REPRODUCTIVE BIOLOGY OF THE ENDANGERED PLANT,
PHLOXHIRSUTA (E.E. NELSON)
by
JULES J. FILIPSKI
A thesis submitted to the Department of Biology
and the Graduate School of Southern Oregon University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
in
SCIENCE
Ashland, Oregon
2005
THESIS APPROVAL PAGE
Approved:
/// 0 /.(5S
Date
Dr. Carol Ferguson, 5Xhair, Thesis Committee
Date
Dr. Steven Jessip, Committee Member
I(
I
/29ROS~
/I
Date
a
DateL/Eg
Dar.Chre
- Wld Comte M
Dr. Charles Welden, Committee Member
Dr.
ii
ACKNOWLEDGEMENTS
The author wishes to thank Dr. Carol Ferguson, who acted as advisor to this
research, and thesis committee members, Dr. Steven Jessup and Dr. Charles Welden.
Dr. Darlene Southworth assisted with pollen identification. Dr. John Sollinger,
and Dr. John Roden offered assistance and support and various aspects of this research.
Steve Kwiatkowski (Biology Lab Supervisor), Harold Berninghausen (Lab Technician),
Martin Lenz, Dr. Bitty Roy (Associate Professor, Center for Ecology and Evolutionary
Biology, University of Oregon), Pete Figura (Environmental Scientist for the California
Department of Fish & Game), and Nadine Kanim (U.S. Fish & Wildlife Service)
deserve recognition for their assistance.
Special thanks are due to Kathleen Donham for her assistance with insect
handling and identification, and for her companionship in the field.
iii
ABSTRACT OF THESIS
REPRODUCTIVE BIOLOGY OF THE ENDANGERED PLANT,
PHLOXHIRSUTA (E.E. NELSON)
by JULES J. FILIPSKI
Phlox hirsuta is an endangered serpentine endemic plant with a limited
distribution in Siskiyou County, California. The overriding goal of this research was to
contribute to a better understanding of the life history, especially the reproductive
strategies, of P. hirsuta. The major objectives of this research were to determine the
mating system of P. hirsuta and the role of pollinators in seed set. No research on the
mating system of P. hirsuta existed prior to this research. A pollinator exclusion
experiment and insect observations were performed at two sites for two consecutive
seasons. Seeds were collected from treatment plants and seed viability and germination
rates were determined. In addition, a hand-pollination experiment was carried out in the
second season in order to better define the mating system(s) used by P. hirsuta. Results
indicate that P. hirsuta is a primarily outcrossing species and produces a high amount of
viable seed, which germinates readily in the lab. The major insect visitors to P. hirsuta
were Lepidopterans but they were infrequently observed on the plants.
iv
TABLE OF CONTENTS
CHAPTER
PAGE
I.
INTRODUCTION .........
II.
METHODS
.....................
1
Site Description . . . . . . . . . . . . . . . . 7
China Hill . . . . . . . . . . . . . . . . 8
Cracker Gulch . . . . . . . . . . . . . . 10
Determining Mating System . . . . . . . . . . . 11
Pollinator Exclusion Experiment . . . . . . . 12
Hand Pollination Experiment . . . . . . . . 14
Insect Observations . . . . . . . . . . . . . . . 16
Pollen Identification . . . . . . . . . . . . . . .1 8
Seed Viability Testing .......
III.
...............
19
RESULTS
Phenology .....
.........
.....
.....
22
Pollinator Exclusion Experiment . . . . . . . . . . 27
Hand Pollination Experiment . . . . . . . . . . . 34
Pollen Identification . . . . . . . . . . . . . .
Insect Observations
36
. . . . . . . . . . . . . . . 40
Seed Traits and Viability . . . . . . . . . . . . . 46
v
CHAPTER
IV.
PAGE
DISCUSSION
Phenology ...........................
....
49
Mating System . . . . . . . . . . . . . . . .
50
Pollen Types . . . . . . . . . . . . . . . . .
62
Seed Traits, Viability, and Germination . . . . . . . 63
Conclusions ...............................
66
V.
LITERATURE CITED . . . . . . . . . . . . . . . . . 69
VI.
APPENDIX A: PLANT COMMUNITIES . . . . . . .
VII.
APPENDIX B: POLLEN REFERENCE COLLECTION
vi
..
76
.
.
79
LIST OF TABLES
PAGE
Table 1. Treatment and Site Comparisons of Treatment and Site Comparisons
of Flower, Fruit, and Seed Set. . . . . . . . . . . . . . . . 33
Table 2. Bloom Periods of Plants in Coincidental Bloom with P. hirsuta at
China Hill. . . . . . . . . . . . . . . . . . . . . . . . 37
Table 3. Bloom Periods of Plants in Coincidental Bloom with P. hirsuta at
Cracker Gulch. . . . . . . . . . . . . . . . . . . . . . . 38
Table 4. Insects Observed Probing P. hirsuta Flowers at China Hill. . . . . . . 42
Table 5. Insects Observed Probing P. hirsuta Flowers at Cracker Gulch. . . . . .42
Table 6. Lepidopterans Collected from China Hill with Phlox Pollen. . . . . . . 45
vii
LIST OF FIGURES
PAGE
Figure 1.
Phlox hirsuta Plants at China Hill, Yreka, California. ... . . . . . .6
Figure 2.
Phlox hirsuta Flowers. . . . . . . . .
Figure 3.
Distribution of Known Occurences of Phlox hirsuta (Yreka Phlox),
Siskiyou County, California (USFWS 2004). . . . . . . . 9
Figure 4.
Percent of Floral Bloom in Unbagged and Bagged Plants
at China Hill in 2004. . . . . . . . . . . . . . . . . 24
Figure 5.
Percent of Floral Bloom in Unbagged and Bagged Plants
at China Hill in 2005. . . . . . . . . . . . . . . . . 24
Figure 6.
Percent of Floral Bloom in Unbagged and Bagged Plants
at Cracker Gulch in 2004. . . . . . . . . . . . . . . .25
Figure 7.
Percent of Floral Bloom in Unbagged and Bagged Plants
at Cracker Gulch in 2005. . . . . . . . . . . . . . . .25
Figure 8.
Percent of Floral Bloom at China Hill and Cracker Gulch, 2004.
.
Figure 9.
Percent of Floral Bloom at China Hill and Cracker Gulch, 2005.
. . 26
Figure 10.
Mean Number of Flowers, Fruits, and Seeds in Unbagged and
Bagged Plant Treatments at China Hill in 2004. . . . . . . 29
Figure 1.
Mean Number of Flowers, Fruits, and Seeds in Unbagged and
Bagged Plant Treatments at China Hill in 2005. . . . . . . 29
Figure 12.
Mean Number of Flowers, Fruits, and Seeds in Unbagged and
Bagged Plant Treatments at Cracker Gulch in 2004. . . . . 30
VI .i
. . . . . . . . . . .6
. 26
Figure 13.
Mean Number of Flowers, Fruits, and Seeds in Unbagged and
Bagged Plant Treatments at Cracker Gulch in 2005. . . . . 30
Figure 14.
Mean Number of Flowers, Fruits, and Seeds in Unbagged Plants
from China Hill and Cracker Gulch in 2004. . . . . . . . 31
Figure 15.
Mean Number of Flowers, Fruits, and Seeds in Unbagged Plants
from China Hill and Cracker Gulch in 2005. . . . . . . . 31
Figure 16.
Mean Number of Flowers, Fruits, and Seeds of Four Treatments in
the Hand Pollination Experiment. . . . . . . . . . . . 35
Figure 17.
Pollen Grains of Phlox hirsuta and Phlox speciosa as Seen Under
a Light Microscope at 63X. . . . . . . . . . . . . . .39
Figure 18.
Pollen Grains of Gilia capitata and Linanthus sp. as Seen Under
a Light Microscope at 63X. . . . . . . . . . . . . .39
Figure 19.
Phlox hirsuta Seedlings in the Southern Oregon University
Greenhouse. . . . . . . . . . . . . . . . . . . . . 48
ix
INTRODUCTION
Phlox hirsuta (E.E. Nelson) (Polemoniaceae) is a semi-woody perennial
endemic to serpentine soils. The Polemoniaceae family consists of 26 genera
(approximately 400 species) of perennials and annuals. Phlox is the largest genus in
the family. The species status of P. hirsuta had been in flux since its first description
in 1899. However, it has been considered a separate species since 1993 (Patterson &
Wilken).
Phlox hirsuta has a cespitose growth habit and ranges in height from 3-12 cm
(Figure 1). The inflorescence is an open to compact cyme with one to several bisexual
flowers. The tubular calyx consists of five green linear sepals that arel0-1 5mm long,
2-3 mm wide, and densely covered with glandular hairs. The corolla is salverform
(having a long slender tube with spreading limbs) and radially symmetrical. Five
petals, white to bright pink in color, unite to form a corolla tube, 10-13 mm long and
1-1.5 mm wide (Figure 2). There is a dense band of cilia at the base of the inner
corolla tube, just above the ovary. Nectar is produced by a glandular disk and collects
in the base of the corolla tube (Grant & Grant 1965).
Five stamens are irregularly inserted in the corolla tube and are not exerted
above the tube entrance. Anthers (1-2 mm long) dehisce longitudinally and produce
yellow pollen. Pollen morphology in the Polemoniaceae family is highly
I
2
variable and species specific. Pollen characteristics of a number of Phlox species have
been described (Stuchlik 1967, Erdtman 1966, Taylor & Levin 1975, Moore et al.
1991). However, P. hirsuta pollen has not been described to date.
The syncarpous (attached to calyx) pistil is 5-7 mm in length (thus located
halfway down corolla tube), and the style terminates with three stigmas (< 1 mm).
The superior ovary (1.5 mm long and 1 mm wide) has three united carpels with one
ovule per carpel. The fruit is a capsule. Detailed species descriptions of P. hirsuta
can be found in The Genus Phlox (Wherry 1955), The Jepson Manual (Patterson &
Wilken 1993) and Appendix II of The Draft Recovery Plan for Phlox hirsuta (Yreka
Phlox) (USFWS 2004).
Phlox hirsuta only grows on serpentine soils. Serpentine soils are derived
from ultramafic parent rocks (serpentinite and peridiotite). There is substantial
variation in mineral and nutrient content in serpentine soils from different sites due to
varied composition of the parent rocks (Walker 1954). Compared to nonserpentine
soils, serpentine soils contain high amounts of magnesium and iron, low amounts of
exchangeable calcium, and high amounts of heavy metals, such as nickel and
chromium (Walker 1954). Serpentine soils generally have low levels of nitrogen and
phosphorous (Gordon & Lipman 1926).
Phlox hirsuta has a small geographic range (vicinity of Yreka, California) and
narrow habitat specificity (restricted to serpentine), and is thus, considered an endemic
taxon. Serpentine endemism is a well-known feature of California flora and many rare
taxa on California Native Plant Society lists are from serpentine habitat (Skinner &
3
Pavlik 1994). An explanation of the evolution, distribution, and management of rare
serpentinite plant species is available in Kruckberg's (1984) monograph on California
serpentines.
While edaphic endemism (restriction to a particular soil type) may play a large
role in the rarity of P. hirsute,there are many causes of rarity among species (habitat
destruction, inability to adapt to geological or climatic changes, depauperate gene
pool, or a limiting aspect of their biology) and in most cases multiple factors
contribute to a species's decline. Therefore, to develop an effective conservation
strategy, it is important to understand causes of rarity as well as the biology of a rare
taxon. For example, if a plant taxon relies on pollinators for reproduction and habitat
destruction factors into its rarity, then the impact of habitat destruction on its
pollinators needs to be addressed as well.
Phlox hirsuta was federally listed as an endangered species in February 2000.
The Draft Recovery Plan for Phlox hirsuta (USFWS 2004) identifies the need for
research on its biology to assist in conservation efforts for the species. It is important
to have an understanding of the ecology of a plant in order to develop an effective
conservation plan that ensures a species's long-term survival. Reproductive biology
can tell us much about how a plant interacts with both the biotic and abiotic
components in its environment. Little information exists on the reproductive biology
of P. hirsuta. The only previous data on P. hirsuta reproduction is Adams' (1987)
work on fruit and flower production. Knowledge of a plant's mating system(s) has
potential application in management plans to conserve genetic variability (Kearns &
4
Inouye 1993) in that it can help determine a plant species's effective population size.
For example, outcrossing plants tend to have a larger effective population size to
attract sufficient pollinators.
Brown (1990) describes five basic types of mating systems among vascular
plants: 1) Predominantly selfing (self-fertilizing, self-compatible)- flowers that can be
fertilized by their own pollen (autogamy) or flowers that can be fertilized by the pollen
from other flowers on the same plant (geitonogamy); 2) Predominantly outcrossing
(self-incompatible, xenogamous)- requires some kind of pollen vector; 3) Capable of
both selfing and outcrossing; 4) Partial apomixis- producing fruit without sex through
agamospermy or vegetative reproduction; and 5) Partial selfing of gametophytes (in
ferns and artificial breeding). Mating systems are influenced by environmental factors
(Barrett 1990) and may vary between populations of a species or among individuals of
a population (Lyons & Antonovics 1991). For example, a population may be
primarily outcrossing at a site where pollinators are available while a population of the
same species at a different site may be selfing, if pollinators are limited (Kearns &
Inouye 1993).
In addition, determining seed viability can help ascertain whether plants are
self-pollinating (self-compatible) and capable of self-fertilization. Chemical tests and
germination tests are used to determine seed viability. A number of other studies have
examined seed germination rates of other Phlox species (Atwater 1980, Baskin 2005,
Kaye 1997, Springer & TyrI 1989, Roy 2004). However, no information on the
germination requirements and germination rates of P. hirsuta existed prior to this
5
research.
