 2007 included dipterans, halictids, eumenines, and masarines. Nearly 25% of all visitors were bombyliids and, for this reason we considered these flies as possible effective pollinators (see section 4.2). Halictids were frequently observed and captured at 24

PAGE 35

the Sunset City site, yet they were infrequent at the other two sites. Since these bees are quite small when compared with the large frequently captured bombids, we either simply overlooked these tiny bees (as was suggested for megachilids) during capturing sessions, or they were missing at both the Phantom Canyon and Locke Park sites in 2007. Unexpectedly, and in contrast to most other Penstemon visitation studies, few P seudomasaris vespoides (masarines) were observed or captured on P. degeneri in 2007 (Table 2). Ps eudomasaris vespoides is a well known oligolectic pollen wasp that feeds on P enstemon pollen and is typically reported to be a frequent visitor to other, similar P enstemon species in western North America (Gess 1996 Beaty et al., 2004, Kimball 2008, Spackman 2004, Tepedino et al., 1999 2006, Wilson et al. 2004). Interestingly, masarines were some of the most frequent visitors at both the Sunset City and Phantom Canyon study sites in 2008 (C.T. English pers. obs.). It is unknown why this variation occurred. 4.2 Effective Pollinators Insect visitors to Penstemon degeneri were considered effective pollinators if they met four predetermined criteria: frequent visitation of P degeneri flowers, contact with stamens and stigmas during their visit, accumulation of moderate to copious amounts of P degeneri body pollen and high fidelity (faithfulness) toP. degeneri pollen. 25

PAGE 36

Many visitors did not fulfill enough of the aforementioned criteria to be considered effective pollinators. For example P. vespoides was not considered an effective pollinator in 2007 due to low visitation rates and small amounts of P. d e generi body pollen (Table 2). I am currently examining additional specimens captured in 2008 to determine their effectiveness in terms of body pollen. Even though bombyliids were frequent visitors, they were not considered effective pollinators in 2007 due to behavior and body pollen observations that indicated the flies were obtaining nectar without accumulating or transferring pollen (C.T English pers obs .). As with P vespoides I am examining bombyliids captured in 2008 for P. degeneri body pollen to determine their effectiveness Halictids were relatively frequent visitors only at the Sunset City site. Unfortunately many of the halictids captured were accidentally washed with ethanol thus eliminating body pollen from these specimens A few Lassioglos s um manitou e /lum were observed to carry enough P degeneri pollen to be considered effective pollinators, yet their numbers were too low to be considered effective pollinators in 2007 (Table 4). Finally a variety of apids, including bees in the genera Anthophora and Bombus (including, B. nevadensis nevadensis, B appositus, and B griseocollis) had high fidelity toward P d e g e neri yet carried very little body pollen (Tables 4 Figures 9, 10, 11). Therefore, masarines, bombyliids, halictids and the various apids previously mentioned were eliminated as effective pollinators in 2007 since they did not meet more than one and in some cases none, of the four criteria. 26

PAGE 37

Of the remaining Bombus and Osmia bees, several Osmia species were the most effective pollinators found on P. d e g e n eri during the 2007 field season. At Locke Park these bees were frequent visitors second only to Bombus centra/is (Table 2) Often 100 % of the O s mia examined carried copious amounts of P d e g e n e ri body pollen (Figure 13) Behavior observations verified that O s mia bees were contacting P d e gen e ri stigmas at the areas they accumulated P d e g e n e ri pollen on their bodies. Additionally Osmia had a high percentage of P d e g e neri pollen in their scopae Some species including 0. br ev is 0. p e n s temonis 0. proxima 0. bruneri and 0. paradisc ia, carried 100 % P. degeneri pollen in their scopae Therefore Osmia bees had the highest fidelity rates of any other bee studied in 2007 (Table 4, Figure 9). These findings support studies by Tepedino eta! (1999 2006) and by Crosswhite and Crosswhite (1966) suggesting that some O s mia bee species may depend entirely on P e n s temon pollen for nutrition Bombus centra/is and B. huntii are also considered effective pollinators since they met all four criteria However even though these two species were the most frequent visitors toP. degeneri in 2007, visitation among sites was highly variable (Table 2). Furthermore, visitation was extremely low in the subsequent 2008 field season suggesting that Bombu s may be both spatially and temporally unreliable Furthermore Bombus species did not meet the other effective pollinator criteria, including P d e g e n e ri body pollen accumulation and fidelity rates with as much strength as O s mia species. For example Bombus carried less P d e g e n e ri body 27

PAGE 38

pollen in comparison to Osmia. Whereas 80% of B. centra/is individuals carried only moderate to copious (150-500 pollen grains) amounts of P. degeneri body pollen, 100% of 0. buchephala 0. bruneri and 0. paradisica bees carried copious (greater than 500 pollen grains) amounts of P degeneri body pollen (Figures 10, 11, 12, 13). Additionally, Bombus species carried less P. degeneri pollen in their scopae and, therefore, had lower fidelity rates in comparison to Osmia species (Table 4, Figure 9). Except for the near absence of P. vespoides, these results meet our expectations, and are in accordance with other effective pollinator studies on rare endemic Penstemons (Tepedino et al. 1999, 2006). Among P. degeneri 's large visitor guild only a few Osmia and Bombus species be considered effective pollinators, although the degree of pollinator specialization is still unknown. These data seem to suggest that P. degeneri is an asymmetric generalist, in which the plants are pollinated by an assortment of specialist and generalist pollinator taxa (Kllnball 2008). In contrast, asymmetric specialists are plants that are pollinated by only a few generalist pollinator taxa (Ashworth 2004). In the face of habitat fragmentation and the possible loss of specialist pollinators, it has been shown that asymmetric generalists and asymmetric specialists have similar responses regarding reproductive success. In theory, even if the specialist pollinators are not present within a fragmented habitat, the generalists assume the major role of pollination (Ashworth 2004). Certainly this study has shown that Osmia species appear to be superior 28

PAGE 39

pollinators in contrast to Bombu s species. Yet, without further study we do not know if and how much reproductive success would be lost in the absence of Osmia pollinators. Additionally, due to spatiotemporal pollination differences described in this study (elevation differences), P degeneri 's habitat is fragmented and subject to variabi lity with respect to the occurrence of specialist pollinators. The relatively low frequency of Osmia at two P. degeneri sites in 2007, may have been compensated for by the frequent occurrence of Bombus generalists. It is noteworthy that P degeneri is able to self pollinate flowers deprived of pollinators were observed to set fruit in 2007 (Carol English pers. obs.). Furthermore, preliminary polle n/ovule ratios for P. degeneri obtained for the Sunset City, Phantom Canyon and Locke Park sites (Angela Schultz and Erin Lowther unpublished data) s uggest that P. degeneri is facultative xenogamous, with a mixed, yet predominantly outcrossing breeding system (Cruden 1977). Therefore, successful fruit set in P. degeneri likely results from a combination of outcrossing effected by bees and self-pollination in the absence of effective pollinators. 29

PAGE 40

4.3 Fut u re St u dies Since completing this study additional insect v isitors have been collected and are currently being examined for body pollen The s e data will contribute to the current research and offer a broader perspective regarding which visitors are consistently and effectively pollinating P d e g e n e ri In addition to focusing on bees the wasps butterflies and bombylliids will also be evaluated for pollinator effectiveness These results will be coupled with additional behavior observations to detennine how many flowers per inflorescence inforescences per plant, and plants per observation area are visited by different insect v isitors (Table 3) Along with identifying the effective pollinators for P. d e g e n e ri it is also desirable to study breeding system and reproducti v e allocation A reproductive assessment was conducted in 2008 while a breeding system study is planned for the 2009 research season. Objectives include estimating overall fruit and seed set during one reproductive season and determining the breeding system used by P d e g e neri for successful fruit set. These results in addition to the pollen/ovule ratios that are being calculated will help to clarify the specific breeding system used by P d e g e n e r i (Tepedino et al. 1999 2006 Cruden 1977 Kearn s 1993) These data will also facilitate conclusions regarding reproductive limitation. If P d e g e n e ri populations are being effectively pollinated by bees then we expect that 30

PAGE 41

there will not be a significant difference between the fruit set of open pollinated controls and hand-pollinated plants. 31

PAGE 42

F i g ur e 1: Locations for three Penstemon degeneri Crosswhite research sites : Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), in Fremont County, Colorado 32

PAGE 43

Table 1: Penstemon degeneri Crosswhite p henology at three Col orado sites in 2007, incl u ding Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP) Fremont County. Elevation is specified in meters. Site and Rosette approximate emergence elevation sse 2133 6 PC 2438.4 LP 2804.16 517 5/17 5 / 22 Floral budding 5 / 23 6/16 6 / 20 Early anthesis 6/16 7 / 1 7 / 6 33 Mid anthesis 6 / 22 7/10 7/13 Late anthesis 6 / 30 7 /17 7 / 20 Matu e frui 817 8 / 3 9 / 4

PAGE 44

F i g u re 2 a-c: Floral morphology and receptivity in P e n s t e mon d e g e n e ri Crosswhite: (a) straight style (b) slightly curved style still situated within groove on upper inside surface of flower and (c) fully curved style almost touching staminode hairs 34

PAGE 45

a) b) c) Figure 3a -c: Pollen grains viewed at 400X from: (a) P enstemon virens P enne l ex Rydb. (b) P. degeneri Crosswhite, and (c) P virgatus A. Gray. 35

PAGE 46

Table 2: All visitors captured on Penstemon degeneri Crosswhite in 2007 at the Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), Colorado sites in 2007. Numbers with a slash indicate females / males Site sse PC LP Megachilidae Hoplitis albifrons argentifrons (Cresson 1864) H. truncata mescalerium (Cockerell1910) H. producta (Cresson 1864) 2 Megachile m e lanophaea (Smith 1853) 1 Osmia brevis (Cresson 1864) 16 0. bucephala (Cresson 1864) 5 0. pentstemonis (Cockerell 1906) 5 O.juxta (Cresson 1864) 6 0. albolateralis (Cockerell 1906) 2 0. proxima (Cresson 1864) 2 0. bucephala (Cresson 1864) 4 0. bruneri (Cockerell 1906 ) 1 0. densa (Cresson 1864) 1 0. nanula (Cockerell 1906) 1 0. paradisica (Sandhouse 1924) 1 Halictidae Agapostemon sp. 2 Lassioglossum sp. 8 Dialictus sp. 3 5 1 Halictus sp. 2 36

PAGE 47

T abl e 2: continued Apidae Melissodes sp. 1 Bombus centra/is Cresson 1864) 69 / 5 37 B.fervidus (Fabricius 1798) 0 / 2 1 B. appositus (Cresson 1878) 2 511 9 B huntii (Greene 1860) 9 8/1 46 B. nevadensis nevadensis (Cresson 1874) 1 16 B.spp. 3 7 B. bifarius (Cresson 1864) 2 1 B morrisoni (Cresson 1864) 1 B griseocol/is (DeGeer 1773) 1 B rufocinctus (Cresson 1863) 1 Anthophora montana (Cresson 1869) 5 4/1 1 A. terminalis (Cresson 1869) 1 3/2 A. ursina (Cresson 1869) 1 Diadasia sp. 3 Apis mellifera (Linnaeus 1758) Colletidae Hyaleus 1 Hesperiidae Papilionidae 2 3 5 Papilio rutulus Di tera 1 Bombyliidae 8 3 5 Syrphidae 4 Eumeninae 1 Masarinae Pseudomasaris vespoides 2 Trochilidae Total Species at each site 23 23 24 Total number of species at all three sites 44 37

