Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
AN ABSTRACT OF THE THESIS OF J. Shelly Steph n Botany Plant and the degree of Master of Science in for Pathology Biosystematic Title: presented Studies of on April 11. 1985. Phacelia capitata (Hydrophyllaceae). a Species Endemic to Serpentine Soils in Southwestern Oregon Redacted for privacy Abstract approved: Renton L. Chambers Phacelia Phacelia capitata magellanica Magellanicae), perennials American this western in secunda P. of North P. (= is member a of wide-ranging, the group (species complex polyploid group a polymorphic America related to the South magellanica). The purpose of was to determine, to the extent possible, the project interrelationships studies using Rruckeberg of P. capitata within the Magellanicae, of its distribution, cytology, morphology, reproductive biology, and population ecology. Phacelia and capitata is endemic to ultramafic (serpentine peridotite) outcrops, substrates, in southwestern Oregon. generally native Field and surveys metamorphic Jackson Counties, revealed 22 extant The species occurs on open serpentine slopes, populations. outcrops, Douglas, Coos, similar or and roadbanks or similarly disturbed sites, in areas with a south to southeast exposure. The plant community with which P. capitata is frequently associated is a Pinus ieffreyi/grass savanna. from five P. capitata sites and two E. corvmbosa Soils were sites the chemically analyzed. soils were materials, and phosphorus and compared were from found very serpentinized parent fully possess to calcium/magnesium low levels low ratios of when to an eighth sample from a non-serpentine soil. sample soil derived At six of these locations metamorphic chemical from one similar rock capitata P. site, A derived from a in appearance to serpentine, had characteristics intermediate between those of serpentine and non-serpentine soils. Chromosome revealed counts from existence the tetraploid one of P. capitata populations diploids (n=11) at 19 of them; of were counts (n=22) populations, 20 these obtained also populations from two containing diploids. Taximetric populations capitata readily mostly capitate verify when heterophylla studies of living plants from 26 Phacelia the morphological distinctiveness of P. compared to P. corvmbosa, P. hastata, and P. ssp. virgata. Phacelia capitata is most distinguished from the other species by its narrow, entire, or silvery-pubescent rosette leaves, its often sub-capitate cyme branches, and its sparse, obscure glandulosity. the Significant seed absence insect visitors. collect of nectar from set in P. capitata does not occur in the When natural pollinators congested flowers, it is likely that self-pollination inflorescence predominant is cross-pollination Magellanicae, between occurs to adjacent of an However, (geitonogamy). some extent in members of the evidenced as flowers hybridization by and introgression at the tetraploid level. The origin problematic. interbred with There may no It the is polyploids possible of capitata P. is P. capitata may have that in the past with P. Jeterophvlla ssp. virciata, or more a of southerly member the complex, P. egena. of also some morphological evidence that P. capitata is have occasionally hybridized with P. hastata. plants have However, been found that suggest any hybrid contact between E. capitata and Z. corymbosa. many At numbers this of locations been has increase an in capitata following habitat disturbance, and P. appears there to have from any characterized its populations. The "released" species the paucity that species evidently possesses poor competitive ability in the more formerly vegetation closed and soils, also appears which to exists on non-serpentine be a poor competitor even in some undisturbed serpentine communities. Owing capitata should very to is be narrow its not favorable endangered maintained geographic response to disturbance, P. with extinction. However, it on a monitoring list, because of its distribution serpentine or sub-serpentine soils. and restriction to Biosystematic Studies of Phacelia capitata (Hydrophyllaceae), a Species Endemic to Serpentine Soils in Southwestern Oregon by J. Stephen Shelly A THESIS submitted to Oregon State University in partial fulfillment of the requirements for the degree of Master of Science Completed April 11, 1985 Commencement June 1985 APPROVED: Redacted for privacy Professor of Botany and Plant Pathology in charge of major Redacted for privacy Chairman of department of Botany and Plant Pathology Redacted for privacy Dean of Gradua School 1 Date thesis is presented Typed by April 11. 1985 J. Stephen Shelly ACKNOWLEDGEMENTS wish I Chambers, thank to his for major professor, Dr. Kenton L. and guidance during all phases of help His encouragement, and his frequent sharing project. this my of knowledge, are greatly appreciated. wish also I committee, Robert Drs. Petersen, thank to other the Frenkel, members Wilson, Mark of my and Roger the careful reading of the thesis which they for did on fairly short notice. specimens were kindly provided by curators of the Loan following of Museum of Natural History, University herbaria: Oregon California (ORE); Herbarium (CAS/DS); Academy University of Sciences/Dudley Washington of (WTU); Humboldt State University (HSC); University of California, Berkeley (UC); and the Jepson Herbarium of the University of California (JEPS). Thomas Moore C. Pathology and gratefully provided the of Oregon the acknowledged. by and research, Support, the travel funds, provided by Dr. Department of Botany and Plant State University Herbarium, are Funds for computer analysis were Committee for Unsponsored Research, Milne Computer Center, Oregon State University. Insect Andrew R. taxa were Moldenke, identified by Drs. John D. Lattin, and William P. Stephen, Department of Entomology, Oregon State University. Assistance with the computer analysis was provided by Drs. C. David McIntire, Robert Frenkel, and Steve Broich. enjoyable day of field work was spent with Mr. Russ An Holmes, Bureau of Land Management (Roseburg), who provided information on locations of Phacelia capitata sites. owe special thanks to LaRea Johnston, for her advice I and expertise included completion assistance also go Bob Special thanks befriended me soon after my who and who and has taught me much about the the state; and Mark Nelson, who provided valuable of field Meinke, work, photographic the of Oregon, in These have especially respects. purchasing my pick-up truck! in to arrival flora many in assistance, made and many interesting Phacelia collections in southwestern Oregon. Without of many thesis students, past and present, this graduate fellow advice, and