Polemoniaceae family members range from predominantly autogamous (do not
require pollen vector) to obligately xenogamous (require pollen vector) (Grant 1961).
In general, Phlox appears to be an entomophilous genus. Bees, hawkmoths, noctuids,
butterflies, hummingbirds, bats, beeflies, and beetles have been implicated as
pollinators of many Phlox species based on observed visitations (Grant 1965, Scott
1997). There are few detailed pollination studies of other Phlox species (WiggamsHarper 2003, Springer & Tyrl 1989). Related studies on other Phlox species include
the affects of habitat fragmentation on reproduction (Heard et al. 1999, Hendrix &
Kyhl 2000, Kyhl 1997), environmental influences on growth and reproduction (Kelly
& Levin 2000, Schwaegerle & Levin 1990), and competition and efficiency of
pollination systems (Levin 1970, Levin 1972, Levin & Berube 1972).
The goal of this research was to better understand the life history of P. hirsuta
in order to help direct conservation and management plans for the species. Little
information on the life history of P. hirsuta existed prior to this research. No research
on the mating system of P. hirsuta existed as well as little information on seed
viability and seedling establishment of P. hirsuta. The major objectives of this
research were to: 1) determine the mating system of P. hirsutaand the role of
pollinators in seed set, 2) characterize the pollen of P. hirsutain order to identify it on
insects collected from the field sites, and 3) identify seed viability to help ascertain the
mating system and germination requirements to aid in future propagation efforts,
6
Figure 1. Phlox hirsuta Plants at China Hill, Yreka, California.
Figure 2. Phlox hirsuta Flowers.
METHODS
Site Description
Phlox hirsuta is found only on serpentine barrens and slopes at elevations
between 850 m and 1500 m above sea level. All P. hirsuta sites are located within the
vicinity of the city of Yreka in Siskiyou County, California (Figure 3). Phlox hirsuta
occurs on five serpentine soil types across its distribution (Soil Conservation Service
1983). The climate is characterized by hot (mean highs of 12-40'C), dry summers and
cool (mean lows of -20-15'C), wet winters. High, gusty winds are common yearround. While these general climatic conditions are similar among sites, microclimates
vary due to local differences in exposure, elevation, vegetation cover, and topography
(USFWS 2004).
The P. hirsuta sites in Siskiyou County, California are referred to as
occurrences because it is not known whether they are truly separate populations (no
gene flow between them) or part of a larger metapopulation (USFWS 2004) (Figure
3). Roads and considerable distance (minimum of 0.89km) separate the P. hirsuta
occurrences making gene flow unlikely. However, gene flow between occurrences
has not been studied.
7
8
Two of the five known P. hirsuta occurrences were chosen as study sites.
These two sites, China Hill and Cracker Gulch, had fairly dense populations of P.
hirsuta, which was desirable so many plants, as well as insect activity between plants,
could be observed simultaneously. Study plots containing at least 100 P. hirsuta
plants were established within the two P. hirsutaoccurrences.
China Hill
The China Hill occurrence, located at 41 °44'32", 122°36'56"; T45N, R7W,
sec. 14, includes 19 hectares on which approximately 3,000 P. hirsuta plants grow in a
scattered distribution. This site is approximately 1.6 km northeast of downtown Yreka
and 3.22 km from the nearest other P. hirsuta occurrence (Figure 3). China Hill
consists of exposed, open serpentine ridges and slopes ranging in elevation from 850
m to 900 m. For a detailed description of the soil types at China Hill, refer to The Soil
Conservation Service (1983) or The Draft Recovery Plan for Phlox hirsuta (USFWS
2004). China Hill is adjacent to an insular portion of Great Basin Province plant
communities (Sagebrush steppe and Juniper savanna) (USFWS 2004).
In March 2004, a 20m x 20m plot was established within the China Hill
occurrence. The elevation at this point is 865 meters. The plot is on the southwest
facing side (aspect 2400) of a ridge, and adjacent to a dirt road. A systematic count
revealed that there are 195 Phlox hirsuta plants within the plot.
9
Figure 3. Distribution of Known Occurences of Phlox hirsuta (Yreka Phlox),
Siskiyou County, California (USFWS 2004).
MYreka
Phlox occurrences
10
The plot does not contain any trees or tall shrubs but woodlands dominated by
Juniperus occidentalis (Western Juniper) and Ceonothus cuneatus (Buckbrush) grow
on surrounding hills. Total vegetative cover in the plot was visually estimated to be
approximately 48%. The dominant shrubs (comprising of 40% cover of the plot) are
Eriogonum sphaerocephalumvar. halimoides (Wild Buckwheat) (35%) and
Chrysothamnus nauseosus (Rabbitbrush) (5%). The dominant herbs (approximately
8% cover) are annual grasses such as Bromus tectorum (2-6%) and Vulpia
microstachys (1-4%), and perennial grasses (4%) such as Elymus multisetus and
Festuca idahoensis. Other commonly occurring herbs include Eriophyllum lanatum
(1%), Eriogonum strictum (1%), Phaceliacorymbosa, and Lomatium macrocarpum.
For a complete list of plants occurring at the China Hill site refer to Appendix A.
Cracker Gulch
The Cracker Gulch occurrence is 11.27 km southwest of the City of Yreka, in
the Yreka Creek drainage, on the south side of Highway 3 (41 °40'4.4", 122°43'2.7";
T44N, R8W, sec. 13) (Figure 3). This occurrence is 0.88 km from a third occurrence,
Soap Creek Ridge (USFWS 2004). The Cracker Gulch occurrence contains
approximately 1000 plants that are scattered within less than four hectares. This site is
characterized by a mixture of more developed serpentine soils. For a detailed
description of the soil types at Cracker Gulch, refer to The Soil Conservation Service
(1983) or The Draft Recovery Plan for Phlox hirsuta (USFWS 2004). The plant
11
community at Cracker Gulch is characterized as Jeffrey Pine (Pinusjeffreyi)
woodland.
The Cracker Gulch occurrence is on an unnamed mountain that juts out in a
northeast direction where it tapers into a rocky outcrop. It has steep side slopes and a
ridge top that flattens into a long bench. A 1Om x 40m plot was established on part of
this bench, where P. hirsuta is most dense, in April 2004. The bench has a slight
convex shape and elevation is approximately 1300 meters. The plot begins
approximately 50 meters from the dirt road. There are an estimated 280 P. hirsuta
small plants in the plot.
Vegetative cover in the plot is 38% and the remainder consists of bare soil and
exposed rock. The canopy cover (15%) is comprised of Pinusjeffreyi (Jeffrey Pine)
(10%), and Calocedrus decurrens (Incense Cedar) (5%). The minimal shrub layer
(2.5% cover) consists of Cercocarpusledifolius and Chrysothamnus nauseosus. The
herb layer (35% cover), while dominated by grasses, is quite diverse. Minuartia
douglasii (3%), Chlorogalum sp. (1%), and Eriogonum stricture (2%) are some of the
more common herbs. Appendix A provides a plant list of this site.
Determining Mating System
For two consecutive seasons (Spring 2004 and 2005), P. hirsuta phenology
(timing of developmental events such as flowering) was recorded, and mating system
was examined by a pollinator exclusion experiment on a randomly chosen sample of
12
40 plants in two plots (China Hill and Cracker Gulch). Insect observations were made
in each plot throughout the blooming period. In addition, a hand-pollination
experiment was performed on a sample of P.hirsutaplants at China Hill in 2005. The
mating system was determined by measuring the reproductive yield (number of fruits
and seeds produced) per plant in the different treatments in both experiments.
Pollinator Exclusion Experiment
To determine the type of mating system in P. hirsute,a pollinator exclusion
experiment was designed. Twenty of the 40 randomly selected plants (referred to as
unbagged plants) were tagged with aluminum tags nailed into the ground next to each
plant. These plants were left accessible to pollinators. The other twenty plants
(referred to as bagged plants) in the plot were similarly tagged, and were fitted with
white nylon mesh pollinator exclusion bags (mesh size of 0.7mm x 0.4mm and
expandable net openings). The bags slid over two wire loops placed perpendicular
into the ground to form a small dome. The bottoms of the bags were secured with
metal stakes and rocks from the site. The mesh bags were put over plants just prior to
the beginning of the bloom period and left on until one week after the bloom period
ended. At China Hill, the bags were on plants from mid March to mid May in 2004
and late February to early June in 2005. At Cracker Gulch, bags were on from early
April until late May in 2004 and mid March to early June in 2005.
13
In 2005, the same plants were used in the samples in the two plots but half the
plants from the bagged treatments were switched out with half the plants from the
unbagged treatments in both plots. This was done to minimize sampling error
(exclude the possibility that only unique individuals, capable of a particular mating
system, were chosen) and to check for bagging effects.
Floral phenology of sampled plants at both sites was monitored and recorded
two to three times a week throughout the flowering season. Number of buds, number
of flowers in anthesis (open with mature pollen released from anthers), number of
capsules, and presence of pollen was recorded for each tagged plant. At each site, the
beginning, ending, and peak of the flowering season were determined by calculating
the percentage of total flowers in bloom from all sampled plants each day phenology
was done. A seasonal phenology curve depicting peak bloom was developed from
these data. Peak bloom is defined as the majority of flowers on the sampled plants in
the stage of anthesis. Data were also gathered on presence of floral damage
(herbivory) and presence of insects on or inside flowers.
At the end of the bloom season, the semi-persistent calyces (or persistent
pedicel scars in some cases) were counted to determine the total number of flowers
produced by each plant. Plants were checked for developing fruits (capsules) upon
removal of exclusion bags. All plants were monitored for fruit development for the
next several weeks. The number of fruits on the plants was counted prior to bagging
them for seed collection. As fruits began to swell and turn from green to brown, small
white mesh bags were fastened on one to several plant stems in order to capture seeds
14
as fruits dehisced. Bags were used to allow seeds to attain full maturity on the plant
and to ensure that all seeds from a plant were collected and counted. Three types of
bags were used for seed collection; 1) Paint strainer bags made of nylon and mesh size
of 0.7mm x 0.4mm were attached to P. hirsuta plant stems by cinching an inlaid
cotton drawstring, 2) Bridal sachet bags made of nylon and mesh size of 0.3mm x
0.3mm were attached to plant stems by an inlaid nylon drawstring, and 3) Homemade
bags made of organdy material and mesh size of 0.2mm x 0.2mm were attached to
plant stems with a wire twist tie.
Seed collection bags were on plants at China Hill for approximately 35 days in
May-June of 2004, and 2005. At Cracker Gulch, bags were on for approximately 25
days in May-June of 2004, and 40 days in June-July of 2005.
Collected seeds were transported to the Southern Oregon University (SOU)
lab, where they were manually cleaned and counted. Seeds from each tagged plant
were placed in paper envelopes and stored in a dry, dark place at room temperature,
approximately 180 - 290 C, for a fourteen-week after-ripening period in 2004 and a tenweek after-ripening period in 2005.
Hand Pollination Experiment
In the 2005 season, 15 plants adjacent to the established plot for the pollinator
exclusion experiment were randomly selected at China Hill for hand-pollination
treatments to help better define the mating system(s) of P. hirsuta. The plants were
15
tagged with aluminum tags. Each tagged plant received four treatments; 1) Unbagged
(open control for self-incompatibility), 2) Bagged (control for self-compatibility), 3)
Bagged & Selfed (receiving pollen from flowers on same plant to test for
geitonogamy- another form of self-compatibility), and 4) Bagged & Outcrossed
(receiving pollen from other plants). For the three bagged treatments, small nylon
bridal sachet bags (mesh size- 0.3mm x 0.3mm) were attached to individual shoots on
the plant for the duration of the bloom period. For the hand-pollination treatments a
small scissors and forceps were used to transfer pollen. The corolla tube was cut
lengthwise, which allowed access to the stigma. In the third treatment, the anthers in
flowers were emasculated (pulled off with forceps), forceps cleaned with ethanol, and
an anther with dehiscing pollen from another flower on the plant was pulled off with
forceps and lightly brushed over the stigmatic lobes of the flower being handpollinated. The same procedure was followed for hand-pollinating the outcrossed
treatment, except that dehiscing anthers from other P. hirstua plants (5m-20m away)
were used.
Phenology (flower set) and reproductive yield (fruit and seed set) was recorded
and seed collection was performed in the same manner as in the pollinator exclusion
experiment.
16
Insect Observations
To obtain some basic information about potential pollinators, insect activity
was observed at the China Hill site three to five times a week and at the Cracker Gulch
site one to two times a week, during the bloom period. Relevant data such as the local
weather conditions and specific insect observations including name of insect, activity
of insect (flying by, basking, probing flowers), and time of occurrence were recorded.
More details were recorded when insects were observed probing P. hirsuta flowers
(i.e. amount of time probing individual flower, whether insect was observed to visit a
consecutive flower on same plant or another plant, and whether insect was collected).
The activity of probing was defined as an insect spending at least five seconds with its
head over a flower corolla tube. In some cases, especially with Lepidopterans, the
probosces were observed as being extended towards the corolla tube.
Observations were performed during four distinct two-hour time periods
(CAM= Crepuscular AM, beginning one hour before sunrise; DAM= Diurnal AM, 10noon; DPM= Diurnal PM, 2-4; CPM= Crepuscular PM, beginning one hour after
sunset) in 2004. These time periods were used to ascertain differences in the level of
insect activity throughout the day as well as to identify potential pollinators that may
only be active during a specific time period. In 2004, observations were made in the
plots and immediately surrounding area. In 2005, insect observations were more
extensive, covering a broader period of time and a larger portion of the P. hirsuta
occurrences at China Hill and Cracker Gulch. The intent was to go to where the
17
insects were most active instead of wait for insect activity to occur in the plot. More
evening observations, using a blacklight (to attract night-flying insects), also took
place in 2005.