PAGE 48

oApidae Diptera o Megachildae o Halictidae Vespidae o Lepidoptera Colletidae o Trochilidae Figure 4: Visitors captured on Penstemon degeneri Crosswhite at three Colorado sites in 2007; organized by family or order. 38

PAGE 49

7 .30% o Bombus centralis/huntii Apidae o Megachilidae. 53% o Halictidae Figure 5: Bees captured on P enstemon degeneri Crosswhite at three Colorado sites in 2007; organized by family or order, with Bombu s centra/is and B. huntii (Apidae) grouped to indicate their relative abundance. 39

PAGE 50

Table 3: Behavior of visitors of Penstemon degeneri Crosswhite at three Colorado sites in 2007. Taxon N Average number Average number Average of plants visited of flowers visited overall per plant length of visit Apidae 174 160 5.3 22.6 min Halictidae 19 2 2 64 sec Megachiledae 7 3 2.5 43 sec Diptera 14 3 3 27 sec Lepidoptera 6 9 2 120 sec Trochilidae 1 10 2 15 min 40

PAGE 51

a) b) c) Fig u re 6ac : Bee behavior on P e n s t e mon d e g e n e ri Crosswhite flowers : (a) halictid (Halictidae) bee entering a flower inverted remaining up to 60 s e conds and actively collecting pollen and possibly nectar; (b) O s mia (Megachilidae) bee about to enter a flower ; (c) Bombu s (Apidae) bee with large scopal (pollen) sac collecting nectar 41

PAGE 52

a) b) c) Figure 7a-c: Progression over 30-45 minutes depicting a female worker bee, Bombus huntii, on Penstemon degeneri Crosswhite collecting and accumulating pollen and nectar in her scopae: (a) early ( 5-l 0 mins) during the visit, with relatively small amounts of pollen in the scopal sac; (b) following approximately 15 mins. of pollen and nectar collecting, with larger scopal sac; (c) late (30-45 mins.), with very large scopal sac making it difficult for the bee to hang onto the flower. 42

PAGE 53

Bee families carrying > 80/o P degeneri Scopae Pollen Apidae 43% Megachilidae 57% Figure 8: Percent megachilid (Megachilidae) and apid (Apidae) bees that carried greater than 80 % P e n s t e mon d ege n e ri Crosswhite pollen in their scopae at three Colorado sites in 2007. 43

PAGE 54

100 90 80 70 60 50 40 30 20 10 0 C Osm i a brevis N=1 0 Osm i a penstemon i s N=3 0 Osm i a proxima N=2 OOsml a bruneri N=1 Osm i a par adisica N=1 0 D i alictis sp N=1 Osm i a densa N=1 OAnthophora termina lis N=5 Dalictus spp N=1 Lassioglossum heterominum N= 1 o Hopln i s producta N=1 0 Bombus spp N=S Bombus central i s N=89 Osmia jux1a N=5 Osmia a l bolateral i s N=2 Bombus griseoco llis N=1 c Bombus apposHus N = 13 0 Bombus hunti i N=96 0 Lassloglossum sisymbril N=4 0 Bombus nevadens i s nevadensis N=4 o Anthopho r a montana N=B 0 Halictus tripartHus N=2 0 Bombus monison ii N=1 0 Bombus fervidus N= 1 Bombus bitaris N=1 c Osm i a nanula N=1 OHopiH is albifrons argentifrons N = 1 0 Hopi H i s truncata mescalerium N=1 c Osm i a bucephala Dialictus sp N=1 Bombus rufocinctus N=1 Figure 9: Percent P enstemon degeneri Crosswhite scopal pollen found on 31 species of female worker bees captured at three Colorado sites in 2007. 44

PAGE 55

Table 4: Hymenoptera species collected from Penstemon degeneri Crosswhite at three Colorado sites in 2007, including: number collected at the Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP) sites, and in total (NC); number collected with pollen loads (NP); and mean percentage (SD) of P. degeneri pollen in the pollen loads (% PD). Includes total number of species captured at each site and time spent capturing. sse PC LP NC NP %PD Megachilidae H oplitis produ c ta (Cresson 1864) 2 2 1 94 H. albifrons argentifrons (Cresson 1864) 1 1 2 1 10 H. tnmcata mescalerium (Cocke rell 191 0) 1 1 1 8 Osmia brevi s (Crcssonn 1854) 1 1 16 18 11 99.8.60 0 pentstemonis (Cockerell 1906) 5 5 3 99.6.58 O juxta Cresson (Cresson 1864) 6 6 5 82.6. 1 0 albolateralis (Cockerell 1906) 2 2 1 83 0 proxima (Cresson 1864) 2 2 2 100 0 bucephala (Cresson 1864) 4 0 bruneri (Cockerell 1906) 1 1 1 100 0 densa (Cresson 1864) 1 1 1 96 0. nanula (Cockerell 1906) 1 1 2 1 20 0 paradisica Sandhouse (1924) 1 1 1 100 0. gaudiosa (Cockerell 1906) 1 1 1 20 M e gachile m e lanoehaea {Smith 1853} Halictidae Agaposte mon sp. 2 Hal ictus tripartitus (Cockerell 1895) 2 2 Lassioglossum s i sy mbrii (Cockerell 1895) 2 2 L. heterorhinum (Cockerell 1895) 6 6 4 67.4 L. manitouellum (Cockerell 1895) 1 1 95 Dial ictus s 1 1 AJ:!idae Apis melifera (Linnacus 1758) Bombus. griseocollis (DeGccr 1773) 1 1 1 81 B appositus Cresson 2 6 9 17 13 81.5 B. bifarius (Cresson 1878) 2 1 3 3 21.6.2 B.nevadensis nevad e nsis (Cresso n 1878) 1 13 14 4 6.3.7 B centra/is (Cresson 1878) 74 37 111 75 82.2.0 B huntii (Greene 1860) 9 8 45 62 46 78.5.3 B. nifocinctus (Cresson 1878) 1 5 45

PAGE 56

Table 4 continued B fervid u s ( Fab r i c iu s 1798) 2 2 B m o r riso nii (Cresso n 1878) 1 1 1 25 B.spp 10 10 9 8814.4 Anthophora Montana (Cresso n 1869) 5 6 1 12 7 61.442.6 A ursina ( C r esso n 1869) 1 1 A t e r min a /i s (Cresson 1869) 1 5 6 4 76. 2 D iadas i a ri n c on i s (Coc k erell 1897) 2 3 Me li ssodes S f!f!: 2 2 Colletidae H y_laeus an n ul a tu s { Linn ac u s 1 7 58} 1 Eumeninae 1 1 2 Masarinae 2 ...... 2 T ota l Sp ecies 17 16 22 T ota l {min/ d a:z:s} 720 / 3 720 / 3 720 / 3 46

PAGE 57

'""" '""' 40% '""' I -. m '""' Bombus centralis Bombus huntii Bombus appositus Anthophora montana Anthophora Bombus nevadensis (N= 8 7 1 4 .7mm) (N= 42 1S.6mm) (N=12 19 .5mm) (N=3 1 7 8mm ) terminalis (N=2 (No7 24.2mm) 13 0mm) Figure 10: Percent apid (Apidae) bees caught at three Colorado sites in 2007 with moderate to copious amounts of P enste mon d egeneri Crosswhite pollen on their body where H = head DT = dorsal thorax VT= ventral thorax and VA=ve ntral abdomen Includes number of individuals captured (N) and average size. 47

PAGE 58

"""' .. ,. ..,. 70% ..,. .. ,. ''"' "" -,.,. .,. Bombus centn.lls (No:Sll, Bombus huntll (Nc8, 15.&mm) Bombus pposttus (N-2 ArYthophor. mont., (N-2 14. 7mm) 111. 5mm ) 17 6mm ) Figure 11: Percent apid (Apidae) bees caught at three Colorado sites in 2007 with copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H = head, DT = dorsal thorax, VT = ventral thorax, and V A = ventral abdomen. Includes number of individuals captured (N) and average size. 48

PAGE 59

120% 100% 80% OH oT 60% OVT DOA vA 40% 20% Figure 12: Percent megachilid (Megachilidae) bees caught at three Colorado sites in 2007 with moderate to copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H = head, DT = dorsal thorax, VT = ventral thorax, and V A = ventral abdomen. Includes number of individuals captured (N) and average SIZe. 49

PAGE 60

120% 100% 80% 60% 40% 20% 1 I OH or DVT DDA vA Figure 13: Percent megachilid (Megachilidae) bees caught at three Colorado sites in 2007 with copious amounts of P enste mon degeneri Crosswhite pollen on their body where H = head, DT= dorsal thorax VT = ventral thorax, and VA=ventra l abdomen. Includes number of individuals captured (N) and average size. 50

PAGE 61

APPENDIX A Fuchsin Gelatin Recipe (as suggested by Beattie A. J. 1971. Pan-Pacifi c Entomologist 47: 82) Ingredients needed: -distilled water, 175 ml -glycerin 150m! -gelatin 50 g -crystalline basic fuchsin as desired enough to make solution the color of a fine claret" -crystalline phenol, 5 g important in humid environments, may be left out if gelatin can be refrigerated. Fuchsin Gelatin Procedure: NOTE: Do not touch gelatin if phenol is used. 1. Add the gelatin to the distilled water in a beaker and heat until the gelatin dissolves 2 Add the glycerin 3 Add phenol, if desired 4. Add basic fuchsin crystals a few at a time until the solution is the color desired. Too light will not stain the pollen, but too dark may mask details of the pollen 5. Filter the solution through cheesecloth 6. Pour into sterile Petri plates that can be covered If phenol is not used refrigerate the plates and slides. Plates will keep about a month without refrigeration. 51

PAGE 62

APPENDIXB Glycerin Jelly Slides (Erdtnab 1960, 1969) 1. Clean two slides and one cover slip per sample (e .g. wash with detergent, rinse in 70% ethanol). 2 If the pollen sample is stored in 70% ethanol, put a drop of liquid glycerin on the slide and mix in a drop of pollen suspension. Let the mixture sit and evaporate for about 10 minutes. 3 Clean the point of a dissecting needle by washing it in alcohol or heating it in a flame. 4. Spear a small piece (about 2mm3 ) of solid glycerin jelly on the point ofthe needle. 5 Touch the piece of jelly to the glycerin-pollen mixture on the slide to pick up pollen on the jelly. 6. Place the piece of jelly with the pollen on it onto the second clean slide, about 2.5cm from one end of the slide (leaving room for a label at the other end). 7 Warm the slide gently on a hotplate or over a flame (Bunsen burner or alcohol lamp) until jelly melts. Do not allow the glycerin jelly to boil, which would introduce air bubbles that are difficult to remove. 8. Place the slide under a microscope and spread the pollen grains out while stirring the jelly carefully with a clean needle. 9. Re-warm the slide gently and lower the cover slip carefully over the molten jelly. The jelly should make a circle 2-3 mm in diameter. 10. While the slide is still warm, place it face down on a stand that supports it at both ends, without letting anything touching the cover slip while the jelly is still molten. Label and refrigerate the slides. 52