companionship encouragement, the These very special could never have been completed. many among friends, Fredricks, Dan Joe Antos, Nancy I am sure, include: Luoma, Teresa Magee, Bill Ripple, Glenn Stewart, and Carolyn Wright. My unselfishly wonderful thanks are especially for Ginger King, who sincere provided patience encouragement, (which I fear I and listened may have with tried on occasion!) to all my concerns over the last six months. This Shelly, thesis is dedicated to my parents, John and Lois for all their love, support, and understanding, and for fostering my interest in the natural world. am grateful for experienced. the gifts of all I have To them I seen and TABLE OF CONTENTS INTRODUCTION 1 TAXONOMIC RELATIONSHIPS 3 ENVIRONMENTAL RELATIONSHIPS Geology Soils Climate Site locations and characteristics Associated vegetation Vegetation history 13 13 16 23 BIOSYSTEMATIC STUDIES Cytology Morphology of the species Taximetric studies Scanning electron microscopy Reproductive biology Flowering and pollination biology Seed production Pollen viability Competitive ability and ruderal response 45 45 48 48 24 32 41 71 78 78 86 90 93 DISCUSSION AND CONCLUSIONS Speciation and evolutionary relationships The endemic status of the species Recommendations for preservation 99 99 110 111 BIBLIOGRAPHY 113 APPENDIX 118 LIST OF FIGURES Figure Page 1. The phacelia maoellanica polyploid complex (species group Magellanicae). 2. Distribution of Phacelia capitata in southwestern Oregon. 15 3. Cluster analysis of Phacelia species. 69 4. Seed of Phacelia capitata. 73 5. Seed of Phacelia corymbosa. 75 6. Pollen and leaf vestiture of Phacelia capitata. 79 7. Stem vestiture of Phacelia capitata and P. corvmbosa. 81 8. Distribution of Phacelia capitata and E. corymbosa in southwestern Oregon. 104 9. Possible interrelationships of phacelia capitata within the species group Magellanicae. 109 8 LIST OF TABLES Table Page 1. Selected soil analyses for eight Phacelia sites in southwestern Oregon. 22 2. Climatic data for selected weather stations in southwestern Oregon. 25 3. Reported locations of Phacelia capitata in Oregon. 27 4. Associated tree and shrub vegetation at Beatty Creek. 37 5. Herbaceous vegetation at Beatty Creek. 38 6. Chromosome numbers for Phacelia species group Magellanicae in southwestern Oregon. 47 7. Study populations of Phacelia corymbosa, P. hastata, and P. heterophylla in southwestern Oregon. 50 8. Means and standard deviations for ten morphological characteristics measured on living plants from 26 Phacelia populations. 52 9. Climax leaf lobing in Phacelia capitata. 56 10. Floral branching below the terminal cymes in Phacelia capitata. 62 11. Comparison of Phacelia capitata, P. corymbosa, P. hastata, and P. heterophylla ssp. virgata in southwestern Oregon. 72 12. Insect pollinators of Phacelia capitata and P. corymbosa. 87 13. Seed production in open-pollinated and enclosed inflorescences of Phacelia capitata. 89 14. Pollen viability in Phacelia species group Magellanicae. 91 15. Ruderal response of Phacelia capitata at Lee Creek. 97 BIOSYSTEMATIC STUDIES OF PHACELIA CAPITATA (HYDROPHYLLACEAE), A SPECIES ENDEMIC TO SERPENTINE SOILS IN SOUTHWESTERN OREGON INTRODUCTION Phacelia belonging group plants perennials, ovary of species in group affinities having morphologically are subgenus of divided a perennial species with species P. secunda (=P. magellanica). group the is wide-ranging a America North American Kruckeberg the P. magellanica polyploid complex (species to Magellanicae), western of capitata which and with leaves with four The members defined as being those which Phacelia the South are biennials or possess entire leaves, or pinnately divisions or leaflets, and an entire ovules, two to each placenta (Heckard, 1960). The members polyploid species pillar occurring including of the species group Maaellanicae form a complex which has, as its foundation, six at the P. capitata. however, Heckard base chromosome number of x=11, In his detailed study of the group, made (1960) no attempt to determine the interrelationships of P. capitata within the complex. Phacelia over to a its capitata small rarity, endangered is restricted to serpentine soils geographic area in southwestern Oregon. Due it has received attention as a potentially species and has been placed on various state and federal lists of such species. 2 This study was undertaken to assess various aspects of the distribution, cytology, ecology, reproductive biology, and morphological characteristics extent the were possible, natural of P. capitata. populations To the of other species in complex (P. corvmbosa, P. hastata, and P. heterophylla) considered interrelationships In addition, be used in of an attempt to determine the P. capitata within the Maciellanicae. it is hoped that the information obtained can to evaluate the endangerment status of P. capitata and to make recommendations for managing the species. 3 TAXONOMIC RELATIONSHIPS The Hydrophyllaceae widespread family. regarding Munz genera and stated that and 300 but genera and species it contains. of that the family contains about 25 More recently, Cronquist (1981) species. Hydrophyllaceae comprises about 20 genera the species. 250 small differ in their opinions suggested (1959) relatively a Taxonomists number the forms Constance and Chuang (1982) refer to 18 genera and approximately 326 species. family The western United by recognized to North with Hydrolea Constance which Nama in habitats of the dry have Of the 18 genera (1963), 16 are primarily confined (exceptions are outlying represented is common States (Cronquist, 1981). America, Wigandia, most is Nama, species Phacelj.a, and in South America, as well from the Hawaiian Islands). pantropical, and Codon is restricted to South Africa. The history, Hydrophyllaceae which was had has reviewed by varied a Constance taxonomic (1963) and is summarized as follows. Jussieu are now (1789) Boraginaceae, Phacelia, and Convolvulaceae. family for five of the genera that assigned to the family. (Hydrophyllum, genera accounted two Brown and ( Hydrolea (1818) Of these, he placed three Ellisia) and in the family Nama) in the family separated the former three as the order Hydrophylleae and the latter two as the Hydroleae. The Hydroleae, as monographed by 4 Choisy contained (1833), Eriodictyon Wigandia; Hvdrolea, Nama, Romanzoffia, and was added later (1846). De Candolle (1845) described the genus Codon. Finally, family and above were united into the genera by Gray (1875). He added the genus later Lemmonia, and divided the family into Since that time, the familial relationships tribes. four the Hydrophyllaceae Tricardia, of all the Hydrophyllaceae have been