Demography, Ecology and Reproductiv, Biology of California ...

Demography, Ecology and Reproductiv, Biology of CaliforniaJewelflower (Caulanthus californicus: Brassicaceae)

Final Report to theState of California Department of Fish and Game

Contract # FG 146 [

Report Prepared by:

Dr. Susan J. MazerBeth A. Hendrickson, MA

Department of Biological SciencesUniversity of California

Santa Barbara, CA 93106

Telephone: 805-893-8011FAX: 805-893-4724

September 15, 1993

INTRODUCTION ........................... ,............................................................................... 1

STUDY SITES .................................................... _........................................................... 3

METHODS ...................................................................................................................... 4

1992 Field season ................ ........... ........................................................................... 4

.1993 Field season ...................................................................................................... 9

STATISTICAL ANALYSES ........................................... _............................. ;................ 12

RESULTS AND DISCUSSION ..................................................................................... 12

Observations on population structure and ecology ................................................... 12

Phenology and reproduction patterns ........................................................................ 13

Seedbank emergence ................. .............................. '.................................................. 15

Seed dormancy .......................................................................................................... 15

1992 Field Study Results ................................................................................................ 17

Comparison of study sites ......................................................................................... 17

Survival ..................................................................................................................... 22

Frequency Distributions of Reproductive Traits ....................................................... 23

Fecundity and factors affecting reproductive output ................................................. 24

Effects of individual plant identity on mean ovule and seed number perfruit ............................................................................................................................ 24

The relationship between plant size and reproduction .............................................. 25

Floral characteristics and reproductive output .......................................................... 26

Adjusted fecundity .................................................................................................... 28

Factors affecting seed number per fruit ..................................................................... 29

Spontaneous Self-Pollination .............................................. ...................................... 31

1993 Field and greenhouse study results ......................................................................... 32

Competition study .................. 2.................................................................................. 32

Causes of mortality in weeded and control plots ...................................................... 34

Reproductive components in weeded and control plots ............................ . ............... 37

General reproductivepatterns in 1993 ....................................................................... 39

Comparision of 1992 and 1993 fruit production ....................................................... 40

Greenhouse pollination studies ......................................................... :....................... 41SUMMARY AND MANAGEMENT RECOMMENDATIONS ................................... 44

Management considerations ...................................................................................... 45

Interspecific interactions ........................................................................................... 47

Giant Kangaroo Rat impacts .i................................................................................... 48Future research ...................................................................................... _................... 49

BIBLIOGRAPHY ................... ........................................................................................ 51

Figures(Figures l-25) .............................................................................................. 54

Appendix A: Maps .................................................................................................... 76

Appendix B:Photographs ......................................................................................... 80

Appendix C: Data _omfield and g_enhouse studies ............................................... 85

Acknowledgements

We would like to thank several people for their assistance in performing field andlaboratory studies and data analysis for this report. Lorne Wolfe conducted fieldstudies in 1992 and prepared two progress reports on that work. Jason Price assistedin field work in 1992 and 1993, performed much of the greenhouse work, and assistedwith data analysis. Mike Brown, Stephen Brown, Steve Silver, and John Damuth alsoprovided field assistance. Thek Moua conducted the greenhouse pollinations andassistance with seed counts and weights. In addition, we are grateful to Russ Lewisand Roy van de Hoek of the Bureau of Land Management and Dean Taylor ofBiosystems Analysis Inc. for their assistance in locating field sites and for sharingtheir expertise on Caulanthus californicus with us, and to Tony Nelson, ChuckWarner, and George Butterworth of The Nature Conservancy for kindly allowing usthe use of facilities at the Painted Rock Reseach Station on the Carrizo Plain.

September 15, 1993 Mazer/Hendrickson 1

INTRODUCTION

The following is the final report concerning the research we conducted on

California Jewelflower (Caulanthus californieus [S. Watson] Payson, Brassicaceae)

in 1992 and 1993 under contract to the California Department of Fish and Game. In

1986 this plant was classified as an endangered native California species under the

, Native Plant Protection Act by the California Department of Fish and Game. It is

also listed as endangered by the US Fish and Wildlife Service under the Endangered

Species Act.

California Jewelflower historically ranged from the southern end of Fresno

County and the western side of Tulare County to Bakersfield in the San Joaquin

Valley; it was also common in the Carrizo Plain and Cuyama Valley (Taylor and

Davilla, 1986; see Map l). Of 40 known historical locations in this range, Taylor and

Davilla found the species to be extant at only two sites, one of which was an artificial

population (the result of a reintroduction effort). Since 1986, Russ Lewis of the

Bureau of Land Management has located about 30 new populations in the Carrizo

Plain, 19 new populations in the Cuyama Valley and 9 populations on BLM lands in

the Kreyenhagen Hills to the north. Taylor and Davilla report that by 1986

agricultural development had caused the elimination of habitat representing 39% of

the historical populations of this species on the floor of the San Joaquin Valley.

Habitat destruction due to agriculture and grazing appears to be the primary cause for

the decline of the Jewelflower throughout its range. Habitat alteration due to the

invasion of non-native species may also be a contributing factor.

Habitat suitable for California Jewelflower is on gently sloping (generally less

than 20%) sites with loamy sand soils. Russ Lewis has noted a frequent association

of the species with soils that have some quartz content in the sand. Associated plant

communities are juniper woodland in the Cuyama Valley and Kreyenhagen Hills and

Ephedra scrub in the Carrizo Plain, both with understory species mostly consisting of

annual grasses and herbs. Common associated species included goldenbush

(Ericameria linearifolia), fiddlenecks (Amsinckia sp.), filaree (Erodium cicutarium),

red brome (Bromus rubens), and Arabian grass (Schismus arabicus). Caulanthus

californicus tends to occur in strongly aggregated colonies covering relatively smallareas.

September 15, 1993 Mazer/Hendrickson "_

California Jewelflower is an annual plant in the IVIustard family. Seedlings

possess highly characteristic 3-lobed cotyledons, which abscise as the basal rosette

forms. The juvenile stage is represented by a three- to eight-leaved rosette, which

lasts several weeks prior to bolting. When bolting, individuals produce one to several

flowering stems that usually branch from the base of the rosette. Dark red-purple

sepals on buds at the tips of the flowering stems later change to green and white as the

buds open. Flowers are white and pendant, and the fruits are long, flattened

(obcompressed) siliques. (see cover illustration from Abrams, 1944). The spherical

seeds possess no physical structures that aid dispersal; as the fruit dries, it splits

longitudinally to release the seeds below the parent plant.

The principal objective of our study was to provide quantitative data

concerning demographic and reproductive processes occurring within natural

populations. The primary objectives of the field studies in 1992 were:

(1) To quantify variation in l_lantsize within and among populations;

(2) To determine whether seeds produced in the field were viable and easily

germinable;

(3) To detect, if present, easily germinable seeds in the seed bank;

(4) To determine whether pollinator visitation is essential for seed production;

(5) To determine natural levels of potential reproductive success (flower production)

and realized reproductive success (fruit and seed production); and

(6) To quantify variation in and correlations among components of reproduction and

the traits that are correlated with reproductive success (e.g., plant size, flower

size, etc.)

Field work in 1993 included:

(I) A study of seedling demography; and

(2) A study of the effects of possible competition from native and non-native annual

grasses and herbs on survivorship and reproduction.

Greenhouse crosses were performed in 1993 to evaluate the effects of self-pollination,

cross-pollination within a population and cross-pollination between populations on

seed production per fruit and seed mass.


STUDY SITES

Field investigations in 1992 were focused on five populations in the Carrizo

Plain and Cuyama Valley: Cuyama, Foster Plot, Russ Plot, Population 4, and Travers

Ranch. We used two of the same populations in the 1993 study (Population 4 and

Foster Plot), in addition to another site (Population 6) selected because it contained a

larger population than was available at other sites. This relatively large population

size was necessary for our study of the possible effects of inter-specific competition

on survivorship and reproduction. The major criterion used in 1992 in selecting

populations as study sites was that the adult flowering population must include at

least 80 individuals; in 1993 we selected sites which had at least 200 flowering

individuals in 1992. Two of the poPUlations in the Carrizo Plain (Foster and Russ

Plot) are completely contained within fences designed to exclude large grazing

mammals. These exclosures were established by Bureau of Land Management

biologists (Russ Lewis and Roy van de Hock). The other four populations were not

enclosed and thus the potential existed for greater levels of herbivory by largemammals.

The locations of the sites are described below and are also shown on the map

in Appendix A (Map 2).

1. Cuyama. This site is located on the east side of Santa Barbara Canyon Road

approximately 2.5 miles south of Foothill Road. Just after the second cattle guard on

Santa Barbara Canyon Road, the population can be located by walking about 200

yards to the east towards the river. In 1992, there were 80 flowering individuals in

this site, clustered on a sloping embankment of a dry streambed leading to the river.

2. Foster Plot. This population is located on the south side of Soda Lake Road 15.4

miles from Highway 166 (or 6.0 miles from the south entrance to the Washburn

Ranch). There were 232 flowering individuals in this site in 1992.

3. Russ Plot. This population is located on the west side of Soda Lake Road. 13.4

miles from Highway I66 (or 2.0 miles south of Foster Plot) there is a dirt road

heading south. Approximately 0.3 mile along this road Russ Plot is located about 300

yards to the west. There were 160 flowering individuals in this site in 1992.


4. Population 4. This population is situated very close to Foster Plot, but it is not

enclosed by a barbed wire fence. There is a narrow dirt road heading southwest from

Soda Lake Road at a junction that is immediately south of Foster Plot. At 0.6 miles to

the west of Soda Lake Road this population is located approximately 400 yards from

the road to the west. There were approximately 200 flowering individuals in this site

in 1992.

5. Travers Ranch. This population is also on the south side of Soda Lake Road close

to the Travers Ranch. At 11.7 miles from Highway t66 take the road heading south

for about 0.5 miles until the end of the fence. The population is on the hillside

approximately 300 yards from the road to the west. There were 108 floweringindividuals in this site in 1992.

6. Population 6. This population is located on the west side of Soda Lake Road. At

16.8 miles north of Highway 166 a small dirt road leads south just before Soda Lake

Road crosses a cattle guard, near a water tank. The road branches almost

immediately, with the right-hand branch following a fence line towards the hills.

After about 0.25 miles on this fight-hand branch, the road and fence make a 90-degree

right turn. The population can be found by walking about 200 yards to the east from

this point towards the crest of the small ridge. Russ Lewis found about 200 floweringindividuals at this site in 1991.

METHODS

1992 Field season

1, _%[JP_LgIL_L.

Transect and plot methods were used to estimate the density of Caulanthus

californicus in both Russ and Foster Plots during the period of peak flowering, March

31-_April 2, 1992. At each site line transects were established at regular intervals and

quadrats were placed at approximately 10 meter intervals along the transect. The

Caulanthus plants within each 1 meter 2 quadrat were counted. Forty quadrats were

sampled at each site.


2. Demo_raohv of Reproduction.

In order to study- the reproductive biology o_"California Jewelflower, a detailed

examination of marked individual plants was conducted in each of three populations

in 1992 (Cuyama, Foster and Russ). During the period of March 31 - April 2, 100

flowering individuals were randomly chosen in both Russ and Foster Plots. Each

plant was marked with a flagging stake and given an identification number. For each

individual we recorded the following data:

(l) number of leaves; (2) number of reproductive stems bearing flowers; (3) numberof flowers.

There were two types of fruit loss or damage noted in these populations in

1992. Abortion of developing' fruits was recorded if normal development was

arrested or if a fruit was shriveled. Such spontaneous fruit abortion could be due

either to pollinator failure or to the unavailability of resources necessary for fruit

development (even if fertilization of ovules was successful). Second, herbivore

damage was recorded if at least some part of a fruit had obviously been eaten.

Several Lepidoptera larvae were found on the outside of some fruits so this seems to

be a likely cause of at least some of the damage and subsequent fruit abortiori. Fruits

that failed to develop were classed as either "aborted" or "damaged" (a fruit could not

represent both classes simultaneously).

The same individuals were resampled at Foster during the period April 11-i2

and the following measurements recorded: (1) length of the two longest leaves; (2)

plant height; (3) number of developing fruits; (4) number of aborting fruits; and (5)

number of fruits with apparent herbivore damage. During this sampling period these

measurements could not be recorded on plants in Russ Plot because of extremely high

levels of rattlesnake (Crotalis viridis) activity in the study population. On this date 80

individuals were marked at the Cuyama population and the same measurements

recorded as at Foster Plot.

One of the observations made at the onset of flowering of Caulanthus

californicus was that individual flower size was extremely variable. Previous studies

on another species in the same family, wild radish (Raphanus sativus, Brassicaceae),

have demonstrated that traits that set the upper bound to male and female

reproductive success (e.g. pollen and ovule production) are positively correlated with

September 15. 1993 Mazer/Hendrickson 6

flower size (Stanton and Preston, 1988; Mazer and Schick, 1991a, 1991b). Similar

patterns are seen in lpomopsi aggregata (Polemoniaceae) (Campbell, 1992). In

addition, Stanton and Preston (1988) observed in R. sativus that mean corolla

diameter is positively correlatied with the number of insect visits to the flower,

suggesting that large-flowered individuals should both receive and export more pollenthan small-flowered individuals. Other workers have also found that flower size in

wild species is positively correlated with seed number per fruit (Kang and Primack,

1991), and that both decline within individuals as they age. This result implies that, as

resources decline over the course of the reproductive season, both flower size and

reproductive output per flower diminish. It seems likely that the variation in flower

size that we observed in C. californicus both within and among individuals in part

reflects the resource status of the individual plants and flowers.

In order to quantify variation in flower size in C. californicus, one flower from

each plant was collected on April 11 for measurements to be conducted later in the

laboratory. Taking advantage of the fact that flowers mature and open in a very

predictable order within inflorescences of Caulanthus (from the basal positions

• [oldest flowers and developing fruits] towards the apical positions [youngest flowers

and buds] of a flowering stem), the bud representing the next flower scheduled to

open was removed from the primary flowering stem and placed in a small plastic

centrifuge tube and then in a cooler with ice. For each flower bud we measured the

following traits: (1)bud length; (2) ovary length; and (3) anther iength. In

addition, five flower buds were collected from 25 randomly selected plants per site in

order to obtain an estimate of variation in ovule number within and between

individuals.

Final reproductive success was estimated using material collected during a

field trip May 2-4. During this trip at least one fruit was collected from each of the

marked plants in Foster and Russ Plot, and Cuyama. The most recently ripened fruit

that had not yet dehisced was placed in a small envelope. The seeds were counted

and weighed in the laboratory to determine the average weight of individual seeds, an

important determinant of early seedling performance in many plant species. In

addition, three fruits were collected from each of at least 25 plants at Travers,

Population 4 and Cuyama in order to determine the amount of variation in seed

number per fruit within individuals and among plants within a sing!e population.

September 15, 1993 Ivlazer/Hendrickson 7

These seeds are now stored under refrigeration at the University of California (Santa

Barbara), and are available for future cultivation or recolonization efforts.

3. Spontaneous Self-Pollination Experiment.

A study was initiated to determine whether pollinator visitation is required for

fruit set, or whether plants are capable of spontaneous self-pollination without the aid

of insect pollinators.

This experiment was conducted in Population 4 and initiated on April 2, 1992.

Twenty plants that had at least two recently opened inflorescences were randomly

selected. A pair of inflorescences of similar age were selected on each plant. One

inflorescence was left untouched: the control treatment or inflorescence. The other

inflorescence was entirely enclosed within a mesh bag (made from white bridal veil)

and tied off on the bottom with a twist tie: the pollinator excuislon treatment. From

the control inflorescence, the next flower bud scheduled to open was removed and

stored in a plastic micro-centrifuge tube for later measurement. The mesh bag

allowed the inflorescence to exchange gases_but did not allow insects to visit and

pollinate the flowers. If California Jewelflower were completely reliant on

spontaneous self-pollination, then fruit production would be the same on both the

control and bagged inflorescences. If bagge d inflorescences (the pollinator exclusion

treatment) failed to set fruit, then it could be concluded that Caulanthus requires

insect visitation to produce fruits. This would imply that pollinators must be

available during the establishment of new ai'tificial or natural populations of

Caulanthus californicus. Our field work was conducted on windy clays during which

pollinator activity appeared to be absent. In order to determine the taxonomic

identity and typical population densities of those pollinators that successfullly service

C. californicus, future field studies should attempt to observe and to collect

pollinators during windless days or during early morning periods during which wind

velocity is low and pollinators are foraging.

4. Seed Bank Study.

One of the factors that can influence the population size of annual plants is the

presence of a seed bank in the soil. Seed dormancy and the presence of a long-lived

persistent seed bank would be one potential mechanism allowing the long term


survival of populations growing in the extremely unpredictable desert-like

environment of the Carrizo Plain (Leck et al., 1989). Even if survivorship or

reproduction of germinated individuals is extremely low (or nonexistent) during a

drought year or series of years, dormant seed in the soil may remain viable until

conditions are suitable for germination, increasing the likelihood of long term

population survival and successful reproduction. There is some evidence that

Jewelflower population dynamics are strongly influenced by the behavior and

presence of buried dormant seed. Sites which contained only a few individuals in the

drought years of 1988 and 1991 were found to have hundreds of plants during 1992

surveys after a high rainfall winter (Russ Lewis, CNDDB forms, 1988 to 1992).

Soil samples for the seed bank study were collected on April 11 and 12, 1992

(prior to dispersal of the 1992 seed crop) at Foster Plot, Cuyama, Travers Ranch and

Population 4. At each site 25 flowering individuals were randomly chosen. Two soil

cores (10 cm diameter x 5 cm deep) were removed from within one meter of each

plant. Both cores from each plant were placed in a single plastic bag. The contents of

each bag were air-dried, weighed, and a 250-gram, approximately one-half inch layer

of soil from each bag was then placed on a one-inch layer of sponge rock in a plastic

dish in the University of California Greenhouse. The dishes were watered and

monitored daily for the presence of emerging Caulanthus californicus seedlings in

addition to other herbaceous species. Seedlings emerging from these samples would

represent the seed bank stored during previous years.

A second soil sampling was conducted at Russ and Foster plots on November

15, 1992, prior to the onset of the fall rains. We established two transects traversing

each population, and collected soil cores at two-meter intervals along the transect. A

total of 33 soil samples were collected (16 from Russ plot and 17 from Foster plot)

and placed in separate plastic bags. Each soil sample was air dried and thoroughly

mixed and two 250-gram samples from each bag were spread out on sponge rock in

plastic dishes as above. Half of samples from each transect was put into a cold room

(5 degrees C) for I month for stratification and one set put directly into the

greenhouse. Seedlings emerging from these samples would represent the stored seed

from past dispersal events as well as the seed from the 1992 spring.


6. Germination studies

Following weighing, seeds collected from field populations were stored in

envelopes in the refrigerator. Approximately 100 seeds that were sown in pots of

University of California arid soil mix and placed under ambient greenhouse

conditions failed to germinate. A variety of treatments was subsequently used to

attempt tO break seed dormancy. These included various methods of scarification to

break the seed coat (the usual site of dormancy mechanisms), stratification (placing

seeds in cold wet conditions for various lengths of time), manipulation of Soil matrix

: potential through varying the relative amounts of sand and soil mix, and different

growth conditions (greenhouse and lathhouse).

Scarification treatments included rubbing seeds with fine (100 grit) sandpaper;

nicking them with a razor blade; and hot water scarification (dropping seeds into

several times their volume of boiling water and then allowing the water to cool).

After scarification, seeds were either planted in pots and watered or placed between

pieces of wet filter paper in closed petri dishes; both pots and petri dishes were then

refrigerated for 11 days. After the cold treatment the petri dish seeds were planted

into pots and all the pots were moved into the greenhouse and watered daily with

rainwater for a period of at least 30 days.

Other seeds were dropped into cold rainwater and allowed to soak for a week

to remove potential germination inhibiting agents. Seeds were then planted into

various soil mixtures, and gibberellic acid (GA3) was added to half the pots in various

concentrations. In order to determine if seeds were viable we sent samples to Ransom

SeedLab forgerminationand seedqualitytests. ,.

1993 Field season

, I. Competition study

One of the potential causes of the decline of Caulanthus californicus

throughout its range in the San Joaquin Valley is competitive exclusion by non-native

species such as Bromus rubens. Taylor (1986) cites a study by Westman et al.

(1985) showing the recent evolution of pollution-tolerant biotypes in Bromus rubens.

He hypothesizes that the puzzling disappearance of Jewelflower during recent

September15.1993 MazedHendrickson10

decades from sites where it had previously persisted in spite of grazing might be clue

to a combination of lack of tolerance for poor air quality and competition with

vigorous, oxidant-resistant Bromus rubens. While we were unable to test the effects

of air pollution on Jewelflower, we did initiate a study to investigate the possible

effects of interspecific competition on the survival and reproduction of the plants.

On February 6, 1993 we established four pairs of 0.25 m2 study plots in the

Foster Plot population. Pairs were matched as closely as possible with respect to the

number of Caulanthus seedlings in each plot and microhabitat characteristics of the

plots. It was intended to establish five pairs of plots, but we could not find a matched

quadrat for the last plot, so it was retained as an additional unmatched control plot. In

each plot we counted the number of Caulanthus seedlings, which on this date were

small rosettes with only a few leaves. Seedlings were marked with toothpicks] The

treatment plot was weeded, removing all plants of other species within the plot and

within about a 10 cm border around the plot. Weeding was done by scraping the dry,

hard soil surface with a knife blade so as to disturb the soil as little as possible. The

plot with the greater number of seedlings was chosen as the interspeeiflc removal

treatment plot, since this would provide a more conservative test of our hypothesis

that the removal of interspecific competition would increase seedling survivorship

and adult reproduction in the Caulanthus plants. If, in spite of higher densities of

Caulanthus, the weeded plots supported larger Caulanthus individuals with higher

survival rates than the unweeded plots, then interspecific competition would clearly

represent an important factor that repressed Caulanthus reproduction and survival.

In a similar manner, between March 4th and 8th we set out 10 pairs of plots at

population 6, and five pairs at population 4.

We returned to the Foster population plots on March 8 and counted survivors

and new recruits. At that time plants were still mostly rosettes, although a few had

started to bolt and produce flower buds. All sites were revisited on March 31 and

April 1, and survivors and recruits were counted. At that time plants were in flower

but most were not yet fruiting. On each plant, we recorded: (1) branch number, (2)

flower number, (3) the number of unopened buds; (4) the number of developing

fruits; and (5) the number of branches on each plant that had been lopped off by

mammalian herbivores. Giant kangaroo rats (Dipodomys ingens) were common at

each site and were probably the main predator, but jackrabbits may also have


contributed some damage. No cattle or sheep grazing occurred at any of the

populations during our study in 1993.

We returned to 'all the sites on April 18 and May 2, recording survivorship for

each quadrat and the number of branches, branches eaten, flowers, buds, developing

fruits, aborted fruits and aborted flowers for each individual within the quadrats.

2. Greenhouse pollination studies

On January 23, 1993 we visited a site about 0.8 miles NW of population 4,-

about 2/3 of the way between population 4 and population 6. This site was later

called Four-Post Valley, in reference to a group of four fence posts that provide a

landmark visible from Soda Lake Road. Russ Lewis had found 978 plants here in

1991. We removed 20. Caulanthus seedlings from this site, using a bulb planter to

remove a plug of soil containing each seedling and transferring the plug to a three-

inch pot. We brought the seedlings back to UCSB and maintained them in a growth

chamber (14 hr of light/10 hr dark; 20 degrees C day; 15 degrees C night). Additional

seedlings were removed on February 6 at population 4 and at a site just 0.4 miles

south of population 4 which we cNled population 5. Ten seedlings were taken fromeach of these sites.