PAGE 63

REFERENCES Ashworth, L., R. Aguilar, L. Galetto, and M. Adrean Aizen. 2004. Why do pollination generalist and specialist plant species show simi lar reproductive susceptibility to habitat fragmentation? Journal of Ecology. 92:717-719 Beatty, B., W. Jennings, and R. Rawlinson. 2004. Penstemon degeneri Crosswhite (Degener's Beardtongue): A Technical Conservation assessment. Prepared for the USDA forest Service, Rocky Mountain Region Species Conservation Project. Denver Colorado. Castellanos, M. C., P. Wilson, S. J. Keller, A. D. Wolfe, and J.D. Thomson. 2006. Anther evolution: pollen presentation strategjes when pollinators differ. American Naturalist. 167 : 288-296. Crosswhite, F. S., C. D. Crosswhite. 1966. Insect pollinators of Penstemon series Graciles (Scophulariaceae) with Notes on Osmia and other Megachilidae. American Midland Naturalist. 76: 450-467. Crosswhite, F. S. 1967. Revision of Penstemon section Penstemon (Scrophulariaceae) II. A Western Alliance in Series Graciles. American Midland Naturalist. 77: 28-41. Crud en, R. W. 1977. Pollen-Ovule Ratios: A conservative indicator of breeding systems in flowering plants. Evolution. 31: 32-46. 53

PAGE 64

Dieringer, G., and L. R. Cabrera. 2002. The interaction between pollinator size and the bristle staminode of Penstemon digitalis (Scrophulariaceae ). American Journal of Botany. 89: 991-997. Gess, S. K. 1996. The Pollen Wasps, Ecology and Natural History of the Masarinae. Harvard University Press. USA. Keams, C. A. 2001. The Natural History of Bumblebees, A sourcebook for Investigations. University Press of Colorado. Boulder, Colorado. Keams, C., and W.lnouye. 1993. Techniques for Pollination Biologists. University Press of Colorado. Niwot, Colorado. Kimball, Sarah. 2008. Links between floral morphology and floral visitors along an elevational gradient in a Penstemon hybrid zone. Oikos. 117: 10641074. Lesica, P., R. Yurkewycz, and E. E. Crone. 2006. Rare plants are common where you fmd them. American Journal of Botan y. 93: 454-459. Michener, C. D. 2000. The Bees of the World. The John Hopkins University Press. Baltimore, Maryland. Spackman Panjabi, S. C. 2004. Visiting insect diversity and visitation rates for seven globally-imperiled plant spec ie s in Colorado's middle Arkansas Valley. Colorado Natural Heritage Program College of Natural Resources, Colorado State University, Ft. Collins, Colorado. Prepared by 54

PAGE 65

the Native Plant Conservation Alliance and the National Fish and Wildlife Foundation Tepedino V.J ., S.D. Sipe s, and T. L. Griswold. 1999 The reproducti v e biology and e f fective pollinator s of the endanger e d beardtongue P en s t e m o n p enlandii. Plant S ys t e mati cs E v olution 219: 39-54 Tepedino V. J., W.R. Bowlin and T. L. Griswold. 2006. Pollination Biology of the Endemic Blowout Penstemon (P ha y d e nii S Wats.: Scrophulariaceae) in Nebraska. Journal of the Torr ey Botani c al So c i ety 133: 548-559 Tepedino V. J., T R. Toler B. A. Bradley, J. L. Hawk, T L. Griswold. 2006. Pollination biology of a disjunct population of the endangered sandhills endemic P enst e mon ha y d e nii S. Wats (Scrophulariaceae) in Wyoming USA Plant E c ology 193 : 59-69 Wilson P ., M C. Castellanos J. N. Hogue J.D. Thomson and W. S. Armbruster. 2004. A multivariate search for pollination syndromes among P e n s t e mons Oiko s 104 : 345-361 Wolfe, A., C. P. Randle S. L. Datwyler J. J. Morawetz N. Arguedas and J. Diaz. 2006. Phylogeny taxonomic affinities and biogeography of P en s t e m o n (Plantaginaceae) based on ITS and cpDNA sequence data Ame ric an Journal of Botan y 93 : 1699-1713 55



PAGE 1

EFFECTIVE POLLINATION IN A RARE COLORADO ENDEMIC, PENSTEMON DEGENERJ By Carol Theresa English B.S., University of California, Santa Cruz 1985 A thesis submitted to the University of Colorado Denver in partial fulfillment of the requirements for the degree of Masters of Science Biology 2008

PAGE 2

This thesis for the Master of Science Degree by Carol Theresa English has been approved by Leo P. Bruederle Michele Engel ? Jennifer Ramp Neale Date

PAGE 3

English, Carol Theresa (M.S. Biology) Effective Pollinators in a rare Colorado endemic, Penstemon degeneri Thesis directed by Associate Professor Leo P. Bruederle ABSTRACT Pens lemon degeneri Crosswhite (Piantaginaceae) is a rare Colorado endemic about which little is known; it is considered threatened and globally imperiled, with few federal or state level conservation regulations in place. Because P. degeneri depends on local pollinators for successful reproduction, determining effective pollinators is an essential step in conservation management for P. degeneri. This research determines visitors and effective pollinators for Penstemon degeneri. Effective pollination was assessed by correlating visitor frequency with visitor behavior, amount and placement of P. degeneri pollen on the visitor's body, and the visitor's fidelity toward P. degeneri. In 2007, three populations of P. degeneri were studied that span an elevation range of7,000-9200' in Fremont County, Colorado. As with other rare endemic Penstemon species in the west, P. degeneri attracts a diversity of visitors. These visitors represent four orders and ten families, including butterflies (Lepidoptera), flies (Diptera), bees and wasps (Hymenoptera), and one hummingbird (Trochillidae) based upon observations conducted during the 2007 field season. Only a few of the many visitors, including Osmia (Megachilidae) and Bombus (Apidae ), were verified as effective pollinators. Unexpectedly the oligolectic pollen wasp, Pseudomasaris vespoides, was nearly absent in 2007, although preliminary results indicate that this

PAGE 4

preliminary results indicate that this pollen wasp was a frequent visitor in 2008. The most effective pollinators on P. degeneri were Osmia bees, yet these bees did not occur with consistency at all three study sites. However, as with the pollen wasp, preliminary results indicate Osmia bees were frequent visitors at each of the sites in 2008. A high percentage of these individuals carried copious amounts of P. degeneri pollen on their bodies, and many had 100% fidelity toward P. degeneri pollen in 2007. Apids, including Bomb us centra/is and B. huntii were also effective pollinators, although less effective than Osmia bees. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. Leo P. Bruederle

PAGE 5

DEDICATION I wish to dedicate this thesis to my husband, Dave Elin, who without his loving support, patience, and many long hours of assistance with photography and fieldwork, I would not have finished this degree:

PAGE 6

ACKNOWLEDGEMENTS I give thanks to my graduate advisor, Dr. Leo Bruederle, who has guided me with great patience and valuable knowledge to reach this goal. I would also like to thank Dr. Vince Tepedino, who answered numerous questions regarding pollination biology. Without his help, this research would not have happened. Additionally, I want to give out great thanks to both Virginia Scott and Terry Griswold, who keyed out all the bee species, and offered field materials in support of collecting and storing the insects.

PAGE 7

TABLE OF CONTENTS Figures ....................................................................................................................... viii Tables ............................................................................................................................. x Chapter 1.0 Introduction ........................................................................................................ 1 1.1 Genus Penstemon .............................................................................................. 1 1.2 Penstemon degeneri ........................................................................................... 4 1.3 Research Objectives ........................................................................................... 6 2.0 Materials and Methods ...................................................................................... 8 2.1 Research Sites .................................................................................................... 8 2.2 Phenology .......................................................................................................... 8 2.3 Effective Pollinators .......................................................................................... 9 2.4 Visitors ............................................................................................................. lO 2.5 Visitor Behavior ............................................................................................... II 2.6 Body Pollen ..................................................................................................... 12 3.0 Results .............................................................................................................. l5 3.1 Phenology ........................................................................................................ 15 3.2 Visitors ............................................................................................................. l6 3.3 Visitor Behavior ............................................................................................... I? 3.4 Body Pollen ..................................................................................................... 19 4.0 Discussion ........................................................................................................ 23 4.1 Visitor Guild .................................................................................................... 23 4.2 Effective Pollinators ........................................................................................ 25 4.3 Future Studies .................................................................................................. 30 Appendix A. Fuchsin Gelatin Recipe and Procedure ............................................................ 51 B. Glycerin Jelly Slides Procedure ....................................................................... 52 References .................................................................................................................... 53 vii

PAGE 8

LIST OF FIGURES 1. Locations for three Penstemon degeneri Crosswhite research ........................ 32 sites: Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), in Fremont County, Colorado. 2. Floral morphology and receptivity in Penstemon degeneri Crosswhite: ........ 34 (a) straight style, (b) slightly curved style still situated within groove on upper inside surface of flower, and (c) fully curved style almost touching staminode hairs. 3. Pollen grains viewed at 400X from: (a) Penstemon virens Pennel ex Rydb.,.35 (b) P. degeneri Crosswhite, and (c) P. virgatus A. Gray. 4. Visitors captured on Penstemon degeneri Crosswhite at three ....................... 38 Colorado sites in 2007; organized by family or order. 5. Bees captured on Penstemon degeneri Crosswhite at three ............................ 39 Colorado sites in 2007; organized by family or order, with Bombus centra/is and B. huntii (Apidae) grouped to indicate their relative abundance. 6. Bee behavior on Penstemon degeneri Crosswhite flowers: (a) halictid ......... .41 (Halictidae) bee entering a flower inverted, remaining up to 60 seconds, and actively collecting pollen and possibly nectar; (b) Osmia (Megachilidae) bee about to enter a flower; (c) Bombus (Apidae) bee with large scopa (pollen) sac collecting nectar. 7. Progression over 30-45 minutes depicting a female worker bee, Born bus ..... .42 huntii, on Penstemon degeneri Crosswhite collecting and accumulating pollen and nectar in her scopae: (a) early (5-10 mins) during the visit, with relatively small amounts of pollen in the scopal sac; (b) following approximately 15 mins. of pollen and nectar collecting, with larger scopal sac; (c) late (30-45 mins.), with very large scopal sacs making it difficult for the bee to hang onto the flower. 8. Percent megachilid (Megachilidae) and apid (Apidae) bees that carried ....... .43 greater than 80% Penstemon degeneri Crosswhite pollen in their scopae at three Colorado sites in 2007. 9. Percent Penstemon degeneri Crosswhite scopal pollen found on 31 .............. 44 species of female worker bees captured at three Colorado sites in 2007. viii

PAGE 9

10. Percent apid (Apidae) bees caught at three Colorado sites in 2007 with ....... .47 moderate to copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsa1 thorax, VT=ventra1 thorax, and VA =ventral abdomen. Includes number of individuals captured (N) and average size. 11. Percent apid (Apidae) bees caught at three Colorado sites in 2007 with ....... .48 copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and V A=ventral abdomen. Includes number of individuals captured (N) and average size. 12. Percent megachi1id (Megachilidae) bees caught at three Colorado sites ....... .49 in 2007 with moderate to copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA =ventral abdomen. Includes nwnber of individuals captured (N) and average size. 13. Percent megachilid (Megachilidae) bees caught at three Colorado sites ........ 50 in 2007 with copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA =ventral abdomen. Includes number of individuals captured (N) and average size. IX

PAGE 10

LIST OF TABLES I. Penstemon degeneri Crosswhite phenology at three Colorado sites in ........... 33 2007, including Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), Fremont County. Elevation is specified in meters. 2. All visitors captured on Penstemon degeneri Crosswhite in 2007 at the ........ 36 Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), Colorado sites in 2007. Numbers with a slash indicate females/males. All visitors are identified to the lowest possible taxonomic level. 3. Behavior ofvisitors of Penstemon degeneri Crosswhite at three Colorado .... 40 Sites in 2007. 4. Hymenoptera species collected from Penstemon degeneri Crosswhite ......... .45 at three Colorado sites in 2007, including: number collected at the Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP) sites, and in total (NC); number collected with pollen loads (NP); and mean percentage (SD) of P. degeneri pollen in the pollen loads(% PD). Includes total number of species captured at each site and time spent capturing. X