debated, surely due to the fact that, the family Constance as (1963) stated, "(t)0 some extent appears to be a collection of morphological and geographical odds-and-ends." Hydrophyllaceae classification under Baillon (1890) submerged the the Boraginaceae, this but Brand (1913) restored the was not accepted. Hydrophyllaceae to family status. More (1976) recently, and difficulties relationships the family it closest Cronquist with (1981) the the illustrate interpreting in of classifications by Thorne contrasting continuing closest the Hydrophyllaceae. familial The former aligns the Boraginaceae, while the latter places the to Polemoniaceae. Cronquist (1981) summarizes his interpretation of the problem: The been Hydrophyllaceae have usually considered to be allied to the Polemoniaceae on the one hand, and to the Boraginaceae, on the other. I agree. For purposes of formal classification, I find it useful to associate the Hydrophyllaceae with the Polemoniaceae, and to put the Boraginaceae in an allied order. This arrangement helps in the construction of orders that have a reasonable degree of morphological coherence. Despite its problematic relationships, the 5 Hydrophyllaceae The members shrubs, and more often sympetalous, cleft nearly variously in the to many as the flowers or dissected, are solitary, or often helicoid cymes, modified, and generally pentamerous; the calyx or more commonly to the base or middle, as alternate and with the corolla lobes, toward the base or well up in the tube, very often flanked by carpels united a pair of small scales; the gynoecium is of two to terminal, form mostly half-inferior, a compound, unilocular superior ovary; the or rarely style is usually deeply bifid, sometimes cleft essentially the base, or undivided; the stigmas are mostly capitate; placentae is compound, so; the corolla is regular or nearly so; the stamens attached to leaves are or wholly opposite, simple cleft, compound; the vestiture; in partly pinnately to by the following traits. seldom (Eriodictvon, Wiciandia) or rough-hairy palmately perfect, are herbs sometimes or entire rarely is are often alternate characterized is are generally two, parietal or sometimes axile; the fruit a capsule, or sometimes irregularly dehiscent or indehiscent (Cronquist, 1981). Constance the genus Phacelia for the Phacelia. and Eutoca. genera Miltitzia. into (1963) the taxonomic history of Bentham (1837) recognized two genera, De Candolle (1845) was responsible Microgenetes (=Euglypta), Cosmanthus, and Gray (1875) combined all of these but Miltitzia Phacelia. cytological summarized and Further classifications, based on other evidence, were subsequently produced 6 by various suggested, (1963) surveys, with students that Howellanthus recently, (=Microcenetes), classification three subgenera Cosmanthus Phacelia and the of (x=11)." by forth Eutoca, genus: and Miltitzia, put can be achieved Most al. (1984) recognize four sections et treatment of extensive cytological arrangement the of (x=8), Cronquist their in basis "...natural a recognition the (x=9), the on Finally, Constance genus. the of However, the Phacelia. Constance Euglypta will be followed here. Phacelia with perennial, borne helicoid compound base; corollas shortly often the flowers are aggregated into a the calyx is divided nearly to the variously blue or purple to pink, are the corolla tube near its base; the styles are to to very deeply cleft; the capsules are loculicidal; leaves wholly phakelos, these Chuang, of annual and array species; and yellow, and tubular or rotate; the filaments are or attached or cymes, (Constance diverse a herbaceous inflorescence; the white, of taprooted in species 200 consists It the largest genus in the family, far approximately 1982). the by is entire to pinnately dissected, and chiefly are alternate. meaning The genus name comes from the Greek fascicle, and referring to the congested inflorescences (Hitchcock and Cronquist, 1973). Phacelia Magellanicae, comprises the capitata in the is a member of the species group subgenus Phacelia. well-known The Magellanicae P. magellanica polyploid complex 7 (Figure This 1). Heckard 1942, Kruckeberg, among 1944, studies by and (Cave 1959; Heckard, 1956; 1950, 1947, studied intensively was cytological after (1960) Constance, complex 1956) that revealed the existence of polyploidy Superimposed on a series of six diploid its members. interfertile complex of tetraploids (n=11) species (n=22) and study, Heckard (1960) stated that each of the basic diploid an is few a species, except similar counterpart intimately hexaploids P. capitata, on associates it tetraploids." polymorphic tetraploid a At the time of his "...morphologically superstructure with the This intergradation 1949), often indecipherable, has which level tetraploid species, (Anderson, sometimes and has the intergrading introgression (n=33). of between accompanied by led to very complex, patterns of variation. Heckard (1960) summarized the problem: For the most part, as is characteristic of introgressing taxa, entire series of populations, which sometimes extend throughout large areas, are involved in the genic recombinations, resulting in a variation pattern that is extremely difficult to analyze. This is especially true in this complex where the initial differences are not of a magnitude to furnish characters which can be conveniently measured. The result of this intergradation is an extensive assemblage of interrelated plants within which lines must be drawn somewhat arbitrarily in order to delimit taxonomic units. As has a result had a of such variation patterns, the Magellanicae protracted taxonomic history. At one time or another, the complex has been said to include five or more species, and as few as three species. In his final DASYPHYLLA F RIGIDA DASYPHYLLA SIMULANS -frMOPOLA ILEPTOSEPALA HEXAPLOID C OMPACTA HASTATA E GERA CALIFORNICA U AR GENTEA TETRAPLMD PATULA IMBRICATA I I CORYMBOSA HET EROPHYLLA MUTABILIS 0 RE ON EN SIS ---] CA PIT AT A PATULA 1 t I MBR CAT A CORYMBOSA IHETEROPHYLL A 17 IBERNARDINA Figure 1. MUTABILIS NE M ORALIS DIPLOID VIRGATA] The Phacelia magellanica polyploid complex (species group Magellanicae). From Heckard (1960). oo 9 taxonomic treatment, (1960) recognized 13 species Heckard in the United States and Canada. The made Roseburg, near annotated Herbarium, egena." In in his mentioned (1956) 1919, as "Phacelia description of P. the resemblance of the new species, but he assigned no name to suggestive features the of P. capitata, it lacks the inflorescence of capitata P. from It was subsequently Co." Although the silvery, entire leaves of this plant it. are Gray forma specimen to Douglas Kruckeberg capitata, of the at