Insects in the two P. hirsuta study sites probing P. hirsuta flowers were
collected using aerial nets, aspirators, or hands to establish insect reference collections
for each site. Collected insects were processed in the standard method described in
Borror & DeLong (1964). All voucher insects collected in 2004 and 2005 were
organized in Cornell drawers and deposited in the SOU Insect Museum. Of the
collected insects, all voucher Lepidopterans (including all those observed probing P.
hirsute), plus other voucher insects observed probing or suspected to have been
probing P. hirsuta (two specimens from the Hymenopteran Order and four specimens
from the Dipteran Order) were checked for Phlox pollen under a dissecting
microscope. To remove pollen, the probosces of 95% of the collected Lepidopterans
were detached from the specimen with tweezers and placed in a watchglass with
ethanol or directly on a slide. Alternatively, pollen from pinned insects (5% of the
Lepidopterans and the six non-Lepidopteran specimens) was removed by dripping
70% ethanol on the specimen's mouthparts held over a watchglass. The resulting
solution in the watchglass was pipeted onto a slide and checked for pollen under a
LeicaTM microscope. Pollen grains were compared to the pollen reference collection to
determine if Phlox pollen was present.
The proboscis of one specimen of each species of Lepidopteran observed
probing P. hirsuta was uncoiled, air-dried, and measured using Spot TM computer
18
software. The probosces lengths were compared to the length of the Phlox corolla
tube to determine if the insect was capable of accessing the nectar at the base of the
corolla tube (which is 10-12 mm long). Images were taken of the elongated proboscis
of each species and visible Phlox pollen on probosces with Leica
TM
and Spot RT TM
digital image software.
Pollen Identification
In order to characterize P. hirsuta pollen and distinguish it from other pollen
carried on insects, P. hirsuta pollen and pollen from all plants in coincident bloom was
collected. Coincidental bloom periods of plants at the China Hill site (Appendix A)
were recorded from plants within a 2000 square meter area surrounding the plot, and
from the adjacent Phlox speciosa site (in which P. hirsuta also grows) located
approximately 1,000 meters to the North at China Hill. Coincidental bloom periods of
plants at the Cracker Gulch site (Appendix A) were recorded from plants within a
1000 square meter area surrounding the plot.
Flowers of plants were collected in the field and stored in paper envelopes.
Anthers were removed from dried flowers in the lab, suspended in a 10% glycerol
solution, and agitated to loosen pollen from the anthers. Under a LeicaTM microscope,
most sculpturing types and other features (size, shape) can be discerned. The resulting
pollen was examined with a compound microscope and photographed with Spot RT
TM
digital image software at 63x in Phase Contrast and Differential Interference Contrast
19
(DIC) modes. Permanent voucher slides were made from dried pollen of each plant
species mounted in Polyvinyl-Lacto-Glycerol (PVLG) and deposited in the SOU
Insect Museum.
Seed Viability Testing
Testing for seed viability includes chemical tests, such as Tetrazolium Chloride
that dyes respiring seeds, and germination tests. Both tests were performed on P.
hirsuta seeds from China Hill. Only the Tetrazolium Chloride test was performed on
seeds from Cracker Gulch due to the small amount of seeds produced there. Sample
sizes of all treatments were dependent on the number of seeds produced by each
treatment at both sites.
In October 2004, 500 seeds were randomly selected from the total of 538 seeds
collected from unbagged plants in the pollinator exclusion experiment at China Hill,
and divided into five treatment groups (Tetrazolium Chloride (TZ), and four
germination treatments) of 100 seeds each. A total of 51 seeds were collected from
Cracker Gulch. Twenty-five seeds from the total of 51 seeds collected from unbagged
plants at Cracker Gulch were randomly selected for the TZ test.
Seeds were tested for viability with Tetrazolium Chloride in the standard
method described in Grabe (1970) and Kearns and Inouye (1993). The seed sample
from unbagged plants at China Hill (100 seeds), the sample from bagged plants at
China Hill (one seed), and the sample from unbagged plants at Cracker Gulch (25
seeds) were placed on Whatman'sTm No. 1 filter paper, moistened with distilled water,
20
in separate sterile, plastic Petri dishes (9cm diameter) for 24 hours at room temperature
(20'C). Petri dish tops were put on and wrapped with parafilm to inhibit evaporation
yet allow some gas exchange. After the imbibing period, each seed was bisected with
a razor blade and placed in dry Petri dishes. The sample from China Hill was
separated into four Petri dishes with 25 seeds each for easier observation. The
bisected seeds were then covered with 0.1% Tetrazolium Chloride solution. The Petri
dishes were left open during the day and checked hourly, adding more solution when
necessary, and covered overnight. Staining on the seeds was analyzed under a
dissecting microscope after four hours and again after 24 hours.
The Association of Official Seed Analysis website (www.aosaseed.com) was
used as a reference in determining the significance of staining patterns. A viable seed
was defined as a seed that respires sufficiently to cause appropriate staining. Seeds
were considered viable if they were stained entirely pink-red or at least had a
completely stained radicle with the majority of the embryo stained. Seeds were
considered not viable if they were entirely green and unstained, had an incompletely
stained radicle, or had purple-blotchy stains (evidence of bruising/damage).
The four germination treatment groups were scheduled to undergo a 12-week
cold/moist stratification period to help break dormancy and then each group was to
receive a second dormancy-breaking treatment if needed (i.e. KMN0 ). Each
3
treatment group of 100 seeds was dispersed in a plastic Ziplock
Tm
bag containing 118
ml of peat moss moistened with 75 ml of distilled water. Closed bags were left at
room temperature for four hours to allow seeds to imbibe. Bags were then laid flat on
21
a wire rack in the laboratory refrigerator (set at 51C but ranged from 3° to 70 C). A
viable seed was defined as a seed from which a visible radicle tip had broken the seed
coat and begun to emerge, thereby initiating the process of germination.
In the Southern Oregon University greenhouse, germinated seeds were planted
in a mixture of 1/2 regular potting soil and
1/2
sand/loam. Seeds were sown in individual
cells of six-pack trays and in 12.7 x 27.9 cm containers approximately l/2 cm apart.
Since bagged plants at China Hill produced more seeds in 2005 (14 seeds)
compared to 2004 (one seed), seed viability in bagged plants could be tested more
thoroughly, to determine if fertilization had occurred. The sample (14 seeds) from
bagged plants and a sample of 14 randomly selected seeds from unbagged plants at
China Hill were placed in plastic Ziplock
Tm
bags containing 118 ml of peat moss
moistened with 75 ml of distilled water. The bags were placed in the laboratory
refrigerator to undergo cold/moist stratification as described above.
RESULTS
Phenology
At China Hill in 2004, flowering of P. hirsuta began during the third week of
March and ended 7.5 weeks later in the second week of May with peak bloom
occurring in early April (Figure 4). In 2005, flowering of P. hirsuta began during the
second week of March and ended ten weeks later during the last week of May with
peak bloom occurring in mid April (Figure 5). In both years, peak bloom was defined
by 40-50% of the total flowers produced, in bloom. The two treatments (unbagged
and bagged) in the pollinator exclusion experiment exhibited very similar phenologies,
beginning, ending, and peaking around the same time. These results indicate that
bagging the plants did not adversely impact flower set, although there appears to be a
slight delay of two to five days in flowering among bagged plants both years. The
phenology curves of the two seasons, 2004 and 2005, at China Hill were similar, with
peak bloom occurring at the same time of year. However, peak bloom and the overall
bloom period lasted longer in 2005.
At Cracker Gulch in 2004, P. hirsuta began flowering in early April and ended
seven weeks later during the third week in May (Figure 6). In 2005, P. hirsuta began
22
23
flowering in early April and ended ten weeks later in mid June (Figure 7). Peak bloom
in 2004, 40% of the flowers on bagged plants and 20% of the flowers on unbagged
plants (two small peaks), occurred in mid-late April. Peak bloom in 2005 represented
27% of the flowers on bagged plants, and 35% of the flowers on unbagged plants.
The phenologies of the unbagged and bagged plants were similar, suggesting no
negative bagging effects on flowering. In 2005, peak bloom was two weeks later than
in 2004, and the overall bloom period lasted longer.
In 2004, P. hirsuta started blooming earlier and peaked earlier at China Hill,
whereas bloom period started later at Cracker Gulch (Figure 8) but ended the same
time as China Hill. Thus, P. hirsuta plants at China Hill bloomed over a longer period
of time than plants at Cracker Gulch in 2004. In 2005, P. hirsutaplants at Cracker
Gulch started blooming two weeks later, peaked five weeks later, and ended two
weeks later than China Hill (Figure 9). The bloom periods lasted approximately ten
weeks at both sites.
24
Figure 4. Percent of Floral Bloom in Unbagged and Bagged Plants at China Hill in
2004.
50
+
% of total
bagged Rs.
40
E
O
-9%
30
of total
unbagged fis.
C
I
3E 20
En
-
- --
-
10
0
Figure 5. Percent of Floral Bloom in Unbagged and Bagged Plants at China Hill in
2005.
50
% of total
bagged flowers
--
E
40
-_
-
+
-
C
30
-9- % of total
Cn
as
3
unbagged
flowers
20
10
0
"I
-,)" 4"(O \
o \ ,(O \
e\
,)\
14111111-,\r)> -\
I
25
Figure 6. Percent of Floral Bloom in Unbagged and Bagged Plants at Cracker Gulch
in 2004.
50
E
-
----- --
,
bagged
40
0
0
-E
-
-
+ % of total
fls.
/ \ % of total
-
30
___
_____
--
_--
unbaggedfs.j
_
20
c
10
0
Figure 7. Percent of Floral Bloom in Unbagged and Bagged Plants at Cracker Gulch
in 2005.
40
S
D
30
\% iof total
bagged
-
0
20
-_
__
-flowers
-
I.=
_
,
X
I0
10
0
N\
\
\,,r"
)\6 ")\<1
\
\"\
c
% of total
unbagged
flowers
26
Figure 8. Percent of Floral Bloom at China Hill and Cracker Gulch in 2004
50
E
Cracker
Gulch
40
+ China Hill
0
=
30
1
20
'
10
30
0n
Ik
N
"I
e
N\f
Figure 9. Percent of Floral Bloom at China Hill and Cracker Gulch in 2005.
F
-__-__
50
+
S
0
40
E-
China Hill
Cracker
g ulc h
30
U)
20
0
10
0
-\\C,
14-0
N",
N\',
0
N\11
e
P
e
4)> 4-4
b\N 0\\"
_
I
27
Pollinator Exclusion Experiment
In 2004, unbagged plants at China Hill produced a mean of 40.36 flowers
(range 0-118), 21.18 fruits (range of 0-116), and 19.55 seeds (range of 0-116); bagged
plants produced a mean of 78.6 flowers (range of 0-275), .05 fruits (range of 0-1), and
.05 seeds (range of 0-1) (Figure 10). A student's t- test (assuming unequal variances)
revealed a significant difference (p< 0.05) in fruit set and seed set between treatments.
In 2005, unbagged plants at China Hill produced a mean of 78.60 flowers (range 7283), 23.15 fruits (range of 0-71), and 20.85 seeds (range of 0-67); bagged plants
produced a mean of 141.90 flowers (range of 5-509), 1.05 fruits (range of 0-6), and
0.7 seeds (range of 0-5) (Figure 11). As in 2004, a student's t- test (assuming unequal
variances) revealed a significant difference (p< 0.05) in fruit, and seed set between
bagged and unbagged treatments. In 2005, there was also significant difference in
flower set between treatments.
At China Hill in 2005, mean flower set per plant in unbagged plants increased
from 2004 but not significantly. A student's t-test (assuming unequal variances)
showed no significant difference in mean flower, fruit and seed set in unbagged plants
and a significant difference (p< 0.05) in flower, fruit and seed set in bagged plants
between 2004 and 2005.
In 2004, unbagged plants at Cracker Gulch produced a mean of 8.7 flowers
(range of 0 to 42), 3.25 fruits (range of 0-21), 2.55 seeds (range of 0-17); bagged
plants produced a mean of 4.60 flowers (range of 0-30), 0.00 fruits, and 0.00 seeds
28
(Figure 12). A student's t-test revealed a significant difference (p< 0.05) in fruit and
seed set between the two treatments. Forty-eight percent of the sampled plants at this
site produced no flowers this season. Similarly, Adams (1987) found 38% of the P.
hirsutaplants at a similar site (Soap Creek Ridge) produced no flowers. In 2005,
unbagged plants at Cracker Gulch produced a mean of 25.60 flowers (range of 0 to
82), 3.90 fruits (range of 0-25), 2.7 seeds (range of 0-19); bagged plant produced a
mean of 43.30 flowers (range of 0-141), 0.00 fruits, and 0.00 seeds (Figure 13). A
student's t-test revealed a significant difference (p< 0.05) in fruit and seed set between
the two treatments.
Although the average number of flowers per plant in unbagged plants and
bagged plants increased significantly (p < 0.05) in 2005, the average number of fruits
and seeds in 2005 did not increase significantly in either treatment at Cracker Gulch.