PAGE 11

1.0 Introduction 1.1 Genus Penstemon Penstemon Mitchell (Plantaginaceae) is the largest endemic genus in North America (ca. 270 spp.), with the majority of species located west ofthe Continental Divide (Wolfe 2006). This diversity is believed to represent a rare example of continental radiation characterized by high neoendemism and rapid speciation (Wolfe 2006, Lesica et al. 2006). Continental evolutionary radiations have been documented worldwide including: southwestern Australia, South Africa, Mediterranean Basin, and western North America (Lesica et al. 2006). Western North America genera believed to exemplify recent radiations include Penstemon (Plantaginaceae), Erigeron L., (Asteraceae), Astragalus L. (Fabaceae), and Eriogonum Mich. (Polygonaceae ). These genera, which are thought to have evolved relatively recently, are widespread and have large numbers of narrow endemics (Lesica et al 2006). Penstemon is hypothesized to have originated in the central Rocky Mountains in the late Tertiary with subsequent diversification due to evolutionary adaptations, including pollinator and niche specialization (Wolfe et al. 2006). Penstemon is a member of Plantaginaceae tribe Cheloneae. Species within this tribe are characterized as having a cymose inflorescence and a sterile staminode (Wolfe et al. 2006). The genus Penstemon is divided into six

PAGE 12

subgenera including Cryptostemon and Dissecti, both of which are monotypic, Penstemon (ca.182 spp.), Habroanthus (ca. 50 spp.), Saccanthara (ca. 28 spp.), and Dasanthera (ca. 9 spp.); these subgenera are further divided into sections and subsections. Penstemon is distributed coast to coast from Guatemala to Alaska, with over I 00 species restricted in distribution to a single state; for example, approximately nine Penstemon species are endemic to Colorado (Wolfe et al. 2006). Most Penstemon species attract many different pollinators including bees, wasps, moths, and flies, while the remainder (ca. 40 spp.) are adapted to attract hummingbirds, specifically (Wolfe et al. 2006). These two syndromes differ with respect to: anther dehiscence patterns, corolla color and shape, and staminode morphology. For example, differences in anther dehiscence have been correlated with pollination efficiency (Castellanos et al. 2006). Bird pollinated flowers release pollen in large quantities over a short amount of time, whereas bee pollinated flowers release pollen in smaller quantities over longer periods of time (Castellanos et al. 2006). Furthermore, bird-pollinated Penstemons tend to have red, narrowly pendant shaped corollas with reduced, glabrous staminodia. In contrast, insect-pollinated Penstemons tend to have blue, pink, or purple corollas, with wide ventral lobes that are used as landing pads, and variously shaped pubescent staminodia that facilitate pollination (Wolfe et al. 2006). The staminode, which is located on the corolla throat directly under the anthers and 2

PAGE 13

stigma, is larger than the fertile stamens and has bristles on the adaxial distal surface (Dieringer et al. 2002). Recent studies on P. digitalis Nutt. have demonstrated that staminode bristles and pollinator size correlate with pollination efficiency (Dieringer et al. 2002). When staminodia are removed, pollen deposition efficiency has been shown to be reduced for small bees in comparison to larger bees (Dieringer et al. 2002). The staminode acts as a "step stool" and allows the dorsal side of small bees to contact the stigma more consistently and efficiently as the bees enter the corolla orifice (Dieringer et al. 2002). The staminodia in other Penstemon species, such asP. palmeri Gray, with their larger corollas and, thus larger pollinators, act as levers. Heavy bees activate the lever and cause the style to drop onto the dorsal side of their body (Dieringer et al. 2002). Recent pollination biology studies on congeners (including P. penlandii Weber, P. debilis O'Kane and Anderson, P. glaber Pursh, P. haydenii Gray, P. harringtonii Wats., and P. strictus Benth) have shown a wide variety of bees, wasps, flies, butterflies, and beetles visiting flowers; effective pollinators for these Penstemon species include bees in the Megachilidae (e.g., Osmia) and Apidae (e.g., Bombus). Pseudomasaris vespoides, a wasp in the Masaridae, exhibits a specialized preference (oligolecty) for Penstemon pollen (Beaty et al. 2004). A pollination study performed on Penstemon series Graciles demonstrated that small bees in the genus Osmia are the primary pollinators for most species in 3

PAGE 14

this alliance. Penstemon hirsutus (L.) Willd., P. tenuiflorus Penn., and P. ok/(!Jhomensis Penn. are exceptions and are pollinated principally by small bees in the genera Hoplitis and Ceratina (Crosswhite and Crosswhite 1966). This same study found evidence supporting that entire subgenera of Osmia are mostly dependent on Penstemon pollen and nectar for nutrition. Furthermore, a small group of species within Osmia subgenus Nothosmia are oligolectic for Penstemon series Graciles alliance Graciles. More recent studies have also shown that Osmia bees are the most effective pollinators for P. pendlandii, P. haydenii, P. harringtonii, and P. /emhiensis, specifically (Tepedino et al. 1999, 2006). 1.2 Penstemon degeneri Penstemon degeneri Crosswhite, commonly referred to as Degener's Beardtongue, is a rare species endemic to south central Colorado (Crosswhite 1967). The Colorado Natural Heritage Program considers P. degeneri to be a species of special concern, and ranks it S2 (vulnerable to extirpation, endangered or threatened in the state) and G2 (globally imperiled due to its regional rare and endemic status). Due to the fact that very little is known about the biology of this species, there are few mechanisms at the federal or state level to regulate its conservation. These regulations may not be adequate to conserve P. degeneri overthelongterm. Penstemon degeneri belongs to subgenus Habroanthus section Habroanthus subsection Humiles series Gracilis alliance 0/iganthi (Beaty et. al 2004). The 4

PAGE 15

five species within alliance 0/iganthi occur in mountainous areas of three western states. These include: P. degeneri and P. griffinii Nels. in southern Colorado, P. injlatus Crosswhite, and P. pseudoparvus Crosswhite in northern and central New Mexico, and P. oliganthus Crosswhite in eastern Arizona (Fig. 1) (Crosswhite 1967). Penstemon degeneri has pale, lavender, zygomorphic corollas that are 14-19 mm long. The corolla throat is obscurely ridged, with sparse, yellow, non glandular hairs; the mouth has a few white hairs; and the staminode is sparsely yellow bearded. The four free stamens are didynamous, terminating in glabrous anthers (Crosswhite 1965). Penstemon degeneri is endemic to Fremont, Teller, and Custer Counties, Colorado, where it occupies habitats varying from pinyon-juniper woodlands (1830 to 2042 m elevation) to montane meadows and ponderosa pine parklands (2073 to 2743 m) (Beaty et al. 2004). In pinyon-juniper woodlands, associated species include open canopies of Pinus edulis Engleman (Pinaceae ), with a mixed understory of shrubs, including Juniperus communis (Smith) Celakovsky (Cupressaceae), grasses, and forbs. In montane meadows and ponderosa parklands, associated species include Danthonia parryi Scribner (Poaceae ), Pinus ponderosa Douglas (Pinaceae ), Populus tremuloides Mich. (Salicaceae ), and Pseudotsuga mensiesii (Mirbel) Franco (Pinaceae). Other associated species include Achillea lanulosa Nutt. (Asteraceae)., Antennaria spp. Gaertner 5

PAGE 16

(Asteraceae), Artemisiafrigida Willd. (Asteraceae), Carex L. spp. (Cyperaceae), Lupinus L. spp. (Fabaceae), Muhlenbergia montana Schreber (Poaceae), Thermopsis montana Nutt. (Fabaceae), and Penstemon virens Pennell (Beaty et al. 2004). The Colorado Natural Heritage Program prepared status reports for P. degeneri in 1981, based on five known populations, and again in 1991, based on nine known populations (Beatty et al. 2004). In 1998, several more populations were found during a floristic survey, bringing the total number of known populations to 14. Since that time, ten of these populations have been re affirmed. Population size varies greatly from 25 to thousands of individuals (Beatty et al2004). Population abundance trends have not been documented, yet some populations appear to have declined severely, with some evidence of herbivory (Beaty et al. 2004). A short term insect visitation study completed in 2004 verified that a variety of insects, including flies, bees and wasps, visit P. degeneri (Spackman 2004), yet it remains unknown which insects are most effective pollinators. 1.3 Research Objectives The objectives of this research are to develop a comprehensive list of floral visitors for P. degeneri and, subsequently, determine which visitors are most effective pollinators across the range in elevation for this species. I hypothesize that a variety of insects will visit P. degeneri, including butterflies (Lepidoptera), 6

PAGE 17

flies (Diptera), and bees and wasps (Hymenoptera). Furthermore, I hypothesize that small bees within genus Osmia, a variety of bumblebee species within genus Bombus, and the oligolectic wasp Pseudomasaris vespoides are the most effective pollinators for P. degeneri (Tepedino et al. 1999, 2006, Wilson et al. 2004). Additionally, effective pollinator species have been shown to vary between years and in different geographic locations (Tepedino et al. 1999, 2006). This variation is suspected to be due to the fact that pollination effectiveness depends on the coordinated timing of floral reproductive maturity and insect activity (Tepedino et al. 1999, 2006). Therefore, I expect effective pollinators for P. degeneri to vary both spatially and temporally. Determining effective pollinators for a rare, endemic and entomophilous plant is important since pollinators are essential for reproductive success. Furthermore, because previous studies have revealed that a co-evolutionary relationship exists between certain hymenopterans and Penstemon, the former also depend on this species for successful reproduction (Tepedino et al. 1999, 2006, Wilson et al. 2004). Therefore, pollinators and their habitat must be conserved along with the plant. Preserving these effective pollinators also preserves a breeding system that maintains the appropriate and healthy levels of genetic diversity for P. degeneri (Kearns 1993). Finally, discovering the effective pollinators for another Penstemon species may help scientists better understand the evolutionary mechanisms leading to speciation in genus Penstemon. 7

PAGE 18

2.0 Materials and Methods 2.1 Research Sites Pollinator studies were carried out at three research sites in Fremont County, Colorado: Sunset City, Phantom Canyon, and Locke Park (Figure 1). These sites represent, more or less, the ecological range of P. degeneri. Sunset City Colorado is pinyon-juniper woodland at 2,300 m elevation, and occurs on land administered by the Bureau of Land Management (BLM). It is located approximately 11 miles east of Canyon City, and six miles south of the Royal Gorge Bridge that spans the Arkansas River at N38 24.780 Wl05 24.894. Phantom Canyon is a montane meadow at 2,700 m elevation, and is located on BLM land approximately 20 miles north-east of Canon City and the Arkansas River at N38 36.821 W 105 08.089. Locke Park is a ponderosa parkland at 3,070 m elevation, and is located in the San Isabel National Forest, approximately 20 miles south of Canyon City and the Arkansas River at N3815.867 Wl05 15.091. All field and lab-work was completed with the help of one to several student field assistants in 2007 and 2008. 2.2 Phenology Between May and September 2007, the following phenological stages were recorded at each of the aforementioned P. degeneri sites: rosette emergence; floral budding; early, middle, and late anthesis; and mature fruit set. Anther dehiscence 8