Plants were kept in the growth chamber until they began to bolt and flower.

At that point they were transferred to the greenhouse and a series of hand-pollinations

were done. Flowers from each plant were hand-pollinated using pollen from either

the same plant or from another plant. Pollen sources were alternated in sequence

along the flowering stem in order to randomize the potential effects of fruit position

(on the stem) on seed set. Following pollination,- each flower was marked with the

identity of the pollen donor using small adhesive tags wrapped loosely around the

pedicel. Crosses were performed using plants from the same population (selfed vs.

outcrossed pollen) and plants from different populations as donors. After the fruitsk

developed they were collected and the number of mature, filled seeds, aborted seeds,

and total seed weight per fruit were recorded. Evidence for inbreeding depression

could include the observation that seeds produced by selfing would be smaller or

exhibit higher abortion rates than seeds produced by outcrossing within or between

populations. Similarly, if outcrossed pollinations between individuals of the same

population were less successful or produced smaller seeds than crosses between


individuals from different populations, inbreeding depression due to the expression of

deleterious alleles exprssed in homozygous form might be the cause. Outbreeding

depression would be detected by lower seed weights or higher abortion rates

following inter-population outcrosses relative to outcrossed pollinations performed

between individuals derived from the same population.

STATISTICAL ANALYSES

-: We used several statistical techniques to analyze the data collected on

California Jewelflower. First, descriptive statistics were calculated to characterize the

mean and variation (standard deviation, coefficient of variation) of the traits we

measured. Second, the variation around the mean of several important traits was

displayed by plotting frequency distributions. Such illustrations of the data allow one

to view how the data are distributed within a population. Third, we used analysis of

variance to determine whether the three populations studied most intensively (Russ,

Foster and Cuyama) differed with respect to the mean values of several traits. Since

some of the data were not normally distributed, we also used non-parametric tests

which do not assume a normal distribution. These included the Mann-Whitney U-test

for comparing two groups, the Kruskai-Wallis test for comparing three groups, and

the Wilcoxon signed rank test for paired comparisons. Fourth, we used regression

analysis to explore the relationship between plant size (height) and various measures

of reproductive success, as well as between floral characters and reproductive

components.

RESULTS AND DISCUSSION

Observations on population structure and ecolo_,v

Caulanthus californicus exhibits a non-random distribution on two levels:

populations occur in extremely discrete patches or colonies throughout the range of

the species, and the distribution of individuals within a population tends to be

clumped. The tendency for the plants to occur in distinct colonies has been noted by

other investigators (Taylor and Davilla 1986, Jepson 1936). The occurrence of

aggregated colonies of individuals suggests that seed dispersal distance is not very


great in this species, although it is also possible that the specific environmental

conditions that promote germination or allow seedling survival are aggregated within

natural habitats.

Overall, the density of plants in 1992 was greater in Foster Plot than in Russ

Plot (5.66 vs 2.43 plants/m2). However many of the quadrats in Foster Plot had no

plants in them. Consequently, the variance among quadrats was also much greater in

Foster Plot (32.06 vs 5.93). The high variation in plant density among quadrats

within a population of California Jewelflower is not surprising given the strongly

aggregated distribution of individuals in both populations. The non-random

distribution of individuals (variance-to-mean ratio greater than 1; Ludwig and

Reynolds, 1988) suggests that external microhabitat factors such as an association

with kangaroo rat disturbance or a nurse plant relationship may be controlling plant

distributions within a site. There is a rather striking trend within sites for plant

densities to be highest on and around large mounds. These mounds are most likely

the result of kangaroo rat activity and perhaps mark the site of a dead Ephedra

individual. The mounds also tend to support vigorous populations of Amsinckia,

Lasthenia an d Orthocarpus.

Phenology and reproduction patterns

Phenological observations are mostly based on the 1993 data set, since in

1992 plants were already flowering by the first census date and we had no

information on earlier phenological events such as seedling emergence or the timing

of bolting. Heavy rainfall and.poor road conditions in the winter of 1992-1993

prevented our visiting the sites as early in 1993 as we would have preferred and we

were not able to establish the time of seedling emergence. Many seedlings and

juveniles (small rosettes with 2 to 6 leaves, 1 to 3 cm in diameter) were present at all

the sites we inspected in late January 1993, and seedlings continued to emerge

through February at the Foster population. Seedlings had probably started to emerge

soon after the first heavy fall rains in the Carrizo Plain.

By the first week of March in 1993 most plants were still in the rosette stage,

but some individuals had started to bolt and to form buds. Interestingly, it appeared

that those plants which were bolting tended to be the ones with smaller rosettes: someindividuals with rosettes about l cm in diameter had started to bolt and form a 2 cm


stalk and tiny buds by March 4 (see Appendix B. Photo 5). Individuals with larger

rosettes tended to remain vegetative longer. Virtually all individuals in the

populations on the Carrizo Plain were in flower and most had some developing fruits

by April 1, 1992, and most were in flower or at least had buds by the same date in

1993, although few plants had developing fruits on that date. Plants appeared to be a

week or two behind in their development in 1993 relative to the same date in 1992.

The largest plants did not have open flowers on April 1, 1993 but had large numbers

of buds. Flowering continued in both years through April, with developing fruits

appearing by mid April during 1993. By mid-April 1993 most of the very small

plants were dead, usually without having produced fruits. By the beginning of May

ripe fruits were present or had already dehisced and flowering was largely over except

in the largest individuals, which continued to produce fresh flowers and buds through

early May in 1993.

Thus, there appeared to be a range of flowering phenologies present within the

population related to plant size, with the smallest plants flowering first and senescing

prior to the time of peak flowering of most individuals, and a few very large plants

starting to flower later in the season and continuing to flower past the peak period for

most of the population. Such a range of size-related flowering times is commonly

observed in annual plants of unpredictable environments (Schmitt, 1983; Dieringer,

1993), although the relative size of early- vs. late-flowering individuals varies among

species. In general, however, the presence of phenologicai variation among

individuals in natural populations probably assures that at least some individuals will

produce fruits even in years where soil moisture is insufficient to allow survival in the

later part of the season. The presence of high levels of variation in the flowering time

among individuals of C. californicus is typical of many annual plant species of desert

enviornments (Went, 1949; Shreve, 1951; Rathcke and Lacey, 1985; Lacey, 1986;

Fox, 1989, I990a, 1990b).

Russ Lewis originally located some of these populations in 1988; during that

year (a drought year) he estimated that 60% of the plants were in flower and 40%

were fruiting on March 22, at least a month earlier than was the case in 1993 (Lewis

1988). In C. californicus populations, those individuals that flower late risk

reproductive failure if soil moisture is insufficient to allow them to complete

reproduction. However, in high rainfall years late-flowering individuals with large

vegetative rosettes develop successfully into large, multi-stemmed plants and


contribute disproportionately (relative to their numbers) to the seed crop of the

population (see fecundity analyses, below).

Seedbank emergence

No emergence of dormant Caulanthus seed was observed in the soil samples

collected in spring of 1992. This may be due to the unusually high rainfall year

which may have promoted the germination of nearly all the dormant seed in the soil,

or it may be that seeds were present but had entered a secondary dormancy which we

were unable to break using various treatments. Soil samples collected in November

of 1992 yielded only one Caulanthus seedling.

Some grass seedlings (Bromus rubens, Vulpia macrostachya and other

annuals) appeared from the soils collected in the spring of 1992. A very few other

seedlings appeared, including Lotus sp., Erodium cicutarium, Lepidium sp., Salsola

kali, and Pectocarya penicillata. Apparently the soil seed bank of many other

annuals had largely germinated prior to the date on which we sampled the soil. On

the other hand, large numbers of seedlings of various species emerged from the soil

samples collected on November 15, 1992. The most common were grass seedlings,

which were abundant in almost all of the 250-gram samples. Erodium cicumrium

was present in virtually every sample. Pectocarya penicillata and Calandrinia ciliata

var. menziesii were found in many of the samples. Less common species were:

Tropidocarpum gracile, Crassula erecta, Lotus wrangelianus, Trifolium graciIentum,

Thelypodium lasiophyllum, Lasthenia californica, Lactuca serriola, Lepidium

lasiocarpum, Amsinckia sp.. Castilleja exerta (formerly Orthocarpus purpurascens),

and Oenothera dentata var. johnstonii.

Seed dormancy

We were generally unsuccessful in germinating seed that we collected in 1992

and attempted to grow in the greenhouse and growth chamber. In total, about 500

seeds were subjected to the various treatments, and only one seed germinated during

these tests (it was part of a control group not treated in any way and simply planted in

a sand tray and placed in the lathhouse). This seedling did not survive past the

cotyledon stage. Samples of the seed were sent to Ransom Seed Lab (P.O. Box 300,

Carpinteria, Ca. 93013; telelphone: (805) 684-3427) in 1993. Ransom Lab


investigators observed little or no germination using different chilling and light

regimes or by removing the mucilaginous coating of the seeds. However, by

"clipping" the seeds (carefully cutting the seed coat with a razor blade to expose the

embryo [while avoiding any damage to the radicle]) -- an event that would occur in

nature only after severe abrasion or recurring temperature and humidity changes --

and adding gibberellic acid (GA3,400 ppm) they were able to force up to 52% of the

seeds to germinate (Table 1). The tetrazolium chloride test for metabolic activity

showed up to 98% staining indicating that almost all the seeds were living, and

germination was 98% for both Foster and Travers Plot seeds using a treatment that

combined clipping, GA3, and chilling for 6 days and light.

Thus, the seed was viable but deeply dormant. It may be that a period of

after-ripening is required as well as appropriate environmental cues for germination.

Another possibility is that there is a requirement for a mycorrhizal symbiont to enable

seedlings to grow in the field (Smith, 1980). Ransom Lab scored seeds as

germinating if the radicle (embryo root) had emerged from the seed coat when the

seeds were placed between moist sheets of unbleached blotter paper in their

incubators. If the seeds we planted in sand and various soil mixes produced a radical

but then died from fungal infections or because they lacked the appropriate

mycorrhizal symbiont we would not have scored them as germinating, since no

cotyledons would have appeared above the soil surface. This could also explain why

the one seedling that did appear did not live very long in the lathhouse.


Table I. Germination treatments and results from Ransom Seed Lab tests.

SeedTreatment %Germ: %Germ:Foster Travers

5 degrees C for 6 d., then 20 degrees C, light (8 hours) and water.Seeds placed on moist dye-tree blotter paper within 4" x 6" plastic 0 0boxes.

20 degrees C, light (8 hrs.) water. Seeds placed on moist dye-freeblotterpaperwithin4"x6"plasticboxes. I 0

20 degrees C, light (8 hrs.),400 ppmGA3. Seedsplacedon moist dye- 5 0free blotter paper within 4" x 6" plastic boxes.

Tetrazolium chloride. 1%* 92* 98*

Removed mucilage. Seeds placed on moist dye-free blotter paperwithin 4" x 6" plastic boxes. 0 0Dried and rewet. Seeds placed on moist dye-tree blotter paper within4"x6"plasticboxes. 0 0Clipped and 400 ppm GA3. Seeds placed on moist dye-free blotterpaperwithin4"x6"plasticboxes. 52 14

Clipped. 400 ppm GA3, chilled at 5 degrees C tbr 6 d. then 20 degrees, 98 98light (8 hours). Seeds placed on moist dye+free blotter paper within 4"x 6" plastic boxes.* For the tetrazolium chloride test, the percentage figure is the percent of seeds which showed stainin(metabolic activity).

Since the seeds show deep dormancy it is probable that seeds do accumulate in the

field to provide a long-lived seed bank in the soil. This would protect the population

against years in which few or no plants survive to reproduce, since stored seed in the

soil would still be present to produce new plants in subsequent years. This strategy is

common in desert annuals and allows them to persist as dormant populations in an

unpredictable environment (Leck et al., 1989).

1992 Field Study Results

Comt_ari_on of study site_;

The measurements from the marked plants at each of the three populations

(100 plants at Foster and Russ Plots; 80 plants at the Cuyama population) were

examined statistically in order to determine whether populations differed with respect

to various traits. Many reproductive and morphological traits differed significantly

among populations (Table 2). Plant height and mean length of the two longest leaves

were greater in Foster than in Cuyama on April 11. Among flower and fruit traits,

bud length did not differ significantly between the two populations, but both anther


length and ovary length were significantly greater in Cuyama than in Foster. So,

while overall plant size was greater at Foster than Cuyama on April 11, 1992, the

Cuyama plants had larger floral parts.

In spite of the larger floral parts represented among plants at the Cuyama

sited, the mean number of mature seeds per fruit was significantly lower at Cuyama

than at the other two populations on May 2, 1992 (Table 3). However, seed abortion

rates (detectable within mature fruits) were very low and similar among the three

populations (Cuyama had a lower mean abortion rate but it was not significantly

different from the other two sites). In spite of exhibiting larger stamens and ovaries

than the other two populations, Cuyama plants produced fruits with fewer seeds perfruit.

Seed abortion levels are estimated as the number of shriveled or clearly

inviable seeds found in a mature fruit; however, seeds and ovules aborted very early

in development would probably not be detectable at that point. Aborted seeds visible

within a mature fruit are those that were probably fertilized but failed to complete

development due to a genetic defect or to an inability to garner sufficient resources.

Early ovule and/or seed abortion is probably an important determinant of seed

number, since ovule numbers per ovary are so much higher and less variable than

seed number per fruit. Based on a comparison of the mean number of ovules per

flower collected from randomly sampled plants on April 11 and of the mean number

of seeds per fruit for marked plants on May 2 (Tables 2 and 3) early ovule and/or

seed abortion appears to have been higher at Cuyama than at Foster (45% vs. 31%).

The comparison of fruits sampled from each of about 25 randomly selected

plants at Travers Plot, Population 4, and Cuyama on May 2 again showed that mature

seed number per fruit differed depending on the population. As when the marked

plants were compared, Cuyama plants had significantly fewer seeds per fruit than the

two Carrizo Plain populations (Table 4). Travers Plot and Population 4 did not differ

significantly in seed number per fruit from Russ Plot and Foster Plot (see Table 3), so

all the Carrizo populations were similar with respect to seed number per fruit.

Individual seeds at Cuyama were significantly lighter in weight than seeds

from the other two populations. Foster and Russ populations had seeds with a mean

weight of 0.764 and 0.77 mg respectively and did not differ significantly from each

September 15, 1993 IVlazer/Hendrickson 19

other, while Cuyama population seeds weighed an average of 0.692 mg, significantly

lower than the other two (Fisher PLSD, p = .0099).

In 1992, fecundity (fruit number per plant times seed number per fruit) was

signficantly higher at Foster and Russ plots than at Cuyama; this was a result of

significantly greater number of fruits per plant as well as more seeds per fruit in the

two Carrizo populations. Percent fruit set (fruits produced per flower) was also

significantly higher at Foster plot than at Cuyama. However, the source of fruit loss

differed among sites. The average number of fruits damaged by insect herbivores

measured on May 2 was much greater in Russ (mean = 2.38) than either Foster (0.79)

or Cuyama (0.74; Table 3)i The mean number of aborted fruits per plant was

significantly higher at Foster than at Cuyama on April 11, and higher than either

Cuyama or Russ on May 2 (Tables 2 and 3).

Table 2. Means and standard deviations (in parentheses) for traits sampled on April 11. Russplot was not sampled on that date due to high rattlesnake activity. Ovule number per ovary wasdetermined from a sample of 3 flower buds per plant for 13 randomly selected plants at Fosterplot and 12 plants at Cuyama. To detect statistically significant differences between Foster andCuyama population means, different statistical analysis techniques were applied as warranted bythe data type. These included analysis of variance (Fisher PLSD) and unpaired t-test fornormally distributed data, and the KruskaI-Wallis test and Mann-Whitney U-test for non-normally distributed data. Numbers of plants examined per trait are indicated by the "n"column following each trait. Statistically significant differences between population means areindicated by p-values <0.05; n.s. =non-significantdifference between populationmeans.

Trait Foster n Cuyama n p

Plant height (cm) 26.5 (9.3) 94 20.2 (5,6) 80 .0001F

Numberoffloweringstems 2.2 (1.8) 96 1.9 (1.4) 80 n.s.

Mean of 2 longest leaves (cm) 28.0 (12.6) 64 2212 (9.3) 78 .0019F

Bud length (ram) 7.2 (2.0_ 84 7.4 (1.4) 61 n.s.

Ovary,length (ram) 1.1 (.8) 82 1.7 (.6) 61 .0001''Iw

Anther length (mm) 1.6 (.7) 82 2.2 (.6) 6l .0001Mw

Ovule number per ovary 62. I (7.9) 13 51.7 (9.3) 12 .0059T

Flwrs + buds + developing frts 29.8 (35.5) 96 25.6 (25.0) 80 n.s.

Number of aborted fruits 3.9 (4.4) 95 1.5 (I.8) 80 .0001_P'V

Number of aborted flowers 1.51 (3.52) 98 .04 (.25) 80 .0001Mw

V = Fisher_PLSD T = unpaired t-test MW= Mann-Whitney U-test

September t5, t993 Mazer/Hendrickson 20

Table 3. Means and standard deviations (in parentheses) for traits on the May 2 sample date formarked plants in each population. Percent fruit set is the number of fruits present on May 2divided by the total number of flowers, buds, developing fruits, and aborted fruits present onApril 11. Fecundity is the number of fruits per plant times the number of seeds per fruit. P-values less than 0.05 indicate that population means differ significantly. The KruskaI-Wailis testdoes not provide pairwise comparisons between means. For normally distributed traits(indicated by F) means with distinct superscript letters differ significantly at the 0.05 level ofsignificance.

Trait Foster n Cuyama n Russ n p

'No. of aborted fruits 1.86 (2.05) 71 1.04 (I.47) 78 .96 (1.34) 85 .0056 KW

No.of fruits 12.7 (15.53 65 8.2 (ll.6"_ 77 11.2 (19.2) 73 .0411Kw

No. of mature seeds per fruit 43.0 a (18.6) 56 28.7 b (16.8) 75 38.2 a (21.0) 61 .0001F

Mean individ, seed wt (mg) .764 a (.172) 55 .692 b (.176"1 76 .77 a (.142) 58 .0099 F

Mean percent seed abortion 1.8 (3.8) 56 1.3 (3.7) 75 1.8 (3.43 60 n.s.

No. of insect damaged fruits .79 (2.753 71 .74 (2.37) 78 2.38 (4.96) 85 0046 KW

Percent fruitset 36.2 (14.9) 65 29.4 (15.1) 77 n.a. .0042 Mw

757 (1294) 56 316 (783) 75 654 (1540) 59 .0006 KwFecundity

F = Fisher PLSD; Kw = Kruskal-Wallis test; r,tw = Mann-Whitney U-test


Table 4. Analysis of variance to detect the effect of population on number of seeds per fruit inTravers Plot, Cuyama population, and Population 4 sampled May 2, 1992. Fruits were sampledfrom approximately 25 randomly chosen plants within each population, p-value less than 0.05indicates that there are significant differences among population means with respect to the meannumber of seeds per fruit.

Analysisof VarianceTable

Source: DF: Sum Squares: MeanSquare: F-test:Between,qroups 2 2309.543 [1154.771 3.597

Within,qroups 158 50727_302 ,321.059 p = ,0297Total 160 53036.845

Population No. of fruits Mean# seeds/fruit': Std.Dev.: Std. Error:

Populationfour 64 40.906 a 16.144 2.018

Travers 55 39.491a 19.559 2.637

Cuyama 42 31.738b 18.248 2.815

• Differentlettersindicatemeansare significantlydifferent(p < .05,FisherPLSD)

Thus there are some clear differences in reproductive traits among populations ..

within a relatively restricted geographical area. Foster and Russ Plots were similar

with respect to fruit and seed production, but the Cuyama population, although only

about 20 miles south of the other two, differed in floral characteristics as well as in

fruit and seed production and seed weight. Foster, Russ, Travers, and Population 4

were all similar to each other with respect to seed number per fruit, and all were

significantly different from the Cuyama population. However, it must be

remembered that these statistical analyses are based on discrete sample dates and the

picture that emerges is like a snapshot of the reproductive output at a particular time:

it is assumed that fruit production did not continue past the last sample date. The

magnitude of the apparent differences in fecundity and floral traits may be partly due

to different timing of the onset of flowering in the different populations.


Higher elevation populations are usually phenologically less advanced than

low elevation populations, and this may explain the observed differences between the

Cuyama and Carrizo populations. The elevation of the Carrizo populations is about

2200 to 2250 feet; the Cuyama population is approximately at 2700 feet. The

proportion of fresh flowers and buds to the total number of flowers, fruits and buds

per plant was higher in plants sampled at Cuyama on April l I than at Foster (77.8%

vs. 46.4%, p = .0001, Mann-Whitney U-test). This higher proportion of fresh flowers

and buds at the Cuyama population indicates that these plants were still flowering

strongly on April 11, when many of the Foster plants were producing primarily fruits.

Figures 7a and 7b display the frequency distributions for the number of flower buds

per plant on this date at both populations; Cuyama had few plants in the lowest

frequency class (most plants had more than 5 buds). Cuyama plants had larg e

numbers of buds but few developing fruits compared to the plants at Foster Plot on

the same date. Thus the fecundity measures for Cuyama on May 2 may not represent

the true seed production of that population, since the plants might have produced

more fruits that season. However, the lower number of ovules per ovary and seeds

per fruit at Cuyama are probably not attributable to the phenological state of the

population, since ovule production per ovary is generally highest during the early

stages of flower production and declines towards the end of the flowering period

(Mazer, Snow and Stanton, 1986).

Survival

The probability of survival in the period from flowering to fruit maturation

was highly site-specific in 1992 (Table 5). Only 72 of the 100 plants that were

initially marked on March 31 in Foster Plot survived until May 2 to produce seed.

Mortality was highest between April 11 and May 2, with 20 plants dying in this

period. In contrast, survival was higher in Russ Plot (85%) than in Foster Plot

between March 31 and May 2 even though these two sites are only two miles apart. It

is difficult to directly compare the survival during the same period in Cuyama since

plants within this site were marked at a later date. However, there was almost no

mortality in Cuyama population between April ll and May 2. These results

demonstrate that mortality in the endangered Caulanthus californicus varies over

extremely small geographical distances. The low mortality at Cuyama during this

period may have been related to the less advanced phenological state of that


population (see above), such that plants were not yet stressed to the point of

premature senescence.

Table 5. A comparison of survival and fruit loss by May 2, 1992 among the three mainpopulations of Caulanthus californicus. Aborted Fruits is the percentage of marked plants thathad at least one aborted fruit. Damaged Fruits is the percentage of marked plants that had atleast one damaged fruit. Survival is the percentage of marked plants that survived fromflowering to fruiting. The level of survival for Cuyama is most likely an overestimate sinceplants were marked well after the onset of flowering.

Population Aborted Fruits Damaged Fruits SurvivalFosterPlot 47 11 72

RussPlot 39 28 85

Cuyama 53 15 -98

Frequency Distributions of Reproductive Traits

Plotting raw data in the form of frequency distributions can be an extremely

helpful way to observe patterns within and differences between natural populations.

In Figures 1 through 8 it is clear that the distributions of several traits differ

markedly between populations. Foster Plot and Cuyama populations are compared

since they were dissimilar with respect to their frequency distributions for some traits.