There was higher reproductive output from unbagged plants at China Hill than
at Cracker Gulch in 2004 (Figure 14) and in 2005 (Figure 15). T-tests indicated a
significant difference (p< 0.05) in flower, fruit, and seed set between the sites in 2004
and a significant difference (p< 0.01) in flower, fruit, and seed set between the sites in
2005.
In sum, P. hirsuta plants at China Hill did not set many fruits or seeds in the
absence of pollinators compared to plants that were accessible to pollinators. At
Cracker Gulch, P. hirsuta did not set any fruit or seed in the absence of pollinators.
29
Figure 10. Mean Number of Flowers, Fruits, and Seeds in Unbagged and Bagged
Plant Treatments at China Hill in 2004. Error bars represent one standard error about
the associated mean. Asterisks indicate significant difference (p< 0.05) between
treatments.
120.......-.
.
* Mean No. of
100
i
z
flowers/plant
-
O Mean No. of
fruits/plant
E Mean No. of
seeds/plant
80
60
40
20
0
unbagged
bagged
Figure 11. Mean Number of Flowers, Fruits, and Seeds in Unbagged and Bagged
Plant Treatments at China Hill in 2005. Error bars represent one standard error about
the associated mean. Asterisks indicate significant difference (p< 0.05) between
treatments.
t
-
180
160 -flowers/plant
140120
-
E Mean No. of
fruits/plant
E 0Mean No. of
100*
6
z
80
1 -
seeds/plant
60
i
40
Mean No. of
*
*
200
unbagged
bagged
30
Figure 12. Mean Number of Flowers, Fruits, and Seeds in Unbagged and Bagged
Plant Treatments at Cracker Gulch in 2004. Error bars represent one standard error
about the associated mean. Asterisks indicate significant difference (p< 0.05) between
treatments.
14 U Mean No. of
flowers/plant
.E Mean No. of
fruits/plant
ElMean No. of
seeds/plant
12
10
z
Z6
6
4*
2
unbagged
bagged
Figure 13. Mean Number of Flowers, Fruits, and Seeds in Unbagged and Bagged
Plant Treatments at Cracker Gulch in 2005. Error bars represent one standard error
about the associated mean. Asterisks indicate significant difference (p< 0.05) between
treatments.
50 45
* Mean No. of
flowers/plant
-
I MeanNo. of
fruits/plant
25 -
Z
ElMeanNo. of
seeds/plant
20
15
12 10
50
*
0
unbagged
bagged
31
Figure 14. Mean Number of Flowers, Fruits, and Seeds in Unbagged Plants from
China Hill and Cracker Gulch in 2004. Error bars represent one standard error about
the associated mean. Asterisks indicate significant difference (p< 0.05) between sites.
6 Mean No. of flowers/plant
50
Z.
ED Mean No. of frits/plant
4 Mean No. of seeds/plant
*
40
6;30
20
10
China Hill
Cracker Gulch
Figure 15. Mean Number of Flowers, Fruits, and Seeds in Unbagged Plants from
China Hill and Cracker Gulch in 2005. Error bars represent one standard error about
the associated mean. Asterisks indicate significant difference (p< 0.05) between sites.
100
90 80
-
*
Mean No. of flowers
Mean No. of fruits/plant
70-
U,Mean No. of seeds/plant
60
~50
i
40
>
l
30
20
0
China Hill
Cracker Gulch
32
The total number of flowers, fruits, and seeds were recorded from the
unbagged and bagged plants at both sites in 2004 and 2005. Although the mean
number of flowers produced on unbagged plants was less than the mean number of
flowers on bagged plants in 2005 (Figs 10-13), a higher percent of unbagged flowers
produced seed at both sites in 2004 and 2005 (Table 1). Forty-seven percent of the
unbagged flowers at China Hill and 29% of the unbagged flowers at Cracker Gulch
produced seeds in 2004 (Table 1). In contrast, bagged plants in 2004 produced a total
of one seed at China Hill, and zero seeds at Cracker Gulch. Similar results occurred in
2005, but fewer flowers set seed (27% at China Hill and 10% at Cracker Gulch) in
unbagged plants than in 2004 (Table 1). In contrast, a higher percent of flowers on
bagged plants (0.49%) at China Hill produced seeds in 2005 than in 2004 (Table 1).
The percent of fruits that then produced seeds was relatively high at both sites
both years. In 2004, a total of 538 mature seeds were recovered from 568 capsules
(95% seed set) from China Hill (Table 1). At Cracker Gulch in 2004, a total of 51
mature seeds were recovered from the 65 capsules (78% seed set) (Table 1). In 2005,
seed set from fruit in unbagged plants was 90% and 69% at China Hill and Cracker
Gulch, respectively. It should be noted that while each flower has the potential to
produce a maximum of three seeds (has a three-carpeled ovary with one ovule per
carpel), flowers from sampled plants produced a maximum of one seed each.
The results (zero to low seed production in bagged plants) from the pollinator
exclusion experiments at both sites during 2004 and 2005 support self-incompatibility
(outcrossing) as the predominant mating system of P. hirsutea
33
Table 1: Treatment and Site Comparisons of Flower, Fruit, and Seed Set.
Total
Total number of
Total number of
number of
fruits (% fruit set)
seeds (% seed set)
flowers
2004
China Hill
(unbagged)
China Hill
(bagged)
Cracker Gulch
(unbagged)
Cracker Gulch
(bagged)
1146
568 (50.0)
538 (47.0)
1572
1 (0.06)
1 (0.06)
174
65 (37.0)
51 (29.0)
91
0 (0.00)
0 (0.00)
1572
463 (29.0)
417 (27.0)
2838
21 (0.74)
14 (0.49)
512
78 (15.0)
54 (10.0)
866
0 (0.00)
0 (0.00)
2005
China Hill
(unbagged)
China Hill
(bagged)
Cracker Gulch
(unbagged)
Cracker Gulch
(bagged)
34
Hand Pollination Experiment
Bagged flowers (treatment 2) and bagged & selfed flowers (treatment 3)
produced no fruits or seeds, while unbagged flowers (treatment 1) and bagged &
outcrossed flowers (treatment 4) produced fruits and seeds (Figure 16). These results
provide additional evidence that P. hirsuta is primarily outcrossing, and may even be
exclusively outcrossing. In particular, self-compatibility is not supported because
flowers that were hand-pollinated with pollen from flowers on the same plant did not
produce seeds.
The total number of flowers produced by individual treatment shoots varied, so
proportions, number of seeds/number of flowers/plant, per treatment, were computed
to ascertain fruit and seed set per treatment. The highest fruit set (number of fruits
produced per total flowers produced) (50%) and seed set (number of seeds produced
per total flowers produced) (40%) occurred in shoots that were hand-pollinated with
outcrossed pollen (treatment 4). The unbagged shoots (treatment 1) had 26% fruit set
and 23% seed set. The bagged shoots (treatment 2) and the bagged & selfed shoots
(treatment 3) both had 0% fruit set and 0% seed set.
One-way Analysis of Variance (ANOVA) indicated a significant difference (F
=
32.49; df = 3, 56; p < 0.0001) between the four treatments. The confidence intervals
(95%) for each mean percent seed set were calculated and compared to determine
which pairs of treatments were significantly different. The bagged and bagged &
selfed treatments were not significantly different but both were significantly different
35
from the unbagged and bagged & outcrossed treatments. Although mean percent seed
set was slightly higher in the bagged & outcrossed treatment than in the unbagged
treatment, the confidence intervals (95%) for the two means (unbagged and bagged &
outcrossed) overlapped slightly, establishing that the difference in seed set between
these two treatments was at most marginally significant.
Figure 16. Mean Number of Flowers, Fruits, and Seeds of Four Treatments in the
Hand Pollination Experiment. Error bars represent one standard error about the
associated mean.
14
* Mean No. Flowers/shoot
El Mean No. fruits/shoot
Mean No. seeds/shoot
12 -.
10-
,2
86
0
0
4
22 2
Unbagged
Bagged
Bagged &
Selted
Bagged &
Outcrossed
36
Pollen Identification
Bloom periods for plants at China Hill (Table 2) and Cracker Gulch (Table 3)
show considerable overlap with the bloom period of P. hirsuta.
Phlox pollen was easily distinguished from all other pollen collected at both
sites (Appendix B) because of its unique morphology. However, under a light
microscope, Phlox hirsuta and P. speciosa pollen are indistinguishable. Phlox hirsuta
pollen grains have a mean diameter of 42 pm (range of 39-45 jm), and are spheroidal,
polyporate (many apertures-circular sunken pores- spaced equidistantly over the
surface), and have a polygonal reticulum (sculpturing elements arranged in a network
that has gaps) that surrounds pores (Figure 17). P. speciosa pollen has a mean
diameter of 35.5 jim (range of 32-39 jim) (Figure 17). In all other respects, P.
speciosa pollen looks identical to P. hirsutapollen under a light microscope.
Gilia capitataand Linanthus sp., also in the Polemoniaceae family, are found
in the China Hill plot and overlap with P. hirsuta in bloom. Gilia and Linanthus
pollen grains are similar in shape (spheroidal) and size (35-40 Pim) to Phlox pollen.
However, they both have distinguishable features on the grain surface. Gilia pollen is
zonocolporate (sunken furrows and pores arranged equidistantly around the equator of
the grain) and the reticulum is striated instead of polygonal (Stuchlik 1967) (Figure
18). Linanthus pollen has a polygonal reticulum that is obscured by pila (bump-like
apertures) or muri (thin furrows) (Stuchlik 1967) (Figure 18).
37
Table 2. Bloom periods of Plants in Coin c~idental Bloom with P. hirsuta at China Hill.
2004
2005
Plant name
A chillea millefolium
Alilium siskiyouense
A msinckia intermedia
Arabis puberula
Astragalus purshiivar. tinctus
Castilleja hispida
Ceonothus cuneat'us
Chrysothamnus nauseosus
Clarkiagracilis
Claytonia exigua
Claytonia lanceolata
Collinsiaparviflora
Crocidium multicaule
Dichelostemma capitatum
Epilobium sp.
Erigeron bloomeri var. bloomeri
Eriogonum sphaerocephalum
Eriophyllum lanatum
Eschscholzia californica
Fritillariapudica
Gilia capitata
Lasthienia californica
Lewisia rediviva
Linanthus sp.
Lithophragmaparviflorum
Lomatium macrocarpum
Lupinus argenteus
Minuartia douglasii
Orobanche uniflora
Phaceliacorymbosa
Phlox hirsuta
Phlox speciosa
Plagiobothryssp.
Sisrynchium douglasii
Viola beckwithii
Start Date
4/14
3/25
4/10
4/8
4/8
4/10
4/4
5/6
4/20
End Date
5/30
4/17
5/30
4/18
4/22
5/30
4/22
5/30
5/30
3/30
4/4
4/6
4/25
4/25
4/17
4/8
4/13
4/30
4/24
5/30
5/30
5/30
5/30
4/20
4/10
4/20
5/30
4/20
5/4
4/6
4/1
4/1
4/20
5/2
4/6
3/22
4/1
4/6
4/18
4/28
5/30
5/30
5/30
5/30
5/13
5/4
5/20
3/15
4/1
Start Date
5/9
3/17
4/10
3/17
3/17
3/31
3/17
End Date
6/10
5/7
5/17
5/21
5/7
5/31
5/12
4/26
2/23
2/23
3/17
2/23
3/23
4/14
4/24
4/28
4/14
4/10
2/23
4/21
5/17
4/26
4/19
3/11
3/11
4/14
4/26
4/10
4/19
2/23
3/5
4/2
2/23
2/23
6/10
4/26
3/31
4/21
4/21
4/21
5/30
6/10
6/25
6/10
6/10
3/23
6/10
5/30
5/17
5/26
5/17
5/10
5/26
6/10
6/10
6/10
6/2
5/3 1
6/10
4/19
4/2
38
Table 3. Bloom periods of Plants in Coincidental Bloom with P. hirsuta at Cracker
Gulch.
2004
Plant name
Achillea millefolium
Allium siskiyouense
Arabispuberla
Calochortussp.
Castillejahispida
Ceonothus cuneatus
Cercocarpusledifolius
Chlorogalum sp.
Claytonia exigua
Collinsiaparviflora
Chrysothamnus nauseosus
Epilobium sp.
Eriogonum strictum
Eriophyllum lanatum
Erysimum captitatum
Fritillariapudica
Gilia capitatum
Lithophragmaparviflorum
Lomatium macrocarpum
Lomatium triternatum
Minuartianuttallii
Penstemon sp.
Phaceliacorymbosa
Phlox gracilis
Phlox hirsuta
Phlox speciosa
Plagiobothryssp.
Thysanocarpuscurvipes
Viola purpurea
2005
Start Date
5/1
4/20
4/6
4/29
4/26
End Date
5/30
5/11
5/2
5/20
5/30
4/20
4/20
5/2
5/11
5/6
4/6
5/30
4/24
4/29
5/30
4/22
5/2
4/20
5/11
5/30
5/30
4/2
5/23
4/26
3/22
4/20
5/20
4/2
5/11
Start Date
5/15
4/28
3/17
5/15
4/18
4/20
4/1
4/18
4/18
4/28
End Date
7/25
5/30
5/21
6/8
6/20
5/30
4/28
5/30
5/30
5/15
4/28
6/4
5/15
5/15
3/17
5/15
4/10
4/10
5/10
5/19
5/30
5/10
3/17
4/1
4/28
4/28
4/1
3/17
6/15
7/25
6/20
6/15
4/18
6/20
5/30
6/20
6/20
6/30
7/25
5/30
4/18
6/15
6/15
6/20
5/l
5/30
39
Figure 17. Pollen Grains of P. hirsuta and P. speciosa as Seen Under a Light
Microscope at 63X
Phlox hirsuta
Phlox speciosa
50 wo
50 pm
Figure 18. Pollen Grains of (ilia capita/aand Linanthus sp. as Seen Under a Light
Microscope at 63X.