PAGE 19

and style movement (curving down) was observed while performing the pollination observation experiments. 2.3 Effective Pollinators Relative pollinator effectiveness for all insect visitors was determined following the techniques ofTepedino et al. (1999, 2006). In so doing, I asked four questions regarding P. degeneri visitors. Which insects are the most frequent visitors to P. degeneri? Do frequent visitors carry relatively large amounts of P. degeneri pollen on their bodies? Does the behavior of the visitors promote the transfer of pollen from their bodies to the stigma? Which female bees have high fidelity toward P. degeneri pollen? To answer these questions I observed the behavior of all insects visiting P. degeneri. I quantified P. degeneri body pollen on all insect visitors that I captured. Subsequently I correlated the body pollen results with the behavior observation results. For example, if a particular bee carried large amounts of P. degeneri pollen on its dorsal thorax, I correlated this result with the part of the body that had been observed contacting the stigma. Finally, I quantified scopa pollen to determine a female bee's degree of fidelity (faithfulness) toward P. degeneri pollen. By integrating and interpreting these results, I subsequently determined the relative effectiveness for all P. degeneri insect visitors in 2007. For example, highly effective pollinators frequently visited P. degeneri flowers, behaved in such a way to promote pollen transfer to the stigma, carried copious 9

PAGE 20

amounts of P. degeneri pollen on their bodies, and carried a high percentage (>80%) of P. degeneri pollen in their scopae. 2.4 Visitors In order to quantify the relative visitation frequency and diversity of all visitors toP. degeneri, insects were captured by passing slowly through a P. degeneri population over a period of two hours and capturin_g any insects seen on or inside P. degeneri flowers. Insects were caught by me or a field assistant using a Bioquip insect net with an 18" handle and a 12" ring, and were then placed in a killing jar. Insects were subsequently pinned, using stainless steel Bohemia Insect Pins, labeled, and stored in a Styrofoam insect box. Some insect species, including butterflies and bombyliid flies, were excluded from frequent capture, as they have long proboscises and can obtain nectar without contacting the reproductive organs; field observations confirmed very little pollen transfer occurred by these insects during visits to P. degeneri. At Sunset City, I captured insects over seven days between June 18 and July 5, for a total of21 hours. Nine of the "catch hours" were in the morning and 12 were in the afternoon. At Phantom Canyon, I captured insects between July 1 and July 17 for a total of three days and 12 hours, including two morning hours and ten afternoon hours. At Locke Park, insects were captured over three days between July 6 and July 18 for a total of 12 hours, with six morning hours and six afternoon hours. 10

PAGE 21

Insects were pinned, labeled using acid-free 25# weight paper, and deposited at the University of Colorado Museum of Natural History. Virginia Scott, Collections Manager for the Entomology Section, completed most of the identifications. A few Osmia and halictid bees were identified to species by Dr. Terry Griswold, Research Entomologist with the USDA-ARS Bee Biology and Systematics Laboratory in Logan, Utah. The data on visitors to P. degeneri were summarized by taxon. In most cases, visitors were identified to species, although some insects were only identified to genus 2.5 Visitor Behavior Insect behavior on P. degeneri both on plants and inside flowers, as well as their movement between plants was recorded. These observations help to discern the extent to which different insect species may be transferring and depositing pollen toP. degeneri stigmas. As mentioned previously, correlating pollinator behavior and body pollen deposition aids in determining effective pollinators for P degeneri. For example, a frequent visitor may or may not behave in such a way that promotes the accumulation and transfer of P. degeneri pollen to other P. degeneri flowers and plants. Pollinator observations at the Sunset City site occurred June 18 to July 5 for a total of six days and 21.25 hours, with 15.5 observation hours in the morning and 13.5 observation hours in the afternoon. At Phantom Canyon, observations 11

PAGE 22

occurred on July 1, 10, and 17 for a total of eight afternoon hours. At Locke Park observations occurred July 7-19 for a total of2.5 hours in an afternoon. With the help of a field assistant, I initially noted the number of plants and flowers being observed, and described the habitat and weather conditions. As an insect approached a P. degeneri flower, the insect was identified to the lowest taxonomic level possible using photographic keys. Additionally, the insect was described noting various physical features, including approximate size, color, and hairiness. The number of plants and the number of flowers visited on each plant were noted. Additionally, I observed insects as they entered P. degeneri flowers and noted whether the insect collected pollen and/or nectar, if the insect deposited pollen on the flower's stigma, and how the pollen was deposited (by which body section). For each taxon (described above), means were calculated for number of plants visited, number of flowers per plant, and length of visit. 2.6 Body Pollen In order to accurately identify P. degeneri pollen on bee bodies, I examined pollen from 24 other species that were observed flowering at the three sites. Anthers from these species were collected, preserved in 70% ethanol, and subsequently dried, during which they dehisced. Basic fuchsin gelatin was prepared following Beattie ( 1971) lab procedures in preparation for pollen staining (Appendix A). The gelatin was cut into small squares and, using forceps, dabbed onto the pollen samples following Erdtnab ( 1960, 1969) (Appendix B). 12

PAGE 23

Each pollen reference slide was labeled, examined at 400x using a compound microscope, and photographed with a digital camera. Only pollen from other Penstemon species appeared similar enough to cause difficulty with identification; however, it was possible to discriminate among the three species, with P. degeneri pollen intermediate in size between P. virens and P. virgatus. (Figure 3a-c ). Furthermore, peak anthesis varied among the three species blooming at my research sites. Insect body pollen was analyzed following Tepedino et al. ( 1999, 2006). All bees, wasps, and flies that had been captured were placed in an insect relaxing chamber containing chlorocresol to relax connective tissue. Subsequently, the length of the insects was measured in millimeters using a Starrett l20Am-l50 metric dial caliper. Each insect's head, thorax, and abdomen were measured separately, and these three measurements were summed to calculate the total insect body length. Body pollen was removed from insects by dabbing each of five body parts (including the head, ventral and dorsal sides of the thorax, and ventral and dorsal abdomen) individually using fuchsin gelatin. Due to copious amounts of pollen, it was often necessary to use the forceps or a probe to gently scrape loose the pollen from insect body parts, especially on the face and caudal thorax, before dabbing the gelatin on the area. After pollen was removed from each section, the forceps and probes were cleaned by flaming before gathering pollen from the next insect. The gelatin from each of the five sections was placed l3

PAGE 24

on a glass microscope slide, labeled, and slowly melted using a Bunsen burner. The melted gelatin was covered with a cover-slip, allowed to cool, labeled, and stored at 4 C for future reference. Pollen was removed from only a few flies and wasps, since very little pollen could be seen on their bodies using a dissecting microscope. Pollen collected from each of the five body sections was identified as either P. degeneri or that of another taxon using the aforementioned reference photographs. For each insect, the total amount of P.degeneri pollen on each body section was then nominally categorized as 0 (0-50 pollen grains), 1 ( 50-150 pollen grains), 2 (150-500) or 3 (>500) and recorded. For each insect visitor that I examined, the percentage of individuals was calculated with respect to the relative proportion of body pollen (e.g., copious) on each body section (e.g., head). Using forceps, scopal pollen (when present) was removed from female worker bees and similarly prepared for microscopic examination. The percentage (mean and standard deviation) of P. degeneri pollen in bee scopae was obtained by averaging three counts of 500 pollen grains. For each count, the pollen was identified as P. degeneri or, when possible, to one of the other species of flowering plants observed at the site. A bee that carried >80% P. degeneri scopal pollen was considered to exhibit high fidelity. 14

PAGE 25

3.0 Results 3.1 Phenology In 2007, young plants were first observed on May 7, 17, and 22 at Sunset City, Phantom Canyon, and Locke Park sites, respectively. Flower buds appeared approximately two weeks later at Sunset City, and approximately four weeks later at both Phantom Canyon and Locke Park (Table 1). Flowering, defined as early, middle, and late anthesis occurred over a period oftwo weeks from June 16-30 at Sunset City, from July 1-17 at Phantom Canyon, and from July 6-20 at Locke Park (Table 1 ). Fruit development, including the time of fruit set (late anthesis) to the time at which fruit matured, took approximately five weeks from June 30 to August 7 at Sunset City, and six weeks from July 17 to August 30, and July 20 to September 4 at Phantom Canyon and Locke Park, respectively (Table). During the first day of an thesis, the longer pair of anthers dehisced first, while the shorter anther pair dehisced subsequently, over a period of one to four hours. Once anthers dehisced, there appeared to be very little pollen remaining. As the anthers dehisced, the style remained straight and hidden within a groove along the upper inside surface of the corolla tube (Figure 2a). Within 24 hours, the style began to curve downward toward the inside, lower surface of the flower toward the staminodium (Figure 2b). Within 48 to 72 hours the style was completely bent, with the stigma nearly touching the dehisced anthers and staminodium (Figure 2c). When deprived of pollinators, approximately 38% of P. degeneri flowers set 15

PAGE 26

fruit. These observations were the result of a breeding system experiment performed in 2007 (Carol English pers. obs.). 3.2 Visitors A total of 44 species comprised the guild of visitors for P. degeneri, including one hummingbird (Table 2, Figure 4). Visiting insects included bees representing the Megachilidae (megachilids), Apidae (apids), Halictidae (halictids), and Colletidae (colletids); flies representing the Syrphidae (syrphids) and Bombyliidae (bombyliids); wasps representing the Masarinae (masarines) and Eumeninae (eumenines); and butterflies, including western swallowtails (Papilio, Papilionidae) and skippers (Hesperiidae ). Bees were the most frequent visitors to P. degeneri and made up approximately 73% of all taxa visiting these flowers. Flies made up 23% of all visitors, while wasps and butterflies made up the remaining 4% (Figure 5). Of the bee taxa, 76% were apids, 15% were megachilids, and 7% were halictids. Two Bombus (bombid) species, Bombus centra/is and B. huntii, represented approximately 70% of the apids and 53% of all bees captured on P. degeneri (Figure 5). Variability among sites existed, both for insect abundance and diversity. For example, even though B. centra/is was the most abundant visitor captured on P. degeneri, this species was not found at Sunset City, and nearly twice as many individuals were captured at Phantom Canyon as compared to Locke Park. 16

PAGE 27

Bombus huntii, the second most common visitor to P. degeneri, was found in low numbers at Sunset City and Phantom Canyon, yet in relatively high numbers at Locke Park (Table 2). Halictid diversity and individual abundance decreased with elevation gain, while megachilid and apid diversity and individual abundance increased with increasing elevation gain (Table 2). Bombyliids were observed in approximately equal numbers at all three sites, yet syrphids only occurred at Sunset City (Table 2). In general, few wasp species were seen on P. degeneri flowers, with masarines occurring in low numbers only at the Sunset City site, and eumenines occurring only at Phantom Canyon and Locke Park (Table 2). 3.3 Visitor Behavior A total of 31 hours was spent observing visitor behavior at the three sites, including 21.5, 7.0, and 2.5 hours at Sunset City, Phantom Canyon, and Locke Park respectively. Flower visitors were identified to genus, family, or order. Apids were observed visiting P. degeneri flowers 174 times. On average during an observation period, apids visited 160 plants and 5.3 flowers per plant, with sustained visits of22.6 minutes (Table 3). Megachilids were observed visiting P. degeneri flowers seven times. On average, megachilids visited three plants and 2.5 flowers per plant, with sustained visits of 43 seconds (Table 3). Halictids were observed visiting P. degeneri flowers 19 times. On average halictids visited two plants and two flowers per plant, with sustained visits of 64 seconds (Table 3). The behavior of dipterans was similar to megachilids and halictids, but visit 17