californica this of a P. capitata-like plant was June 22, 1916 by Peck (6022, WILLU), from a "(d)ry on slope collection first species. The possible interrelationships with this, and other problematic specimens Roseburg-Myrtle the Creek area, will be discussed in the final section of this thesis. Other collections made Douglas in 1916 by Peck (6033, "(d)ry and 6035, "Reston, Douglas bank, Roseburg, Co.", both WILLU), were also discussed by Kruckeberg (1956) in relation seem Heckard P. capitata. to best to Co.", However, these two specimens fit the description of P. hastata offered by The Reston vicinity was searched during my (1960). study, but only individuals of P. hastata were located. The April Co., 13, first collection of "pure" P. capitata was made on 1934, Oregon," "2 by specimen went capitata Kruckeberg miles Eastwood unidentified in ed. south of Myrtle Creek, Douglas and until in Howell (1474, CAS). it This was annotated as P. 1955, by both Heckard and 10 Kruckeberg. On Eastwood Howell from and "Myrtle originally the labeled its name such. as capitata, relocate date Douglas Heckard "(p)ossibly Oregon," Co., P. leucophvlla. resemblance similar as the above collection, obtained another specimen (1475, CAS) Creek, acknowledged it same to they Once again, Kruckeberg P. capitata but did not annotated in 1960 as P. it a tetraploid form?" specimens which in the field. I was unable to A hybrid origin for such a plant is possible. The next (20324, UC), "stony slope went P. "good" collection was made in 1939 by Peck presumably from the eventual type locality, a near Bridge, unidentified until Coos Co." This specimen also it was annotated by Kruckeberg as capitata in ed. in 1954, and he eventually designated it as a topotype. A DS) specimen in 1942, labeled this collection of recognize apitata by commented genotypes locate found from by Constance and Rollins (2956, Mt. Nebo near Roseburg, was originally leucophvlla. P. reminiscent not collected was Like Eastwood and Howell 1475, acknowledged as morphologically E. capitata by Kruckeberg, but again he did it as such. Heckard, (pers. It was annotated in 1961 as P. and although it was a tetraploid, he comm.) that the plant of E, capitata and P. corymbosa. may combine I was unable to individuals similar to this in the field, the plants on hastata. Mt. Nebo being most readily referable to P. 11 The next collections County Douglas ORE), (5314, ORE). of papitata P. were made in by Detling, first in 1942 from Red Mountain and These then in 1949 from Nickel Mountain (6339, specimens were also originally labeled as E. leucophylla, until annotated being as capitata P. by Kruckeberg in 1961. The 1951 collection holotype (Kruckeberg miles 2703, WTU), from a "serpentine roadcut, 2 Coos 42, illustrated 1951, years five serpentine has outcrops northwestern later (Kruckeberg, 1956). been sporadically collected Counties, well as Since Douglas south-central in Jackson It was described and Oregon." County, capitata P. capitata was made in E. of Bridge, above Coquille River and along State east Highway of on and being as re-collected from the type locality in Coos County. description his In compared to it leucophvlla. synonymy under one made of other in close two However, comparisons and of the species, Kruckeberg (1956) relatives, P. corymbosa and P. P. leucophylla was Heckard (1960). placed in Further P. hastata P. capitata with P. corymbosa, P. hastata, by member of the complex, P. heterophvlla, are later section on the morphology of the species, a and a description of P. capitata is presented. his In California made within "(n)o studies, members of Heckard (1960) the Magellanicae. attempt...to depict the emphasized the Consequently, he interrelationships the complex of Phace ia capitata Kruckeberg..." The 12 goals of my study were to consider certain aspects of the biosystematics, in order to ecology, ascertain, relationships within hoped the that and distribution of P. capitata, to the extent possible, its the species group Magellanicae. It is study will contribute understanding of this complex group. to a better 13 ENVIRONMENTAL RELATIONSHIPS Geology The geology the Klamath Mountains in southwestern of Oregon and northwestern California is extremely complex and poorly understood. found the Klamath Mountains, near the Oregon-California in The border. thought age This these highly metamorphosed rocks is section presents a review of the terminology serpentine Mountains of be at least 425 million years (Alt and Hyndman, to 1978). of Some of the oldest rocks in Oregon are substrates, province, and a outline an brief the of summary Klamath of the current theories as to how the serpentine areas became emplaced. Traditionally, define The types of ultramafic all them. term which in the silicates word "serpentine" has been used to rocks and the soils weathering from encompasses all those rock *ultramafic" the elemental composition largely consists of mineral olivine. igneous rocks iron The and most from magnesium, derived common ultramafics the are the peridotite and dunite, and their metamorphic derivative, serpentinite (Kruckeberg, 1969). The along the bordered south Oregon-California the on by portion the Klamath Mountains, of which the Siskiyou Mountains the of Oregon north by state line are a part, are the Oregon Coast Range, on the California Coast Range, and on the east by a the Coast Cascade Range Range. The northern contact with extends along a line southwest of 14 the Bandon the Roseburg rocks vicinity. The contact with younger volcanic the Cascades runs south from the Roseburg area to of southeast Medford, to area, to the Rogue River, and then northeast to Redding, California. Mountains the mountains near Siskiyou Pass, over The southern border of the Klamath traverses then southwest, northwest, thence almost to the California-Oregon state line (McKee,1972). ultramafic The (and corvmbosa P. Curry, Josephine, (Figure 2). found in Jackson, and Coos, throughout scattered Douglas Counties most extensive, contiguous outcrops are southern Curry and Josephine Counties (i.e., the Peridotite expansive are Oregon) in The Josephine sheet). To the north these of outcrops, the ultramafic areas become more widely distributed, in size, and separated by wider areas smaller non-ultramafic of outcrops occupied by Phacelia capitate There substrates. is an overall southwest-northeast trend in the alignment of the outcrops. The of geology Oregon, the North began Era), and floor rest precursors the 200 continues on it a continental million today. years ago (early Mesozoic Muddy sediments on the ocean which in turn rest on basalt lava flows, green rock that forms when heavier rocks absorb are and the Pacific Ocean floor continent about serpentinite, beneath the result of a prolonged collision between is American which the Klamath Mountains, and indeed all of water. peridotite. margin, These heavier serpentinite As the seafloor slides under the muddy sediments, and 15 DIPLOID * TETRAPLOIDS PRESENT n UNKNOWN 0 TYPE LOCALITY UNVERIFIED SITE CUMATE STATION CALIFORNIA Figure 2. Distribution of Phacelia capitata in southwestern Oregon. Ultramafic formations outlined (Wells and Peck, 1961). 