LinanthS Vp.
40
Insect Observations
In 2004, a total of 918 insects were observed within 128 observation hours at
China Hill. A number of different insects from four main insect orders (Hymenoptera,
Lepidoptera, Diptera, and Coleoptera), as well as insects from other orders, were
observed visiting P. hirsuta during all four observation time periods (Crepuscular AM,
Diurnal AM, Diurnal PM, and Crepuscular PM). Fifty-one percent of the total insects
were observed during Diurnal PM, presumably due to more conducive air
temperatures. Twenty-six percent and 15% of the insects were observed in Diurnal
AM and Crepuscular PM, respectively. Eight percent of the insects were observed
during Crepuscular AM. Overall, insect visitation was relatively infrequent and
sporadic, as was visitation specifically to P. hirsute. Four percent (36) of the total
insects were observed probing P. hirsuta flowers (Table 4). The majority of insects
probing P. hirsuta flowers were Lepidopterans. A total of 95 insects were observed
within 29 observation hours at Cracker Gulch in 2004. Two percent (2) of the total
insects were observed probing P. hirsuta flowers (Table 5).
In 2005, observations occurred over broader ranges of time and longer periods,
but focused more on afternoon and evening hours when more insects were observed in
2004. In 180 hours of observation, a total of 1603 insects were observed at China Hill
and 8% (127) of these were observed probing P. hirsuta flowers (Table 4). Most
insects observed probing P. hirsuta were Lepidopterans. In 2005 at Cracker Gulch,
41
415 insects were observed over 41 hours. Only 1.2% (5) of these insects were
observed probing P. hirsuta (Table 5).
In 2004, the most common Lepidopterans probing P. hirsuta were Sphingids,
while in 2005 Nymphalids were the most prevalent. Insects probing P. hirsuta at
Cracker Gulch included Lepidopterans, Hymenopterans, and Dipterans. Insects that
consecutively probed P. hirsuta (went from one P. hirsuta plant or flower directly to
another) included Vanessa cardui (Nymphalidae), Hyles lineata (Sphingidae),
Proserpinusclarkiae (Sphingidae), and Bombus sp. (Apidae). The two Sphingids
were observed visiting only P. hirsuta flowers, and no other plant species.
42
Table 4. Insects Observed Probing P. hirsutaFlowers at China Hill.
Species and Number
Family/Order
observed
2004
1 Erynnis sp.
1 Subfamily: Hesperiinae
I Tribe: Melitaeini Grote
(Checkerspot)
5 Vanessa cardui
2 Papiliozelicaon
1 Hyles lineata
15 Proserpinusclarkiae
6 Taxa
4 Bombussp.
Hesperiidae/Lepidoptera
Hesperiidae/Lepidoptera
Nymphalidae/Lepidoptera
Nymphalidae/Lepidoptera
Papilionidae/Lepidoptera
Sphingidae/Lepidoptera
Sphingidae/Lepidoptera
Sphingidae/Lepidoptera
Apidae/Hymenoptera
2005
5 Subfamily: Hesperiinae
110 Vanessa cardui
I Pontiasp.
1 Colias sp.
4 Hyles lineata
2 Proserpinusclarkiae
4 Bombus sp.
Hesperiidae/Lepidoptera
Nymphalidae/Lepidoptera
Pieridae/Lepidoptera
Pieridae/Lepidoptera
Sphingidae/Lepidoptera
Sphingidae/Lepidoptera
Apidae/Hymenoptera
Table 5. Insects Observed Probing P. hirsuta Flowers at Cracker Gulch.
Species and Number observed
Family/Order
2004
1 Subfamily: Hesperiinae
1 Bombus sp.
Hesperiidae/Lepidoptera
Apidae/Hymenoptera
2005
1
1
2
1
Taxon
Vanessa sp.
Taxa
Bombus sp.
Sphingidae/Lepdidoptera
Nymphalidae/Lepidoptera
Syrphidae/Diptera
Apidae/Hymenoptera
43
The proboscis lengths of the main Lepidopteran species (Proserpinusclarkiae1 1mm, Hyles lineata- 35mm, Vanessa cardui- 10-12mm, Subfamily: Hesperiinae1Omm) observed probing P. hirsuta were all long enough to reach the base of the 1012mm corolla tube (where nectar is located).
In 2004, all of the collected Lepidopterans were examined for pollen.
Individual pollen grains were counted and the overall relative pollen load was
calculated as follows: low (1-20 pollen grains), medium (20-50 grains), and high (over
50 grains). Proserpinusclarkiae and Hyles lineata all carried high amounts of Phlox
pollen on their probosces, regardless of whether they were collected from P. hirsuta
plants (Table 7). In all cases, P. clarkiae and H. lineata did not have pollen from any
other plant species. Hesperiajuba(Subfamily: Hesperiinae) also carried Phlox pollen
but the loads were relatively low (Table 7). Other Lepidopteran species examined,
one Erynnis sp., one Nymphalis californica, two Pontiasp., one Pierissp., and one
unidentified moth, carried no Phlox pollen.
In 2005, 100 specimens, including all collected Lepidopterans, three Dipterans,
and three Hymenopterans, were examined for pollen, and thirty-eight of these were
confirmed to have Phlox pollen (Table 7). As in 2004, the Sphingids collected were
carrying Phlox pollen. Almost 50% percent of the Vanessa cardui collected were
carrying Phlox pollen. Although more V carduiwere collected, the pollen load was
always smaller than the load on Sphingids. All four specimens from the subfamily
Hesperiinae, two of which were collected on P. hirsuta, had a low Phlox pollen load.
Other Lepidopterans carrying Phlox pollen included one Erynnis propertius, one
44
Papiliozelicaon (Anise Swallowtail), two Pontiasisymbrii, one Pieris rapae, one
Noctuid moth (Cissusa discreta), and three Geometrid moths (Drepanulatrix
unicalcarariaand Plataeatrilinearia)(Table 7). Of these, only Papiliozelicaon had a
high Phlox pollen load. Other Lepidopterans that were examined but found not to
carry Phlox pollen included one Blue (Hesperidae), one Nymphalis californica, and 18
moths (primarily Noctuidae). The only other insect carrying Phlox pollen besides
Lepidopterans was Bombus vosnesenskii. This specimen was carrying four Phlox
pollen grains in addition to pollen from Lomatium macrocarpum and Phacelia
corymbosa. Other insects that were collected from P. hirsuta plants, included one fly
(Muscidae) and two syrphid flies, and one small bee, none with Phlox pollen.
Few insects were observed probing P. hirsuta at Cracker Gulch, and
consequently, few specimens were collected and data were combined for 2004 and
2005. One specimen, Hesperiajuba,was found with a low Phlox pollen load.
Lepidopteran specimens examined and not carrying Phlox pollen included two
Hesperia Columbia, four Nymphalis californica,four Euphydras editha, one Pontia
sp., and two Noctuids (Autographa californica). In addition to Lepidopterans, two
syrphid flies that were collected while on P. hirsuta were examined but not found to
be carrying Phlox pollen.
45
Table 7. Lepidopterans Collected from China Hill with Phlox pollen. Pollen loadlow (1-20 grains), medium (20-50 grains), high (over 50 grains).
IINo. of
Species
Family
Collected
No. with Pollen load
specimens
while on
Phlox
examined
P. hirsuta? pollen
,
I
I.
2004
Hesperiajuba Hesperiidae
1
Hyles lineata
Froserpinus
4 ______
4
-4
.4
2
.4
Sphingidae
Sphingidae
Yes(l)
No(M)
No
Yes
+
2
5
.-
2
low
I
I-
2
5
*
high
high
-
clarkiae
2005
Drepanulatrix Geometridae
unicalcararia
.
Plataea
Geometridae
trilinearia
Atalopedes
Hesperiidae
campestris
Hesperiajuba Hesperiidae
Hesperia
Hesperiidae
Columbia
Erynnis
Hesperidae
propertius
Cissusa
Noctuidae
discreta
.
Vanessa
Nymphalidae
cardui
Vanessa
Nymphalidae
cardui
Papilio
Papilionidae
zelicaon
Pontia
Pieridae
sisymbrii
..
Pieris rapae
Pieridae
-+
.4
.
2
No
2
ilow
1
No
1
low
4 ___________
-
No
4 ___________
t
.4
L
2
2
Yes
No
-
2
2
1
4
1
low
low
-
No
4- ___________
No
+
9
Flow
-
L
1
Yes
i
medium
4
4i
8
1
mediumhigh
lowmedium
high
-2
low
.-
l
21
No
6
1 ________-
4
2
No
10
No
1
.
Hyles Lineata Sphingidae
Proserpinus
Sphingidae
1ilow
4
4
-
.1
1
+
+
clarkiae
1.
-4
.
4 _____
-44-
F.
Ii_
I
I
12
3
I
I
t
No
No
Yes(2)
No(l)
4- _________________________
I
1
2~
3
low
medium
high
46
Seed Traits and Viability
In May of 2004, ten mature P. hirsuta capsules were collected from plants
outside of the study site at China Hill. The capsules were examined under a dissecting
microscope. Eight of the ten capsules contained one seed each and two were empty.
Phlox hirsuta seeds averaged 4.5mm in length (3 .5-5.5mm range) and 2.5mm in width
(1.5-3mm range). A linear-shaped embryo filled the center of the oblong seed (Baskin
2001). The green embryo was surrounded by white tissue, most likely a reduced
endosperm (Atwater 1980).
Seed viability was tested with Tetrazolium Chloride in 100 of the 538 seeds
collected at China Hill and 25 of the 51 seeds collected at Cracker Gulch in 2004.
Sixty-six percent of the P. hirsuta seeds from China Hill and 76% of the seeds from
Cracker Gulch were viable.
The seeds in the germination treatments (a total of 400 seeds from China Hill,
100 seeds per treatment) had slightly higher germination rates than viability rates of
seeds in the TZ treatments. On the 9th day of the cold/moist stratification period,
several seeds in all four bags were observed to have emerging white radicle tips.
Seeds continued to germinate for several more days. On days 21 and 28, seed viability
was assessed by counting those seeds that had begun to germinate. A total of 337
seeds out of 400 germinated, an 84% germination rate in the sampled seeds from
China Hill.
47
Of the 337 germinating seeds planted out in the SOU greenhouse, 266 (79%)
emerged from the soil and grew into seedlings but only 199 of these survived the next
six weeks, and developed one or two sets of true leaves. Seed coats stuck to the
cotyledons of many emerging seedlings, impairing their growth and causing some
fatalities (Figure 19). Damp-off disease, a common fungal pathogen affecting
seedlings in greenhouses, killed 35 seedlings. In December of 2004, Marla Knight,
botanist on the Klamath National Forest, took the 199 seedlings to a professional
nursery. Care of the seedlings and future plans for their planting out have been taken
over by the National Forest and the USFWS.
In 2005, small samples of seeds from bagged plants and unbagged plants at
China Hill underwent the same cold/moist stratification as in 2004. Radicle tips began
to emerge from seeds beginning on the ninth day of stratification (as in 2004). Seeds
were counted on day 22, at which time, 42% of the seeds from bagged plants
germinated and 71 % of the seeds from unbagged plants germinated.
48
Figure 19. Phlox hirsuta Seedlings in the Southern Oregon University Greenhouse.
Photo by Dr. Carol Ferguson.
DISCUSSION
Phenology
The bloom period of P. hirsuta lasted six to twelve weeks. Plants were in
several phenological stages simultaneously (i.e. buds, flowers in anthesis, and
immature fruits). Observations of several individual flowers showed they remained in
anthesis for several days (Ferguson 2004). The long bloom period may assure
pollinator visitation in plants not capable of selfing by buffering periods of low insect
activity (i.e. during adverse weather conditions). In 2004 and 2005, China Hill and
Cracker Gulch experienced strong gusty winds accompanied by low temperatures not
conducive to insect flight. In addition, the weather was cooler and rainier in 2005 and
the bloom period began earlier and lasted longer than in 2004. The bloom periods of
plants in coincidental bloom with P. hirsuta (Tables 2 and 3) show that several other
plant species have bloom periods lasting six weeks and longer.
In most plants, flowering time is primarily determined by a genetically programmed
response and/or by environmental cues such as day-length, temperature, rainfall, and
plant size or density (Kelly & Levin 2000). If a plant is insect-pollinated, the bloom
period needs to coincide with the availability of the appropriate pollinator (Fenner
1985). Synchronized blooming among plants of the same species may be
49
50
crucial for plants dependent on outcrossing but synchronized blooming with other
plant species may create competition for pollinators. Based on field observations, P.
hirsuta competes with several other plants (Allium sp., Lomatium macrocarpum) for
the pollen vectoring services of various Lepidopteran species (Tables 2 and 3). Thus,
while the long bloom period of P. hirsuta is advantageous for attracting pollinators,
there is likely competition for pollinators among several long-blooming species at
China Hill and Cracker Gulch.
Peak bloom occurred later at Cracker Gulch than at China Hill both years and
there was more variability (unbagged plants had two small peaks in bloom in 2004)
between years at Cracker Gulch, suggesting that factors such as elevation, weather
patterns, plant size and population size may influence the flowering time of plant
populations as has been found in studies of other Phlox species (Kelly & Levin 2000,
Hendrix & Kyhl 2000).