PAGE 28

length was shorter, averaging 27 seconds per plant (Table 3). Butterflies were observed six times, averaging nine plants and two flowers per plant, with sustained visits of two minutes (Table 3). Only one hummingbird was observed visiting flowers; it returned for two 15 minute visits, averaging ten plants per visit and two flowers per plant (Table 3). As mentioned previously, insect behavior also consisted of observing the specific behavior of insects inside and near P. degeneri flowers. As bombids entered the flowers, their heads and caudal thorax contacted both the stamens and stigma, which was correlated with a noticeable change in buzzing. These bees were actively gathering nectar, yet passively gathering pollen. After exiting flowers, they spent many seconds cleaning pollen off their face near the antennae and eyes. They visited and collected nectar from flowers in an area for up to 35 minutes, at both Phantom Canyon and Locke Park. Scopal sacs could be observed increasing in size, until finally the bee would fall toward the ground while attempting to land on a flower (Figure 7a-c). The bees would than fly quickly and directly in a straight path toward the southeast. Another bombid would arrive shortly and begin to work the flowers, seemingly avoiding previously visited flowers. At Phantom Canyon and Locke Park, numerous Geranium caespitosum and Oxytropis /ambertii plants were blooming, along with the P. degeneri. Many bombids were observed landing briefly on G. caespitosum flowers, then quickly 18

PAGE 29

flying back to P. degeneri flowers. Bombids were observed foraging on both 0. lambertii and P. degeneri flowers, yet usually on one or the other species, without switching between the two species in a single foraging trip. Halictids would flip upside down after entering P. degeneri flowers, and stay inside the flowers for up to a minute at a time, while collecting pollen and possibly nectar. The ventral thorax and abdomen would contact the stigma (Figure 6a). Megachilids entered a flower with the dorsal abdomen contacting the stigma. After exiting the flower, they sometimes entered 1-2 additional flowers on the same plants, and then quickly flew off or landed on the ground to groom their bodies (Figure 6b ). 3.4 Body Pollen The pollen wasp Pseudomasaris vespoides was observed and captured very few times and carried little body pollen. Additionally, zero to very few pollen grains were detected on insects including eumenines, collectids and dipterans; therefore, only apids, megachilids, and halictids were examined in detail for P. degeneri body pollen. For most bombids, a majority of the pollen was deposited on the head, with decreasing amounts on the thorax and abdomen, respectively; that was not the case for B. centra/is, in which most of the pollen was deposited on the dorsal thorax (Figure 10). Approximately 73% of B. centra/is bees, the most abundant species captured on P. degeneri, carried moderate to copious amounts of pollen on their heads, while 78% carried moderate to copious amounts 19

PAGE 30

of pollen on their dorsal thorax (Figure 1 0). Forty-six percent maintained copious amounts of P.degeneri pollen on their dorsal thorax, while only 28% maintained copious amounts of pollen on their heads (Figure 11 ). Fewer than 55% of B. huntii bees, the next most abundant species, carried moderate to copious amounts of pollen on their heads, while 27% individuals carried moderate to copious amounts of pollen on their dorsal thorax (Figure 10). Few ofthese individuals maintained copious amounts of pollen on their heads and dorsal thorax, with none in the moderate category (Figures 10, 11). Fifty-three and 36% ofthe 12 Bombus appositus individuals captured, carried moderate to copious amounts of P. degeneri pollen on their heads and dorsal thorax, respectively, with very few individuals carrying copious amounts of pollen on their bodies (Figures 10, 11 ). Other apids, including, B. nevadensis nevadensis, Anthophora Montana, and A. terminalis, carried very little P. degeneri pollen on their bodies (Figures10, 11). Often 100% of the megachilids, the second most abundant family of bees captured on P. degeneri, carried moderate to copious amounts of pollen on their heads, dorsal thorax, and ventral abdomens, in contrast to Bombus (Figures 12, 13). Many of these individuals, often 100%, also carried copious amounts of P. degeneri pollen on their bodies when the moderate category was eliminated (Figures 12, 13). Approximately 60-70% of Osmia brevis, the most abundant megachilid found on P. degeneri, carried moderate to copious P. degeneri pollen on their dorsal thorax and heads, and ventral abdomens (Figure 12). Although 20

PAGE 31

this sample is small, all other megachilids, including Osmia juxta, 0. buchephala, 0. proxima, 0. bruneri, 0. paradisica, and Hoplitus algentifrons, carried moderate to copious amounts of P. degeneri pollen on various body parts (Figure 12, 13). One halictid Lasioglossum sisymbrii, carried moderate to copious amounts of P. degeneri pollen on their bodies, at least in half of the individuals from which data were collected. Fifty-seven percent of all female worker bees that carried greater than 80% P. degeneri scopal pollen were megachilids. The remaining 43% were apids (Figure 8). Bombus centra/is, the most abundant bee species on P. degeneri, carried 3100% P. degeneri scopal pollen (mean= 82.2 28.0, N = 75) (Table 4). The second most abundant species on P. degeneri, B. huntii, carried 12-99% P. degeneri pollen (mean= 78.5% 29.3, N = 46) (Table 4, Figure 9). Forty-two Osmia bees, representing 11 species, made up the next most abundant group of bees captured on P. degeneri. These individuals had scopal loads that ranged from 8-100% P. degeneri pollen (mean= 86.3% 26.8, N = 30) (Table 4, Figure 9). The three most abundant Osmia species (O.brevis, 0. penstemonis, and 0. juxta) carried a high percentage of P. degeneri scopal pollen at 99.8% 0.60, 99.6% 0.58, and 82.6% 35.1, respectively (Table 4, Figure 9). Infrequently captured species, such as 0. albolateralis, 0. proxima, 0. bruneri, 0. densa, and 0. paradisica also carried high percentages of scopa pollen (Table 4, Figure 9). In general, less than expected numbers of halictid bees were captured on P. 21

PAGE 32

degeneri, with Lassioglossum sisymbrii representing the most abundant halictid captured; they carried 4-93% P. degeneri scopal pollen (mean= 67% 42.4, N = 4)) (Table 4, Figure 9). 22

PAGE 33

4.0 Discussion 4.1 Visitor Guild As expected, this research conducted during the 2007 field season revealed that a large variety of insects visited P. degeneri. These visitors represented four orders and ten families, including butterflies (Lepidoptera), flies (Diptera), bees and wasps (Hymenoptera), and one hummingbird (Trochilidae) (Table 2). These results corroborate the findings of other pollination studies, which have repeatedly shown that Penstemon species with blue to purple corollas and wide ventral lobes are primarily visited and pollinated by a wide variety of insects (Beaty et al. 2004, Kimball2008, Spackman 2004, Tepedino et al. 1999, 2006, Wilson et al. 2004). Furthermore, effective pollinator studies by Tepedino et al. ( 1999, 2006) on rare and endemic congeners have shown that only a few bees among the many insect visitors are effective pollinators. Again, the results of my research corroborate the findings ofTepedino et al (1999, 2006), and suggest that even though P. degeneri has a large assortment of insect visitors, only a few bee taxa are effectively pollinating flowers. Tepedino et al (1999, 2006) found that the frequency of visitation was directly correlated with effective pollination, which is also supported herein. Bombus and Osmia bees were the most frequent visitors to P. degeneri in 2007, yet their frequency varied greatly among sites (Figure 5, Table 2). For example, Bombus centra/is was only captured at Phantom Canyon and Locke 23

PAGE 34

Park. Since B. centra/is is known to occur typically above 2133m elevation, its absence from lower elevation sites is not surprising (Kearns 1993, V. Scott, pers. Comm.). And while I expected to encounter Osmia bees at similar frequencies across the elevation range, they were observed and captured predominantly at the Locke Park site in 2007. As there was a relatively high frequency of both blooming plants and large bee visitors, especially in genus Bombus (Table 2), I may have simply overlooked the smaller Osmia bees during my fieldwork at the other sites, which was conducted earlier in the season and before my skills had improved. Interestingly, preliminary results from 2008 reveal a relatively high frequency and diversity of Osmia bees on P. degeneri at all three study sites. This variation is not surprising and, to some extent, matches results from other studies on effective pollination (Tepedino et al. 1999, 2006). Whereas our study found Bombus to be the most frequent visitor followed by Osmia, Tepedino et al. (1999, 2006) found Osmia to be the most frequent visitors to Penstemon pendlandii and P. haydenii. However Bombus were not frequently captured by Tepedino et al. (1999, 2006) and, therefore, relative frequency of Bombus to Osmia was not calculated. Less frequently encountered insect visitors to P. degeneri in 2007 included dipterans, halictids, eumenines, and masarines. Nearly 25% of all visitors were bombyliids and, for this reason, we considered these flies as possible effective pollinators (see section 4.2). Halictids were frequently observed and captured at 24

PAGE 35

the Sunset City site, yet they were infrequent at the other two sites. Since these bees are quite small when compared with the large frequently captured bombids, we either simply overlooked these tiny bees (as was suggested for megachilids) during capturing sessions, or they were missing at both the Phantom Canyon and Locke Park sites in 2007. Unexpectedly, and in contrast to most other Penstemon visitation studies, few Pseudomasaris vespoides (masarines) were observed or captured on P. degeneri in 2007 (Table 2). Pseudomasaris vespoides is a well known oligolectic pollen wasp that feeds on Penstemon pollen and is typically reported to be a frequent visitor to other, similar Penstemon species in western North America (Gess 1996, Beaty et al., 2004, Kimball 2008, Spackman 2004, Tepedino et al., 1999, 2006, Wilson et al. 2004). Interestingly, masarines were some of the most frequent visitors at both the Sunset City and Phantom Canyon study sites in 2008 (C.T. English pers. obs.). It is unknown why this variation occurred. 4.2 Effective Pollinators Insect visitors to Penstemon degeneri were considered effective pollinators if they met four predetermined criteria: frequent visitation of P. degeneri flowers, contact with stamens and stigmas during their visit, accumulation of moderate to copious amounts of P. degeneri body pollen, and high fidelity (faithfulness) toP. degeneri pollen. 25

PAGE 36

Many visitors did not fulfill enough of the aforementioned criteria to be considered effective pollinators. For example, P. vespoides was not considered an effective pollinator in 2007 due to low visitation rates and small amounts of P. degeneri body pollen (Table 2). I am currently examining additional specimens captured in 2008 to determine their effectiveness in terms of body pollen. Even though bombyliids were frequent visitors, they were not considered effective pollinators in 2007 due to behavior and body pollen observations that indicated the flies were obtaining nectar without accumulating or transferring pollen (C.T. English pers. obs.). As with P. vespoides, I am examining bombyliids captured in 2008 for P. degeneri body pollen to determine their effectiveness. Halictids were relatively frequent visitors only at the Sunset City site. Unfortunately many of the halictids captured were accidentally washed with ethanol, thus eliminating body pollen from these specimens. A few Lassioglossum manitouellum were observed to carry enough P. degeneri pollen to be considered effective pollinators, yet their numbers were too low to be considered effective pollinators in 2007 (Table 4). Finally a variety of apids, including bees in the genera Anthophora and Bombus (including, B. nevadensis nevadensis, B. appositus, and B. griseocollis) had high fidelity toward P. degeneri, yet carried very little body pollen (Tables 4, Figures 9, l 0, II). Therefore, masarines, bombyliids, halictids and the various apids previously mentioned were eliminated as effective pollinators in 2007, since they did not meet more than one, and in some cases none, of the four criteria. 26