Plant height, which is an estimate of overall plant size, is distributed normally; the

majority of plant individuals are relatively close to the mean, and distributed

symmetrically about the mean (Figures la and lb). In other words, there are similar

numbers of small and large plants while most are intermediate in size. A similar

distribution is seen for seed number per fruit (Figures 4a and 4b). In contrast, flower

and fruit production are highly skewed (Figures 2a, 2b, 3a, and 3b); the majority of

individuals produce relatively few flowers and fruits while _t small number of

individuals account for the majority, of flower, fruit and seed production. This sort of

pattern is extremely common among monocarpic plant species in which the majority

of offspring that constitute the next generation are derived from a nonrandom portion

of the previous year's adult population (Stanton, 1985).

The shape of the frequency distributions can illustrate the difference between

two populations for some of the traits shown in Tables 2 and 3. The distribution of


plant height at Foster plot appears exhibits a peak at approximately 25 cm, and there

is a long tail of taller individuals to the right of the peak (Figure la). At Cuyama the

distribution is more clustered around the mean of approximately 20 cm, with no very

tail individuals (Figure Ib). Cuyama individuals more frequently had very few fruits

per.plant, and only one plant had a large number of fruits (Figures 2a and 2b). The

range of seed number per fruit was similar in the two populations (Figures 4a and

4b), but more fruits from Cuyama contained lower numbers of seeds. In both

populations there was a peak in ovule number per fruit at about 50 to 60 (Figures 5a

and 5b), but Cuyama has a larger number and proportion of ovaries with fewer

ovules than did Foster Plot.

The frequency distribution of the number of aborted and damaged fruits per

plant was skewed so that most individuals had about one aborted fruit while some

plants had five to ten aborted fruits (Figures 8a and 8b). The percentage of plantswithin the three sites that had at least one aborted or herbivore-damaged fruit in 1992

is presented in Table 5. The probability that an individual suffered some level of

fruit abortion was highest in Cuyama (53%), lowest in Russ (39%) and intermediate

in Foster (47%). However, the number of aborted fruits per plant was highest at

• Foster population, as seen in Tables 1 and 2 above. In contrast, herbivore-damaged

plants were most common in Russ (28%), least in Foster (11%), and intermediate in

Cuyama (15%).

Fecundits, and factors affecting, reproductive output

Effects of individual plant identity on mean ovule and seed number per fruit

We observed a great deal of variation among individuals within California

Jewelflower populations with respect to reproductive traits such as seed number per

fruit, floral traits, and other components of reproduction. Presumably genetic and/or

environmental factors contribute to this variation. Since these factors operate on an

individual plant basis, traits would be expected to be less variable within individuals

than among individuals. We examined the variation within and among individual

plants of two traits relevant to fecundity: ovule number per ovary and seed number

per fruit.


Ovule number per ovary was determined in up to 5 buds per plant collected on

April t 1 from each of 12 randomly selected individuals at Cuyama and 13 individuals

at Foster Plot. Individual plants at both populations differed significantly from each

other with respect to mean ovule number per ovary (p = .0001, Fisher PLSD, Figure

9). That is, buds from the same plant tended to be more similar to each other in ovule

number than to buds from other plants.

Seed number per fruit was determined in three fruits per plant collected on

May 2, 1992 from 25 randomly selected plants each at Travers, at Cuyama, and at

Population 4. Again, individual plant identity was a significant factor determiniiag

mean seed number per fruit; fruits from the same plant tended to be more similar to

each other with respect to mature seed number than to fruits from different plants

(Table 6, p = .0001, Fisher PLSD).

Table 6. Analysis of variance for the effect of individual plant identity on seed number per fruitfor pooled plants from Travers, Cuyama, and Population 4. Three fruits were collected fromeach randomly selected plant. There were 161 fruits collected from 57 plants total.

One Factor ANOVA X 1 : ID# Y 1 : Seed #

Analysisof VarianceTable

Source: OF: Sum Squares: Mean Square: F-test:Between,qroups 58 33900.511 605.366 3.29

Within ,qroups 104 19136.333 184.003 p = .0001Total 160 53036.845

The relationship between plant size and reproduction

As shown above, ovule and seed number are traits which are relatively

consistent within an individual but differ between individuals. What are some of the

factors that might produce differences among individuals with respect to these and

othei" traits? Habitat variation on a small scale (such as variation in micro-topography

or soil texture or quality) could affect plant vigor, which in turn may affect fruit and

seed production. Genetic differences among plants could also directly or indirectly

result in differences in plant vigor. One measure of plant vigor is plant size.

Regression analysis was used to explore the relationship between plant size and

several reproductive factors.


Plant height was an extremely reliable predictor of the reproductive output of

individuals of Caulanthus californicus, as shown by the results of the regression

analyses in Figures 10 - 15 (figures show data from Foster Plot except where

important differences exist in the regression for a trait at Cuyama, then both

populations are illustrated). Plants that were relatively tall produced more flowers,

buds, fruits and seeds per fruit (Figures 10 - 13) than shorter plants. The regressions

fit a polynomial curve better than a straight line; this may reflect the fact that plant

volume (including root mass and the ability to garner soil resources) increases

geometrically rather than linearly as plants get larger. Fruit set (the number of fruits

on May 2 divided by the total number of flowers, fruits and buds on April 11) was

independent of plant height (our estimate of plant size). Thus the probability of a

flower developing into a fruit was not dependent on plant height (Figure 14).

Different plants of the same height varied with respect to fruit set (the proportion of

flowers that successfully develop into fruits) from ~0.10 - 0.60 or higher.

There was no relationship between plant size and mean individual seed weight

at Foster Plot (r2 = .045, p = .3131). Mean individual seed weight per fruit was also

independent of the number of mature seeds per fruit in both populations. It seems

that in these plants seed weight (an indicator of seed quality) is conserved even when

stress or other factors cause increased ovule abortion. When resources are limiting

plants produce fruits with fewer seeds but no reduction in seed quality. However

there was a significant positive regression between plant size and mean individual

seed weight at Cuyama (Figure 19, r2 = .176, p = .0009). Thus, at Cuyama larger

plants produced heavier seeds than smaller plants,

Floral characieristics and reproductive output

Interestingly, larger plants produced larger buds (Figure 15). Bud size varied

almost four-fold among plants, even though the buds were chosen in a systematic way

to ensure that they were in similar stages of maturity (see Methods section). Such

high variation in flower size among plants is unusual in flowering plant species, but

may reflect differences in resource availability among flowers and individuals. These

differences in resources allocated to developing flowers may be related to the age of

the individual relative to the time of first flowering, to the quality of the microsite in

which an individual is growing, or they may be genetically based. Flower size is

September I5, 1993 Mazer/Hendrickson 27

usually conserved phenotypically (relative to total plant size), presumably due to the

necessity to attract specific insect pollinators that may avoid flowers that are of an

unusual size or flowers that they cannot physically accomodate. The increase in bud

length with plant size appears to reach a maximum as the curve levels off at bud

lengths of about 9 ram. Because larger flowers have larger ovaries (and presumably

more ovules; but see below) than small flowers, the average number of seeds per fruit

would also be expected to be positively related to plant size. The positive correlation

between plant size and seed number per fruit (Figure 12), however, could also be

mediated by increased pollinator visitation (and increased ovule fertilization) to larger

plants, to increased resource garnering ability of larger plants, or to lower rates of

ovule and/or seed abortion in ire'get plants relative to smaller plants.

There were some clear differences between Foster Plot and Cuyama with

respect to floral characteristics and size correlations. Bud length and ovary length

were significantly and positively related to the number of seeds per fruit at Foster Plot

(Figures 17a and 18a). At Cuyama, the relationship between bud length and seed

number was significant although the correlation was weaker (r2 = .256, p = .0003,

Figure 17b). Furthermore, at Cuyama ovary length was not significantly correlated

with seed number (Figure lgb); interestingly, ovary length at this site varied

independently of bud length while the Foster plants showed a highly significant

positive correlation between these factors (Figures 16a and 16b). The lack of

correlation between bud length and ovary and anther length at Cuyama is somewhat

puzzling. Although the plants were at an earlier phenological stage at Cuyama than at

Foster (see above), almost all of them did have open flowers. The buds chosen for

measurements were the next ones scheduled to mature on the stem, so it is unlikely

that this is a result of the buds being in very different stages of development. Also,

ovaries and anthers were significantly longer at Cuyama than at Foster (Table 2),

which would not be the case if the Cuyama flowers were less mature. Ovule number

per ovary in randomly selected plants was significantly lower at Cuyama than at

Foster in spite of the longer ovaries measured in the marked plants at Cuyama on the

same date (Table 2).

An unexpected observation was that the frequency distributions of bud, anther

and ovary length and ratio of anther-to-ovary length appear to show a bimodal

distribution in both populations (Figures 20 - 23), although the sample size is too

small to be sure that the pattern is a general one. Also, regression of the ratio of


anther-to-ovary length on ovary length shows two distinct groups (Figure 24), one

with small ovaries (< 1 ram) and anthers 1.5 to 3 times as long as the ovaries, and one

with ovaries about 1.5 to 3 mm long and anthers of roughly the same length. This

kind of distribution would not be expected if the ovaries were gradually increasing in

length with age or with increasing plant size; it seems to indicate that there are two

different types of flowers present. This might be the case for examp.le, if plants

produced flowers with different sexual functions: with smaller and effectively male

flowers producing primarily pollen, and larger hermaphroditic flowers that produce

seeds and pollen. It has been shown in other plant species that relative allocation to

•male and female functions can vary with stresses such as drought, poor soil or cold

temperatures, shifting the sexual expression towards maleness (Charnov, 1982; Mazer

and Schick, I991a, 1991b).

If such a pattern exists in California Jewelflower it would be interesting to

know if both types of flower are produced on a single plant or if different plants

produce different types of flowers, and whether the cause is genetic or environmental.

More sampling is needed to confirm this pattern and explore the consequences to

fecundity and outcrossing rates in these populations.

Adjusted fecundity

Fecundity estimates for the three populations shown in Table 3 were based on

the number of seeds in a single fruit taken from each individual. Since seed number

per fruit was less variable within an individual than among, individuals (Table 6) and

since the regression of seed number per fruit on plant height was positive and highly

significant (Figure 12), we used the polynomial regression for each population to

calculate a fitted estimate of seed number per fruit for each plant based on plant

height (Foster and Cuyama populations). The "adjusted fecundity" of an individual

was the product of this fitted seed number per fruit and the total number of fruits

produced by the individual. Mean fecundity estimates were then compared between

populations (Table 7). Fecundity estimated this way should be more reliable on a

whole plant basis than using a single fruit from the plant to estimate the total number

of seeds produced by that plant. Adjusted fecundity is somewhat lower than that

calculated using a single fruit per plant (Table 3) but the differences between

populations remain, with Cuyama producing significantly fewer mean seeds per plant

September 1.5, 1993 IVlazer/Hendrickson 29

than Foster population. Since we did not have plant heights for Russ Plot we could

not calculate an adjusted fecundity for that site. but other measures of fecundity

indicate that it is similar to Foster Plot and higher than Cuyama (Table 3).f

Table 7. Comparison of calculated fecundityestimates for Foster and Cuyama populations.

3roup: No. of plants Meanadj. fecundity Std.Dev.: Std. Error:

FOSTER 55 7! 1.6738" 1073.288 144.722

CUYAMA 75 278.118" 435.779 50.319

*Significantlydifferentmeans(p= .()003,Mann-WhitneyU-test)

Factors affecting seed number per fruit

Interestingly, the regressions of seed number per fruit on various plant traits

differed between the Foster and Cuyama populations. The Foster plants showed a

much tighter relationships between seed number per fruit and plant height and

between seed number per fruit and leaf length. Among plants from Cuyama, the

correlations, while still significant, had a lower r2 value. This suggests that at Foster

plot resources are closely linked with seed set (Figure 12, r2 = .47), while at Cuyama

the relationship is less predictable (r 2 = 217). At Cuyama, factors other than resource

availability may be more important in determining the number of seeds per fruit at

Cuyama. Other such factors may include pollinator availability or behavior.

Recalling that in 1992 we demonstrated that C. californicus shows little if any

spontaneous self-pollination in the absence of pollinators (see the section on the

pollinator exclusion study), it is reasonable to imagine that differences between

populations in mean seed number per fruit and/or in the relationship between plant

size and seed number per fruit could be mediated by pollinators. If, t'or example, in

the Cuyama population: (a) seed set per fruit is pollen-limited (determined by the

amount of pollen received rather than the amount of maternal plant resources

available), (b) pollinator abundances are lower than at Foster plot, or (c) pollinator


visitation rates are independent of plant size, then the lower seed number per fruit and

the independence between seed number and plant size exhibited at Cuyama could

reflect differences in pollinator service.

The mean number of seeds per fruit was significantly lower at Cuyama than at

the other two populations (see Table 3). Also, fruit set (the number of whole and

insect-damaged fruits on May 2 divided by the number of flowers, buds, developing

and aborted fruits on April 11) was significantly lower in Cuyama than in Foster plot.

This is evidence supporting the hypothesis that pollinators may limit fruit set to a

higher degree at Cuyama than at Foster. One would expect fruit set to be lower at

Cuyama if pollinators were less abundant than at Foster, since plants apparently do

not spontaneously self-pollinate to any great extent. On the other hand, there was a

higher absolute number of fruit abortions per plant on both April 11 and May 2 at

Foster population (Tables 2 and 3), and the abortion rate (number of aborted fruits

divided by the total number of developing and aborted fruits) was higher at Foster

plot than at Cuyama on both dates (although only significantly so on April 11; see

Table 7 below). This would indicate that some other factor in addition to or instead

of pollinator availability may be responsible for the difference between these two sites

with respect to fruit abortion rates.

The general picture that emerges is reminiscent of the pattern observed with

respect to the effects of herbivory during the 1993 field study (see results in next

section). Plants at Cuyama were shorter than those at Foster plot; at the same time

they had fewer aborted fruits, as did the plants which had suffered herbivory in the

weeded plots. The removal of older fruiting branches by herbivores reduces the

number of developing fruits and aborted fruits per plant. This could also explain the

weaker relationship between plant size and the mean number of seeds per fruit at

Cuyama: shorter plants were not necessarily less vigorous since some of them may

have been short due to clipping of their branches by herbivores. In fact, pruning of

branches has been shown in some species to increase allocation to female functions

(seed production) relative to pollen production (Charnov, 1982). Unfortunately

herbivory due to mammals was not noted in our observations of plants at either site

during the 1992 study, so we cannot evaluate this possible explanation.

I.


Table 7. Rates of fruit abortion in plants of Foster and Cuyama populations on April It andMay 2 1992. Rate of abortion is defined as the number of aborted fruits on a particular datedivided bv the total number of fruits (developing plus aborted fruits).

Mean rate of fruit abortion on April 11, 1992

Group: No. of plants Meanabort,rate Std.Dev.: Std. Error:

FO_It=lq 93 .398" .276 .029

CUYAMA 70 .269* .274 .033

"Significantlydifferentmeans(Mann-WhitneyU-test,p = .0022)

Mean rate of fruit abortion on May 2, i992

Group: No.of plants Meanabort,rate Std.Dev.: Std.Error:

FOSiER 65 .183" .178 .022

CUYAMA 78 .155" .194 .022

RUSS 75 .166" .22 .025

• No significantdifferencesamongmeans(Kruskal-Wallistest, p = .3131)

Spontaneous Self-Pollination

The results of the I992 experimental manipulation to determine if C.

californicus is capable of setting seed in the absence of pollinators yielded highly

significant results (Table 8). On average, inflorescences that were enclosed in mesh

bags produced less than one fruit. In contrast, control inflorescences produced

approximately 7 fruits each. Thus, reproductive success in this species is highly

dependent on the activity of pollinators.

Despite bearing relatively showy purple and white flowers, Caulanthus

populations did not attract many pollinators during the day. In fact, only two

bumblebee (Bombus) individuals were observed visiting flowers during the entire

1992 study: one was seen in Population 4 and the other was observed at Foster.

However, given that the plants did not readily self-pollinate when insect pollinators

were excluded, the insects were probably there but were not seen due to the timing of


observations. That is, pollinators active primarily during the early morning and

evening hours would likely not be observed by investigators present during the

middle part of the day.

Table 8. Results of the pollinatorexclusionexperiment. Valuesare the mean:1:s.d. of thenumberof fruits producedby two differentgroupsof inflorescences.See MethodsSectionfordetails. The twomeansaresignificantlydifferentfromeachother(t=7.12,P<<.001).

Treatment Mean # developing fruits+ s.d. Sample size

Baggedinflorescence 0.95+ 1.82 20Controlinflorescence 6.92+ 3.30 20

1993 Field and greenhouse study results

Comt_etition study

Comparison of weeded and control plots yielded some surprising results. We

were unable directly to test our hypothesis that removal of competitors, and a release

from interspecific competition, would alone increase survival and reproduction in C.

californicus. The reason for this is that plants in the treatment plots suffered

• dramatically greater herbivory than plants in control plots. Apparently they were

more visible to herbivores when not surrounded by other plants and were

preferentially eaten• Survivorship among pre-reproductive plants was much lower in

weeded plots than in control plots at Foster population during the first month after

weeding (Mann-Whitney U-test, p = .0146; see Tables 9 and 10).

We do not have survivorship data from other sites during this time period

since the plots were installed later at the other sites. Recruitment tended to be higher

(but not significantly) in the control plots (a mean of 4.2 seedlings per plot) than in

the weeded plots (.75 seedlings per plot); this may have been a result of previously

concealed seedlings "appearing" at a later sample date when the plants were larger

and easier to find among the other vegetation.


Table 9. Comparison of mortality in weeded and control plots at Foster population betweenFebruary.6 and March 4, 1993.

Group: No. of plots Mean# dead Std. Dev.: Std. Error:

Control . .447 .


Measuring mortality as a rate (dividing the number of dead seedlings by the

number of seedlings in the plot at the first sample date) produced similar results: a

mean of 46% of the seedlings died in the weeded plots between the first and second

sample dates at Foster population, while in the control plots only 3.3% of the

seedlings died. These differences were significant (Mann-Whitney U-test, p = .0207).

It is possible that the increased insolation in the weeded plots relative to the control

plots resulted in higher soil temperatures and higher rates of seedling dessication and

death.

Table 10. Comparison of mortality rates in weeded and control plots at Foster populationbetween February 6 and March 4, 1993.

Group: # of plots Meanmort. rate Std.Oev.: Std. Error:

Weeded 4 .46" .279 .139

Control 5 .033" .075 .033

• Significantlydifferentmeans(Mann-WhitneyU-test,p = .0207)

For all sites, the cumulative mortality rate (cumulative number of deaths over

total cumulative number of plants in a plot) was compared in weeded vs. control plots


(Table 11). Cumulative mortality rates were calculated from the first sample date

(seedlings) through April 18, when plants were in full flower. The weeded plots had

higher mortality than control plots (46% vs. 25%, p = .0364. Mann-Whitney U-test).

There was no significant treatment effect on cumulative mortality rate within any one

population; only when data were pooled to include the quadrats from all three

populations were significant effects of treatment (weeded vs. control plots) detected.

Table Ii. Cumulative mortality in weeded and control plots through April 18.

Cumulative mortality in control plotsMoan: Std.Dev.: Sld. Error: Variance: Coati Var.: Count:

" 124,- 1 83 1.066 1064 1,02.30,119Minimum: Maximum: Range: Sum: SumofSqr.: # Missin,q:

IO .833 .833 4.686 2.304 11Cumulative mortality in weeded plots

Mean: Std.Dev.: Std.Error: Variance: Coef.Var.: Count:

14,3- 1.3,1 1.073 1.097 16797,I1,Minimum: Maximum: Ran,qe: Sum: Sum of Sqr.: # Missinq:

I0 I1 I1 18.342I,.,08 12* Significantly different means (Mann-Whitney U-test, p = ,0364)

Causes of mortality in weeded and control plots

We could not obtain reliable estimates of the number of plants removed by

herbivores vs. the number dying of other causes, however the degree of herbivory on

the surviving plants suggests herbivory may have been a major cause of mortality for

those plants which disappeared between sampling dates. On April 18, 52% (81 of

156) of the plants in the weeded plots had suffered herbivory with a mean of 2.8

branches eaten per plant, while only 29% (50 of 174) of the plants in the control plots

had suffered herbivory, with a mean of 1.7 branches eaten per plant (Tables 12 and

13). These differences between treatments were significant (Mann-Whitney U-test,

p=.005). Thus plants in weeded plots experienced herbivory more often and more

severely than plants in control plots. This was confirmed by observations of plants in

the weeded plots sometimes being virtually chewed down to the ground level, while

plants in adjacent control plots were almost untouched (for photos, see Appendix B).


Table 12. Comparison of the mean number of branches eaten per herbivore-damaged plant inweeded and control plots (only plants with one or more branches eaten are included_. All threestudy sites are combined.

Group: # plantsdamaged Mean# bran.eaten Std. Dev.: Std. Error:

Icontrol 11.7- 11.216 .172

"Significantly differentmeans(Mann-WhitneyU-test,p = .005)

Table 13. Comparison of the mean number of branches eaten per plant in weeded and controlplots. All plants in all three study sites are combined.

Group: No. of plants Mean# bran.eaten Sld. Dev.: Std.Error:

Weeded 156 1.42g* 2.46 :1g7Centre 174 .489" 1.007 .076


It should be noted that out of 89 non-reproductive plants which died between

the sample dates at the beginning and the end of March, the remains of only 2 were

found and could be determined to have died from some sort of environmental stress.

The other 87 simply disappeared, suggesting loss through laerbivory. Between March

31 and April I8, there was apparently more stress-related mortality, with the remains

of 35 dead (dessicated) plants still present out of 73 deaths. Out of the 106 plants

which died between April 18 and May 2, the remains of 31 were still present. The

apparently greater losses to herbivory for plants in the rosette stages (prior to March

31) may simply be due to the fact that it is more difficult to locate dead rosettes than

September 15, 1993 Mazer/Hcndrickson 36

plants which were reproducing prior to death (rosettes are more likely to "disappear"

than are adult plants, regardless of the cause of death).

There was no significant effect of weeding on stress-related mortality detected

on .April 18; standing dead plants were equally common in weeded and in control

plots. Plants which died and remained in the plots tended to be the smallest plants in

the plots. Usually these were plants with one stem less than 10 cm tall and with no

more than one or two flowers and a few small buds. None of these plants appeared to

have produced any fruits. Thus differences in mortality among plots can be assumed

to be due to greater herbivory rates and/or higher stress in the weeded plots. Higher

levels of interspecifie competition in the control plots did not result in the

expected reduction in survivorship and reproduction.

Comparisons of the three populations showed there was a population effect on

the mortality rate in weeded plots in March. On March 31, the mortality rate in

weeded plots at Population 4 was 59.3%, while in the other two populations (Foster

and population six) the rates were 8.3 and 13.5% (Table 14). Herbivore activity was

extremely high at population 4 during this period (for photos, see Appendix B).

Cumulative mortality in weeded plots did not differ significantly among populations.

Mortality in control plots did not differ significantly among populations for any

census interval or cumulatively.


Table 14. Comparison of mortality rates in weeded plots in three populations (Foster,Population 4, and Population 6) during March 1993. The Kruskal-Wallis test indicates thatthere were significant differences among populations with respect to the mean mortality rateobserved within the weeded plots.