Mating System
In the pollinator exclusion experiment, bagged plants at Cracker Gulch
produced no seeds both years. Bagged plants at China Hill produced a very small
number of seeds (one in 2004 and fourteen in 2005). Even though a t-test indicated a
significant difference between seed set in bagged plants between the two years, this
difference was not as significant overall when compared to the greater difference
between seed set in bagged plants and unbagged plants (538 seeds from unbagged
51
plants in 2004 and 417 seeds from unbagged plants in 2005). In the hand-pollination
experiment, only unbagged flowers (treatment 1) and the bagged & outcrossed flowers
(treatment 4) produced seeds. Based on these results, P. hirsuta is primarily selfincompatible and requires a pollen vector between plants.
The morphology of P. hirsuta flowers further supports self-incompatibility.
While some of the anthers are physically located close enough to the stigma for selfpollination to occur, the flowers are protandrous, meaning the anthers mature prior to
the stigma becoming receptive. Protandry is believed to be a mechanism to inhibit
self-compatibility because the stigma is less likely to receive pollen from the same
flower since most pollen is gone by the time it is receptive (Kearns & Inouye 1993).
The more pollinators that visit a flower, the less pollen is available to pollinate its own
flower by the time the stigma becomes receptive (Levin 1972).
How, then, is the small amount of seed production in bagged plants at China
Hill explained? It may indicate that P. hirsuta is capable of selfing, perhaps under
certain conditions. Alternatively, insects may have vectored pollen from other plants
either through the mesh of the exclusion bags or under the bags. Insects were
observed under the bags during the course of the study.
Self-compatibility is weakly supported by the small number of seeds produced
in bagged plants at China Hill in the pollinator exclusion experiment, but not
supported by the results of the hand-pollination experiment (no seeds produced in
flowers that were bagged nor in flowers that were hand-pollinated with pollen from
flowers on same plant). However, selfing could be occurring in some P. hirsuta
52
plants. Brown's (1990) mating system types are believed to exist (in species) on a
continuum. For example, some species are facultatively xenogamous; they can
outcross when pollinators are available and self-fertilize when there is a lack of
pollinators (Goodwillie 2001). The mating system of a plant species can vary between
years or between populations, and even individuals within a population may vary in
the amount of selfing or outcrossing (Lyons & Antonovics 1991). For example, a
plant flowering near either end of the phenology curve may be predominantly selfpollinated while plants flowering during the peak are predominantly outcrossed. Also,
rare color morphs may receive fewer pollinator visits and thus, self-pollinate more
(Brown & Clegg 1984). The low frequency of insect visits observed on P. hirsuta
flowers over two seasons (165 insects at China Hill and 5 insects at Cracker Gulch)
suggests that P. hirsuta may be partially self-compatible. Self-compatibility is
considered an adaptation to sparse or ineffective pollinators and as a means of
escaping competition for pollinators (Grant & Grant 1965, Proctor & Yeo 1973).
If P. hirsuta were self-compatible to some degree, there still needs to be a
mechanism for the transfer of pollen to the stigma within the flower or between
flowers on the bagged plant in order for seed to set. Small insects could vector pollen
on plants under exclusion bags (Kearns & Inouye 1993). In this study, insects, such as
thrips (Thysanoptera) (1-2mm wide), small beetles (Chrysomelidae) (3-5mm),
leafhoppers (Membracidae), and spiders were observed inside bagged plants. In fact,
a small amount of fruit predation occurred in both bagged plants and unbagged plants
(7.5 % of the plants) both sites, particularly China Hill, in 2004. In all cases, a
53
prematurely browned ovary/capsule shell was found with a neat round hole in it. On
one occasion, a small, green Lepidopteran larva was observed chewing into P. hirsuta
ovaries.
Insects on bagged plants may have occasionally transferred enough pollen to
promote self-fertilization in bagged plants, if indeed P. hirsutais capable of selfing.
However, the results of the hand-pollination test (no seeds produced on bagged shoots)
do not support this. Another possibility is that seeds produced inside bagged plants
were not a result of selfing but rather a result of insects moving between plants
through the exclusion bags and vectoring pollen from another plant (outcrossing). The
mesh size of the exclusion bags (0.7mm x 0.4mm) was small enough to exclude most
insects but may have still allowed minute insects (such as thrips) access. Another
possible explanation for seed production inside bags could be Lepidopterans probing
flowers through the mesh.
The fact that some insects were present on bagged plants and observed
crawling in and out of flowers yet few seeds were produced, is strong evidence that P.
hirsuta is primarily self-incompatible or at least facultatively xenogamous.
Facultative xenogamy refers to flowers that can be cross-pollinated if pollinators are
present but will self-pollinate, if they have the capacity to do so, in the absence of
pollinators (Kearns & Inouye 1993).
Apomixis refers to reproduction without sex either through agamospermy
(non-sexual formation of an embryo) or vegetative reproduction (Kearns & Inouye
1993). Since no pollen is required, plants that are apomixic are capable of producing
54
seeds from bagged and emasculated (anthers removed) flowers. The results of the
hand-pollination experiment do not indicate P. hirsuta employs apomixis as a mating
system because no seeds were set in the bagged shoots or the bagged & self-pollinated
shoots.
P. hirsuta appears to require a pollen vector to set fruit. Animals, wind, and
water can carry pollen. This study did not address whether wind vectored P. hirsuta
pollen. Springer & Tyrl (1989) inferred wind to be a vector inducing selfing in their
study on Phlox oklahomensis, a species that grows in similar environmental conditions
as P. hirsuta. The insect-exclusion cages used in their study might have reduced wind
velocity necessary for flower agitation to induce self-pollination. It is doubtful that
wind serves as an effective "internal" pollen vector to induce self-pollination in P.
hirsuta because the hand-pollination experiment did not reveal a capacity for selfing in
P. hirsuta, even when flowers were hand-pollinated with pollen from other flowers on
the same plant. Moreover, it is unlikely P. hirsuta is anemophilous (outcrossing
occurs via wind-pollination). The pollen of wind-pollinated plants has a smooth, dry
surface while entomophilous (insect-pollinated) plants have sticky, highly ornamented
pollen (Proctor & Yeo 1973). Phlox hirsutapollen grains have sticky oil bodies on
their surface, which inhibits dispersal by wind.
Based on experimental results, field observations, pollen type, and inspection
of insects for pollen, insects are inferred to be the pollen vectors of P. hirsuta. Grant
(1961) stated that most Phlox species are outcrossers pollinated by Lepidopterans. He
based his inferences on observations and morphological characteristics of flowers,
55
such as corolla tube length matching proboscis length of different species. For
example, a number of Lepidopterans, including Monarch (Danaus), Skipper (Polites),
Sulfur, Swallowtail (Papilio),and a diurnal moth (Scepsis) were observed visiting
Phlox glaberrima, P. divaricataand P. bifida. Plitmann & Levin (1990) inferred 24
Phlox species to be predominantly xenogamous (self-incompatible) and Lepidopteranpollinated based on pollen-ovule ratios and stigmatic pollen germination percentages.
Few pollination studies have been done on Phlox species. Wiggam-Harper
(2003) found a moth, Hemaris diffinis (Sphingidae), to be the most effective pollinator
of P. divaricata. Phlox divaricatawas shown to be primarily xenogamous but have
some minimal capability for autogamy (Wiggam-Harper 2003). A study on P.
oklahomensis (Springer & Tyrl 1989) revealed its primary mating system to be
autogamy (selfing). Insect visitation to this Phlox species was rare but one Sphingid
moth, Hemaris diffinis, was found to be carrying P. oklahomensis pollen.
The results from this study indicate that Lepidopterans are the primary
pollinators of P. hirsuta. Lepidopterans observed visiting P. hirsutaand also found to
be carrying Phlox pollen grains on their probosces include Proserpinusclarkiae
(Sphingidae), Hyles lineata (Sphingidae), Vanessa cardui (Nymphalidae), and
members of the subfamily Hesperiinae (Hesperiidae). Phlox hirsuta appears to have a
facultative plant-pollinator relationship. That is, it does not depend on just one insect
species for pollination services.
Some Sphingids visited Phlox hirsutaplants consecutively, a behavior that
maximizes pollen transfer. The number of Sphingids observed visiting P. hirsutain
56
this study may not accurately portray frequency of visitation. These moths have
rapidly beating wing movements while they feed, making them more difficult to see.
In addition, they are diurnal and crepuscular feeders, and they may feed more at
crepuscular times when it is more difficult to see them. Two of the three Hyles lineata
specimens observed in 2005 were observed during crepuscular evening periods.
In 2005, Vanessa cardui (Nymphalidae) was observed probing P. hirsuta
much more frequently than in 2004. A population explosion of V. cardui occurred in
2005. Thus, this increase in V carduivisitation may reflect the greater overall number
of V. cardui observed in northern California and southern Oregon in 2005. Some V
cardui individuals visited P. hirsutaplants consecutively but the majority appeared to
visit Phlox sporadically and interchangeably with a host of other plant species. Pollen
checks support the observation results; pollen from other plants was found on many
specimens and Phlox pollen loads were lower (range of 4-50 pollen grains) than those
on Sphingids. In sum, based on observations of insect behavior (but not frequency of
visits) and checks for pollen load, V cardui appear to be an effective, but inconsistent
pollinator of P. hirsuta.
Members of the subfamily Hesperiinae (Skippers) appear to be occasional
pollinators of P. hirsuta. Although several specimens were observed probing P.
hirsuta and most of those collected were carrying Phlox pollen, the pollen load was
very low (range of 2-10 pollen grains). However, it is possible that these specimens
were carrying more pollen grains on their faces since their proboscis was just barely
57
long enough to access nectar. Their faces would have been submerged in anthers as
they fed.
Other insects observed probing P. hirsuta or carrying Phlox pollen (Table 7)
are not believed to contribute greatly to the pollination of P. hirsuta but rather, based
on observations are likely to be occasional or incidental pollinators.
It should be noted that the pollen load measurements made in this study were
estimates based on the number of grains from a proboscis, cut at the base and viewed
on the microscope slide. Levin & Berube (1972) utilized this method, and stated that
some loss of pollen grains probably occurred during proboscis coiling and recoiling
prior to examination.
General insect activity at the sites between the two years was variable.
Preliminary observations suggest that insect activity is related to weather conditions.
Daily and yearly fluctuation in insect visitors has been noted on other Phlox species
growing in similar environmental conditions (Springer & TyrI 1989, Wiggam-Harper
2003). In general, more observation hours took place in 2005, and consequently, more
insects were observed. In addition, the population explosion of Vanessa cardui in
2005 accounted for a larger number of observations.
This study gathered preliminary data on the possible pollinators of P. hirsuta.
A likely extension of this would be to determine which pollinators are most effective.
The most common method to determine pollinator efficiency is to measure stigma
pollen loads after pollinator visitation. However, this method involves careful control
and manipulation of the plants (i.e. often growing them in greenhouses and aborting
58
flowers to measure pollen load) and given that P. hirsuta is an endangered species, it
may not be feasible.
In both years, plants at Cracker Gulch produced a lower percentage of
fruits/seeds per flowers than plants at China Hill. Adams (1987) also found this to be
the case when he studied P. hirsuta populations at China Hill and Soap Creek Ridge (a
higher elevation site similar to Cracker Gulch). A complex array of factors influence
fruit set in plants, including population size/density, plant size, geography,
microclimate, climate, and insect activity patterns (Fenner 1985, Kearns & lnouye
1993, Kyhl 1997, Schwaegerle & Levin 1990, Wolf 2001). Cracker Gulch is a smaller
population made up of smaller-sized plants than at China Hill. Parameters (diameter,
height) of plants in the pollinator exclusion experiment were measured both years and
the diameter of individual Cracker Gulch plants was
1/4
to
12
the size of China Hill
plants. Larger plants have been shown to have a greater percent seed set, thus
contributing more to production of offspring (Dudash 1991). Studies have shown
lower rates of pollen exchange and lower numbers of pollinators in smaller or
noncontiguous populations (Kyhl 1997, Rathcke & Jules 1993, Wolf 2001).
Population size and density may play an important role in attracting pollinators and
promoting pollen exchange (Kearns & Inouye 1997, Paschke et al. 2002, Rathcke &
Jules 1993). Plants in the center of a dense population may be mostly outcrossed;
differently sized populations may differ in primary mating system used (Kyhl 1997) or
differences occur between populations of outcrossing species if pollinators are
common at one location but rare in another (Goodwillie 2001, Kearns & Inouye 1993).
59
In addition, the suite of floral visitors tends to change along an ecogeographical
transect so two populations may have different pollinators (Grant & Grant 1965).
Cracker Gulch is also at a higher elevation and consequently experiences a harsher
climate, which may inhibit pollinator visitation. The insect observations in this study
indicate that insect activity appears to be much lower at Cracker Gulch than at China
Hill. It is possible that environmental conditions, such as resource availability, soil
type, temperature, and light, are more favorable for P. hirsuta at China Hill.
Fruit set in P. hirsutaplants at both sites was fairly low both years. In 2004,
fruit set was 50% and 37% at China Hill and Cracker Gulch, respectively. In 2005,
fruit set was 29% and 15% at China Hill and Cracker Gulch, respectively. Fruit set
can be much higher, 80-100%, in plant populations that are well adapted to their
environmental conditions. Adams (1987) found a small sample of P. speciosa,
another Phlox species that grows at China Hill and Cracker Gulch, to have over 90%
fruit set. Springer & Tyrl (1989) found open-pollinated P. oklahomensis to have
81.4% fruit set.