PAGE 37

Of the remaining Bombus and Osmia bees, several Osmia species were the most effective pollinators found on P. degeneri during the 2007 field season. At Locke Park, these bees were frequent visitors, second only to Bombus centra/is (Table 2). Often 100% of the Osmia examined carried copious amounts of P. degeneri body pollen (Figure 13). Behavior observations verified that Osmia bees were contacting P. degeneri stigmas at the areas they accumulated P. degeneri pollen on their bodies. Additionally, Osmia had a high percentage of P. degeneri pollen in their scopae. Some species, including 0. brevis, 0. penstemonis, 0. proxima, 0. brnneri, and 0. paradiscia, carried I 00% P. degeneri pollen in their scopae. Therefore Osmia bees had the highest fidelity rates of any other bee studied in 2007 (Table 4, Figure 9). These findings support studies by Tepedino et al (1999, 2006) and by Crosswhite and Crosswhite (1966) suggesting that some Osmia bee species may depend entirely on Penstemon pollen for nutrition. Bombus centra/is and B. huntii are also considered effective pollinators, since they met all four criteria. However, even though these two species were the most frequent visitors to P. degeneri in 2007, visitation among sites was highly variable (Table 2). Furthermore, visitation was extremely low in the subsequent 2008 field season, suggesting that Bombus may be both spatially and temporally unreliable. Furthermore, Bombus species did not meet theother effective pollinator criteria, including P. degeneri body pollen accumulation and fidelity rates, with as much strength as Osmia species. For example, Bombus carried less P. degeneri body 27

PAGE 38

pollen in comparison to Osmia. Whereas 80% of B. centra/is individuals carried only moderate to copious (150-500 pollen grains) amounts of P. degeneri body pollen, I 00% of 0. buchephala, 0. bruneri and 0. paradisica bees carried copious (greater than 500 pollen grains) amounts of P. degeneri body pollen (Figures 10, 11, 12, 13). Additionally, Bombus species carried less P. degeneri pollen in their scopae and, therefore, had lower fidelity rates in comparison to Osmia species (Table 4, Figure 9). Except for the near absence of P. vespoides, these results meet our expectations, and are in accordance with other effective pollinator studies on rare endemic Penstemons (Tepedino et al. 1999, 2006). Among P. degeneri 's large visitor guild only a few Osmia and Bombus species can be considered effective pollinators, although the degree of pollinator specialization is still unknown. These data seem to suggest that P. degeneri is an asymmetric generalist, in which the plants are pollinated by an assortment of specialist and generalist pollinator taxa (Kimball 2008). In contrast, asymmetric specialists are plants that are pollinated by only a few generalist pollinator taxa (Ashworth 2004). In the face of habitat fragmentation and the possible loss of specialist pollinators, it has been shown that asymmetric generalists and asymmetric specialists have similar responses regarding reproductive success. In theory, even if the specialist pollinators are not present within a fragmented habitat, the generalists assume the major role of pollination (Ashworth 2004). Certainly this study has shown that Osmia species appear to be superior 28

PAGE 39

pollinators in contrast to Bombus species. Yet, without further study we do not know if and how much reproductive success would be lost in the absence of Osmia pollinators. Additionally, due to spatiotemporal pollination differences described in this study (elevation differences), P. degeneri 's habitat is fragmented and subject to variability with respect to the occurrence of specialist pollinators. The relatively low frequency of Osmia at two P. degeneri sites in 2007, may have been compensated for by the frequent occurrence of Bombus generalists. It is noteworthy that P. degeneri is able to self pollinate-flowers deprived of pollinators were observed to set fruit in 2007 (Carol English pers. obs.). Furthermore, preliminary pollen/ovule ratios for P. degeneri obtained for the Sunset City, Phantom Canyon and Locke Park sites (Angela Schultz and Erin Lowther unpublished data) suggest that P. degeneri is facultative xenogamous, with a mixed, yet predominantly outcrossing breeding system (Cruden 1977). Therefore, successful fruit set in P. degeneri likely results from a combination of outcrossing effected by bees and self-pollination in the absence of effective pollinators. 29

PAGE 40

4.3 Future Studies Since completing this study, additional insect visitors have been collected and are currently being examined for body pollen. These data will contribute to the current research and offer a broader perspective regarding which visitors are consistently and effectively pollinating P. degeneri. In addition to focusing on bees, the wasps, butterflies, and bombylliids will also be evaluated for pollinator effectiveness. These results will be coupled with additional behavior observations to determine how many flowers per inflorescence, inforescences per plant, and plants per observation area are visited by different insect visitors (Table 3). Along with identifying the effective pollinators for P. degeneri, it is also desirable to study breeding system and reproductive allocation. A reproductive assessment was conducted in 2008, while a breeding system study is planned for the 2009 research season. Objectives include estimating overall fruit and seed set during one reproductive season, and determining the breeding system used by P. degeneri for successful fruit set. These results, in addition to the pollen/ovule ratios that are being calculated, will help to clarify the specific breeding system used by P. degeneri (Tepedino et al., 1999, 2006, Cruden 1977, Keams 1993). These data will also facilitate conclusions regarding reproductive limitation. If P. degeneri populations are being effectively pollinated by bees, then we expect that 30

PAGE 41

there will not be a significant difference between the fruit set of open pollinated controls and hand-pollinated plants. 31

PAGE 42

Figure 1: Locations for three Penstemon degeneri Crosswhite research sites: Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), in Fremont County, Colorado. 32

PAGE 43

Table 1: Penstemon degeneri Crosswhite phenology at three Colorado sites in 2007, including Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), Fremont County. Elevation is specified in meters. Site and Rosette approximate emergence elevation sse 2133.6 PC 2438.4 LP 2804.16 517 5/17 5/22 Floral budding 5/23 6/16 6120 Early anthesis 6/16 7/1 7/6 33 Mid anthesis 6/22 7/10 7/13 Late anthesis 6/30 7/17 7/20 Matu e frui 8/7 8/31 9/4

PAGE 44

Figure 2a-c: Floral morphology and receptivity in Penstemon degeneri Crosswhite: (a) straight style, (b) slightly curved style still situated within groove on upper inside surface of flower, and (c) fully curved style almost touching staminode hairs. 34

PAGE 45

a) b) c) Figure 3a-c: Pollen grains viewed at 400X from: (a) Penstemon virens Pennel ex Rydb., (b) P. degeneri Crosswhite, and (c) P. virgatus A. Gray. 35

PAGE 46

Table 2: All visitors captured on Penstemon degeneri Crosswhite in 2007 at the Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP), Colorado sites in 2007. Numbers with a slash indicate females/males. Site sse PC LP Megachilidae Hoplitis albifrons argentifrons (Cresson 1864) 1 H. truncata mescalerium (Cockerell1910) I H. producta (Cresson 1864) 2 Megachile melanophaea (Smith 1853) 1 Osmia brevis (Cresson 1864) 1 16 0. bucephala (Cresson 1864) 5 0. pentstemonis (Cockerell 1906) 5 O.juxta (Cresson 1864) 6 0. albolateralis (Cockerell 1906) 2 0. proxima (Cresson 1864) 2 0. bucephala (Cresson 1864) 4 0. bruneri (Cockerell 1906) 1 0. densa (Cresson 1864) 1 0. nanula (Cockerell 1906) 1 0. paradisica (Sandhouse 1924) 1 Halictidae Agapostemon sp. 2 Lassioglossum sp. 8 Dialictus sp. 3 5 Halictus sp. 2 36

PAGE 47

Table 2: continued Apidae Melissodes sp. 1 Bombus centra/is Cresson 1864) 69/5 37 B. fervidus (Fabricius 1798) 0/2 1 B. appositus (Cresson 1878) 2 5/1 9 B. huntii (Greene 1860) 9 8/1 46 B. nevadensis nevadensis (Cresson 1874) 1 16 B. spp. 3 7 B. bifarius (Cresson 1864) 2 1 B. morrisoni (Cresson 1864) B. griseocollis (DeGeer 1773) B. rofocinctus (Cresson 1863) 1 Anthophora montana (Cresson 1869) 5 4/1 1 A. terminalis (Cresson 1869) 1 3/2 A. ursina (Cresson 1869) 1 Diadasia sp. 3 Apis mellifera (Linnaeus 1758) Colletidae Hyaleus Hesperiidae Papi1ionidae 2 3 5 Papilio rotulus Diptera 1 Bombyliidae 8 3 5 Syrphidae 4 Eumeninae Masarinae Pseudomasaris vespoides 2 Trochilidae 1 Total Species at each site 23 23 24 Total number of species at all three sites 44 37

PAGE 48

oApidae D ip! era o Megachildae o Halictidae Vespidae o Lepidoptera Colletidae o Trochilidae Figure 4: Visitors captured on Penstemon degeneri Crosswhite at three Colorado sites in 2007; organized by family or order. 38

PAGE 49

7.30% Bombus centralis/huntii Apidae o Megachilidae. 53% o Halictidae Figure 5: Bees captured on Penstemon degeneri Crosswhite at three Colorado sites in 2007; organized by family or order, with Bombus centra/is and B huntii (Apidae) grouped to indicate their relative abundance. 39

PAGE 50

Table 3: Behavior of visitors of Penstemon degeneri Crosswhite at three Colorado sites in 2007. Taxon N Average number Average number Average of plants visited of flowers visited overall per plant length of visit Apidae 174 160 5.3 22.6 min Halictidae 19 2 2 64 sec Megachiledae 7 3 2.5 43 sec Diptera 14 3 3 27 sec Lepidoptera 6 9 2 120 sec Trochilidae 1 10 2 15 min 40

PAGE 51

a) b) c) Figure 6a-c: Bee behavior on Penstemon degeneri Crosswhite flowers: (a) halictid (Halictidae) bee entering a flower inverted, remaining up to 60 seconds, and actively collecting pollen and possibly nectar; (b) Osmia (Megachilidae) bee about to enter a flower; (c) Bombus (Apidae) bee with large scopal (pollen) sac collecting nectar. 41

PAGE 52

a) b) c) Figure 7a-c: Progression over 30-45 minutes depicting a female worker bee, Bombus huntii, on Penstemon degeneri Crosswhite collecting and accumulating pollen and nectar in her scopae: (a) early ( 5-l 0 mins) during the visit, with relatively small amounts of pollen in the scopal sac; (b) following approximately 15 mins. of pollen and nectar collecting, with larger scopal sac; (c) late (30-45 mins.), with very large scopal sac making it difficult for the bee to hang onto the flower. 42

PAGE 53

Bee families carrying > 80/o P. degeneri Scopae Pollen Apidae 43% Megachilidae 57% Figure 8: Percent megachilid (Megachilidae) and apid (Apidae) bees that carried greater than 80% Penstemon degeneri Crosswhite pollen in their scopae at three Colorado sites in 2007. 43