Kruskal-Wallis X 1 : POP Y 2 : mort r_atoMar

13: 2

# Groups 3#Cases 19

H 8.324 _p= .0156

H correctedforties 8.596 p = .0136

# tiedgroups 2

Population No.ofplotpaim Meanmob.rate Std.Dev.: Std.Erro_

Foster 4 .083 .167 .083

Four 5 .593 .339 .151

Six 10 .135 , .138 .043

Reproductive components in weeded and control plots

As described above, herbivory in weeded plots was much greater than in

control plots. This produced effects on reproductive components, since the

herbivores completely removed flowering branches. Over all populations, the

number of branches removed was higher in weeded than in control plots when we

compared plot means on theApril 18 sample date (see Table 15). The mean number

of whole (intact) branches per plant per plot was significantly lower in weeded plots,

while the mean number of branches eaten by herbivores was significantly higher. The

mean number of developing fruits per plant was significantly lower in the weeded

than in the control plots, despite the fact that fruit and flower abortion was

significantly higher in the control plots. This last effect can be explained by the

higher herbivory rate in weeded plots, where mature flowering br_inches with greater

numbers of fruits, aborted fruits and flowers appeared to be preferentially removed by

herbivores. Plants which hadbranches eaten early in the season produced new

flowering stems rapidly, so even though plants in the weeded plots had more branches


eaten and fewer branches, the total number of flowers and buds cumulatively

produced per plant was roughly equal to those in the control plots. The trends within

each population were similar to that in the pooled set of data; however, within Foster

plot and Population 4 differences were not statistically significant between treatments,

although the number of plots was small.

Table 15. Comparison of trait means between weeded plots (n = 18plots) and control plots (n =18 plots) over all populations on April 18, 1993. Numbers in parentheses are standarddeviations; p-values are reported for significantly different means (Wilcoxonsigned rank test).

Traits Weeded Mean Control Mean p

Number of whole branchesperplant 1.09 (.77) 1.67 (I. 13) .0097

Number of branches eaten per plant 1.45 (1.54) .617 (.757) .021

Number of flowers per plant 5.20 (6.54) 5.10 (6.49) n.s.

Number of buds per plant 10.58 (10.44) 8.49 (7.29) n.s.

Number of developin[ fruits per plant 2.92 (5.58) 5.32 (7.64) .0052

Number of abortedfruitsperplant .30 (.73) 1.25 (1.65) .0047

Number of abortedflowers per plant .28 (.50) 1.30 (1.75) .0219

Qualitatively similar patterns were seen on the May 2 1993 sample date. The

number of branches per plant was roughly equal between weeded and control plots,

probably due to the ability of the damaged plants to produce new stems very rapidly

(Table 16). The number of flowers and buds per plant tended to be higher on the

weeded plants, while the number of developing fruits tended to be lower. Aborted

fruits and flowers were again fewer on the plants in weeded than control plots,

although this was significant only for aborted flowers.


Table 16. Comparison of trait means between weeded plots (n = 16 plots) and control plots (n =16 plots) over all populations on May 2, 1993. Numbers in parentheses are standard deviations;p-values are reported for significantly different means (Wilcoxon signed rank test).

Traits Weeded Mean Control Mean p

Number of whole branches per plant 2.17 (1.67) 2.15 (1.75) n.s.

Number of branches eaten per plant 3_91 (3.00) 1.51 (1.28) .0061

Numberof flowersper plant 3.52 (4.54) 1.42 (3.06) n.s.

Numberof budsperplant 7.76 (9.8l) 3.36 (8.57) .0258

Number of developin[ fruits per plant 6.17 (6.69) 8.68 (9.34) n.s.

Number of aborted fruits per plant 1.43 (2.18) 4.06 (4.66) n.s.

Number of aborted flowers per plant 1.82 (3.46) 6.63 (8.25) .0392

General reproductive patterns in 1993

Of the 573 plants recorded in 1993 in all the populations, 160 (28%) had died

without reproducing by April 18 (it was assumed that plants which disappeared

before April 18 did not produce fruits, since no mature fruits were found on any

plants in any of the populations on this date). Between April 18 and May 2 i06

plants died, 28 of which were still present but had produced no fruits. On May 2 we

made observations on 201 plants including standing dead plants (some of the very

dense plots were subsampled, as it was not possible to sample every plant due to time

constraints). Of the plants sampled, 91 (45%) bore no developing fruits, and 39%

bore no fruits or open flowers and were unlikely to produce any fruits before

senescence (Figure 25). Thus, out of the total 573 seedlings followed over the course

of the study in all the plots, at least 329 (57%) failed to reproduce. Many of these

were plants which germinated and grew to reach the age of flowering but still failed

to produce any seed.

Some mortality can be attributed to the weeding treatment, since mortality

was higher in the weeded plots than in the controls. However, even within the control

plots, 60 out of 268 plants (22%) died by April 18, and an additional 60 died by May

2, for a total of 44% dead without reproducing. Out Of 99 plants including 14

standing dead individuals on which we counted flowers, fruits and buds, 31 (31.3%)

had no developing fruits or open flowers and were unlikely to produce any fruits

September 15, 1993 MazedHendrickson 40

before senesence. Thus even among the control plots at least 137 out of 268 plants or

51% probably died before producing any seed at all, We observed that many of the

plants which died without reproducing were the smaller plants that bolted relatively

early, as noted above in the section on phenology.

it is perplexing to note that in a year with particularly high rainfall such a

large proportion of the population failed to produce seeds. Would these plants have

successfully reproduced under other conditions? Again, we can refer to the

hypothesis noted above, that some plants may produce only very small flowers which

do not go on to become fruits but do contribute to the next generation through their

pollen. Thus, some of the plants which did not produce fruits may nonetheless have

had some reproductive success as pollen donors. Further examination of this

possibility could be a subject of a future research project.

It could be that the smaller plants were stressed due to shading by the tall, lush

vegetation produced by the high rainfall. No significant differences were found

between weeded and control plots in the number of standing dead individuals on

April 18 or May 2, as would be expected if competitive interactions influenced

survival. However, it was extremely difficult to locate standing dead individuals

within the control plots since the vegetation was so dense; this might have contributed

to a low estimate of plants dying from competitively induced stress in those plots.

Comparision of 1992 and 1993 fruit production

Although we cannot directly compare fecundity estimates between 1992 and

1993 due to a lack of data concerning seed number per fruit in 1993, a comparison of

fruit production at Foster Plot in the two years reflects the differences in the sampling

techniques used. In 1992 plants produced a mean of 12.7 fruits per plant for 65 plants

sampled (Table 2); in 1993 the mean was 8.2 fruits per plant (n = 66 plants, including

7 standing dead plants). It is possible that plants were simply larger and produced

more fruits in 1992 than in 1993; however it seems more likely that sampling all of

the plants within each 0.25 m2 plot in 1993 resulted in the inclusion of more of the

smaller individuals that were difficult to see at first glance, and which produce few or

no fruits. This kind of intensive sampling may give a more realistic picture of the

population as a whole. The "random" selection and marking of individuals employed

in 1992 may have failed to include the smallest (but still abundant) flowering

September 15, 1993 Mazer/Hendrickson 4l

individuals (see photo 5, Appendix B for an example of a tiny precociously floweringindividual seen in 1993).

Greenhouse ootlination studies

It has been suggested (Price and Waser 1979, cited in Barrett and Kohn, 1991)

that because of limited dispersal and adaptation to local environmental conditions

there is an optimal outcrossing distance for many plant species. Crosses between near

neighbors would suffer from inbreeding depression because plants are likely to be

related, while crosses from widely separated individuals would suffer from

outbreeding depression because of the disruption of favorable gene complexes

following the mating between individuals adapted tO the particular environmentalconditions at their home site.

The occurrence of populations of California Jewelflower in discrete, widely

separated patches, the clumped distribution of plants within a population (see

observations on ecology and population structure, above), and the low levels of seed

dispersal would all tend to increase inbreeding and adaptation to local environmental

conditions. If pollinators minimize the distance of inter-plant flights, then most plants

would receive pollen from their nearest neighbors, which because of the lack of seed

dispersal mechanisms would also tend to be their relatives. Over time_ inbreeding

causes deleterious genes to be purged from the gene pool, since plants which received

a double copy of a deleterious gene would be less likely to reproduce. Thus, if the

existing population structure of C. californicus is similar to the conditions under

which the plants evolved (that is, if populations have typically been small and

patchily distributed, allowing the purging of deleterious genes) we would not expect

to see much inbreeding depression in this species. However, if deleterious alleles

have not had sufficient time to be purged from local populations, or if outcrossing in

natural populations is sufficiently high to allow deleterious alleles to increase in

frequency by remaining "masked" in heterozygote form with dominant non-

deleterious alleles, then we would expect to see the detrimental effects of inbreeding

depression following self-pollination.

There is some evidence for an optimal outcrossing distance in California

Jewelflower. Inbreeding depression did not appear to be severe, suggesting that many

deleterious alleles (at least those expressed during seed development) have been


purged from natural populations. Selfing in greenhouse crosses did not appear to

result in decreased seed number or increased seed abortion rate (see Table 17).

However, there was an effect on seed weight, with mean individual seed weight

significantly lower in selfed fruits than outcrossed fruits (0.623 mg vs. 0.676 mg). In

addition, comparison of outcrossed matings between individuals from the same

population to outcrossed matings between individuals from different populations

provide some evidence of outcrossing depression. The abortion rate was significantly

higher in fruits resulting from between-site crosses (17.6% vs. 6.5%) than in fruits

from outcrossed pollinations in which the pollen donor came from the same

population as the maternal plant (Table 18). No other trait measured was

significantly different between within- and between-site crosses, although there is a

tendency for between-site crosses to produce fruits with fewer seeds (L7.9 vs. 22.7)

and lower seed weight (.642 mg vs..676 mg) than within-site crosses.

These results provide support for the postulate that there exssts an optimal

outcrossing distance in Caulanthus californicus, since there was some inbreeding

depression expressed following selfing and some outcrossing depression exhibited by

crosses outside the population. The optimal crosses were those outcrosses performed

between individuals within the same population. Other studies using plants from

more widely separated sites would be useful in further examining this breeding

system.

It is important to note that the differences we observed between fruits and

seeds resulting from self-pollination vs. from out-crossing suggest that these small

greenhouse populations did include genetically distinct individuals. If the greenhouse

populations (representing different sites of seedling collection) included no genetic

variation, we would not expect to detect any phenotypic differences between fruits

and seeds resulting from self-pollinations and outcross pollinations. The fact that the

identity of the pollen donor (self vs. within-site outcross vs. between-site outcross)

did affect the quality of the seed produced implies that the individuals used as pollen

donors did differ genetically.


Table 17. Comparison of selfed and outcrossed fruits (within-population crosses only) fromgreenhouse pollination studies.

Treatment Number Mean seed Mean rate of Mean individual

of fruits number per seed abortion seed weight (mg)fruit perfruit(%)

Selfed 60 19.5(13.5) 11.4 (28.1) .623 (.16)*

Outcrossed 150 21.7 (15.7)" 6.5 (20.0) .676 (.144)*

* Significantly different means (Mann-Whitney U-test, p = .0388

Table 18. Comparison of mean seed abortion rates for within-site vs. between-site crosses.Within-site crosses include only crosses between individuals (self-pollinations are not included inthis comparison).

Group: No. of fruits Meanseedabort, rate Sld. Dev.: Std.Error:

within site 154 .085 ,2 .016betweensites '34 .1761 .291 ,05

• Differencesbetweenmeansare significant(p = .0016,Mann-WhitneyU-test)

It was observed that seed number per fruit in all the greenhouse-grown

individuals was much lower than that in field-collected fruits regardless of the

pollination treatment (compare Tables 3 and 17). Mean individual seed weights

were also somewhat lower, especially when compared to the Carrizo populations

where these greenhouse plants originated. Plants grown in the greenhouse (where

light levels are low compared to the field) appeared etiolated, and flowers were small

compared to field-grown plants. This observation suggests that. environmental

variation in resource levels in the field may account for the observed variation in

flower size in the field.

Several factors may account for the apparently lower resource status of

greenhouse-raised plants. First, lack of sunlight in the greenhouse could be partially

responsible for the relatively low number of seeds per fruit and a low seed weight.

That is, low seed number per fruit could be due to the production in the greenhhouse


of relatively small flowers with small ovaries and few ovules. In addition, rates of

early ovule and seed abortion may have been higher in the greenhouse than in the

field, resulting in reduced seed production per fruit in the greenhouse. Second, the

relatively small pot size of the cultivated plants (plants were raised in the greenhouse

in a-inch pots) may have contributed to their small size, Third, the seedlings

collected from the field and raised in the growth chamber experienced long days

relative to the concurrent daylengths in the field. The growth chamber may have

promoted precocious flowering before the rate of root growth and/or of

photoassimilation had reached levels that normally promote or allow bolting in the

field. Such precocious flowering may have contributed to the production of the small

flowers, low seed numbers, and small seeds observed in the greenhouse.

SUMMARY AND MANAGEMENT RECOMMENDATIONS

Populations of California Jewelflower (Caulanthus californicus) in the

Carrizo Plain and Cuyama Valley were found to occupy relatively small, discrete

areas. Individuals within the populations exhibit a clustered or aggregated

distribution. Since there appears to be little seed dispersal, it is possible that

populations of California Jewelflower have always been relatively small and largely if

not totally reproductively isolated, leading to the formation of populations highly

adapted to a particular site and to frequent inbreeding. There appeared to be some

degree of reproductive isolation between relatively close populations (outbreeding

depression), and also some inbreeding depression when plants were selfed. Crosses

that produced the highest quality (largest) seeds were outcrossed pollinations between

individuals from the same population. There was little spontaneous self-pollination in

the field in the absence of pollinators, so the breeding system can be assumed to

involve outcrossing to some degree. More information is needed about the long-term

effects of inbreeding and outbreeding depression in this species, particularly if

revegetation efforts include the establishment of new populations from seed sources

of limited genetic variation. Owing to time and personnel constraints, we wereunable to determine whether the differences in seed. mass we observed between selfed

and outcrossed pollinations resulted in differences in progeny performance following

germination.


Seeds of California Jewelflower showed a high degree of dormancy, and the

formation and persistence of a seed bank is probably a key factor enabling these small

populations to maintain their genetic diversity and to persist in an unpredictable

environment (Leck et al., 1989). The living population is likely to be both much

larger and more genetically diverse than is apparent from the aboveground population

because the soil seed bank may persist for years and contain many generations ofseeds.

California Jewel flower seedlings appeared to emerge following the first heavy

winter rains in December, and remained in a rosette stage until March, when

flowering began. Initiation of flowering varied among individuals, with some very

small plants bolting at the beginning of March and some of the largest plants still not

having open flowers by April i. Variation in flowering phenology could be related to

genetic or environmental factors, and probably serves to ensure that at least some

(early flowering) individuals are able to set fruit in years when the soils dry out early,

while in more favorable years large plants that delay flowering are able to produce

particularly large numbers of seeds.

Populations of California Jewelflower showed within- and between-site.)

variation in a number of floral and reproductwe traits. Populations in the Carrizo

Plain were similar to each other with respect '_omost traits and all differed from the

Cuyama Valley population in a number of floral traits and in reproductive output.

Some of the differences may have been due to the slightly later flowering Of the

higher elevation population at Cuyama. Other potential causes for differences

observed between populations with respect to reproductive output include geographic

variation in the composition, abundance, or behavior of insect and mammalian

herbivores and/or the abundance and behavior of pollinators. Abiotic environmental

differences between sites may also contribute to the observed differences between

populations in these two areas.

Mana_,ement considerations

If reintroduction of the species to sites within its historical range is to be

considered as one of the components of a recovery plan, the above observations raise

important considerations. Reintroductions may fail if:


1. The site selected does not possess the particular abiotic (soil, aspect, temperatures

etc.) and biotic (pollinators) habitat characteristics required for species to establish

or to reproduce;

2..The introduced population originates from a site which is far away from the

reintroduction site, and which is characterized by a distinct temperature and/or

rainfall regime. The environmental cues necessary for germination or flowering

of the introduced population may not exist at the new site.

3. The genetic diversity of the introduced seedbank is insufficient, so that inbreeding

reduces plant performance over time;

4. The introduced seedbank is not sufficiently diverse to produce a population

capable of adapting to the new site and surviving there over the long term; and

5. The introduced seedbank contains individuals from widely separated populations

and the resulting offspring suffer from the effects of outbreeding depression.

One strategy to minimize the potential traps of inbreeding and outbreeding

depression and to increase the chances of producing a population diverse enough to

allow adaptation to a site would be to collect seed from several populations as near as

possible to the reintroduction site. The populations should all be relatively close

together (within a mile or less if possible). If clear morphological differences

between populations are observed, this may indicate strong genetic divergence of

populations. In this case, to avoid possible outbreeding depression, test-crosses

between individuals from the alternative populations may be conducted to determine

whether they are genetically compatible. Alternatively, seed from only one

population source could be used for each colonization trial.

Compromises will be inevitable when planning a seed collection strategy prior

to recolonization or revegetation efforts with C. californicus. If more than one

population is used as a seed source, one risks some level of genetic incompatibility

(outbreeding depression) following matings in the field of individuals representing

distinct populations. However, since seeds from several sites would probably be

genetically more diverse than seed from a single source, the probability that some

seed would be adapted to the uncolonized site may be increased relative to single-


source seed collections. If the strategy used aims to collect seeds that are as

genetically diverse as possible, then habitats from which seeds are collected should be

as different as possible with respect to slope, aspect, soil type and vegetation: Seed

from each population should be collected over a period of 3 - 5 years and stored until

use, in order to sample the diversity present in the seedbank of each source

population. Only one fruit per individual should be taken per year, and as many

different individuals as possible should be sampled within each population (while

avoiding inappropriate impacts to the population by collecting no more than 5% of

the seeds from any particular site in any one year). Test crosses between individuals

representing distinct habitats can be conducted in the field (transporting pollen

between sites in glass vials) or in the greenhouse to determine the level (if present) of

genetic incompatibility between populations.

lnterspecific interactions

Although our study provided no direct evidence for interspecific competition

having a negative effect on survival or reproduction, our observations suggest that it

may play a factor in some situations. Introduced annual grasses such as Brornus

rubens were present at all the study sites, but did not dominate any of them as Bromus

rubens does at some of the historical sites for this species in the San Joaquin Valley

visited by Taylor and Davilla (1989). In one of the control plots where there was a

particularly high cover of annual grasses many of the Caulanthus plants died after

bolting but before producing any fruit. Most of the other C. californicus plants

appeared to be stressed, and it is possible that this was a result of competition with the

grasses.

It should also be noted that the effects of interspecific competition may not be

highly detrimental in high rainfall years such as the two years of this study.

Resources would tend to be more limiting in a low rainfall period, and that is when

competitive interactions are more likely to be important.

Removal of potential interspecific competitors by weeding did produce a

substantial and unexpected effect due to the predation of branches and fruits by

mammalian herbivores (giant kangaroo rats and rabbits) on the weeded plots. Plants

which were in the unweeded plots had lower mortality and lower probability of

predation, probably because they were less obvious to the animals. They also had


greater reproduction, since herbivores removed entire flowering branches as they

reached maturity in both the control and weeded plots, but they did this with greater

frequency in the weeded plots.

Although we were not able to directly investigate the presence or effects of

interspecific competition in California Jewelflower populations our results do have

some implications for other aspects of management. Grazing by cattle or sheep

during the rosette stage of the plants is unlikely to result in much damage by direct

herbivory; however if the overall plant cover is reduced by grazing then bolting

Jewelflowers will be more obvious and subject to herbivory from small mammals and

cattle. Herbivores can effectively prevent fruit production by repeated grazing, since

usually the entire inforescence is removed at once and a new flowering branch must

be initiated from the base of the plant, which in turn is then removed as it matures.

Heavy grazing, particularly between February and mid-May, is likely to be

detrimental to California Jewel flower populations. Light grazing during the period

before the plants have bolted (bolting may start between late February and April,

depending on the year and the site) may be an acceptable practice, but more

information is needed on the impacts of large grazers before guidelines can be set. In

particular, the effects of grazers on soil texture (due to the compacting effects of

heavy animals) and subsequent germination and seedling survivorship must beassessed.

Giant Kangaroo Rat impacts

Co-management for this and other endangered species in the Carrizo Plain

raises some interesting problems. Our weeding study indicated that small mammal

herbivores can have a substantial impact on California Jewelflower reproduction by

repeatedly removing inflorescences as they mature, retarding or even preventing the .

formation of mature fruits by some plants. One of these mammalian predators is the

giant kangaroo rat, itself an endangered species and probably one of the principal

herbivores on California Jewelflower at the Carrizo Plain sites. Kangaroo rat

populations appeared to be increasing rapidly over the last two years and impacts to

Jewelflower populations may become serious if the area is overpopulated by the

kangaroo rats on a long-term basis. Increasing densities of predators on kangaroo rats

and the normal fluctuation of small mammal population sizes will probably alleviate


such problems in the long term if the natural ecological balance is maintained in the

system.

Future research

Several questions raised by this study deserve closer attention in future research with

this species. Projects could include:

1. Conduct detailed studies of the abiotic environmental factors that permit

successful persistence of Caulanthus califon_icus populations. What do all sites

share with respect to: yearly mean temperature; temperature maxima and minima;

rainfall; soil type; soil texture: slope and aspect?

2. Identify primary pollinators of Caulanthus californicus to ensure that new sites of

population establishment will contain sufficiently abundant natural or introduced

populations of them.

3. Cultivate Caulanthus californicus colonies in the greenhouse to produce an

experimental seed bank. Pilot studies to determine the ability to establish new

colonies using these cultivated seeds should be undertaken. The success of this

effort would reduce the need to remove seeds produced in the field to provide a

seed source for new populations.

4. Conduct more thorough and systematic sampling of flowers to determine if there

is actually dimorphism in flower type; and, if so, whether the two types of flowers

occur within an individual or on different plants.

5. Compare outcrossing rates and male and female reproductive success between the

two types of flowers.

6. Examine potential environmental and genetic correlates of floral dimorphism.

7. Perform further outcrossing experiments using plants from more distantly related

populations (eg. Cuyama, Kreyenhagen Hills, and Carrizo Plain populations) to

determine whether distant outcr0sses consistently result in increased or decreased


performance. Other measuresof offspring fitness should be employed as well as

seed counts and weights: germination rates, establishment, growth rates and

reproduction of offspring should all be measured.

8..Conduct greenhouse studies to determine potential interference of Bromus rt,bens

on California Jewelflower, and possible interactions with air pollutants.


BIBLIOGRAPHY

Abrams, L. 1944. Illustrated flora of the Pacific States. Stanford University Press,Stanford, CA.

Barrett, S. C. H. and J. R. Kohn. 1991. Genetic and evolutionary consequences ofsmall population size in plants: Implications for conservation. In: Geneticsand Conservation of Rare Plants. D. A. Falk and K. E. Holsinger (Eds.).Oxford University Press, Oxford.

Campbell, D. R. 1992. Variation in sex allocation and floral morphology inIpomopsis aggregata (Potemoniaceae). American Journal of Botany. 79:516 - 521.

Charnov, E. L. 1982. The Theory of Sex Allocation. Monographs in PopulationBiology. Princeton University Press, Princeton, New Jersey.

Dieringer, G. 1991. Variation in individual flowering time and reproductive successofAgalinisstrictifolia (Scrophulariaceae). American Journal of Botany. 78:497 - 503.

Elias, T. S. (Ed.) 1987. Consen'ation and Management of Rare and EndangeredPlants. California Native Plant Society, Sacramento, CA.

Fox, G. A. 1989. Consequences of flowering time variation in a desert annual:adaptation and history. Ecology 70:i294 - 1306.