16 occasionally peridotite ranges slab slabs and ultramafic soft along of Serpentinite is and easily deformed that it can be pressed up cold faults. Under pressure, it takes on the "consistency in The 1972). (McKee, bodies ultramafic intrusion of many Oregon occurred during southwestern Jurassic Period, between 190 and 136 million years ago. However, and the margin, continental the mass (Alt and Hyndman, 1978). toothpaste" the against When such a rocks can be squeezed through fault zones in the accumulating so are added to growing coastal serpentinite, pressed containing seafloor as the ancient Klamath Mountains. such is bedrock of may be remobilized with further compression, they some location bodies as' may recently have as been the moved to their present Eocene Epoch, approximately 30-38 million years ago (Baldwin, 1976). Soils The effects of serpentine soils on the physiognomy and composition physical review soil vegetation of and chemical have been properties related to various such substrates. of A of the properties of these soils, and the results of sample characterize analyses, the are edaphic presented in order to environments of P. capitata, and the more southerly serpentine species, P. corvmbosa. Understanding vegetation numerous has the been attempts to soils on influence of serpentine problematic in the past, because of find one single property of these 17 soils which plant life. "(t)his are universally explains unique impact on their the words of Proctor and Woodell (1975), In has led to unnecessary controversy, for serpentines exceedingly variable." defeats variability This efforts to generalize about the serpentine phenomenon. Physically, rocky (i.e. matter and serpentine Woodell, these and soil nutrient levels may The properties should they 1954; found are rarely However, diverse parent of soils of unfavorable overriding for plant importance; soil influence (Walker, 1969; Proctor and Woodell, 1975). It is to note that, while serpentine soils have often equated no 1954). has been that, while physical discriminating the Kruckeberg, soils reduced in considered as part of the complex of serpentine provides interesting been be result can low clay texture, (Walker, in exist conclusion making chemistry exceedingly physical characteristics gravelly topography) materials. growth, mantle, steep conditions factors Undesirable 1975). and moisture from azonal soils)(Walker, 1954; Proctor (i.e., shallow content, vary lithosols), to having a high clay and organic , content (i.e., soils with xeric soil conditions, Whittaker (1960) significant serpentine and differences in soil moisture between non-serpentine soils in measurements from the Siskiyou Mountains. Proctor chemical serpentine and Woodell (1975) evaluated at length the characteristics most often cited as the causes for soil infertility and toxicity. These have 18 included low phosphorous, especially low levels potassium; and nickel, levels the of chromium, calcium of nitrogen, nutrients major high levels of heavy metals, and cobalt; and frequently, coupled high with levels of magnesium. The vary effects with major nutrient levels on plant growth of In some cases, the species under consideration. addition of increase in plant growth, while in others no improvement is seen are variability subscribe largely to serpentine plant demonstrations of the soils and of plant responses to that low major nutrient levels the chemical stringency of serpentine for example One in which nutrient addition increased is detailed by Turitzin (1982), in a study of growth nutrient view the account soils. the best the of These differences in 1975). Proctor and Woodell claim that few workers would now them. In one of to serpentine soil leads to an K Woodell, and (Proctor response and P, N, limitations on a serpentine area in California. assessing the effects of adding nutrients to the soil on growth of two annual grasses (Bromup mollis and Vulpia microstachys), were the effect it limiting was found that nitrogen and phosphorous nutrients, while added calcium had no on the productivity of these species. however, possible variability in the As mentioned, responses of other species in a given area should be kept in mind. A large body of literature has developed regarding the biological effects of high concentrations of heavy metals 19 which frequently situation are highly is due to pH (i.e., metal and weathering, and and modification other plant same Woodell, 1975). plants of to heavy In summarizing the and Woodell conclude that, 1) there Proctor growth, evidence for many serpentine soils and plants that nickel other response the by the toxicity metal the effects of nickel, chromium, and cobalt on on good and lower uptake of organic and variability (both inter- elements, in of characteristics soil other potential), redox (Proctor research content. factors, including diversity of parent intraspecific) metals to several of metal heavy their in the ecological significance of heavy metals in compounds, presence is variable origin rock Again the complex, and soils from different sites very is Variability characterize serpentine soils. is an important factor, and equally good evidence in cases it is of little or no importance; 2) the that considerations nickel, chromium there but generally is less evidence that important in serpentine soils; and 3) there is is frequently about variability apply to chromium as little evidence for cobalt being as important as nickel or chromium in such soils. The soils status has also infertility. often been characteristic low of been A calcium and magnesium in serpentine of major interest in assessing their low Ca/Mg ratio (i.e., less than 1.0) has cited as influencing the plant most important growth; adverse in other cases, Ca levels alone are emphasized, and in yet others, high 20 consideration major event, these any In 1975). of the opinion that both Ca and Mg should have are authors Woodell, and (Proctor Mg account any in soil serpentine of infertility. The results of White (1971) adequately demonstrate the a series of soil-chemical and vegetation analyses detailed in serpentine Ca Ca/Mg low a distributions levels the also the found that within the study