PAGE 54

100 110 eo 70 eo 50 40 30 20 10 0 Osmia brevis NE10 Osmia penstemonis N-3 C Osmia pnlldrM N=2 COsmla bruner! N=1 Osmia pel8dlslcll N=1 Dl811dls sp N=1 Osmia dens8 N-1 CAnlllophon tennlnalls N=5 oallctus spp N l.asloglossum helerorhinum N= 1 [] HopiKis pn>ducta N=1 eombus spp N-e Bombus Dllllralls N=ll Osmla Juxia N-5 Osmia albolat8!alis N=2 Bombus grlsaocollis N=1 Bombus appositus N=13 c Bombus huntll N=ll8 c Lassloglossum slsymbr11 N=4 [] Bombus nevadensis nevadensls N=4 II Anlhophora montana N-8 Hallctus lr1pertllus N=2 Bombus monlsonll N CBombus fervldus N=1 Bombus biflllis N=1 Osmia nanula N-1 HopiKis alblfrons argenlifrons N=1 c HopiKis truncma mescalertum N=1 Osmla bucephala Dl811ctus sp N=1 eombus rufocinctus N-1 Figure 9: Percent Penstemon degeneri Crosswhite scopal pollen found on 31 species of female worker bees captured at three Colorado sites in 2007. 44

PAGE 55

Table 4: Hymenoptera species collected from Penstemon degeneri Crosswhite at three Colorado sites in 2007, including: number collected at the Sunset City (SSC), Phantom Canyon (PC), and Locke Park (LP) sites, and in total (NC); number collected with pollen loads (NP); and mean percentage (SO) of P. degeneri pollen in the pollen loads (% PO). Includes total number of species captured at each site and time spent capturing. sse PC LP NC NP %PD Famill: Megachilldae Hoplitis producta (Cresson 1864) 2 2 I 94 H. albifrons argentifrons (Cresson 1864) 2 I 10 H. truncatamescalerium (Cockerell1910) I I 8 Osmia brevis (Cressonnl854) 16 18 II 99.8.60 0. pentstemonis (Cockerell 1906) 5 5 3 99.6.58 O.juxta Cresson (Cresson 1864) 6 6 5 82.6.1 0. albolateralis (Cockerell 1906) 2 2 I 83 0. proxima (Cresson 1864) 2 2 2 100 0. bucephala (Cresson 1864) 4 0. bruneri (Cockerell 1906) I I I 100 0. dens a (Cresson 1864) I I I 96 0. nanula (Cockerell 1906) I 2 I 20 0. paradisica Sandhouse (1924) I I I 100 0. gaudiosa (Cockerell 1906) I I I 20 Megachi/e melanoe_haea 1853} I Famill: Halictidae Agapostemon sp. 2 Halictus tripartitus (Cockerell 1895) 2 2 Lassioglossum sisymbrii (Cockerell 1895) 2 2 L. heterorhinum (Cockerell 1895) 6 6 4 67.4 L. manitouellum (Cockcrcll1895) I I I 95 Dialictus sp_. I I Famill: Apis melifera (Linnaeus 1758) I Bombus. griseocol/is (DcGcer 1773) I I I 81 B. appositus Cresson 2 6 9 17 13 81.5 B. bifarius (Cresson 1878) 2 I 3 3 21.6.2 B.nevadensis nevadensis (Cresson 1878) I 13 14 4 6.3.7 B. centra/is (Cresson 1878) 74 37 Ill 75 82.2.0 B. huntii (Greene 1860) 9 8 45 62 46 78.5.3 B. rufocinctus (Cresson 1878) I I 5 45

PAGE 56

Table 4 continued B.fervidus (Fabricius 1798) 2 2 B. morrisonii (Cresson 1878) 1 1 1 25 B.spp 10 10 9 88.4 Anthophora Montana (Cresson 1869) 5 6 I 12 7 61.4.6 A. ursina (Cresson 1869) I I A. terminalis (Cresson 1869) 5 6 4 76.2 Diadasia rinconis (Cockerell 1897) 2 3 Melissodes Sf!J!: 2 2 Colletidae Hr_laeus annulatus 1758} Eumeninae I 2 Masarinae 2 ..... 2 Total Species 17 16 22 Total 720/3 720/3 720/3 46

PAGE 57

Bomtu; BarnlluiiUIIi Borntu........ AIOiphOr'l,....... Anltlaptwn lamlu 14.7rrwn) (N-42, 15 ......... ) (N-12. 11.5mm) (N-3, 17.8mm) t.rmtnlll (N-2, (N-7, 24.2mm) 13.Dmm) Figure 10: Percent apid (Apidae) bees caught at three Colorado sites in 2007 with moderate to copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA=ventral abdomen. Includes number of individuals captured (N) and average size. 47

PAGE 58

a.mtu ....... (N-SI, Born1u hldll (N"'. 1S.IIIwn) ...._IIJIPDIIU (*'2. AnllaptWii ...._ (N-2. 1.?nn) 11.5mm) 11-lrml) Figure 11: Percent apid (Apidae) bees caught at three Colorado sites in 2007 with copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and V A=ventral abdomen. Includes number of individuals captured (N) and average size. 48

PAGE 59

120'11 ,_ .H oT OVT ODA .v ... -20'11 0% Figure 12: Percent megachilid (Megachilidae) bees caught at three Colorado sites in 2007 with moderate to copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and V A=ventral abdomen. Includes number of individuals captured (N) and average SIZe. 49

PAGE 60

100% 1 [ 1 Figure 13: Percent megachilid (Megachilidae) bees caught at three Colorado sites in 2007 with copious amounts of Penstemon degeneri Crosswhite pollen on their body, where H=head, DT=dorsal thorax, VT=ventral thorax, and VA=ventral abdomen. Includes number of individuals captured (N) and average size. 50

PAGE 61

APPENDIX A Fuchsin Gelatin Recipe (as suggested by Beattie, A. J. 1971. Pan-Pacific Entomologist 47: 82) Ingredients needed: -distilled water, 175 ml, -glycerin, 150ml, -gelatin, 50 g, -crystalline basic fuchsin as desired enough to make solution the "color of a fine claret" -crystalline phenol, 5 g important in humid environments, may be left out if gelatin can be refrigerated. Fuchsin Gelatin Procedure: NOTE: Do not touch gelatin if phenol is used. 1. Add the gelatin to the distilled water in a beaker and heat until the gelatin dissolves 2. Add the glycerin 3. Add phenol, if desired 4. Add basic fuchsin crystals a few at a time until the solution is the color desired. Too light will not stain the pollen, but too dark may mask details of the pollen 5. Filter the solution through cheesecloth 6. Pour into sterile Petri plates that can be covered. If phenol is not used, refrigerate the plates and slides. Plates will keep about a month without refrigeration. 51

PAGE 62

APPENDIX B Glycerin Jelly Slides (Erdtnab 1960, 1969) 1. Clean two slides and one cover slip per sample (e.g. wash with detergent, rinse in 70% ethanol). 2. If the pollen sample is stored in 70% ethanol, put a drop of liquid glycerin on the slide and mix in a drop of pollen suspension. Let the mixture sit and evaporate for about 1 0 minutes. 3. Clean the point of a dissecting needle by washing it in alcohol or heating it in a flame. 4. Spear a small piece (about 2mm3 ) of solid glycerin jelly on the point of the needle. 5. Touch the piece of jelly to the glycerin-pollen mixture on the slide to pick up pollen on the jelly. 6. Place the piece of jelly with the pollen on it onto the second clean slide, about 2.5cm from one end of the slide (leaving room for a label at the other end). 7. Warm the slide gently on a hotplate or over a flame (Bunsen burner or alcohol lamp) until jelly melts. Do not allow the glycerin jelly to boil, which would introduce air bubbles that are difficult to remove. 8. Place the slide under a microscope and spread the pollen grains out while stirring the jelly carefully with a clean needle. 9. Re-warm the slide gently, and lower the cover slip carefully over the molten jelly. The jelly should make a circle 2-3 mm in diameter. I 0. While the slide is still warm, place it face down on a stand that supports it at both ends, without letting anything touching the cover slip while the jelly is still molten. Label and refrigerate the slides. 52

PAGE 63

REFERENCES Ashworth, L., R. Aguilar, L. Galetto, and M. Adrean Aizen. 2004. Why do pollination generalist and specialist plant species show similar reproductive susceptibility to habitat fragmentation? Journal of Ecology. 92: 717-719 Beatty, B., W. Jennings, and R. Rawlinson. 2004. Penstemon degeneri Crosswhite (Degener's Beardtongue): A Technical Conservation assessment. Prepared for the USDA forest Service, Rocky Mountain Region, Species Conservation Project. Denver, Colorado. Castellanos, M. C., P. Wilson, S. J. Keller, A. D. Wolfe, and J. D. Thomson. 2006. Anther evolution: pollen presentation strategies when pollinators differ. American Naturalist. 167: 288-296. Crosswhite, F. S., C. D. Crosswhite. 1966. Insect pollinators of Penstemon series Graciles (Scophulariaceae) with Notes on Osmia and other Megachilidae. American Midland Naturalist. 76: 450-467. Crosswhite, F. S. 1967. Revision of Penstemon section Penstemon (Scrophulariaceae) II. A Western Alliance in Series Graciles. American Midland Naturalist. 77: 28-41. Cruden, R. W. 1977. Pollen-Ovule Ratios: A conservative indicator of breeding systems in flowering plants. Evolution. 31: 32-46. 53

PAGE 64

Dieringer, G., and L. R. Cabrera. 2002. The interaction between pollinator size and the bristle staminode of Penstemon digitalis (Scrophulariaceae ). American Journal of Botany. 89:991-997. Gess, S. K. 1996. The Pollen Wasps, Ecology and Natural History of the Masarinae. Harvard University Press. USA. Keams, C. A. 2001. The Natural History of Bumblebees, A sourcebook for Investigations. University Press of Colorado. Boulder, Colorado. Keams, C., and W. Inouye. 1993. Techniques for Pollination Biologists. University Press of Colorado. Niwot, Colorado. Kimball, Sarah. 2008. Links between floral morphology and floral visitors along an elevational gradient in a Penstemon hybrid zone. Oikos. 117: 10641074. Lesica, P., R. Yurkewycz, and E. E. Crone. 2006. Rare plants are common where you find them. American Journal of Botany. 93: 454-459. Michener, C. D. 2000. The Bees of the World. The John Hopkins University Press. Baltimore, Maryland. Spackman Panjabi, S. C. 2004. Visiting insect diversity and visitation rates for seven globally-imperiled plant species in Colorado's middle Arkansas Valley. Colorado Natural Heritage Program College of Natural Resources, Colorado State University, Ft. Collins, Colorado. Prepared by 54

PAGE 65

the Native Plant Conservation Alliance and the National Fish and Wildlife Foundation. Tepedino, V.J., S.D. Sipes, and T. L. Griswold. 1999. The reproductive biology and effective pollinators of the endangered beardtongue Penstemon penlandii. Plant Systematics Evolution 219: 39-54 Tepedino, V. J., W.R. Bowlin, and T. L. Griswold. 2006. Pollination Biology of the Endemic Blowout Penstemon (P. haydenii S. Wats.: Scrophulariaceae) in Nebraska. Journal of the Torrey Botanical Society 133: 548-559 Tepedino, V. J., T. R. Toler, B. A. Bradley, J. L. Hawk, T. L. Griswold. 2006. Pollination biology of a disjunct population of the endangered sandhills endemic Penstemon haydenii S. Wats. (Scrophulariaceae) in Wyoming, USA. Plant Ecology 193: 59-69 Wilson, P., M. C. Castellanos, J. N. Hogue, J. D. Thomson, and W. S. Armbruster. 2004. A multivariate search for pollination syndromes among Penstemons. Oikos 104: 345-361 Wolfe, A., C. P. Randle, S. L. Datwyler, J. J. Morawetz, N. Arguedas, and J. Diaz. 2006. Phylogeny, taxonomic affinities, and biogeography of Penstemon (Plantaginaceae) based on ITS and cpDNA sequence data. American Journal of Botany 93: 1699-1713 55