Fox, G. A. 1990a. Components of flowering time vaariation in a desert annual.Evolution 44: 1404- 1423.

Fox, G. A. 1990b. Drought and the evolution of flowering time in desert annuals.American Journal of Botany 77:1508 - 1518.

Hoover, R. F. 1970. The Vascular Plants of San Luis Obispo County, California.University of California Press, Berkeley CA.

Kang, H. and R. B. Primack. 199 I. Temporal variation of flower and fruit size inrelation to seed yield in Celandine poppy (Chelidonium majus, Papaveraceae).American Journal of Botany. 78:711 - 722.

Lacey, E. 1986. Onset of reproduction in plants: size- versus age-dependency.Trends in Ecology and Evolution 1:72 - 75.

Leck, M. A., V. T. Parker, and R. L. Simpson. 1989. Ecology of Soil Seed Banks.Academic Press, San Diego, CA.

Lewis, R. I988, 1991. California Natural Diversity Data Base forms describing newCaulanthus californicus populations in the Carrizo Plain, Cuyama Valley andKreyenhagen Hills.

Ludwig, J. A. and J. F. Reynolds. I988. Statistical Ecology. John Wiley & Sons,New York.


Mazer, S. J. and C"T. Schick. 199 la. Constancy of population and genetic parameters forlife-history and floral traits in Raphanus sativus I. Norms of reaction and the natureofgenotype by environment interactions. Heredity 67: 143-156.

Mazer, S. J. and C. T. Schick. 1991b. Constancy of population and geneticparameters for life-history and floral traits in Raphanus sativus II. Effects ofplanting density on phenotype and heritability estimates. Evolution 45: 1888-1907.

Mazer, S.J., A. A. Snow and M. L. Stanton. 1986. Fertilization dynamics andparental effects upon fruit develo.pment in Raphanus raphanisrrum:consequences for seed size variauon. American Journal of Botany 73: 500-511.

Munz, P. A. and D. D. Keck. 1959. A California Flora. University of CaliforniaPress, Berkeley, CA.

Price, M. V. and N. M. Waser. 1979. Pollen dispersal and optimal outbreeding inDelphinium nelsonii. Nature 277: 294-298.

Rathcke, B. and E. P. Lacey. 1985. Phenological patterns of terrestrial plants.Annual Review of Ecology and Systematics. I6:179 - 214.

Schmitt, J. I983. Individual flowering pbenology, plant size, and reproductivesuccess in Linanthus adrosaceus, a California annual. Oecologia 59:135 -140.

Shreve, F. 1951. Vegetation of the Sonoran Desert. Carnegie Institution ofWashington Publication 591, Washington.

Smith, C. F. 1976. A Flora of the Santa Barbara Region, California. Santa BarbaraMuseum of Natural History, Santa Barbara, CA.

Smith, J. P. and K. Berg. I988. Inventory of Rare and Endangered Vascular Plantsof California. Special Publication No. 1 (4th edition), California Native PlantSociety, Sacramento, CA.

Smith, S. S. E. 1980. Micorrhizas of autotrophic plants. Biological Review 55:475 -510.

Stanton, M. L. I985. Seed size and emergence time within a stand of wild radish(Raphanus raphanistrum L.): the establishment of a fintess hierarchy.Oecologia 67:524 - 531.

Stanton, M. L. and R. W. Preston. 1988. Ecological consequences and phenotypiccorrelates of petal size variation in wild radish, Raphanus sativus(Brassicaceae). American Journal of Botany. 75:528 - 539.

Taylor, D. W. and W. B. Davilla. 1989. Status survey of three plants endemic to theSan Joaquin Valley. Prepared for: Office of Endangered Species, US Fishand Wildlife Service, Sacramento, CA.


Twisselman, E C. 1967. A'Flora of Kern County California. Wasman Journal ofBiology25:1 - 395.

Young, J. A. and C. G. Young. Collecting, Processing and Germinating Seeds ofWildland Plants. Timber Press, Portland, OR.

We'nt, F. W. 1949. Ecology of desert plants. II. The effect of rain and termperatureon germination and growth. Ecology 30:1 - 13.

54

20"

16"

14"

12"

10 i' ,• _ ] , ,

0 8 _ i - •• diI:i;i

6" :I :dl?_Id_:?:"

4" [-. t ._ --

2" ,_,_i_;±

010 20 30 40 50 60 70

Plant Ht (cm)

Figure la. Frequency distribution for plant height in Foster plot on April 11, 1992.

20"

121

10

o 8U

6"

4" I '

0 10 20 30 40 50 60 70

Plant Ht (cm)

Figure lb. Frequency distribution for plant height in Cuyama population, April 11, 1992.

55

0 i i i i i i i i i i i i i i i i i i i i i

45

40'

35"

30"

25"

151 _KK_:

5 _ _

0 10 20 30 40 50 60 70 80 90 100

Number of fruits per plant

Figure 2a. Frequency distribution for number of developing fruits per plant at Foster Plot on May 2, 1992.

50'

45

40"--

35

30"

2s-

L) 201

15"

,51

00 10 20 30 40 50 60 70 80 90 100

Number of fruits per plant

Figure 2b. Frequency distribution for number of developing fruits per plant at Cuyama population, May 2,1992.

56

45

40 1

35" _

0 25 50 75 ,100 125 150 175 200 225 250

Total number of flowers, developing fruits & buds per plant

Figure 3a. Frequency distribution of total flowers, developing fruits and buds per plant at Foster Plot onApril 11, 1992.

45 I , t , t , I , t , r , t , I t , I , I

40

35

30

= 20

!_ _, _-.

t_

0 25 50 75 100 125 150 175 200 225 250

Total number of flowers, developing fruits & buds per plant

Figure 3b Frequency distribution of total flowers, developing fruits and buds per plant at Cuyama populationon April 11, 1992.

57

4Ill,I,1,1,1,1,111,1,1

12

10 ¸

8̧

4' _}i.

I

0 10 20 30 40 50 60 70 80 90 100

Number of seeds per fruit

Figure 4a. Frequency distribution of number of seeds per fruit at Foster Plot on May 2, 1992.

4 i , i i i , i i i i i i i i i , i i i , i

12]

10] I _.

IoNINI rN0 10 20 30 40 50 60 70 80 90 100

Number of seeds per fruit

Figure 4b. Frequency distribution for number of seeds per fruit at Cuyama population on May 2, 1992.

58

16'

14"

,12'

10'

00 10 20 30 40 50 60 70 80 90 100

Number of ovules per ovary

Figure 5a. Number of ovules per ovary in Foster Plot plants on April 11, 1992.

16"

10-

6-

o 4"

0 10 20 30 40 50 60 70 80 90 100

Number of ovules per ovary

Figure 5b. Number of ovules per ovary in Cuyama population plants on April 11, 1992.

59

70"

60

J

OL)

0 10 20 30 40 50 60 70 80 90 100

Number of open flowers per plant

Figure 6a. Frequency distribution of number of open flowers per plant at Foster Plot on April 11, 1992.

70"

5O

40

o 30

20

10

00 10 20 30 40 50 60 70 80 90 100

Number of open flowers per plant

Figure 6b. Frequency distribution of number of open flowers at Cuyama population on April 11, 1992.

oO

01,111,1,111,111,111,1

60 ¸

50 ¸

40 ¸ _

: 30G_

20

0 10 20 30 40 50 60 70 80 90 100

Number of flower buds per plant

Figure 7a. Frequency distribution of number of flower buds per plant at Foster Plot on April I I, 1992.

80"

5014ot

20 :-'

10 __

0 N_0 5 10 15 20 25 30 35 40 45 50

Number of aborted fruits per plant

Figure 7b. Frequency distribution of number of flower buds per plant at Cuyama population on April II,1992.

61

01*lltltltl*lllllll lit

60"

50"

4O]_

l i_l.Q _

20"_ _

10 .... _

N0 10 20 30 40 50 60 70 80 90 100

Number of flower buds per plant

Figure 8a. Frequency distribution of number of aborted fruits per plant on May 2, 1992 at Foster Plot.

80"

O

30

20

10

00 5 10 15 20 25 30 35 40 45 50

Number ofabortedfruits per plant

Figure 8b. Frequency distribution of number of abo_ed _ui_ per plant at Cuyama population on May 2,1992.

62

Mean ovule number per plant

Error Bars: ± 1 Standard Error(s) ,9O

8O

7O

6O

50

40

3O

20

10

0

Individual Plant Identity

Figure 9. Means and standard errors for ovule numbers in buds from randomly sampled plants (5 buds perplant) at Foster Plot and Cuyama population on April 11, 1992.

63

160 ' ' ' ' y = 13.068- 1.823x + .07x 2 '

140" o/

1201 .

I00" _80" _ o

6o- o "y/40" • _O,_ a__20" O _,_/ " _'O

O__aa o o

0 10 20 30 40 50 60

Plant Ht (cm)

Figure 10. Number of live flowers and buds per plant vs. plant' height in Foster Plot on April I l, 1992 (r2 =.681, p = .00Of).

y = 9.69 - 1.03x + .035x 2100

90 •

80

70 •

60

o_ 50C'4"_ 40' •

•=- 30 Nb •

:_ 20 • •

10' III1•

O' -m.

-100 10 20 30 40 50 60

Plant Ht (cm)

Figure If. Number of fruits on May 2, 1992 at Foster Plot vs. plant height (r 2 : .547, p = .0001)

64

100 " ' ' ' y' =:24"918 + 3.?69x_.O29x 2

90 0 0

80

70 o o o _o60 o oo oo ...9_° _'_0 y/ Oo

°.oo o-

_. 30 .,,"°°'_°o° " oI 20: / o o

10" "0 0

010 15 20 25 30 3; 4o 4; 50 55 60Plant Ht (crn)

Figure 12. Number of seeds per fruit on May 2, 1992 at Foster Plot vs. plant height on April 11 (r_ = .471. p =.0001).

_, _¢=,15.23.298 - 150.941x + 3.945x2. ,9000 ' ' ' '

80001 A

70001

6000 Z

5ooo. A

_ 4000"0300 "

2000 •

I000

, , . , , , . ,

10 15 20 25 30 35 40 45 50 55 60

Plant Ht (cm)

Figure 13. Fecundity (fruit number x seed number per fruit vs. plant height for Foster Plot (r2 = .594, p =.O00l).

65

y = .00 ,2x+ .316, ,r2 =,.0141.2 ' ' ' ' ' '

1 V

V

.8 V_

v "_ vv v

.2 vw V'v .

0 v. v. . vao 30 40 5o 60Plant Ht (cm)

Figure 14. Relationship between plant height on April 11, 1992 and fru!t set on May 2, 1992 at Foster Plot (r2= .014, p = .6389).

y =,-.33,4 + .4,26x • .005,x212 ......

,..;...,.8 • •_• oe8 • _•

J_pw lg _

_D4Z• O0_ •6jw_ --

m 4

2

010 : 20 30 40 50 60

Plant Ht (cm)

Figure 15. Bud length vs. plant height at Foster Plot on April II, 1992 (r2 = .483, p = .0001). At Cuyama therelationship between bud length and plant height is also significant (r2 = .446, p = .0001).

66

y = 1.007 - .275x + .037x2I i i i I • • . , . , .

3 [][] D n []

'2. t [] [] [][] /

[] [] [] _/ []

_ 1.51 0 _ /_ Db

o | •o__nu _[]0- _ -/

03 4 5 6 7 8 9 10 11 12

Bud lengthFigure 16a. Ovary. length regressed on bud length in Foster Plot plants on April 11, 1992 (r2 = .42, p = .0001)

3.5 ' ' Y=''78'5+._91,_-.q33x2i i I i

2.5 []

[] [] DDO

_ 2' [] []m _

_1.5

0 1 [] S [] i

.5 [] [] _ [] []

04 5 6 7 8 9 10 11

Bud length

Figure 16h. Ovary length regressed on bud length at Cuyama population, April I1, 1992 (r 2 --.058, p =.1741).

67

lOG Y ='19"2+9"094x''122x 2 . , .

90' [] []

80'

.-. 70 [] /cq [] [] []60' []

so []

"_ _ []30" m m

20" _ [] []

101 :I []

0 , [].... D,4 5 6 7 8 9 10 11 12

Bud length

Figure 17a. Regression of seed number per fruit on May 2 vs. bud length of next bud to mature on April I1;

Foster Plot plants only. (r2 = .371; p = .0001)

90 ' ' ' ' y '= -35.872 + 11.6,68x : .3,7Yx2, ,

80 a

70 []

_. 50 • [] []

mad= [] [] _.-_u/. 4030 :_ /a :

• [] [] ci: []

10' [] ma mE [] []

04 5 6 7 8 9 10 11

13ud length

Figure'17b. Regression of seed number per fruit on bud length ai Cuyama (r2 = 156, p = .0003)

68

100 ' ' ' Y ='20.023 + 25.1,58x - 3.986x 2 .

901 V v

80"

70" V• v v v

c_ V v V

50_ vVv/...,._.vv Vv

•-_'_'_"40_30 V_-_ "_ v V"v v v

20 v v

10 v

0 .7 v0 .5 1 1.5 2 2.5

Ovary length

Figure lga. Regression of seed number per fruit on ovary length, Foster Plot (r2 = .33, p = .0001).

y = 28.859 - 8.299x + 4.467x 290 ' ' ' ' ' ' ' ' ' ' ' ' '

80 vV

70

_-=,_60 Vv Vv_ 50'--.. V

-_ 40''- V

-p 30'

20, "v -_ v-,_v vV

10' w v v v v vv

00 .5 I 1.5 2 2.5 3 3.5

Ovary lengthFigure 18b. Regression of seed number per fruit on ovary length showed no significant relationship in

Cuyama population (r2 =.087 p = ,0853)

69

y = .454 + .011x + 7.742E-5x 2

1

.8, ,

"U(D .6"

:t |

7.5 10 12.5 1; 17.5 20 22.5 25 27.5 30 32.5

Plant Ht (cm)

Figure 19. Relationship of mean individual seed weight to plant height in Cayama population (r 2 = .176, p =.0OO9)

70

I0

9

8

7

_ 6e_

_ 5

.Q

E= 3'Z

2'

1'

O'0 2 4 6 8 10 12 14

Bud length

Figure 20a. Frequency distribution of bud lengths in Foster Plot on April 11, 1992.

0 I i i I , i i I , i , i i J

9'

8

7'(/3

.(3

5"

4".C3

"_ 3"

Z

I. !]0 2 4 6 8 I0 12 14

Bud length

Figure 20b. Frequency distribution of bud lengths at Cuyama population on April 11, 1992.

71

14

10"

_ 8"

_ 6"_ ,

Z 4"

00 .5 1 1.5 2 2.5 3 3.5 4

Anther lengthFigure 21a. Frequency distribution of anther length at Foster Plot on April ll, 1992.

F a I n I n I * I i I n I a I I I

14 ¸

00 .5 1 1.5 2 2.5 3 3.5 4

Anther length

Figure 2lb. Frequency distribution of anther length at Cuyama population on April 11, 1992.

72

n I n I n I I I a I i I n I n I

14"

'12"

I0" _t_

6

2 _i

.,j , , . , ,

0 .5 1 1.5 2 2.5 3 3_5 4

Ovary length

Figure 22a. Frequency distribution of ovary length in buds from Fo_ter Plot, April 11, 1992.

n I i I n I I I n I n I n I I I

14

12

10 ¸

n

6̧ ,

4 •_ _,_ - _i., _ _

, ' . , .

0 .5 1.5 2 2.5 3 3.5 4

Ovary length

Figure 22b. Frequency distribution of ovary length in buds from Cuyama population, April 11, 1992.

73

14 I , I , I , I , I I t I , I ,

12: l10"

8"

6

4

2

00 .5 1 1.5 2 2.5 3 3.5 4

anther/ovary length

Figure 23a. Frequency distribution of ratio of anther length to ovary length at Fester Plot on April 11, 1992.

14

12

10

8

6

4

,

0 .5 1 1.5 2 2.5 3 3.5 4

anther/ovary length

Figure 23b. Frequency distribution of ratio of anther length to ovary length at Cuyama population on April11, 1992.

74

3

2.5

e"

1.50

1

,N

00 .5 1 1.5 2 2;5 3

Ovary length

Figure24a. Regressionof anther:ovarylengthratiotoovarylengthin FosterPlot.

Be n I i I n I I n I n I *

• • ° •2,5

1.5

=

'0_0 .5 1 1.5 2 2.5 3 3.5

Ovary lengthFigure 24b. Regression of anther:ovary length ratio to ovary length in Cuyoma population.

75

00 I ,I,l,l,l,l,l,l,l_lll

90"

8O

70

6O

5O

Q40

30

20

0 n . . .,--m,,.-- .,rn_-_.-_ r . _ ",-10 0 10 20 30 40 50 " 60 70 80 90 100

dvlping fruits

Figure 25. Frequency distribution of number of developing fruits per plant on May 2, 1993. Both weeded andcontrol plots in all three populations are included.

.. _ • ..._:,.._,_,,-,.- .. 07

_4. e , - ._ ._-L J.I,.._.l X_ '_ g ,

-""_?_ _" :-. ., ...........r,- v- ¢_ .

• _F_o_J." _,,:" "-"" r- ,.u=,,:,,,. ,,;'. ; "_'.v "_-_.":

G : l . _ ..'_ _m.jgj, r ,_.

.'1

--" /.. lu _" _'_.....• : ,,., . q '" .... , ;,_=.-_ .

.t. a_" P "_ 1_ • T U g " ,A.• i

" '_' 8 " / I. _r. _10 Z_ -o K I N G S I -"-.

%-" /" 34 ._ I "_: -- "-'." ,_._.._ _ . _/ "% - , .

"% _'oo.:-h:._. °''°""" "' ' " " --: '., - • 2 ,,,. ._j..f.,

:- ,,_ .: _ ..... _

_""'1'"* _ _ 4° "1"__p _ __J.__. _ _ _ .

• . I O,c_,__, 11 _. . i, • -| '" "_}-'! • 14

I 30 .% A ;r•u • % le 15_ 21A _ jL A N".• _ 17 - ';

• .: 1 20 "_"iIS _ B I S P.Q " _ i

• ' • A 39 , p,' "-'1_-_ _" .

,. ..-_._j e __....-_'_'_ A _ •_ J 38 " 8akersfleid

• _'" -- "".*: " :° .... -:_ l_s-_

• _ 0 12%F_ur¢ 2. M_ of the sou,.hem San 27 r- "-,.,

/oaquin VNI_7 d_icting the dL_'ibu¢ion ,_. G "--'Ix

• . I _ . .-_'_ Vcmura . . ,

Map 1. Historical distribution of Caulanthu_ cMifor_icus (from Taylor and Da_illa,1986)

Map 2. Carrizo Plain populations of Caulanthus californicus (from Lewis 1991) andlocations of study sites.

Map 3. Cuyama Valley populations of Caulanthus californicus (from Lewis 1991)and location of Cuyama population study site.

081

Photographs 1 and 2. Caulanthus californicus habitat views. Photo 1 (top):population 4 site (note pink flag in lower left corner). Photo 2 (bottom): Population 4

site looking SW; shrubs are Ephedra californica.

0S2

Photographs 3 and 4. Caulanthus californicus inflorescences. Photo 3 (top):Inflorescence with buds and open flowers. Photo 4 (bottom): open flowers.

083

O84

PhozoBraDh 6. Weeded (right) and con=rol elozs a= po-u!acion 4.

0_5

Appendix C

Field and greenhouse data

Caulanthus californicus Cuyama population: Lab & Field data for 1992

See text for sampling methods.

x_

O$'7

• . ._ . CI_I

i

oo,oooooooo,ooooooooooooooo,oCM _%1

o::oooo.ooooo oooololo ooooooo ooooooo_ ,

ii ii_ " _' i _ 'ii

_ ...... o _o_o_.. _.=_.= _e_i

'_- , _.._. _. _ _._.,.-_._._ _: . ._.

1

-} • _=_....... . _=o_o°_'"=_°.o_o_ooo_o_o_ooo_oo_=o_=_o__= _ ___._._J_

N I_1_1_

i,,_1_,_ : : _ _ _ _ _ _, _ "_"_ _"_ _ _ _ _"_ _ _ _ _ _ _ _1_ "_ _

v.,l<.>o <.>_<.,<.,<.,o<_o_o_oo<.> o<.><.>o_oo_o_<.><.>__>oI-I<,1<,1.,I,,o,,-=o:._<'=_.:>-=.._o.<,<,<,<,<,<,,,<,<'<'"_°" ,=1__.._<"<'"<,<,,> <,1<,_1;

088

09()

089

. . _o!=_ _ _==_=_= _ =_ m_ =_= ._

i -

oomo_moooooooooooooooooooooooooooo_

._ooooo:_1o_o_o:ooooo:::ooo:oo::_

iiii ........ ......... ......... ......._ __ • _ _ •_

#__'_=='_='_=_°°_'__=_'_ __

_p:_,_.. _. o _ _ _o_ •

-_ _<<_<:_<_X < X'_i _

091

Caulanthus californicus Foster Plot: Lab & Field data for 1992


092

C310_._ . _- r_ a_ _.o e_!_ ' r-- ,-_0 r_ _- _ _n or--

_ _ n. _-. _o. o_ _,. _. • _. • .i,,o,,,,. • _,_ ......

N

N "-N

• " u_ _ _d _ c_ _i _d _,/ a_,c_ o_¢d r_ ¢d ,: _di_ c _ r< "

_©.,_. • • I_ •_._ _-_ _" _1_ _ _,-¢,i_ ,-:i_ ,-.,- "_ "_i ,.-,- _ _ .......

_.._ __._,,., ___ ,,., ,,,, ..,,,,,,,, _'N I

_ _0 _)00L_ 0 _ _0 mO 0:0 iOL ! 0 0 0 _ 0 0 0 !

_,, ._,,., ,,.,,,,.,,.,oo9._. o9 o

093

i I_O 0 o _ _ O o _ _ u Q _00 _ 00 _ • • _QO _ • Q _OQ • • • Q _ •

_ IO O _ O 0 Q Q Q _ Q O, i _ _ 0 _ 0 _ _ 0 _ _ _ 0 Q]_ O _ _ _ _ O _ _ _ O _ Q _ _ _

i _ _i _ _ ° _ _Q __ __ _ _

094

095

i i=_o°o_°_°=o=o_ooo=°=o==o=loooo°oo°_o_i==o_o

N

097

oo oolooii.io•.iooi0 0 0 0 0 0 0 0 • 0 aO CO _ 0_B

i

098

Caulanthus californicus Russ Plot: Lab & Field data for 1992


0_9

iI ° °°° °° I....

I

_1_ r.__,t-_,-_,_ • _'_'" _,_ • _'_ • _-°'_'_'_.- Z_ • _: .i_

lt...............................tl........,......,...............,..........I......I

_!iI.......................................

__l .... i............................ i..........

°....................."...............I1

100

..................................... i...i•

°1......°°°°..........................I°

iiiiili iiQ 0 • _--

O.Q

O_

E

101

i 0 _ .e 0 • Q _ _ • _ • Q _ • _ 0 • • • Q _ 0 _ie • O _ _ a _ • • • • _ • ,. Q • _ _

_" "_'°'°_'°_'"_°_"°" °'° ...... " i'"

ldl..................... '.......'.....I....I""" ........................

_ •

i

'.................................... I""

!02

103

104

ioeoooQejaeoej!oeoee

I

eeeo eeeeeeeo

ooooOeeeeauJe

iiiiiilIiiiiiiiiiolleoello@oQoo eoeoooo

............. i....