overall He concludes that earlier Ca should be modified are As emphasized earlier, however, perhaps best restricted species to of plant distributions in the area studied. importance of high levels of magnesium is acknowledged by Proctor and Woodell (1975). observations to predominant influence of consistently high results interpretations The low and Fe, these even White infertility based on low Ca/Mg serpentine of Mg, Ni, and Cr. of explain principally influenced by concentrations of high incorporate the not could species Cr, in that order. and explanations ratio, It was also found observed. individual of were transects Ni, alone ratio discontinuities vegetation Mg, In some areas, the highest (1954). occurred on serpentine soils. levels that No substantiation was found for the Kruckeberg and (1954) of which had earlier been espoused by Walker theory" Ca areas contact non-serpentine and Oregon. southwestern "low He conducted of the serpentine soil phenomenon. complexity that, 1) many They cite serpentine-endemic plant have a requirement for high amounts of this element 21 optimum for serpentine growth, soils non-serpentine requirement and have been species. might of magnesium in levels the 2) shown be to to many toxic In their words, a "high magnesium be of more general importance in explaining the restriction of some plants to serpentines." The of most reasonable conclusion is that the stringency serpentine chemical with correlated is not with any single or physical property of such soils in general, but suite of factors, mainly chemical in nature. a factors the soils may vary depending These on the composition and age of parent rocks, topography, climate, and the biota of any given area. extent of Chemical their factors may vary, as well, in the interactions with each other (Proctor and Woodell, 1975). Table samples State conducted by the Soil Testing Laboratory at Oregon University. The extraction and analysis methods are outlined Hill 1 presents the results of analyses of eight soil I by site derived but capitata metamorphic are darker and hastata) appear to be P. rocks Soils at the Boomer that are similar blackish-orange in color. to The sample is from a non-serpentine area adjacent to the serpentine samples and Gardner (1978). (E. from serpentine Tiller Berg outcrops are all along soils Elk derived Creek. from The remaining six fully serpentinized parent materials. serpentine The levels of phosphorus soils are all and low Ca/Mg characterized ratios. by low The sub- Table 1. Species and population Selected soil analyses for eight Phacelia sites in southwestern Oregon. Soil type pH P (ppm) K Ca (ppm) meq/100g Mg meq/100g Ca/Mg ratio Cu (ppm) Fe (ppm) % Organic matter P. capitata Beatty Creek serpentine 7.1 4 35 2.1 12.3 0.17 1.1 74 4.7 2.3 1.8 124 5.4 82 17.6 7.7 2 140 3.1 20.0 0.16 2.1 64 4.0 7.5 1 70 1.7 20.4 0.08 2.4 84 1.9 serpentine 6.7 4 31 1.7 10.4 0.16 0.76 86 5.0 Eight Dollar Mtn II serpentine 6.9 1 39 2.5 23.3 0.11 1.7 94 3.3 Grave Creek serpentine 7.2 1 39 2.0 11.5 0.17 1.1 40 2.6 non-serpentine 6.8 34 43 6.3 1.2 5.25 1.5 142 2.4 sub-serpentine 6.2 19 Bridge serpentine 6.9 Elk Creek serpentine Peel Boomer Hill I P. corymbosa P. heterophylla Elk Creek 23 sample serpentine typical values has serpentine reported higher Ca/Mg ratio than the more a but soils, is within the range of it by Proctor and large amounts of (1975) for such Woodell soils. Although serpentine do not The soils indicate highest iron are often found in and Woodell, 1975), the results (Proctor for the samples used in this study. this concentrations iron of are found in the sub-serpentine and non-serpentine samples. The do remaining analyses (pH, K, Cu, and organic matter) not indicate serpentine any samples Serpentine soils between the consistent differences and non-serpentine from the the P. capitata sample. and P. corvmbosa sites are quite similar in chemical composition. Climate The climate Mountains, effects western Oregon, west of the Cascade maritime, being influenced by the moderating is the Pacific Ocean. of Cascades of largely The mountain barrier of the prevents any continental influence in the western portion of the state. Along Annual 2 ° belt the coast, temperatures climate the is from 1270 elevation of the Coast mountains to the ocean. equable. from a minimum of approximately range C, to a maximum of approximately 20°C. ranges quite mm to Range 2540 and Freezing Rainfall in this mm, depending on the the nearness of the weather and snow are 24 rare. Inland from precipitation annual the coast, there is a general decline in amounts, variation. temperatures and show a greater These trends are influenced by higher elevations in certain areas. The the inland Coast and generally which regions Cascade may from Normal uncommon. ranges are 483 mm, is with 33 ° C, precipitation annual of 38°C not extremes the valleys in from 305 mm to 940 mm; in the mountains the amounts greater. valleys temperature in July, normal maximum temperatures to 28 ° C minimum normal the approximately -1° C; range by mild winters and by summers in be quite warm and nights moderately cool. January, In a climate which is have Ranges, characterized days southwestern Oregon, between of the Some average again occurs each winter; in the snowfall annual amount ranges from 178 mm to considerably with more in the mountains (U.S.D.A. Weather Bureau, 1930). Climatic southwestern data Oregon for are selected summarized weather in stations Table in The 2. locations of these stations are indicated in Figure 2. Site locations and characteristics Methods. Phacelia Oregon A capitata State Threatened, compilation was University and of 13 known locations for made from herbarium collections at (OSC), and from the Oregon Rare, Endangered Plant files housed there. Table 2. Station Name & Elev. (m) Gold Beach Climatic data for selected weather stations in southwestern Oregon, 1951-80 (National Oceanic and Atmospheric Administration, 1982). Precipitation Averages Temperature Averages (°C) (mm) Daily Maximum January July Daily Minimum January July Annual January July January July 8.3 15.0 12.1 19.6 4.6 10.4 2099.8 369.3 9.3 4.4 22.0 8.3 32.4 0.4 11.5 820.7 164.1 5.3 5.0 20.1 8.9 28.9 1.1 11.3 802.9 151.1 4.1 5.1 20.2 9.1 29.0 1.1 11.4 847.1 151.4 4.8 15 Grants Pass 282 Riddle 202 Roseburg 142 26 Location information Management Land was also obtained from two Bureau of botanists, Mr. Russ Holmes, Roseburg, and Ms. Joan Seevers, Medford, Oregon. Field April were conducted April-June and 1983, during May-July 1982, During initial field 1984. as many locations as possible, in both serpentine surveys, and studies non-serpentine surveyed areas determine to were Oregon, southwestern