The results from this study suggest P. hirsuta reproductive output may be
pollen-limited and/or resource-limited. Results from the hand-pollination experiment
indicate that both pollen limitation and resource limitation affect reproductive yield in
P. hirsuta. Seed set in bagged flowers hand-pollinated with outcrossed pollen was
17% higher (though not significantly) than in unbagged, naturally pollinated flowers,
providing evidence of pollen limitation in this species. That is, flowers appear to have
the potential and resources to produce more seeds than they do under natural
60
conditions but are limited by the amount of pollen vectored by animal visitors. The
amount and quality of pollen received by a plant influence the number and quality of
seeds produced (Zimmerman & Pyke 1988). Pollination failure may be due to either
the scarcity of pollinators, pollinator inefficiency (i.e. not carrying a sufficient pollen
load, not visiting consecutive plants of same species), or poor pollen quality. If poor
pollination (pollen limitation) is responsible for low seed set, hand-pollination can
cause as much as a tenfold increase in seed set (Burd 1994). Wilson et al (1979)
tested pollen limitation in Phlox divaricata, and found 58% of the plant's flowers set
seed in nature and 82% of the flowers set seed when hand-pollinated. Hand-pollinated
flowers of the serpentine endemic, Helianthus exilis, produced 11.7% more seeds than
naturally-pollinated flowers (Wolf 2001). Linanthusparviflorus, a self-incompatible
member of the Polemoniaceae family experienced a 9-117% increase in seed set when
hand-pollinated (Goodwillie 2001). Polemoniumfoliosissimum, another
Polemoniaceae family member, experienced a 10% increase in seed set among plants
that were hand-pollinated (Zimmerman & Pyke 1988). The studies indicate a positive
effect of extra pollen on the reproductive output of the plant species, and a possibility
of pollen limitation. It is also possible that the number of seeds produced by a plant
species is limited by both the amount of pollen received and by the availability of
other resources (Zimmerman & Pyke 1988).
Hand-pollinating can result in seed set of 80-100% if the plant species is not
resource limited. The low seed set (40%) in the bagged & outcrossed treatment
indicates Phlox hirsuta may also be limited in the amount of resources it can allocate
61
to reproduction. Resource limitation is a common occurrence in nutrient-poor soils
such as serpentinite (Gordon & Lipman 1926, Mason 1946a, Walker 1954). Resource
deficiency can limit reproductive capacity. For example, Cochleariapyrenaica,a
native plant found on ultramafic soils, set more fruit and seed when major nutrients
were added (Nagy & Proctor 1997). Resource deficiency can also be revealed when
the pollination factor is removed (pollination is maximized), and hand-pollinated and
naturally pollinated plants produce a similar seed yield (Fenner 1985).
Alternatively, the lower than expected seed set in the hand-pollinated sample
could have been caused by inadequate hand-pollination technique or by negative
effects of bagging individual shoots for hand pollinating. Contrary to the bagging
method used in the pollinator exclusion experiment, this bagging method entailed
securing a small bag around the stem of one shoot. The small bag may have altered
the microclimate, perhaps holding in more heat and humidity. A negative bagging
effect was believed to influence the low seed set in hand-pollinated inflorescences of
Cochleariabavarica (Paschke et al. 2002).
Therefore, although it cannot be concluded from this study how much P.
hirsuta is affected by pollen or resource limitation, they appear to play roles in the
reproductive capacity of P. hirsuta. Further studies investigating factors causing poor
fruit set are necessary.
62
Pollen Types
G. Erdtman (1966) describes Polemoniaceae pollen as variable, ranging from
porate (pores) to forate (furrows) in aperture type and suboblate to spheroidal in shape.
However, pollen from members of the Phlox genus is more similar, having polyporate
apertures (pores scattered over the surface), reticulate sculpture, pore numbers in
approximately equal numbers, and an average size of 40 pim (Taylor & Levin 1975).
Pollen of P. hirsuta could not be distinguished from pollen of another Phlox species,
P. speciosa. Phlox speciosa is present at both study sites but only in small numbers on
the edge of the P. hirsuta occurrences. However, most insects were collected in solid
(not intermixed) patches of P. hirsuta, and it was inferred that Phlox pollen found on
collected insects was P. hirsuta pollen. None of the Lepidopterans believed to be
pollinators of P. hirsuta were observed visiting P. speciosa. However, more studies
are needed to verify this. Also, scanning electron microscopy (SEM) could be used to
elucidate the differences between pollen of the two Phlox species.
Taylor & Levin (1975) illustrated how pollen size among species of the
Polemoniaceae influences what animals can carry it. According to their data, P.
hirsuta, with a mean diameter of 42 pim, falls within the category that is most likely to
be pollinated by a moth, bee, beefly, or butterfly. Potential pollinators within the
Polemoniaceae family include butterflies, moths, bees, beetles, flies, hummingbirds,
and bats (Taylor & Levin 1975). The diversity of pollinators may reflect the
morphological diversity of pollen in the Phlox family. Other floral traits, besides
63
pollen size, affect what animals vector pollen. Taylor's data shows that butterfly
mediated pollen is the smallest size. There does not appear to be a statistical
correlation between pollen sculpturing types and mating systems. For example, P.
drummondii and P. cuspidata have similar pollen sculpture yet the former is selfincompatible and Lepidopteran pollinated and the latter is self-compatible (Taylor &
Levin 1975). Thus, it cannot be inferred that P. hirsuta and P. speciosa employ the
same mating system or the same pollen vectors.
Seed Traits,Viability, and Germination
The seed traits of P. hirsuta are, as most seeds from the Polemoniaceae family,
in the non-endospermic seed group. This seed type has a reduced endosperm or lacks
endosperm altogether, is 2-8 mm in size, has large cotyledons and an embryo that
occupies most of the seed (Atwater 1980). Seeds are protected by thick woody coats,
which are readily permeable to water but remain semi-permeable or impermeable to
other chemical substances. In addition, seeds in this group may contain chemical
inhibitors, which play an important role in controlling seed germination in the field.
The results of the Tetrazolium Chloride treatment and the germination
treatment in 2004 suggest that a high number of seeds produced by P. hirsuta plants
are viable, and confirm that effective fertilization of flowers occurred under natural
conditions. Two-thirds to 3/4 of the total seeds produced by P. hirsuta plants were
viable in the Tetrazolium Chloride test. While there appears to be a higher percentage
64
of viability at Cracker Gulch than at China Hill in the Tetrazolium Chloride test,
further investigation is necessary to determine if P. hirsuta plants at Cracker Gulch
produce a greater percentage of viable seeds than other sites. The unequal sample
sizes (100 seeds from China Hill and 25 seeds from Cracker Gulch) and small sample
size at Cracker Gulch limit comparison between the two sites.
While TZ tests often reveal higher viability than germination tests due to
unknown germination requirements, seed dormancy and other difficulties with
germination in the lab (Kaye 2005, Grabe 1970), this was not the case with P. hirsuta
seeds in this study. It is possible that some of the viable seeds were not identified in
the TZ test analysis.
The main reason the germination test was done in 2005 was to compare
viability levels between seeds from unbagged plants and bagged plants. Since the
bagged seeds produced a total of 14 seeds, the small sample size (14 seeds each) limits
inferences that can be made from the results. In addition, it is unknown whether the
seeds from bagged plants were produced by selfing or by insect vectors outcrossing
pollen.
The results of the germination treatment indicate that Phlox hirsuta seeds
possess a nondeep physiological dormancy, which is broken by warm summer
temperatures (Baskin 2005). Herbaceous perennials from the temperate
steppe/grassland zone (such as China Hill) are known to have seeds that are dormant
at maturity (Baskin & Baskin 2001). Most of these perennials have seeds that require
cold stratification (due to physical dormancy) while others (P. hirsuta) germinate after
65
dry storage at room temperatures, indicating that after-ripening occurred to break the
nondeep dormancy (Baskin & Baskin 2001). Seed storage in paper bags at room
temperature (20'C or higher) for two to six months, as P. hirsuta seeds in this study,
simulates an after-ripening period (Drake et al. 1998).
Phlox hirsuta seeds appear to come out of dormancy during the warm, dry
summer and require cool temperatures and moist conditions, and most likely
germinate in autumn in the field. Baskin (2005) found a similar dormancy and
germination pattern in Phlox bifida. In another study, neither Phlox drummondi nor
Phlox maculata seeds were found to be dormant (Atwater 1980). Seeds of these
species had an 82% germination rate and a variable germination rate, respectively, at
15-20'C with fungicide treatment in a 21-28 day period. In contrast, seeds from other
species of Phlox have been found to possess deeper dormancy. For example, P.
diffusa, a high elevation species, had a germination rate of just 30% after 16 weeks of
cold stratification (Kaye 1997). Phlox oklahomensis was found to be dormant and
require at least two weeks of cold/moist stratification to achieve a maximum
germination rate of 60% (Springer & Tyrl 1989).
The fact that P. hirsutagerminated at a low temperature of 3° C suggests, in its
natural habitat, this species germinates in fall or early winter. This might explain why
no one has observed seedling establishment in the field (Adams 1987, USFWS 2004).
Alternatively, seedlings may not be observed because of seed/seedling predation, lack
of overwintering survivors, or other factors. Although P. hirsuta seeds do not appear
to have innate deep dormancy there may be enforced dormancy- a condition when
66
viable seeds don't germinate because of some limitation in the environment (Roberts
1972). Future studies would be worthwhile to elucidate the factors in P. hirsuta seed
establishment and the level of reproductive fitness (producing a new generation that
goes on to be reproductively successful itself) in P. hirsuta.
Conclusions
The results from this study provide some baseline information on the
reproductive biology of P. hirsuta. Phlox hirsuta appears to employ a predominantly
outcrossing mating system, and has a facultative pollinator relationship with several
Lepidopterans. Areas of future study of the mating system include testing stigma
receptivity and pollen viability, and examining correlations between insect activity,
weather patterns, and P. hirsutaphenology. Also, a sequential bagging experiment
could be done to determine if there is a time within the bloom period that most
pollination occurs. Another area of future study would be to investigate the floral
scent biology of P. hirsuta. When the chemical components that make up a flower's
fragrance are known, they can be linked to the suite of insects likely to be attracted to
those compounds.
Reproductive yield (fruit and seed set) of P. hirsuta was relatively low, which
indicates that pollen and/or resource availability may be limited. In addition, flowers
produced only one seed each in this study, although they are morphologically suited to
produce three seeds each. Fertilized fruits yielded a high percentage of viable seed,
67
which germinated readily in the lab. Further studies could reveal whether factors,
such as low pollen load delivery, seed/seedling predation, or limited resources
negatively impact seed production and/or germination in the field.
This research may help direct conservation efforts. According to Bond's
(1994) vulnerability index for plants, P. hirsuta has two of the four characteristics,
self-incompatibility and propagation by seed, which put a species most at risk of
extinction. In addition, the risk of pollinator or disperser failure, dependence on
pollinators, and dependence on seeds for reproduction are all factors to consider in
assessing risk of extinction (Bond 1994). Since a propagation effort is limited by the
amount of available of serpentine habitat on which there is minimal competition from
other plant species, an important management goal for P. hirsuta would be to conserve
not just the plant, but the plant-pollinator relationship.
The effective population size of a population is the size at which reproductive
fitness is ensured through sufficient pollination, fertilization, and a diverse gene pool.
Outcrossing plants usually require a larger effective population size to attract
sufficient pollinators and to preserve genetic diversity. Based on the results from this
study, P. hirsuta is predominantly xenogamous, which means that in order to conserve
a diverse gene pool, the conservation of large populations or several smaller
subpopulations should be a priority. Small populations of outcrossing plants are more
susceptible to inbreeding depression (Kearns & Inouye 1997).
The fact that P. hirsuta has a facultative plant-pollinator relationship is to its
advantage. However, the low frequency of insect visits to P. hirsuta may be indicative
68
of pollinator scarcity. A scarcity of pollinators may result in more seed set through
self-pollination, which results in less vigorous offspring (Kearns & Inouye 1997). In
addition, xenogamous plant species that are infrequently visited by pollinators have
been shown to experience a marked decline in pollination when their habitat becomes
fragmented (Rathcke & Jules 1993). Pollinator habitat should be protected and in the
case of P. hirsuta, plant species that are hosts to Lepidopteran larvae should be
conserved.
LITERATURE CITED
Adams, 0. October 1987. Phlox hirsuta habitat and species viability inventory.
Unpublished report prepared for the Klamath National Forest.
Atwater, B.R. 1980. Germination, dormancy and morphology of the seeds of
herbaceous ornamental plants. Seed Science & Technology 8: 523-573.
Barrett, S.C.H. 1990. The evolution and adaptive significance of heterostyly. Trends in
Ecology and Evolution 5(5): 144-148.
Baskin, C.C. 2005. Professor at the University of Kentucky, Environmental &
Evolutionary Biology program. Lexington, Kentucky. Personal
Communication. (email 2/20/05).
Baskin, C.C. and J.M. Baskin. 2001. Seeds: ecology, biogeography and evolution of
dormancy and germination. Academic Press. San Diego, California.
Bond, W.J. 1994. Do mutualisms matter? Assessing the impact of pollinator and
disperser disruption on plant extinction. Philosophical Transactions of the
Royal Society, London B Biological Sciences 344:83-90.
Borror, D.J. and D.M. DeLong. 1964. An Introduction to the Study of Insects. rd Ed.
3
Holf, Rinehart, & Winston, Inc.
69
70
Brown, A.H.D. 1990. Genetic characterization of plant mating systems. Pgs. 145-162
in A.H.D. Brown, M. Clegg, A. Kahler, and B. Weir, eds. Plant population
genetics, breeding, and genetic resources. Sinauer Associates, Sunderland,
Massachusetts.