105

Caulanthus californicus Foster Plot & Cuyama populations: Ovule counts, 1992


106IO # population ovule # (per ovary) ID _ population ovule # (perova,y)

1 1 Foster 53 64 13 Foster _ 78 .....- " 2' 1 Foster 66 65 13 Foster 76

3. • 1 Foster 71 66 Cuyama 4841 1 Foster 75 67 Cuyama 495 t Foster 77 68= Cuyama 476 2 Foster 49 69 _ 1 Cuyama 517 2 Foster 54 701 I Cuyama 55

8 2 Foster 57 71i 2 Cuyama 489 2 Foster 58 721 2 Cuyama 48

1O, 2 Foster 59 731 2 Cuyama 5511 3 Foster 56 74 2 Cuyama 4912: 3 Foster 61 75 2 Cuyama 48131 3 Foster 62 76 3 Cuyama 6214_ 3 Foster 77 3 Cuyama 78

15_ 3 Foster 78 3 Cuyama 6016i 4 Foster 71 79 3 Cuyama 70171 4 Foster 62 80 3 Cuyama 6718i 4 Foster 51 81 4 Cuyama 53191 4 Foster 57 82 4 Cuyama 55

20 i 4 Foster 83 4 Cuyama 5421 I 5 Foster 57 84 4. Cuyama 65221 5 Foster 65 85 4 Cuyama 6023 5 Foster 56 86 5 Cuyama 6024 5 Foster 69 87 5 Cuyama . 5725 5 Foster 88 5 Cuyema 5326 6 Foster 54 89 5 Cuyama 56

• 27 6 Foster 59 90 5 Cuyama 5728 6 Foster 51 91 6 Cuyama 5729 6 Foster 50 92 6 Cuyama 5630 .6 Foster 52 93 6 Cuyama 51

31 7 Foster 56 94 6 Cuyama 5232 7 Foster 61 95 6 : Cuyama 5833 7 Foster 63 96 7! Cuyama 6234 7 Foster 57 97 7 _ Cuyama 6535 7 Foster 57 98 7I Cuyama 6036 8 Foster 76 99 7 Cuyama 6637 8 Foster 77 100 7 Cuyama 6438 8 Foster 71 101 8 Cuyama 5339 8 Foster 78 102 8 Cuyama 5240 8 Foster 76 103 8 Cuyama 594.1 9 Foster 55 104 8 Cuyama 5742 9 Foster 62 105 8 Cuyama 59

43 9 Foster 57 106 9 Cuyama 4144 9 Foster 66 107 9 Cuyama 5045 9 Foster 60 108 9 Cuyama 3946 10 Foster 51 109 9 Cuyama 4447 10 Foster 62 110 9 Cuyama48 10 Foster 47 1'I1 10 Cuyama 33

49 10 Foster 53 112 10 Cuyama 4950 10 Foster 56 113 10 Cuyama

51 11 Foster 62 114 10 Cuyama 4652 11 Foster 64 115 10 . Cuyama ' 34

53 11 Foster 64 116 11 Cuyama 3154 11 Foster 117 11 Cuyama 3855 11 Foster 118 11 Cuyama 5156 12 Foster 61 119 t1 Cuyama 5357 12 Foster 54 120 11 Cuyama 4.9

58 12 Foster 60 121 12 Cuyama 27•59 12 Foster 122 12 Cuyama 3060 12 Foster 123 12 Cuyama 38

61 13 Foster 81 124 t 2 Cuyama 3462 13 Foster 80, 125 12 Cuyama 5463 13 Fester 82

107

Caulanthus californicus Population 4, Travers Ranch, & Cuyama populations:

Seed number data, 1992


10ID # population # of seeds (per fruit) mean seed # per fruit

1 1 four 26 25.332 1 four .273 1 four 234 2 four 40 43.675 2 four 676 2 four 24 •7 3 four 41 54.338 3 four 47 •9 3 four 75 °

10 4 four 55 51.0011 4 four 4712 5 four 23 24.6713 5 four 46 •14 5 four: 5= •15 6 four 35 ' 31.6716 6 four 43, °17 6 four 17 I16 7 four I 26 ! 25.0019 7 four 1820 7 four 3121 6 four 15 34,0022 8 four 45 *23 8 four 4224 9 foul 53 50.0025 9 four 5026 9 four 4727 10 four 68 56.6726 10 four 53 •29 10 four 4930 11 four 47 43,0031 11 four 5032 11 four 32 •33 12 four 39 33.6734 12 four 2935 12 four 3336 13 four 55 46.6737 13 four 26 •38 13 four 59 •39 14 four 19 30,6740 14 four 20 •41 14 four 5342 15 four 32 26.6743 15 four 2044 15 four 2845 16 four 21 23.0046 16 four 2047 16 four 2846 17 four 41 37,5049 17 four 3450 16 four 42 42.0051 18 four 42 °52 18 four 4253 19 four 69 72,3354 19 four 7355 19 four 7556 20 four 56 59.0057 20 four 6256 20 four 59

ID # population # of seeds (per fruit) mean seed # per fruit I 0

59 21 four 49 "48.6760 21 four 52 •61 21 four 4862 22= four 46 41.6763 22 ! four 4664 221 four 33 •65 23i Travers 52 40.0066 23 j , Travers 45 •67 23 Travers 23 •68 24 Travers 25 34.0089 24 Travers 40 °70 24 Travers 37 •71 25 Travers 73 76.6772 25 Travers 90 •73 25 Travers 6774 26 Travers 22 24.0075 26 Travers 24 °76 26 Travers 26 °77 27 Travers 43 44.0078 27 Travers 5479 27 Travers 35, °80 28 Travers 69 ! 53.0081 28 Travers 52 °82 28 Travers 381 °83i 29 Travers 9 41.3384 29 Travers 84 °85_ 29 Travers 31 •86i 30 Travers 34 41.33871 30 Travers 37881 30 Travers 53891 31 Travers 34 38.6790i 31 Travers 4291I 31 Travers 40 °921 32 Travers 52 56.3393 32 Travers 88 •94 32 Travers 5595 33 Travers 15 11.6796 33 Travers 8 •97 33 Travers 1296 34 Travers 33 34.6799 34 Travers 29

100 34 Travers 42101 35 Travers 67 54.33102 35 Travers 61 •1031 35 Travers 351041 36 Travers 38 51.00105i 36 Travers 64 °106 ; 37 Travers 25 20.67107! 37 Travers 16 °108 i 37 Travers 211091 36 Travers 41 25.001101 38 Travers 25111 : 38 Travers 9 °1121 39 Travers 52 47.00113 " 39 Travers 42 °114 40 Travers 51 27.33115, 40 Travers 24116, 40 Travers 7 °

LD# population # of seeds (per fruit) mean seed # per fruit 1 1 0

117 " 41 Travers 57 33.67 ....118 41 Travars 22119 41 Travers 22120 42 Cuyama 14 21.33121 42 Cuyarna 33 •122 42 Cuyama 17123 43 Cuyama 29 16,67124 43 .Cuyama 12 •125 43 Cuyama 9 °126 44 Cuyama 26 24.00127 44 Cuyama 26128 44 Cuyama 20 °129 45 Cuyama 42 32.00130 46 Cuyama 16 *131 45 Cuyama 38132 46 Cuyama 18 19.67133 46 Cuyama 25134 46 Cuyama 16 •135 47 Cuyama 9 28.67136 47 Cuyama 23 •137 47 Cuyama 54 •138 48 Cuyama 53 53.00139 49 Cuyama 33 28.00140 49 Cuyama 22141 49 Cuyama 29 °142 50 Cuyama 24 25.50143 50 Cuyama 27144 51 Cuyama 15 15.00145 52 Cuyama 22 20.33146 52 Cuyama 24 •147 52 Cuyama 15148 53 Cuyama 75 65.00149 53 Cuyama 84 °150 53 Cuyama 36 °151 54 Cuyama 19 27,67152 54 Cuyama 37 °153 54 Cuyama 27154 55 Cuyama 60 61.00

155 55 Cuyama 62 ,_156 56 Cuyama 48 37.3157 56 Cuyama 45158 56 Cuyama 19159 57 Cuyama 59 43.33160 57 Cuyama i 56161 57 Cuyama i 16

III

Caulanthus californicus Foster Plot: Field data for April 1, 1993


11:quad # treatment pop date whole branches eaten branches # of flwrs # of buds dvlping fruits fruit aborts Ilwr aborts

1 1 Control Foster- 411 3 1 .-:.:1- 29 0 0 .... 0

2 1 Control Foster 4/1 1 0 1 9 0 0 03 1 Control Foster 4/1 1 0 1 9 0 0 0

4 1 Control Foster 411 1 0 0 7 0 0 0

5 1 Control Foster 411 1 0 I 0 0 0 0 0

6 2 Control Foster 4/1 1 0 I 0 2 0 0 0

7 2 Control Foster 4/1 1 0 i 0 1 0 0 0

8 2 Control Foster 411 " 1 0 _ 0 0 0 _0 0

9 2 Control Foster 4/1 0 1 ; 0 0 0 0 0

10 2 Control Foster 4/1 I 0 0 0 0 0 011 2 Control Foster 411 0 1 0 0 0 0 0

12 2 Control Foster 4/1 0 1 0 0 0 0 013 2 Control Foster 4/1 0 3 0 0 0 0 014 2 Control Foster 4/1 0 1 0 0 0 0 0

15 2 Control Foster 4/1 0 1 0 0 0 0 0

16 2 Control Foster 411 0 1 0 0 0 0 017 2 Control I Foster 4/1 0 1 0 0 0 0 0

18 2 Control I Foster 411 0 1 0 0 0 0 0

19 3 Control i Foster 411 3 1 17 29 10 0 0

20 3 Control i Foster 4/1 1 0 3 5 1 0 0

21 3 Control , Foster 411 1 1 5 9 2 0 0

22 3 Control _Foster 4/1 1 1 0 3 0 0 0

23 3 Control Foster 411 0 1 0 0 0 0 0

24 3 Control Foster 4/1 1 0 2 4 0 0 0

25 3 Control Foster 4/1 1 0, 0 3 0 0 026 :4 Control Foster 4/1 1 0 _ 3 5 0 0 0

27 4 Control Foster 411 1 0 _ 3 4 0 0 0

28 4 Control Foster 411 1 0 I I 4 0 0 029 4 Control Foster 4/1 1 0 I 3 3 1 0 0

30 4 Control Foster 4/1 1 0 I 2 4 0 0 031 4 Control Foster 4/1 1 0 2 3 0 0 0

32 4 Control Foster 4/1 1 0 0 3 0 0 0

33 4 Control Foster 4/1 1 0 1 3 0 0 0

34 4 Control Foster 4/1 I 0 0 1 0 0 035 4 Control Foster 411 2 0 13 19 9 0 0

r

36 4 Control Foster 4/1 1 0 1 1 0 0 0

37 4 Control Foster 4/1 1 0 1 2 0 0 0

38 4 Control Foster 4/1 1 0 1 1 0 0 0

39 4 Control Foster 411 4 0 20 38 tl 0 040 4 Control Foster 4/1 1 0 0 1 0 0 0

41 4 Control Foster 411 1 O 3 4 2 0 0

42 4 Control Foster 411 1 0 5 8 9 0 0

43 4 Control Foster 411 1 0 0 1 0 0 0

44 . 4 Control Foster 4/1 1 0 1 4 0 0 045 4 Control Foster 4/1 1 0 3 5 2 0 0

46 4 Control Foster 411 1 0 0 2 0 0 047 4 Control Foster 411 1 0 2 3 1 0 0

48 4 Control Foster 411 1 0 I 1 1 0 0 0

49 4 Control Foster 4/1 1 0 I 2 2 0 0 050 4 Control Foster 4/1 1 0 I 2 3 1 0 051 4 Control Foster 4/1 t 0 I 2 3 0 0 0

52 4 Control Foster 411 1 0 i 2 5 0 0 "0

53 4 Control Foster 411 1 0 : 3 6 1 0 0

54 4 Control Foster 411 5 0 30 34 16 0 0

55 4 Control Foster 411 3 0 17 24 15 0 0

56 4 Control Foster 4/1 3 0 16 19 18 0 057 4 Control , Foster 4/1 t 0 1 2 0 0 0

58 4 Control Foster 4/1 t 0 2 3 3 0 O

59 4 Control Foster 411 1 0 2 4 4 0 060 4 Control Foster 4/1 1 0 8 7 7 0 0

61 4 Control : Foster 4/1 1 0 5 5 5 4 0

62 4 Control Foster 4/1 1 0 4 5 3 0 063 4 Control Foster 4/I 3 0 18 25 8 0 064 4 Control Foster 4tl 1 0 3 4 2 0 0

65 4 Control Foster 411 1 0 3 6 4 0 0

11"quad # treatment pop date whole branches eaten branches # of flwrs # of buds dvlping fruits fruit eoorts ftwr aborts

_66. _ _ 4: _..ContrgJ Fqster 4/1 1 0 I 3 0 0 067 4 Control Foster 411 1 0 4 7 3 0 0

68 4 Control Foster 411 1 0 1 3 I 0 0

69 4 Control Foster 4/1 1 0 2 3 1 0 0

70 5 Control Foster 4/1 3 0 8 22 6 0 0

71 5 Control Foster 411 t 0 2 5 0 0 0

72 1 Weeded :ester 4/1 0 1 0 0 0 0 0

73 1 Weeded Foster , 4/1 0 1 0 0 0 0 074 2 Weeded Foster : 411 0 1 0 0 0 0 0

75 2 Weeded Foster i 411 0 1 0 0 0 0 0

76 2 Weeded Foster I 4/1 0 1 0 0 0 0 077 2 Weeded i Foster I 411 0 1 0 0 0 0 078 2 Weeded ; Foster I 4/1 0 1 0 0 0 0 0

79 3 Weeded Foster I 4/1 2 0 0 4 0 0 0

80 3 Weeded I Foster i 411 1 0 2 9 0 0 081 3 Weeded, Foster i 411 3 0 3 28 0 0 0

82 3 W,_-"t_'_ _Foster i 411 1 1 2 2 2 0 0

83 4 Weeded _Foster I 411 4 4 3 53 0 0 0

84 4 Weeded Foster I 411 0 0 0 0 0 0 0

85 4 Weeded. Foster I 4/1 1 0 1 7 0 0 086 4 Weeded : Foster I 4/1 1 0 3 5 0 0 0

87 4. Weeded ; Foster I 411 2 3 2 23 0 0 0

88 4 Weeded i Foster ! 411 2 1 4 19 0 0 0

89 4 Weeded Foster I 4/1 3 2 2 34 0 0 090 4 Weeded Foster I 4/1 4 4 8 111 0 0 0

91 4 WA_._I_.-(_Foster I 411 t 1 0 4 0 0 0

92 4 Weeded Foster I 411 3 5 7 51 0 0 093 4 Wee, led. Foster I 411 3 2 2 24 0 0 0

94 4 Weeded =Foster i 411 7 7 11 75 1 O 0

114

Cattlanthus californicus Foster Plot: Fielddata for April 18, 1993


quad # treatment pop date whole branches eaten branches # of ilwrs # o! buds dvlping lruits fruit aborts flwr aborts

1 I 1 Control Foster 4/18 1 1 7 13 0 0 I " _-02 _ t Control Foster 4/18 1 2 3 9 0 0 _ 03! I Control Foster , 4/18 1 0 18 25 3 O i 24= 1 Control Foster ; 4/18 1 0 5 10 2 O! 0

51 I Control Foster 4/18 0 0 0 0 0 0 0

61 2 Control Foster 4/18 1 0 1 3 0 1 17 _ 2 Control Foster 4/18 1 0 1 3 0 0 0

81 2 Control Foster 4/18 1 0 4 5 0 0 09, 2 Controt -'ester 4/18 0 0 0 O 0 0 0

10_ 2 Control Foster 4/18 O 0 0 0 0 0 0

11 ; 2 Control Foster 4/18 1 I 2, 1 0 O O12' 2 Control Foster 4/18 1 1 0 I 13 0 0 0

13 i 2 Control _Foster 4118 I 0 1 I 5 0 O 0141 2 Control Foster 4118 I 0 2 ! 8 1 0 015; 2 Control _Foster 4/18 1 0 1 I 12 2 1 1

16i 2 Control :Foster 4/18 1 0 0 I 3 0 0 0

171 3 Control Foster 4118 6 1 10 I 35 31 2 O181 3 Control Foster 4/18 1 0 0 0 1 3 2

191 3 Control Foster 4/18 1 0 0 3 1 0 020 3 Control Foster 4/18 2 0 0 0 5 3 0

21 3 Control Foster 4/18 1 1 1 1 0 0 022 3 Control Foster 4/18 0 I 0 0 0 0 0

23 3 Control Foster 4/18 0 0 0 0 0 0 0

24 3 Control Foster 4/18 1 O 0 0 1" 2 525 3 Control Foster 4/18 1 O 0 3 1 O 0

26 4 Control Foster 4118 1 0 0 0 1 1 227 4 Control Foster 4118 I 0 0 1 0 0 0

28 4 Control Foster 4118 1 0 0 0 1 3 3

29 4 Control Foster 4/18 1 0 0 0 1 3 3

30 . 4 Control Foster 4/18 1 0 0 1 O, 0 0

31 4 Control Foster 4/18 1 0 0 0 2 0 132 4 Control Foster 4/18 1 0 0 0 3 I 0 3

33 4 Control Foster 4118 1 0 0 0 : 1 ' 1 3

34 4 Control Foster 4t18 0 1 0 0 0 0 0

35 4 Control Foster 4/18 1 0 0 0 2 _ 4 6

36 4 Control Foster 4/18 0 1 0 0 0 0 037 4 Control Foster 4/18 1 0 0 0 2' 3 2

38 4 Control Foster 4/18 4 0 2 13, 32 ' 30 14

39 4 Control Foster 4118 3 0 1 5 17' 22 14

40 4 Control Foster 4/18 1 0 0 1 1 0 041 4 Control Foster 4118 1 0 0 0 I 0 2

42 4 Control Foster 4/18 1 0 0 1 2 2 248 4 Control Foster 4/18 1 0 0 0, 2, 3 4

44 4 Control Foster 4/18 t 0 0 5 14 _ 13 5

45 4 Control Foster 4118 1 0 0 0 O, 1 0

46 4 Control Foster 4/18 1 0 0 0 I 1 3

47 4 Control Foster 4/18 1 0 0 1 0 0 0

48 4 Control Foster 4/18 1 0 0 4 1 10 049 4 Control Foster 4/18 1 0 0 0 O 1 0

50 4 Control Foster 4118 1 0 0 O 8 8 9

51 4 Control Foster 4118 1 0 1 1 0 0 0

52 4 Control Foster 4118 1 0 0 0 1 1 2

531 4 Control Foster 4118 1 0 0 3 1 0 0541 4 Control Foster 4118 1 0 0 0 3 11 6

55, 4 Control Foster 4118 1 0 0 0 2 5 5561 4 Control Foster 4118 3 0 0 0 14 20 20

57 I 4 Control Foster 4118 3 0 0 0 17 23 17881 4 Control Foster 4/18 3 0 5 14 23 10 6

591 4 Control Foster 4/18 1 0 0 I 0 2 1 260 = 4 Control Foster 4/18 3 0 71 10 24 8 5

61 = 4 Control Foster 4118 1 0 0 I 0 1 0 5

621 " 4 Control Foster 4118 1 0 0 I 0 1 4 4

631 1 Weeded Foster 4118 1 0 0 i 3 0 O O

641 1 Weeded Foster 4118 0 0 0 I 0 0 0 0

651 2 Weeded Foster 4/18 1 0 0 _ 5 0 0 0

quad # treatment pop date whole branches eaten branches # of flwrs # of buds dvlping fruits fruit aborts flwr aborts ] ]" _

66 2 Weeded, Foster 4t18 -0 O. O 0 "_cO -0 " O67 2 Weeded I Foster 4/t8 0 0 ; 0 0 O 0 0

68 2 Weeded Foster 4/18 0 0 O 0 ; 0 0 O

69 2 Weeded Foster 4118 0 0 I O 0 0 O 070 2 Weeded : Foster 4118 0 O i O 0 0 0 0

71 3 Weeded i Foster 4/18 1 0 ! 4 6 ! 1 1 2

72 3 Weeded I Foster 4/18 1 1 I 7 29 I 0 0 0

73, 3 Weeded Foster 4/18 1 1 I 23 39 I 3 2 2

74, 3 Weeded. Foster 4/18 1 1 4 8 1 0 075 : 4 Weeded Foster 4/18 4 0 46 34 22 2 1761 4 Weeded Foster 4/18 1 I 0 2 0 0 0

77! 4 Weeded Foster 4118 1 I 12 10 2 0 0

78i 4 Weeded Foster 4/18 1 1 5 12 2 1 0

79 I 4 Weeded Foster 4118 2 0 12 20 4 O 0I

801 4 Weeded Foster i 4/18 4 0 13 24 10 2 0

81 i 4 Weeded Foster ' 4/18 1 0 0 3 0 0 082 4 Weeded Foster 4/18 I 1 0 5 0 0 : O

83 4 Weeded Foster 4/18 2 2 17 20 3 0 2

84 4 Weeded Foster 4/18 1 0 0 13 0 0 " 0

85 4 Weeded Foster 4/18 1 1 1 5 0 0 0

86 4 Weeded Foster 4118 2 2 14 26 7 0 087 4 Weeded Foster 4/18 t 0 0 2 O O 0

88 I 4 Weeded Foster 4118 0 I O 0 0 0 O 0

117

Caulanthus californicus Foster Plot: Field data for May 2, 1993

Seetextforsamplingmethods.