of the location and status of P. 22 verified populations of P. capitata populations. Results. capitata sites which were Five Of the were intensively studied (Table 3), 17 relocations of previously observed localities. new locations for the species were found (Boomer Hill I, Cow Creek I and II, Doe Creek, and The Drew). Five not previously located (Table Two Mountain) were entering private the cases, system and Butte (Brushy inaccessible due and Nickel restrictions to on land; the latter location is the site of Corporation visible is Rattlesnake along these mine. In both of these P. capitata is likely to be present, since suitable habitat town of Nickel Hanna sites for P. capitata were studied during the course of this project or 3). reported Creek which of Peel. is from surrounding areas. The site is part of an extensive population located along the Little River near the Individuals of P. capitata can be seen many portions of the Little River Road in this area, the two sites selected for study (Little River and Table 3. Site Reported locations of Phacelia capitata in Oregon. All locations are in Douglas County except Goolaway Gap (Jackson County) and Bridge (Coos County). Population number T R Sec Location Elev.,Aspect Slope Years Ownership (m) (CLUSTER) A. VERIFIED LOCATIONS Beatty Creek A090 30S 6W 31 0.4 km N of Cow Creek Rd Beatty Creek RNA 320 E,SE,W 55-65 1979 Dwan & Balk 1982-4 Shelly BLM & private Bilger Creek A050 29S 5W 3 BLM rd 29-5-11.0, 2.3 km NW of Bilger Creek Rd 490 S 50 1982 Holmes 1983-4 Shelly BLM & private Boomer Hill I A080 30S 6W 5 BLM rd 30-6-4.2, 8.1 km SW of Interstate Hwy 5 on Boomer Hill Rd. 760 WSW 35 1983-4 Shelly BLM 30S 6W 9 BLM rd 30-6-4.1, 0.5 km from BLM rd 29-6-34.0 850 ESE 40 BLM 1978 Crowder 1984 Shelly & Holmes Boomer Hill II Bridge A210 29S 11W 35 N of St Hwy 42, 3.2 km E of Bridge 100 S 50 1951 Kruckeberg 1980 Sundberg 1982-4 Shelly private Callahan Creek A180 31S 2W 9 Umpqua NF rd 3230, 0.5 km W of St Hwy 227 490 SE 75 1979 Fosback 1983-4 Shelly private & USFS Cow Creek I A120 30S 7W 35 Cow Ck Rd, 0.5 km W of Doe Ck Rd 300 WNW 65 1983-4 Shelly BLM Cow Creek II A130 30S 7W 34 NW side of Cow Creek, 1.2 km SW of Doe Ck Rd 300 SE 60 1983-4 Shelly private Cow Creek Turn-off A160 30S 7W 36 Cow Ck Rd, 4.8 km W of Glenbrook Loop Rd 230 S 5 1979 Dwan & Balk 1984 Shelly private Doe Creek A15A 30S 7W 22 Doe Ck Rd, 1.9 km N of Cow Ck Rd 400 0 1984 Shelly private Table 3. Site Population number Continued. T R Sec Location Elev. Aspect Slope Years Ownership (m) (CLUSTER) Doe/Thompson Ridge A150 30S 7W 14 BLM rd 30-7-14.0, 3.2 km NW of Doe Ck Rd 640 ESE,SE 40 1982 Holmes private 1982 Shelly & Holmes The Drew A190 31S 2W 21 Umpqua NF rd 3130E, 3.1 km from NF rd 3130 810 ESE 25 1982-4 Shelly private Elk Creek A170 30S 2W 3IS 2W 34 St Hwy 227, 1.6-4.0 km SE of Tiller 375 W,WSW 60-80 1972 Heckard 1979 Fosback 1983-4 Shelly private, BLM, USFS 32S 3W 31 0-0.8 km W of Goolaway Gap 930 E,SE 20-30 1981 Seevers 1982-4 Shelly ELM & USFS Goolaway Gap 3 32 Lee Creek A030 28S 4W 15 BLM rd 28-4-15.0, 0.5 km NW of Lee Ck Rd 415 S,E 0-70 1977 Fosback 1982-4 Shelly BLM Little River A010 26S 3W 28 Co Rt 17, 5.9 km SE of Glide 270 SE 35 1976 A & F Parker 1982-4 Shelly private Myrtle Creek North A060 29S 5W 20 Co Rt 14, 1.6 km N of St Hwy 99, NW of Myrtle Ck 245 SSE 70 1952 Howell 1983-4 Shelly private Peel A020 26S 3W 35 BLM rd 26-3-34.2, 0.5 km E of Peel 295 S 25 1977 Knouse & Moore 1982-4 Shelly private 32S 3W 27 SW flank of Red Mtn, 0.8 km N of Jackson-Douglas County line 930 E,SE 0-20 1942 Detling USFS 1984 Shelly & Nelson Red Mountain Rice Creek A070 29S 7W 35 BLM rd 29-7-36.0, 1.2 km W of Rice Ck Rd 710 ESE 20-30 1978 Crowder private 1984 Shelly & Holmes Salt Creek A140 30S 7W 36 BLM rd 30-7-36.0, 0.5 km NE of Cow Ck Rd 340 S 15 1982 Holmes private 1984 Shelly & Nelson Weaver Road AO6A 29S 5W 40 Weaver Rd, 2.9 km SW of Interstate Hwy 5 210 SE 55 1966 Ornduff 1984 Shelly private Table 3. Site Population number Continued. T R Sec Location Elev. Aspect Slope Years Ownership % (m) (CLUSTER) B. UNVERIFIED LOCATIONS Brushy Butte Myrtle Creek South 28S 4W 18 29S? 5W ? Ca 9 km S of Dixonville N,NW ? 1979 Miles private, BLM? "2 mi S of Myrtle Creek" 180? ? ? 1934 Eastwood & Howell private? SE? 1949 Detling private 50 1982 Holmes private ? 1983 Norris USFS 610 Nickel. Mountain - 30S 6W 21 Nickel Mountain, ca 6.5 km W of Riddle Rattlesnake Creek - 26S 3W 34 Little River Rd, 8 km from St Hwy 138 320 Red Mt Lookout - 32S 3W 23 Red Mt lookout, and N along ridgetop 1460 365? S,SW ? 30 Peel) are on each end of this series of populations. I was not aware of the Red Mountain Lookout site until after the 1984 field Eastwood miles and season. Howell The (1474, historical CAS) is labeled as being "2 of Myrtle Creek, Douglas County." S. collection by This vicinity was surveyed, but no plants were located. The entire capitata occurs southern Coos, southwestern the for southeast of Phace ia and western Douglas, Jackson Counties, in The northernmost location occurs along the Little River in Douglas Cedar this of range southernmost location is near Goolaway Gap, studied easternmost northwestern and westernmost 1956); south-central in species The sites geographic Oregon (Figure 2). County. The known Springs location and is Mountain, in Jackson County. the most isolated of the 22 is also the type locality (Rruckeberg, site is near Bridge, in Coos County. The location is along Elk Creek, just southeast of Tiller in Douglas County. are The serpentine the northeasternmost Oregon (Figure located level 2). on slopes areas, or outcrops In where P. capitata is found such substrates in southwestern general, the sites studied are with southerly exposures; a few are in on slopes with west, northwest, or east exposures. The steepness of these sites is quite varied, with ranging at slopes which Bridge, P. where from 0 to 80%. The lowest elevation capitata occurs is at the type locality near a portion of the site is at approximately 31 100 The m. Goolaway site, highest elevation site is 930 m at study Gap, although the unverified Red Mountain Lookout km northeast of Goolaway Gap, apparently occurs 7.1 at an elevation of 1460 m. is of great interest to field inventory studies of It P. capitata response and that at 18 have the of been The ecological It is sufficient here to state that, Mountain, partially least activities Only railroads. of this will be discussed sites studied, the serpentine habitats 22 at quarrying construction and the at disturbed, Beatty of Creek, mainly by roads and Bridge, Red and Rice Creek sites are the habitats largely or completely undisturbed; there is at the Bridge site. livestock in ramifications later