Brown, B.A. and M.T. Clegg. 1984. Influence of flower color polymorphism on
genetic transmission in a natural population of the common morning glory
Ipomoea purpurea. Evolution 38(4): 796-803.
Burd, M. 1994. Bateman's principle and plant reproduction: the role of pollen
limitation in fruit and seed set. Botanical Review 60: 83-139.
Drake, D., Ewing, K. and P. Dunn. 1998. Techniques to promote germination of seed
from Puget Sound prairies. Restoration & Management Notes 16(1): 33-40.
Dudash, M.R. 1991. Plant size effects on female and male function in hermaphroditic
Sabatia angularis(Gentianaceae). Ecology 72(3): 1004-1012.
Erdtman, G. 1966. Pollen morphology and plant taxonomy: Angiosperms. Hafner
Publishing Company, New York.
Fenner, Michael. 1985. Seed Ecology. Chapman and Hall Ltd., London, Great Britain.
Ferguson, Carol. 2004. The Importance and identification of insect pollinators of the
Yreka Phlox (Phlox hirsutaE.E. Nelson). 1st Progress Report. Unpublished.
Goodwillie, C. 2001. Pollen limitation and the evolution of self-compatibility in
Linanthus (Polemoniaceae). International Journal of Plant Science 162(2):
1283-1292.
71
Gordon, A. and C.B. Lipman. 1926. Why are serpentine and other magnesian soils
infertile? Soil Science 22: 291-302.
Grabe, D.F. 1970. Tetrazolium testing handbook for agricultural seeds. Contribution
No. 29 to the Handbook on Seed Testing, prepared by the Tetrazolium Testing
Committee of the Association of Official Seed Analysts.
Grant, V. 1961. The diversity of pollination systems in the Phlox family. Recent
Advances in Botany (Toronto) I:55-W.
Grant, V. and K.A. Grant. 1965. Flower Pollination in the Phlox Family. Columbia
University Press. 180 pp.
Heard, S.B., Campbell, M.A., Bonine, M.L., and S.D. Hendrix. 1999. Developmental
instability in fragmented populations of Prairie Phlox: a cautionary tale.
Conservation Biology 13(2): 274-281.
Hendrix, S.D. and J.F. Kyhl. 2000. Population size and reproduction in Phlox pilosa.
Conservation Biology 14(1): 304-313.
Kaye, T.N. 1997. Seed dormancy in high elevation plants: implications for ecology
and restoration. p. 1 15-120. In: A. Liston, R.M. Love, D.L. Luoma, R.J.
Meinke, and M.V.Wilson (eds). Conservation and management of native plants
and fungi. Native Plant Society of Oregon, Corvallis, Oregon.
Kaye, T.N. 2005. Executive Director of the Institute for Applied Ecology. Corvallis,
Oregon. Personal Communication. (email 2/20/05).
Kearns C.A. and D.W. Inouye. 1993. Techniques for pollination biologists.
University Press of Colorado. Niwot, Colorado. 583 pp.
72
Kearns C.A. and D.W. Inouye. 1997. Pollinators, flowering plants, and conservation
biology. Bioscience 47(5): 297-307.
Kelly, M.G. and D.A. Levin. 2000. Directional selection on initial flowering date in
Phlox drummondii (Polemoniaceae). American Journal of Botany 87(3): 382391.
Kruckberg, A.R. 1984. California serpentines: flora, vegetation, geology, soils, and
management problems. University of California Press. 180 pp.
Kyhl, J.F. 1997. The effects of habitat fragmentation on Phloxpilosa. M.S. thesis.
University of Iowa, Iowa City.
Levin, D.A. 1970. The Exploitation of pollinators by species and hybrids of Phlox.
Evolution 24:367-377.
Levin, D.A. 1972. Pollen exchange as a function of species proximity in Phlox.
Evolution 26:251-258.
Levin, D.A. and D.E. Berube. June 1972. Phlox and Colias: The Efficiency of a
Pollination System. Evolution 26: 242-250.
Lyons, E.E. and J. Antonovics. 1991. Breeding system evolution in Leavenworthia:
breeding system variation and reproductive success in natural populations of
Leavenworthia crassa (Cruciferae). American Journal of Botany 78(2):270287.
Mason, H.L. 1946a. The Edaphic factor in narrow endemism. I. The nature of
environmental influences. Madrono 8: 209-226.
73
Moore, P.D., J.A. Webb, and M.E. Collinson. 1991. Pollen Analysis. 2nd Ed. Oxford
Press. Maiden, MA.
Nagy, L. and J. Proctor. 1997. Plant growth and reproduction on a toxic alpine
ultramafic soil: adaptation to nutrient limitation. New Phytologist 137: 267274.
Paschke, M., Abs, C., and B. Schmid. 2002. Effects of population size and pollen
diversity on reproductive success and offspring size in the narrow endemic
Cochleariabavarica (Brassicaceae). American Journal of Botany 89(8): 12501259.
Patterson, R. and D.Wilken. 1993. Phlox. In: Hickman J. (ed.). The Jepson manual:
higher plants of California. University of California Press, Berkeley,
California. 1,400 pp.
Plitmann U. and D.A. Levin. 1990. Breeding Systems in the Polemoniaceae. Plant
Systematics and Evolution 170: 205-214.
Proctor, M. and P. Yeo. 1973. The pollination of flowers. Taplinger Publishing Co.
New York, New York.
Rathke, B.J. and E.S. Jules. 1993. Habitat fragmentation and plant-pollinator
interactions. Current Science 65(3): 273-277.
Roberts, E.H. 1972. Viability of seeds. Syracuse University Press.
Roy, Bitty. 2004. Associate Professor at the Center for Ecology & Evolutionary
Biology, University of Oregon. Personal Communication.
74
Schwaegerle, K.E. and D.A. Levin. 1990. Environmental effects on growth and fruit
production in Phlox drummondii. Journal of Ecology 78: 15-26.
Scott, J.A. 1997. Butterflies of North America: a natural history and field guide.
Stanford University Press, Palo Alto, California.
Skinner, M.W. and B.M. Pavlik. 1994. California Native Plant Society inventory of
rare & endangered plants of California. Special Publication No. 1. 5th Ed.
California Native Plant Society.
Soil Conservation Service. 1983. Soil Survey of Siskiyou County, California, Central
Part. U.S. Department of Agriculture, Washington, D.C.
Springer, T.L. and R.J. Tyrl. 1989. Distribution, habitat, and reproductive biology of
Phlox oklahomensis Wherry (Polemoniaceae). Proc. Okla. Aca. Sci. 69: 1521.
Stuchlick, L. 1967. Pollen morphology in the Polemoniaceae. Grana Palynologica 7:
146-240.
Taylor, T.N. and D.A. Levin. 1975. Pollen morphology of Polemoniaceae in relation
to systematics and pollination systems: scanning electron microscopy. Grana
15: 91-112.
U.S. Fish and Wildlife Service. 2004. Draft Recovery Plan for Phlox hirsuta (Yreka
Phlox). Portland, Oregon.
Walker, R.H. 1954. Factors affecting plant growth on serpentine soils. .Part II. In. The
ecology of serpentine soils. Ecology 35(2): 275-286.
75
Wiggam-Harper, S.D. 2003. The pollination and floral scent biology of Phlox
divaricataL. (Polemoniaceae). M.S. thesis. Kansas State University,
Manhatten, Kansas.
Willson, M.F., Miller, L.J., and B.J. Rathcke. 1979. Floral display in Phlox and
Geranium: adaptation aspects. Evolution 33: 52-63.
Wherry, E.T. 1955. The Genus Phlox. Morris Arboretum Monographs III. Morris
Arboretum of the University of Pennsylvania. Philadelphia, Pennsylvania.
Wolf, Amy. 2001. Conservation of endemic plants in serpentine landscapes.
Biological Conservation 100: 35-44.
Zimmerman, M. and G.H. Pyke. May 1988. Reproduction in Polemonium: assessing
the factors limiting seed set. The American Naturalist 131(5): 723-738.
APPENDIX A
PLANT COMMUNITIES
76
77
China Hill Plant Community
(*= in plot)
Trees
Juniperus occidentalis
Shrubs
Ceonothus cuneatus
Chrysothamnus nauseosus*
Eriogonum sphaerocephalumvar.
halimoides *
Herbaceous plants
Achillea millefolium
Allium sp. (siskiyouense orfalcifolium) *
Amsinckia intermedia
Arabis puberula*
Astragaluspurshii var. tinctus *
Blepharipappusscaber
Castillejahispida
Clarkia gracilis
Claytonia exigua
Claytonia lanceolata
Collinsiaparviflora
Crepis occidentalis
Crocidium multicaule *
Dichelostemma capitatum
Epilobium sp. *
Erigeron bloomeri var. bloomeri
Eriogonum strictum var. proliferum*
Eriophyllum lanatum var.
achillaeoides*
Eschscholzia californica*
Fritillariapudica*
Gilia capitatum*
Lasthenia californica
Lewisia rediviva*
Linanthus sp. (ambiguous?)
Lithophragmaparviflora
Lomatium macrocarpum*
Lupinus argenteus
Minuartiadouglasii*
Orobanche uniflora*
Phaceliacorymbosa*
Phlox hirsuta*
Phlox speciosa
Plagiobothryssp. *
Sisyrinchium douglasii
Swertia albicaulis
Thysanocarpus curvipes*
Uropappuslindleyi (or Microseris
sp. ?)
Viola beckwithii*
Grasses
Achnatherum thurberianum
Bromus madritensisssp. rubens *
Bromus tectorum *
Elymus elymoides
Elymus multisetus*
Festuca idahoensis*
Poa bulbosa*
Poa sp. *
Vulpia microstachys*
78
Cracker Gulch Plant Community
(*= in plot)
Trees
Calocedrus decurrens*
Juniperus occidentalis
Pinusjeffreyi*
Pseudotsugamenziesii
Quercus garryana
Phlox gracilis
Phlox hirsuta*
Phlox speciosa
Plagiobothryssp. *
Plectritismacrocera
Thelypodium sp.
Thysanocarpus curvipes *
Viola purpurea*
Shrubs
Ceonothus cuneatus
Cercocarpusledifolius*
Chrysothamnus nauseosus*
Herbaceous Plants
Achillea millefolium*
Allium sp. (siskiyouense orfalcifolium) *
Arabis puberula*
Calochortus tolmei*
Castilleja hispida
Chlorogalum sp. *
Claytonia rubra
Collinsiarattanii
Epilobium sp. *
Eriogonum strictum *
Eriophyllum lanatum*
Erysimum capitatum
Fritillariaaffinis
Fritillariapudica*
Fritillariarecurva
Gilia capitata*
Lithophragmaparviflora
Lomatium macrocarpum
Lomatium triternatum
Lupinus sp.
Minuartianuttalli*
Penstemon sp. *
Phaceliacorymbosa*
Grasses
Agropyron spicatum
Bromus tectorum *
Elymus glaucus*
Festuca idahoensis*
Poa sp.
Vulpia microstachys var. pauciflora
APPENDIX B
POLLEN REFERENCE COLLECTION
79
I
Pollen Reference Collection
Slide
Number
1
2
3
4
5
6
7
10
I1I
------------------12
13
14
15
16
17
18
19
20
21
22
23
24
25
Plant Name
SiteCH (China Hill,
CG (Cracker Gulch)
Alilium siskiyouense
CH, CG
Arabis puberula
CH, CG
Astragaluspurshii
CH
7
1
Castilleiahispida
CH-ICG
CG
CH
Ceonothuscuneatus
CH
Chiorogalum sp.
CG
Clark~ia gracilis
CH
Crocidium mulhicaule
CH
Dichelostemmacapiatwum
CH
Epilobium sp.
~CH, -CG
Erigeron
bloomeri7-I
CH
t
Eriophyllum lanatum
CH, CG
Eriogonumsphaerocephalum I CH
Eschscholziacaifornica
CH
Gilia capitata
Linanthus sp.
CH
Lomatium macrocarpum
CH, CG
Lupinus argenteus
CH
Minuart'iadouglasii
CH
Phaceliacorymbosa
CH, CG
Phlox hirsuta
LCH,:C:G~
Phlox speciosa
CH, CG
Plaiobothryssp.
CH, CG
Sisrynchiumdouglaszi
Violabeckwithii
-
-
*Plants in coincidental bloom with P. hirsuta were chosen for this reference set. This
set does not include all plants at the 2 sites, only blooming plants within 100 meters of
the plots and plants at the P. speciosa location (at China Hill).
. .....
I
2
1) Allium siskiyouense
Allium siskiyouense
soPM
2) Arabispuberula
I
.1mW
-
A
Oi.~ ~~
~m
3) Astragaluspurshii var. tinctus
4) Castillejahispida
4
5) Ceonothus cuneatus
6) Chlorogalum sp.
5
7) Clarkiasp.
~:
. -I
8) Crocidium multicaule
r
~
6
9) Dichelostemma capitatum
10) Epilobium sp.
7
11) Erigeronbloomeri
12) Eriophyllum lanatum
8
13) Eriogonum sphaerocephalum
14) Eschscholzia californica
9
15) Gilia capitata
16) Linanthus sp.
10
17) Lomatium macrocarpum
18) Lupinus argenteus
11
19) Minuartiadouglasii
20) Phaceliacorymbosa
12
21) Phlox hirsuta
Phlox hirsuta
50 Wn
22) Phlox speciosa
Phlox spectosa
50 pm
13
23) Plagiobothryssp.
24) Sisrynchium douglasii
Sisytinchium douglasli
60 pm
14
25) Viola beckwithii
Viola beckwithii
55PM