1 1quad# treatment pop date wholebranches eaten branches # of flwm # of buds civlpingfruits fruit aborts flwraborts ,IL JI.L

1 1 Control Foster 512.... 2 3 6 _ 12 8 i 4 - _02 1 Control Foster 5/2 17 10 43 119 42 i 16 73 1 Control Foster 5/2 3 0 0 9 14 I 9 44 1 Control Foster 5/2 3 0 0 0 0 I 11 25 2 Control Foster 5/2 2 2 3 5 I 0 06 2 Control Foster 5/2 1 2 0 2 0 0 07 2 Control Foster 5/2 2 0 3 2 1 28 2 Control Foster 5/2 2 1 1 4 11 2 19 2 Control Foster 5/2 0 4 0 0 0 0 0

10 2 Control Foster 512 1 1 0 0 2 1 011 2 Control Foster 5/2 2 0 1012 2 Control Foster 5/2 0 0 0 0 0 0 013 2 Control Foster 5/2 2 1 0 0 0 0 1014 3 Control Foster 5/2 8 0 13 24 41 16 915 3 Control Foster 5/2 2 0 ; 0 0 4 0 016 3 Control Foster 5/2 1 0 I 0 0 2 1 017 3 Control Foster 5/2 1 0 I 0 0 1 0 t

18 3 Control Foster 5/2 1 0 I 0 0 I 0 319 3 Control Foster 512 0 0 I 0 0 0 0 120 4 Control Foster 5/2 1 0 I 0 0 0 0 121 4 Control Foster 5/2 t 0 0 0 1 1 022 4 Control Foster 5/2 1 0 0 0 1 2 , 223 4 Control Foster 5/2 t 0 0 0 2 2 424 4 Control Foster 5/2 1 0 0 0 0 0 125 4 Control Foster 512 1 0 0 0 3 0 326 4 Control Foster 5/2 1 0 0 0 1 1 127 4 Control Foster 5/2 1 0 0 0 1 1 428 4 Control Foster 5/2 1 0 0 0 0 0 129 4 Control Foster 512 1 0 0 0 2 4 530 4 Control Foster 5/2 1 0 0 0 1 0 131 4 Control Foster z 5/2 1 0 0 0 2 2 432 4, Control Foster 5/2 t 0 0 0 1 2 333 4 = Control Foster 5/2 2 0 O 0 13 12 1334 4 Control Foster 5/2 1 0 0 0 1 0 335 4 Control Foster 512 1 _ 0 0 0 1 1 336 4 Control Foster 512 1 0 0 0 8 8 337 4 Control Foster 512 1 0 0 4 1 0 038 4 Control Foster 5/2 5 0 0 0 52 10 2239 4 Control Foster 5/2 1 _ 0 O 0 0 1 240 4 Control Foster 512 1 I 0 0 0 1 3 5

41 4 Control Foster 5/2 1 I 0 0 0 1 1 042 4 Control Foster 5/2 1 0 0 0 2 2 443 4 Control Foster 5/2 1 0 0 0 1 0 344 4 Control Foster 5/2 t 0 0 0 3 10 045 4 Control Foster 5/2 4 0 0 0 27 16 646 4 Control Foster 5/2 3 0 0 0 24 16 1647 4 Control Foster 5/2 0 0 0 0 0 0 • t48 5 Control Foster 512 3 0 0 10 29 7 649 1 Weeded Foster 5;2 t 3 0 3 0 0 050 2 Weeded Foster 5/2 2 3 5 7 0 0 0

51 2 Weeded Foster 5/2 2 O 1 4 0 0 052 3 Weeded Foster 5/2 4 3 10 32 8 2 0

53 3 I Weeded Foster 5/2 9 7 43 79 42 4 254 3 I Weeded Foster 5/2 0 2 0 0 4 0 0

I

55 3 Weeded Foster 512 0 2 0 0 2 0 056 4 Weeded Foster 5/2 10 2 0 0 46 20 3357 4 Weeded Foster 5/2 9 10 62 53 40 5 058 4 I Weeded Foster 5/2 1 4 0 0 4 0 059 4 Weeded Foster 5/2 5 1 0 0 9 7 2160 4 Weeded Foster 5/2 t 5 0 0 0 28 0 061 4 Weeded Foster 5/2 6 9 0 0 26 21 3162 4 Weeded Foster 5/2 0 3 .0 0 0 0 O63 4 Weeded Foster 5/2 1 0 0 0 2 2 064 4 Weeded Foster 5/2 1 0 0 0 2 3 065 4 Weeded Foster 5/2 5 0 0 0 17 6 22

II'.quad # _reatment pop date whole branches eaten branches # of flwrs # of buds dvlping fruits I fruit aborts llwr aborts

66 4 Weeded Foster 5/2 6 0 0 -- 0 2 I -0 067 4 Weeded Foster 5/2 12 0 0 0 4 I 0 0

120

Caulanthus californicus Population 4: Field data for April 1, 1993


121quad # treatment pop date whole branches eaten branches # of flwrs # of buds dvlping fruits fruit aborts flwr abort;

1 1 Control Four 4/1 I 0 3 9 5 4 0 " •2 1 Control Four 4/1 1 0 3 2 4 5 0

3 1 Control Four 4/1 2 0 9 13 8 1 04 1 Contro_ Four 4/1 1 0 7 10 9 1 0

5 2 Control Four 4/1 5 1 19 56 3 0 06 2 Control Four 4/1 1 0 6 11 2 0 0

7 2 Contmt Four 4/1 4 ; t 5 43 0 0 08 3 Control Four 4/1 3 _ 2 4 38 0 0 09 3 Control Four 4/1 I = 0 5 9 2 0 0

10 3 Control Four 4/1 4 I 2 8 25 2 0 011 3 Control Four 4/1 4 i 1 13 37 4 0 012 3 Control Four 4/1 1 I 0 5 10 0 0 O13 3 Control Four 4/1 1 i 0 ¢ 8 0 0 014 3 Control Four 4/1 4 ! 2 2 26 4 0 015 3 Control Four 4/1 5 i 4 6 34 0 0 016 3 Control Four 4/1 2 I 0 1 6 0 0 017 3 Control Four 4/1 1 2 : 0 0 0 0 0

18 3 Control Four 4/1 1 0 0 0 0 0 019 3 Control Four 4/1 1 1 _ 0 0 0 0 020 3 Control Four 4/1 1 0 _ 0 0 0 0 021 3 Control Four 4/1 4 1 ! 4 19 0 0 022 3 Control Four 4/1 1 0 0 0 0 0 023 3 Control Four 4/1 1 1 i 0 0 0 0 024 3 Control Four 4/1 1 1 ! 0 0 0 0 025 3 Control Four 4/1 f 0 i 0 0 0 0 026 4 Control Four 4/1 2 0 i 0 3 0 0 027 4 Control Four 4/1 1 0 i 0 2 0 0 028 4 Control Four 4/1 1 0 I 3 8 0 0 029 4. Control FouP 4/1 1 0 I 3 5 0 0 030 4 Control Four 4/1 1 0 I 1 1 0 0 0

31 4 Control Four 4/1 1 0 I 1 5 0 0 032 4 Control Four 4/1 1 0 ! 1 3 0 0 033 4 Control Four 4/1 1 0 I 1 5 0 0 034 4 Control Four 4/1 1 0 I 0 22 0

35 4 Control Four 4/1 1 0 I 4 10 036 4 Control Four 411 2 1 I 0 3 037 4 Control Four 4/1 3 1 I 0 7 038 4 Control Four 4/1 1 0 i 1 5 039 4 Control Four 411 1 0 ! 1 " 5 040 4 Control Four 4/1 1 0 I 0 0 041 4 Control Four 411 1 0 = 0 1 042 4 Control Four 4/1 2 0 ! 0 0 0 "43 4 Control Four 4/1 2 1 , 0 6 044 4 Control Four 4/1 1 1 0 0 0

45 5 Control Four 4/1 1 0 0 6 0 0 046 5 Control Four =4/1 0 1 0 0 0 0 047 5 Control Four i 4/1 1 1 0 3 0 0 0

48 5 Control Four 4/1 1 0 2 8 0 0 049 3 Weeded Four I 4/1 1 0 0 1 0 0 050 3 Weeded Four I 411 1 0 0 0 0 0 0

51 3 Weeded Four 4/1 3 0 0 8 0 0 052 3 Weeded Four 4/1 1 0 1 3 0 0 0

i

53 3 Weeded Four 4/1 I 0 2 14 0 0 054 3 Weeded Four 4/1 2 2 0 7 0 0 055 3 Weeded Four 4/1 2 1 0 8 0 0 056 3 Weeded Four 4/1 1 0 0 7 0 0 057 3 Weeded Four 4/1 4 2 0 18 0 0 058 3 Weeded Four 4/1 1 1 0 0 0 0 059 3 Weeded Four 4/1 1 0 2 8 0 0 060 3 Weeded Four 4/1 3 2 1 12 0 0 061 3 Weeded Four 4/1 1 0 0 10 0 0 0

62 3 Weeded Four 4/1 1 0 3 8 0 0 0

122quad # treatment pop date whole branches eaten branches # of flwm # of buds civlping fruits fruit aborts Ilwr aborts

-63 3 Weeded Four 4/1 1 1 0 - :11 0 0 064 3 Weeded Four 4/1 2 0 0 8 0 0 065 3 Weeded Four 4/1 7 7 0 50 0 0 066 3 Weeded Four 4/1 7 6 4 56 0 0 067 3 Weeded, Four 4/1 1 0 0 0 0 0 068 3 Weeded Four 411 1 1 0 5 0 0 069 3 Weeded Four 4/1 8 6 0 64 0 0 070 3 Weeded Four 4/1 1 I 0 0 2 _ 0 0

71 3 Weeded Four 4/1 2 1 0 13 0 0 072 3 Weeded Four 4/1 1 1 0 O 0 0 073 3 Weeded Four 4/1 5 2 3 73 0 0 074 4 Weeded Four 411 1 0 0 1 0 0 075 4 Weeded Four 4/1 1 1 0 0 0 0 076 4 Weeded Four 4/1 2 2 0 0 i 0 0 077 4 Weeded Four 4/1 1 0 0 0 0 0 078 4 Weeded Four, 4/1 0 1 0 0 0 0 079 4 Weeded Four 4/1 1 0 0 3 0 0 080 4 Weeded Four 4/1 1 0 0 0 0 0 081 4 Weeded Four 4/1 1 0 0 1 0 0 0

82 4 Weeded Four 4/1 1 0 0 2 0 0 I 0

83 4 Weeded Four 4/1 1 0 0 I 0 0 0 084 4 Weeded Four 4/1 I 0 0 0 0 0 085 4 Weeded Four 411 t 0 0 0 0 0 086 4 Weeded Four 4/1 1 0 0 0 0 0 0

87 4 Weeded Four 4/1 1 0 0 5 0 0 088 4 Weeded Four 4/1 1 0 0 5 0 0 089 4 Weeded Four I 4/1 1 0 0 1 0 0 090 5 Weeded Four i 4/1 1 1 0 0 0 0 091 5 Weeded Four 4/1 2 0 0 0 0 0 0

92 5 Weeded Four 4/1 1 0 0 1 0 0 093 5 Weeded Four 4/1 1 1 0 0 0 0 094 5 Weeded Four 4/1 7 3 0 20 0 0 0

123



12Lquad# treatment pop date whole branches eaten branches # of flwm # of buds dvlping fruits fruit aborts f[wraborts

I

1 Con(fol *Fourl 4/18 1 1 'O 0 .... 8 I O2 1 Control Four 4/18 O O 0 S 21 83 1 Control Four 4/18 2 0 0 0 2 4 124 1 Control Four 4/18 O 0 0 0 0 0 05 2 Control Four 4/18 8 2 52 99 28 1 16 2 Control Four 4/18 12 3 38 78 47 2 07 2 Control Four 4/18 5 0 20 34 14 1 0

8 3 Control Four 4/18 4 3 13 29 I 0 0 1

9 3 Control Four 4/18 1 1 2 5 I 0 0 010 3 Control Four 4/18 1 3 O 8 3 3 011 3 Control Four 4/18 1 4 0 12 0 0 012 3 Control Four 4/18 5 4 28 39 19 0 013 3 Control Four 4/18 3 6, 17 22 11 0 014 3 Control Four 4118 I 0 : 5 11 1 0 0

15 3 Control Four 4/18 1 1 i 7 9 0 0 016 3 Control Four 4118 2 0 0 6 2 0 017 3 Control Four 4118 1 0 0 5 0 0 218 3 Control Four 4118 2 1 2 12 0 0 019 3 Control Four 4/18 0 1 0 0 0 0 020 3 Control Four I 4/18 1 0 1 , 6 5 3 2

21 3 Control Four, 4/18 3 1 21 I 20 30 0 0

22 4 Control Four 4/18 2 5 0 I 11 0 0 023 4 Control Four 4/18 1 0 2 2 1 0 0

24 4 Control Four 4/18 1 1 O 0 0 0 025 4 Control Four 4/18 1 0 0 0 0 0 026 4 Control i Four 4/18 1 0 0 0 0 0 0

27 4 Control Four 4/18 1 0 0 3 1 0 0281 4 Control Four 4/18 1 O 0 2 1 0 4291 4 Control Four 4/18 1 0 0 4 2 0 330 4 Control Four 4/18 2 1 0 2 0 0 031 4 Control Four 4/18 1 0 0 4 8 0 032 4 Control Four 4/18 0 3 0 0 0 0 O33 4 Control Four 4/18 2 1 0 2 0 0 034 4 Control Four 4118 0 1 0 0 0 0 035 4, Control Four 4/18 1"i 0 0 2 0 0 036 4 Control Four 4/18 1 i 1 0 0 0 0 0

37 4 i Control Four 4/18 2 I 2 0 7 0 0 038 4 I Control Four 4/18 1 0 2 4 0 0 0

39 4 ! Control Four- 4/18 0 2 0 0 0 0 .040 4 I Control Four 4/18 0 1 0 0 0 0 041 4 Control Four 4/18 1 1 1 1 0 0 0

42 4 Control Four 4/18 1 3 0 2 0 0 043 4 Control Four 4/18 1 0 0 4 0 _ 0 044 5 Control Four 4/18 1 0 1 0 0 = 0 0

45 5 Control Four 4/18 0 0 0 0 0 I 0 0

46 5 Control Four 4/18 0 0 0 0 0 I 0 O47 1 Weeded Four 4t18 1 2 7 0 0 I 0 048 3 Weeded Four 4/18 1 0 0 40I 0 0

49 3 Weeded Four 4/18 0 2 0 0 0 I 0 050 3 Weeded Four 4/18 I 1 0 5 0 I O 0

51 3 Weeded Four 4/18 I 0 0 2 0 I .0 252 3 Weeded Four 4/18 I 0 0 2 0 0 053 3 Weeded Four 4/18 1 0 1 1 0 0 054 3 Weeded Four 4/18 1 3 0 7 O 0 0

55 3 Weeded Four 4/18 1 1 0 10 0 0 056 3 Weeded Four 4/18 0 3 0 0 0 0 057 3 Weeded Four 4118 1 1 2 9 0 O 0

58 3 Weeded Four 4/18 1 3 0 8 "0 0 059 3 Weeded Four 4/18 0 0 0 0 0 0 O60 3 Weeded Four 4/18 3 5 0 20 0 0 061 3 Weeded Four 4118 1 1 0 0 0 O O

62 3 Weeded Four 4118 1 0 0 2 0 0 0

125_quad# treatment pop date whole branches eaten branches # of flwrs # of buds dvrpingfruits fruit aborts flwraborts

63 3 Weeded Four 4/18 1 4 O 8 0 0 064 3 Weeded Four 4/18 1 3 0 7 0 0 068 3 Weeded Four 4/18 1 2 0 3 O 1 066 3 Weede_ Four 4/18 5 14 0 39 0 0 0

87 3 Weeded Four 4/18 5 15 O 25 0 0 168 3 Weeded Four 4/18 1 1 0 7 0 0 0

69 3 Weeded Four 4/18 1 2 0 2 0 0 070 3 Weeded. Four 4/18 0 5 0 0 0 0 0

71 3 Weeded Four 4/18 0 O 0 0 O 0 072 3 Weeded Four 4/18 5 14 14 18 5 0 07:3 4 Weeded Four 4/18 I O 1 2 O 0 074 4 Weeded Four 4/18 I 0 O 1 I 0 O 0

75 4 Weeded Four 4/18 1 1 0 3 = 0 O O78 4 Weeded Four 4/18 1 0 O 5 0 0 077 4 Weeded Four 4/18 1 1 2 0 : 0 0 078 4 Weeded Four 4/18 1 2 0 2 0 0 0

79 4 Weeded Four 4/18 - 1 2 O 7 0 0 080 4 Weeded Four 4/18 1 O 0 0 0 0 5

81 4 Weeded Four 4/18 1 O I O 0 0 0 282 5 Weeded Four 4/18 1 0 i O 5 0 0 083 5 Weeded Four 4/18 5 5 4 97 0 0 0

8, , weededFour4/18 ol OI OI 8 o 0 o85 5 Weeded ,Four 4/18 0 I 0 I 0 I 0 0 0 0

126

Caulanthus californicus Population 4: Field data for May 2, 1993


].2quad# treatment pop date whole branches eaten branches # of flwrs # of buds dvlping fruits fruit aborts flwraborts

1 1 Control Four 512 2 0 0 0 0 • 0 02 1 Control Four 5/2 1 0 0 0 0 0 03 2 Control Four 5/2 0 3 0 0 0 0 04 2 Control Four 5/2 2 1 0 0 4 9 16

5 2 Control Four 5/2 1 4 0 0 14 12 176 3 Control Four 5/2 0 2 0 0 0 0 07 3 Control Four 5/2 2 4 6 0 6 4 148 3 Control Four 5/2 0 4 0 0 0 0 09 3 Control Four 5/2 1 4 0 7 t 0 O.

10 3 Control Four 512 1 0 0 0 1 0 0

t 1 3 Control Four 5/2 1 1 1 0 0 0 0t 2 3 Control Four 5/2 0 0 0 0 0 0 013 3 Control Four 5/2 0 0 0 0 0 0 014 3 Contmt Four 512 0 0 0 0 0 0 0

15 3 Control Four 5/2 0 0 0 0 0 0 016 3 Controt Four 5/2 0 0 0 0 0 0 017 3 Control Four 5/2 0 0 0 0 0 0 0

18 3 Control Four 5/2 0 0 0 0 0 0 019 4 Control Four 5/2 3 9 0 16 0 0 020 4 Control Four 5/2 2 0 0 0 0 0 021 4 Control Four 5/2 0 2 0 0 0 0 0

22 4 Control Four 5/2 1 2 0 2 0 0 023 4 Control Four 5/2 2 I 4 8 0 0 124 4 Control l Four 5/2 0 4 1 0 0 0 025 4 Contml i Four 5/2 1 8 1 5 0 0 026 1 Weeded _ Four 5/2 1 2 3 4 3 0 027 3 Weeded ! Four 5/2 2 3 0 12 0 0 028 3 Weeded = Four 5/2 0 2 0 0 0 0 029 3 Weeded, Four 512 1 0 0 1 0 0 030 3 Weeded Four 5t2 3 9 0 13 0 0 031 3 Weeded Four 5/2 0 2 0 0 0 0 032 3 Weeded Four 5/2 I 4 0 6 0 0 0

33 3 Weeded Four 5/2 1 3 I 0 4 0 0 034 3 Weeded Four 5/2 0 5 I 0 0 0 0 035 3 Weeded Four 5/2 0 31 0 0 0 0 0

36 3 Weeded Four 5/2 0 0 I 0 0 I 0 0 037 3 Weeded Four 5/2 0 0 I 0 0 I 0 0 0

38 3 Weeded Four 5/2 0 0 I 0 0 I 0 0 039 3 Weeded Four 5/2 0 0 I 0 0 I 0 0 0

40 3 Weeded Four 5/2 0 0 I 0 0 I 0 0 I 041 3 Weeded Four 5/2 0 0 I 0 0 I 0 0 I 042 3 Weeded Four 5/2 0 0 = 0 0 i 0 0 I 043 3 Weeded Four 5/2 0 0 0 0 0 I 0 044 3 Weeded Four 512 0 4 0 0 0 I 0 045 3 Weeded Four 5/2 0 0 0 0 0 ' 0 046 3 Weeded Four 5/2 0 0 0 0 0 0 047 3 Weeded Four 5/2 0 0 0 0 0 0 048 3 Weeded Four 5/2 0 0 0 0 0 0 0


51 3 Weeded Four 5/2 0 3 0 0 0 0 052 3 Weeded Four 5/2 5 13 8 27 32 0 053 4 Weeded Four 5/2 0 2 0 0 0 0 054 4 Weeded Four 5/2 1 3 0 2 0 0 0


57 5 Weeded Four 5/2 1 1 I 3 0 0 058 5 Weeded Four 5/2 13 16 31 98 4 2 059 5 Weeded Four 5/2 0 3 0 0 0 0 0

128



quad# treatment pop date whore branches eaten branches # of flwrs # of buds dvlping fruits fruit aborts tlwr aborts ]" _ 9

1 2 Control Six 4/1 2 2 0 9 0 0 0 . .

2 2 Control Six 4/1 0 2 "0 0 0 0 0

3 2 Weeded Six 4/1 2 2 0 10 0 0 04 2 Weeded Six 411 5 5 0 4 0 0 0

5 2 Weeded Six 411 1 1 0 0 0 0 0

6 2 Weeded Six 411 1 1 0 0 0 0 0

7 3 Control Six 4/1 1 0 1 3 1 0 08 3 Control Six 4/1 1 0 3 6 2 0 0

9 3 Control Six 4/1 1 0 I 3 0 0 0

10 3 Control Six 411 1 0 t 6 2 0 011 3 Control Six 411 1 0 1 1 0 0 0

12 3 Control Six 411 1 0 1 2 0 0 013 3 Control Six 411 1 0 6 6 6 0 0

14 3 Control Six 4/1 4 0 29 44 8 0 0


17 3 Control Six 411 t 0 2 2 0 0 0


20 3 Control Six 4/1 1 0 1 2 0 0 0

21 3 Control Six 4/1 1 0 0 1 0 0 0

22 3 Control Six 4/1 1 0 3 7 0 0 0


25 3 Control Six 4/1 1 0 1 2 0 0 0


28 3 Weeded Six : 4/I 6 1 0 41 0 0 029 3 Weeded Six 4/1 8 1 8 29 1 0 0

30 3 Weeded Six 411 2 0 0 11 0 0 0

31 3 Weeded Six, 4/1 1 0 0 5 0 0 0

32 3 Weeded Six ! 411 1 1 0 2 0 0 0

33 3 Weeded Six i 4/1 6 2 3 19 0 0 0

34 3 Weeded Six I 4/1 3 2 0 15 0 0 0

35 3 Weeded Six I 4/1 4 1 4 11 1 0 036 3 Weeded Six 4/1 1 0 0 6 0 0 0

37 3 weeded Six 4/1 6 0 16 71 6 0 0

38 3 Weeded Six I 4/1 1 0 ' 2 9 0 0 0

39 3 Weeded Six _ 4/1 1 0 1 2 0 0 040 3 Weeded Six I 411 1 0 0 0 0 0 0

41 3 Weeded Six 4/1 1 0 0 5 0 0 0

42 3 Weeded Six _ 411 1 O 0 0 0 0 0 . .43 3 Weeded Six, 4/1 1 0 0 0 0 0 0

44 3 Weeded Six 4/1 1 0 0 0 0 0 045 3 Weeded Six , 4/1 4 0 4. 34 0 O 0

46 3 Weeded Six 411 1 0 0 0 0 0 0

47 3 Weeded Six 4/1 1 0 0 0 0 0 0

48 3 Weeded Six 4/1 4 0 6 30 2 0 0

49 4 Control Six 4/1 2 2 0 0 0 0 0

50 4 Weeded Six 411 4 0 3 41 3 0 0

51 4 Weeded Six 4/1 3 1 1 16 0 0 0

52 4 Weeded Six 411 6 1 7 41 4 0 053 5 Control Six 4/1 2 0 6 16 2 0 0

54 5 Control Six 4/1 1 0 0 5 0 0 0

55 5 Control Six 4/1 1 0 2 7 0 0 056 5 Control Six 4/1 I 0 0 7 3 0 0

57 5 Control Six 411 4 0 7 14 2 0 058 5 Control Six 411 1 0 0 2 0 0 0

59 5 Control Six 4/1 3 3 0 10 0 0 0

60 5 Control Six 411 1 0 4 19 0 0 0

61 5 Control Six 411 1 0 0 8 0 0 062 5 Control Six 4/1 3 0 5 26 1 0 0

63 5 Control Six 4/1 1 0 1 2 0 0 0

64 5 Weeded Six 411 1 0 0 5 0 0 0

65 5 Weeded Six 4/1 2 1 0 11 0 0 • 0

130quad it -treatment pop date whole branches eaten branches # of flwm it of buds dvlping fruit_ i fruit abqrts flwr aborts

661 5 Weeded -" Six 4/I 0 0 0 0 0 -. 0 "_ 067 ! 5 Weeded Six 411 2 1 0 21 0 0 068 5 Weeded Six 411 8 2 4 44 0 0 0

69 5 Weeded Six 411 0 0 0 0 0 0 0

70 5 Weeded Six 411 1 13 0 6 0 0 0

71 J 5 Weeded Six 411 0 1 0 0 0 0 072 5 Weeded Six 411 1 0 0 3 0 0 073 5 Weeded Six 4/1 0 0 0 0 0 0 0

74 5 Weeded Six 4/1 1 0 1 6 0 0 0

75 5 Weeded Six 4/1 1 0 0 3 O 0 0

76 5 Weeded Six 4/1 0 0 0 0 0 0 077 5 Weeded Six 4/1 1 1 0 7 0 0 0

78 5 Weeded Six 4/1 4 3 8 T7 0 0 0

79 5 Weeded Six 4/1 0 4 0 0 0 0 0

80 5 Weeded Six 4/1 1 3 0 6 0 0 0

81 5 Weeded Six 4/1 2 1 0 13 0 0 0

82 6 Control Six 4/1 1 0 5 6 3 0 083 6 Control Six 411 1 0 2 4 0 0 084 6 Control Six 411 1 0 0 3 0 0 0

85 6 Control Six 411 1 0 2 5 2 0 0

86 6 Control Six 4/1 1 0 5 6 4 0 0

87 6 Control Six 4/1 1 0 2 3 . 0 0 0

88 6 Control Six 4/1 1 0 1 1 0 0 089 6 Control Six i 4/1 1 0 ; 0 1 0 0 0

90 6 Control Six 411 I 0 I 3 7 0 0 0

91 6 Control Six 411 I 0 I 0 6 0 0 0

92 6 Control Six 4/1 I 0 I 0 1 0 0 093 6 Control Six 4/1 1 0 I 1 1 0 0 094 6 Control Six 4/1 2 0 6 26 2 0 0

95 6 Control Six 411 4 0 21 23 13 0 0

96 6 Control Six 4/1 3 0 18 54 3 0 0

97 6 .Weeded Six 4/1 1 0 0 3 0 0 098 6 Weeded I Six 4/1 I 1 0 0 0 0 0

99 6 Weeded Six 411 1 0 2 6 2 0 0100 6 Weeded Six 4/1 3 1 2 20 0 0 0

101 6 Weeded Six 411 1 0 4 9 2 0 0

102 6 Weeded Six 4/1 1 I 0 0 1 0 0

103 6 Weeded Six 4/1 2 0 6 9 4 0 0

104 6 Weeded Six 4/1 2 0 7 50 7 0 0105 6 Weeded Six 4/1 1 0 I 1 0 0 0

106 6 Weeded Six 411 1 0 0 5 0 0 0107 6 Weeded Six 411 3 0 12 29 8 1 0

108 6 Weeded Six 4/1 1 0 2 5 1 0 0

109 6 weeded Six 4/1 3 0 11 21 5 1 0

110 6 Weeded Six 4/1 2 1 0 4 0 0 0

111 6 Weeded Six 4/1 1 1 I 0 0 0 0

112 6 Weeded Six 4/1 1 1 .1 0 0 0 0113 6 Weeded Six 411 2 0 2 8 0 0 0


116 6 Weeded Six 4/1 1 0 1 3 0 0 0

117 7 Control Six 411 2 1 0 17 0 0 0118 7 Control Six 4/1 2 4 0 27 0 0 0

119 7 Control Six 411 3 5 0 39 0 0 ,0

120 7 Control Six 4/1 1 I 0 5 • 0 0 0

121 7 Control Six 4/1 1 2 0 12 0 0 0122 7 Weeded Six 4/1 2 1 1 8 0 0 0

123 7 Weeded Six 4/1 2 1 I 6 0 0 0

124 7 Weeded Six 4/1 3 2 0 5 0 0 0125 7 Weeded Six 411 4 3 0, 6 0 0 0

126 7 Weeded Six 4/1 1 1 0 0 0 0 0127 7 Weeded Six 4/1 1 1 0 ' 0 0 0 0

128 8 Control Six 4/I 1 0 0, 3 0 0 0

129 8 Control Six 411 1 1 0 ! 0 0 0 0

130 8 Control Six 4/1 1 1 0 ! 0 0 0 0

131quaa # treatment pop dale whole branches eaten branches # of llwrs i # of buds draping truits fruit aborts flwr abort. _

131 8 Control Six 411 0 1 1 0 0 0

132 8 Control Six 4,/1 4 3 3 8 1 0 0,,

133 8 Control Six 4/1 0 0 1 1 0 0

134 8 Weeded Six 4/1 0 1 0 0 0 0 0

135 8 Weeded Six 411 1 0 3 0 0 0136 8 Weeded Six t 4/I 2 3 2 17 0 0 (]

137 8 Weeded Six I 411 0 0 1 0 0 Q

138 8 Weeded Six I 411 3 3 'Q 18 0 0 Q139 8 Weeded Six I 4/1 2 3 0 18 0 0 0

140 9 Control Six 411 1 0 2 4 j 0 0 0141 9 Control Six 4/1 4 I 4 13 ! 0 0 0

142 9 Control Six 4/1 1 0 2 5 ; 0 0 0

143 9 Control Six 4/1 4 1 8 29 0 I 0 0144 9 Weeded Six 4/1 3 5 0 20 0 I 0 0

145 9 Weeded Six 411 1 1 0 5 0 : 0 O

146 9 Weeded Six '4,1 0 0 0 0 O, 0 0

147 9 Weeded Six 411 1 0 0 5 0 I 0 0148 9 Weeded Six 411 .. 1 0 0 8 0 I 0 0149 9 Weeded Six 411 0 0 0 0 0 i 0 0

150 9 Weeded Six 4/1 0 0 0 0 0 0 0

151 9 Weeded Six 411 0 0 0 0 , 0 0 0152 9 Weeded Six 411 0 0 0 0 0 0 0

153 10 Control Six 4/1 3 1 t 17 0 0 0

154 10 Control Six 4/1 1 0 0 1 0 0 0

155 10 Control Six 4/1 3 3 0 0 0 0 0

156 10 Control Six 4/1 1 1 0 0 0 0 0

157 I0 i Weeded Six 4/1 1 11 0 0 0 0 0158 10 I Weeded Six 411 3 11 0 10 0 0 0

10 I Weeded Six 411 4 1 I 0 17 0 0 0159i,oo ,°Weeded,,,,,, , t, 0 0 0 0 0

161 10 Weeded Six 411 1 1 I 0 0 0 0 0

132



quad # treatment pop date whole branches eaten branches # of flwrs # Of buds dvlping fruits fruit aborts flwr aborts 1, 3 3

1 1 Control Six 4118 4 O 18 24 32 0 2

2 1 Control Six 4118 2 O 3 9 13 7 03 1 Control Six 4118 '13 0 35 42 104 _ 6 17

4 1 Control Six 4118 0 0 0 6 ; 2 8

5 1 Control Six 4118 3 O 19 15 10 2 1

' '6 2 Control Six 4/18 2 2 :3 0 0 O

7 2 Control Six 4/18 0 4 0 0 0 O 08 3 Control Six 4/18 0 0 0 0 O 0

9 3 Control Six 4118 0 O O 1 1 2

10 3 Control Six 4/18 O 0 0 t 3 5

11 3 Control Six ' 4/18 0 1 O 0 O 012 3 Control Six 4118 0 0 0 3 0 6

13 3 Control Six 4/18 0 1 I 8 0 0 2

14 3 Control Six 4/18 4 0 0 i 10 28 21 1615 3 Control Six 4118 0 0 I 0 2 0 418 3 Control Six 4/18 O 0 i 0 7 5 61:7 3 Control Six 4/18 1 0 1 2 0 0 0

18 3 Control Six 4118 1 0 0 1 1 0 019 3 Centre Six 4118 6 1 32 45 50 0 0

20 3 Control Six 4/18 1 O 1 6 ! 4 1 i 1

21 3 Control Six 4/18 1 0 0 1 i 0 0 = 0

2'3 3 Control Six 4118 1 0 0 1 0 0 ' 0

23 3 Control Six 4/18 4 0 23 34 19 5 1

24 3 Control Six 4/18 5 0 12 283 26 23 32'5 4 Control Six 4/18 t 1 4 0 0 O

26 5 Control Six 4118 2 0 3 8 12 8 2

2"t 5 Control Six 4118 1 0 4 4 2 O O28 5 ContrOl Six 4118 t 0 1 2 0 0 0

29 5 Control Six 4118 1 0 O 0 0 0 0

30 I 5 Control Six 4118 3 0 15 16 26 0 103'7 I 5 Control Six 4/18 2 O 7 13 14 1 O

321 5 Control Six 4/18 2 0 7 15 10 4 0

33 5 Control Six 4118 O 1 0 0 O 0 0I

34 5 Control SZx 4/18 1 I 0 _ 1 3 0 0 1

35 5 Control Six 4t18 1 I 0 ; 2 7 3 0 0

36 6 Control Six 4118 17 0 64 155 60 0 O37 6 Control Six 4/18 5 0 10 7 32 14 10

38 6 Control Six 4/18 1 0 O 5 (] 0 0

39 6 Control Six 4/18 1 0 0 0 0 0 040 5 Control Six 4/18 t 0 0 0 0 0 0

41 6 Control Six 4/18 1 0 0 0 0 O 0

42 6 Control Six 4118 1 0 0 0 0 0 0

43 6 Control Six 4/18 1 0 2 5 0 3 044 6 Control Six I 4/18 3 0 8 22 6 1 0

45 6 Control Six _ 4/18 1 1 0 0 1 0 0

48 6 Control Six 4/18 1 0 0 0 5 3 6

47 8 = Centre| S_x 4/18 1 0 0 4 2 2 148 6 Control Six 4/18 1 0 O 5 0 O O

49 6 Control Six ! 4/18 1 0 0 0 4 0 350 7 Control Six 4118 3 0 25 24 12 0 0

51 7 Control Six 4/18 4 1 37 30 13 0 0

52 7 Control Six 4/18 4 0 45 29 8 0 0

53 7 Control Six I 4/18 4 1 7 21 4 0 054 7 _ Control Six 4118 3 0 15 17 6 0 0

55 8 Control Six 4118 1 0 1 4 0 0 0

58 8 Control Six 4/18 2 1 7 7 0 0 0

57 8 I Control Six 4/18 2 3 8 12 15 2 058 8 I Control Six 4/18 1 0 1 2 0, 0 059 8 Control Six 4/18 1 0 O 3 1 0 0

60 9 Control Six 4118 1 I 0 1 0 O 0

61 9 I Control Six 4118 2 1 9 16 281 2 162 9l Control Six 4/18 1 0 0 4 2 ! 1 0

63 9 I Control Six 4/18 2 1 4 13 14 , 5 0

84 tO Control Six 4/18 4 t 12 34 t, 0 5

65 10 Control Six 4118 1 O 1 2 0 I 0 0

134quad # treatment pop date whole bran@es eaten branches # of flwrs # of buds idvlping fruits fruit aborts Ilwr aborts

66 10 Control Six 4118 5 1 _ 18 31 i. 5 0 0


69 1 Weeded Slx 4/18 4 2 29 45 47 5 2

70 1 Weeded Six 4118 2 1 26 35 30 6 5

71 t Weeded Six 4118 5 2 61 98 59 8 3


74 1 Weeded Six 4118 2 0 ' 7 6 5 0 0

75 1 Weeded Six 4/18 1 1 I 4 2 0 0

76 1 Weeded Six 4/18 4 0 32 58 17 2 1

77 2 Weeded Six 4/18 0 0 0 0 0 0 078 2 Weeded Six 4/18 0 0 0 0 0 O 0

79 2 Weeded Six 4118 0 0 0 0 0 0 0

80 2 Weeded Six 4/18 1 4 2 8 0 0 081 3 Weeded Six 4/18 I 3 3 4 0 0 0

82 3 Weeded Six 4/18 1 1 I 4 0 0 0

83 3 Weeded Six 4/18 1 1 0 0 2 0 0


86 3 Weeded Six 4/18 4 0 7 8 17 2 0

87 3 Weeded Six 4118 I 2 0 2 0 0 0

88 3 Weeded Six 4/18 1 2 0 4 0 0 0


91 3 Weeded Six 4/18 0 0 0 O 0 0 0

92 3 Weeded Six 4/18 0 0 0 0 0 0 093 3 Weeded Six 4/18 0 0 0 0 0 0 O,

94 3 Weeded : Six 4/18 1 1 0 0 3 0 0

95 3 Weeded : Six 4/18 1 0 0 2 0 0 0

96 3 Weeded = Six 4/18 1 0 0 3 0 0 0

97 3 Weeded = Six 4/18 2 0 7 9 10 1 098 3 Weecled, Six 4118 1 1 0 2 0 0 0

99' 3 Weeded, Six 4/18 1 1 15 23 6 0 0

100 3 Weeded I Six 4/18 3 0 17 19 12 0 0

101, 3 Weeded I Six 4/18 1 0 5 13 0 0 0

102' 4 Weeded I Six 4/18 1 6 0 0 5 0 0103 = 4 Weeded I Six 4/18 1 8 4 12 6 1 0

104i 4 Weeded i Six 4/18 0 5 0 0 0 0 0

105 5 Weeded I Six 4118 1 3 0 3 0 O 0106 i 5 Wee(led I Six 4/18 0 t 0 0 0 0 0

107 5 Weeded ! Six 4118 I 4 0 1 0 0 0

108 i 8 Wee(led I Six 4118 2 3 39 34 7 0 01(39 _ 5 Weeded I Six 4118 2 5 9 13 4 1 0

110' 5 Weeded I Six 4118 1 1 1 7 0 O 0

111 _ 5 Weeded Six 4/18 1 2 0 1 0 0 0112 I 5 Weeded Six 4/18 O O 0 0 0 0 2

1131 5 Weeded I Six 4118 1 2 4 8 O 0 0

114 5 Weeded _ Six 4118 1 3 0 7 0 0 0


117 5 Weeded Six 4/18 0 3 0 0 0 O 0118 5 Weeded Six 4/18 I 2 0 2 0 0 0

119 5 Weeded Six 4/18 0 4 0 0 0 0 0


122 6 Weeded Six 4118 1 0 2 4 1 0 0

123 6 Weeaed Six 4/18 1 0 2 6 1 0 0124 6 Wee_led Six 4118 1 0 3 7 9 0 0125 6 Weeaed Six 4118 1 0 0 0 1 0 0

126 6 Weeded Six 4/18 1 0 0 0 1 0 0

127 6 Weeded Six 4/18 1 0 7 4 8 0 0

128 6 Weeded Six 4118 7 0 35 37 48 0 0


quad # treatment pop date whole branches eaten branciles # of flwrs # of buds dvlping fruits fruit aborts rlwr aborts ]. 3 5

131 6 Weeded Six 4118 1 O 0 O 7 2 3 --132 6 Weeded Six 4118 1 0 1 5 3 0 0

133 6 Weeded Six 4118 1 0 2 1 0 0 O

134 6 Weeded Six 4/18 2 0 3 4 I 2 0 1

135 6 Weeded Six 4/18 3 0 28 56 i 29 1 1

136 6 Weeded Six 4118 1 O 0 4 I 3 0 j 0137 8 Weeded Six 4118 2 1 9 18 28 2 I 0

138 7 Weeded Six 4118 1 2 9 14 0 0 _ 0

139 7 Weeded Six 4118 1 1 5 9 4 0 ' 0140 7 Weeded Six 4/18 1 O 2 37 4 0 O

141 7 Weeded Six 4/18 1 1 5 5 6 0 O

142 7 Weeded I Six 4118 0 0 0 0 0 0 O

143 8 Weeded I Six 4/18 1 3 25 49 3 0 O144 8 Weeded = Six 4/18 1 2 4 20 O 0 0

145 8 Weeded Six 4/18 0 0 O 0 0 0 0

146 8 Weeded Six 4/18 2 3 6 17 5 0 2147 8 Weeded Six 4/18 1 O 3 5 O 0 0

148 .J] Weeded Six 4/18 1 0 2 8 0 0 O149 9 Weeded Six 4/18 1 0 0 3 0 0 0

150 9 Weeded Six 4/18 1 0 0 2 0 0 0

151 9 Weeded Six 4/18 1 3 0 4 0 0 0

152 9 Weeded Six 4/18 0 0 0 0 0 0 0

153 10 Weeded Six 4118 1 0 13 31 0 O 0154 10 Weeded Six 4118 2 1 17 19 5 O 0

155 10 Weeded Six 4118 1 O 0 1 O O O

15,8 10 Weeded Six 4118 1 0 2 9 0 0 0157 10 Weeded Six 4118 1 0 1 3 0 0 0

136

Caulanthus californicus Population 6: Field data for May 2, 1993


quad # treatment pop date whole branches eaten branches # of flwrs # of buds dvlping fruits fruit aborts flwr aborts 1 _._

2 Control Six 5/2 1 2 0 0 1 0 0

2 2 Control Six 5/2 0 5 0 0 0 0 0

3 3 Control Six 5/2 1 1 3 5 0 0 04 3 Control Six 512 4 0 14 0 36 1 36


7 3 Control Six 5/2 4 0 0 0 18 4 30

8 4 Control Six 5/2 1 1 0 3 1 3 0

9 5 Contm_, Six 512 1 1 0 0 3 0 010 5 Control Six 512 2 0 0 0 14 6 t2

11 5 Control Six 5/2 I 0 O 0, 18 2 7

12 5 Control Six 5/2 1 0 0 0 ! 2 4 I13 5 Control Six 5/2 0 0 0 0 i 0 0 0

14 6 Control Six 5/2 1 0 4 6 I 3 0 0

15 6 Control Six 5/2 3 1 5 0 11 2 5

16 6 Control Six 5/2 4 7 0 0 61 0 44

17 6 Control Six 5/2 1 1 0 0 17 0 3

18 7 Control . Six 5/2 4 3 0 0 22 12 6

19 7 Contrul Six 512 8 2 0 0 47 14 2620 ' 7 Control Six 512 12 2 3 9 84 35 29

21 7 Control Six 5/2 1 _ 1 1 3 3 0 I

22' 7 Control Six 5/2 0 I 0 0 0 0 0 023 ! 8 Control Six 5/2 1 2 0 0 4 4 5241 8 Control Six 5/2 1 0 ! 0 0 0 0 3

251 8 Control Six 5/2 2 3 I 0 0 12 10 2261 10 Control Six 5/2 4 2 j 0 0 14 15 22

271 10 Control Six 5/2 4 1 I 0 0 13 15 23281 2 Weeded Six 5/2 0 5 0 0 0 0 029 3 Weeded Six 5/2 8 7 18 29 48 3 0

30 3 Weeded Six 5/2 0 6 0 0 0 0 0

31 3 Weeded Six 5/2 0 3 0 0 0 0 0

32 3 Weeded Six 5/2 0 3 0 0 0 0 0

33 3 Weeded Six 5/2 0 4 0 0 0 0 034 3 Weeded Six 5/2 0 1 0 0 0 _ 0 0

35 3 Weeded Six 5/2 1 9 1 0 2 ! 2 4

36 3 Weeded Six 5/2 2 3 3 6 1 I 0 037 , 3 Weeded Six 512 1 4 0 0 2 J 0 0

38 3 Weeded SEx i 5/2 13 14 18 56 10 0 0

39 3 Weeded Six , 5/2 4 3 0 0 14 5 0

40 3 Weeded Six I 5/2 1 4 0 0 3 0 041 3 Weeded Six i 5/2 1 4 0 0 0 0 3

42 4 Weeded Six I 512 4 13 11 27 6 0 0

43 4 Weeded Six I 5/2 6 17 26 34 15 0 044 4 Weeded Six i 512 2 10 4 10 0 0 0

45 5 Weeded Six J 5/2 0 5 0 0 0 0 0

46 5 Weeded Six I 5/2 3 13 0 11 0 0 047 5 Weeded Six j 5/2 2 6 0 3 3 0 0

48 5 Weeded Six i 5/2 0 5 0 0 0 0 0

49 5 Weeded Six i 5/2 2 9 5 11 0 ! 0 0

50 5 Weeded Six i 5/2 10 11 0 0 99 30 6951 5 Weeded Six ! 5/2 0 0 0 0 0 I 0 0

52 6 Weeded Six, 5/2 0 I 0 0 1 I 0 0

53 6 Weeded Six 5/2 1 0 0 0 0 iI

0 3

54 6 Weeded Six = 5/2 2 0 O 0 11 0 8

55 6 Weeded Six ; 5/2 9 1 7 0 54 2 3456 6 Weeded Six : 5/2 1 0 0 0 10 2 557 6 Weeded Six = 5/2 0 1 0 0 3 0 0

58 6 Weeded S x ! 512 0 4 0 O 27 2 3

59 6 Weeded Six 5/2 ' 1 1 0 0 2 2 560 6 Weeded S x I 5/2 3 4 8 20 57 2 8

61 7 Weeded Six I 5/2 1 4 0 4 0 0 062 7 Weeded S x I 5/2 1 1 0 0 2 0 0

63 7 Weeded Six 512 t I 0 0 3 0 0

64 7 Weeded Six 512 1 4 0 0 1 0 0

65 8 Weeded Six 5/2 6 5 10 22 45 22 4

138

Caulanthus ca!ifornicus Foster Plot, Population 4, & Population 6:

Field survivorship data for 1993


139

140

141

Caulanthus californicus Within and Between site Greenhouse crosses, 1993

See text for explanation.

142.

1,13

144

145

146

14"/

UNIVERSITY OF CALIFORNIA, SANTA BARBARA

BERKELEY • DAVIS * [RVINE • LOS ANGELES • RIVERSIDE • SAN DIEGO • SAN FRANCISCO _/ SANTA BARRARA SANTA CRUZ

DEPARTMENT OF BIOLOGICAL SCIENCES SANTA BARBARA, CALIFORNIA 93106-9610PHONE: (805) 893-3511

FAX: (805) 8934724

Mazer: (805)-893-8011 £_ _. L

17, 1993Gmail: mazer @li fesci.uesb.edu aeptemuer

Ms. Sandy MoreyNatural Heritage DivisionCalifornia State Department of Fish and Game1.416 Ninth Street

Sacramento, CA 95814

Dear Sandy,

Enclosed you will find the final report describing our field and laboratorywork, results, and recommendations concerning Caulanthus californicus . Thisreport fulfills our commitment to Contract FG 1461.

If you have any questions concerning details of the methods or resultsdescribed in this or in our previous reports on woolly threads and the kern mallow,please don't hesitate to call or FAX me at the phone numbers listed above. I amsending a copy of this report to Rodd Goss and to Ivana No61, a new District Botanistof the BLM. She recently graduated with a Master's degree from UCSB and is nowstationed at the BLM office in Bakersfield.

Let's keep in touch,

Sincerely,

Susan J. Mazer

Associate Professor of Biology

UNIVERSITY OF CALIFORNIA, SANTA BARBARA

BERKELEY * DAVIS • IRVINE * LOS ANGELES * RIVERSIDE * SAN DIEGO * SAN FRANCISCO _I SANTABARBARA SANTACRUZ

DEPARTMENT OF BIOLOGICAL SCIENCES SANTA BARBARA, CALIFORNIA 93106-9610PHONE: (805) 893-3511

FAX: (805) 893-4724

Mazer: (805)-893-801 IILmaik [email protected] September 29, 1993

Ms. Sandy MoreyNatural Heritage DivisionCalifornia State Department of Fish and Game1416 Ninth StreetSacramento, CA 95814

Dear Sandy,

Enclosed is an unbound copy of the Caulanthus californicus report. Just soyou know, I have no funds available for this xeroxing (nor any help to do it), so allthe time and copying charges come out of my pocket!!

I look forward to hearing from you soon,

Sincerely,

Susan J. Mazer

Associate Professor of Biology

Demography, Ecology and Reproductiv, Biology of California ...

Documents