section. a a strong favorable shows to disturbance at many locations. evolutionary in species the abundance so some light grazing by Phacelia capitata increases greatly in disturbed areas, that many of its populations have been "released" from any paucity that may have characterized them in earlier times. tendency shared is with most of This weedy the other taxa of the species group Magellanicae (Heckard, 1960). the south of its range, P. capitata is replaced on To serpentine soils corymbosa. The in by the closely related species P. latter species, in addition to occurring the Siskiyou Mountains region of Oregon and California, grows in location the in north Coast Ranges of California and in one Butte County, California, in the Sierra 32 The Nevada. approach most Oregon, Here, closely the in km the gap. northwestern in two species Jackson County, these of area southwest of Cedar Springs Mountain. stretch of the West Fork Evans Creek Road, west of the Phacelia corymbosa is found approximately 8.9 km Nowhere (see These two locations southwest, along Grave Creek. separated are ranges the Goolaway Gap site for E. capitata occurs along a 0.8 to geographic were by an these of non-serpentine substrates. area species found in biotic sympatry two Figure 8, page 104). The Grave Creek location for P. corvmbosa represents locations for this species; the other is on Iron Mountain, in southern Coos one and the of northern two northernmost known Counties Curry (Baker, 1956). Associated vegetation The have spectacular influences that ultramafic substrates on great plant the interest taxonomic dwarfed, is often the vegetations some one, effects unique, words and of and are of physiognomic and is common. Species composition endemism and range disjunction are serpentine of ecologists Serpentine vegetation is frequently xerophytism characteristic In These importance. and community both to biosystematists. which inhabits them has been of life floras Whittaker (Kruckeberg, 1969). (1954), most serpentine throughout the world appear to be dominated by or some combination, of three growth forms: 33 coniferous trees, plants." He Siskiyou in states that the stunted, open appearance of serpentine and size exclusion broad-leaved trees. status as menziesii Pseudotsuga of reduction or opening, vegetation results from the reduction numbers such shrubs, and "grass-like sclerophyllous equivalents shrubby to and the the of In soils of less favorable nutrient there serpentines, lowering, trend towards a of the plant community reduction and is toward the lower strata (Whittaker, 1954). in the (1954, In a associations occurring on ultramafic soils plant The Siskiyou Mountains have been studied by Whittaker White (1971), and Atzet and Wheeler (1984). 1960), his gradient analysis studies, Whittaker found that, on gradient xeric, fairly from more mesic, low elevation sites to more higher elevation sites, the vegetation changes from open Chamaecyparis Pseudotsuga menziesii intermediate forest-shrub monticola-P. to very through stands, stands of monticoladistinctive Pseudotsuga-Pinus lambertiana-P. ieffrevi-Calocedrus decurrens, open decurrens. lawsoniana-Pinus stands This of jeffreyi P. latter and occasional g. community is termed a woodland, "temperate savannah," or pine steppe (Whittaker, 1954). White serpentine (1971) substrates characteristics successional are observed being status. in plant five southwestern related Gentle, to associations Oregon, elevation on their and rocky slopes above 1200 m distinguished by the open character of the vegetation. 34 Important species sandbergii, Ceanothus include Allium species, Poa idahoensis, Aspidotis falcifolium, cuneatus. vegetation Festuca and densa, At middle elevations (860-980 m) the retains its including open Ceanothus character, but cuneatus, increase shrubby in abundance. Below 860 according to intermediate, Vulpia or climax. young, i.e., status, Within the youngest successional species are predominant (Epilobium minutum, microstachys, heterophylla, These vegetation transects were grouped successional annual areas, the m, Aira Githopsis species are Agoseris caryophyllea, speculariodes, and Madia exiqua). considered to indicators of open be mineral soils. successionally the In cuneatgs becomes perennial intermediate areas, Ceanothus important, grasses and with along forbs an ($tipa increase in lemmonii and Eriophyllum lanatum) and the fern species Aspidotis densa. The climax described Pinus as an open at elevations grassland below 860 m is savannah with scattered ieffrevi and occasional Calocedrug decurrens. was notable a herbaceous ($tipa and community stratum lemmonii, Festuca Monardella lanulosa). absence is dominated Danthonia spp.) villosa, Notable and the by cuneatus. perennial The grasses californica, Roeleria nitida, forbs Polvgala in Ceanothus of There (Ranunculus californica, climax occidentalis, and situation Achillea was the 35 absence of species indicative Aspidotis densa. There was a low abundance of mineral open of due soil, the to limited amount of such habitat (White, 1971). Atzet and ieffrevi Wheeler (1984) further described the Pinus Series ieffreyi/grass Association approximately with Siskiyou the in 1390 m. approximately Mountains. The P. defined as occurring below is The climax dominant is P. ieffreyi, 55% average of herb and grass cover species. Vegetation collected plant community adjacent center type occurs. to, Table data Quantitative were order to describe the associated species and in generally see sampling. This with which capitata Pbacelia study was conducted within, and the Beatty Creek Research Natural Area (RNA; The site is located approximately in the 3). of the known geographical range of the species, and the plant community is undisturbed. Methods. circular The data macroplots. east-facing slope within RNA. the west-facing Two along were collected macroplots the west in five 1000 m2 were situated on an side of Beatty Creek, The other macroplots were located on the slope above the east macroplot the side of the creek, outside the RNA. Within collected: and of following data were m2 canopy coverage of tree and shrub species, Pinus for number each ieffreyi and Pseudotsuga menziesii, the trees greater than 3 m in height and the number 36 of trees and less The slope than or equal to 3 m in height. The average aspect of each macroplot were also noted. elevation of the area is approximately 320 m. Within microplots were each total), (40 positioned angles macroplot at eight placed were The radii on each of 4 radii. 2 angles to each other, and at 45° 90° to the up-downhill axis of each macroplot. radius, the microplots 0.25 On each were placed on the uphill side, 6 and 12 m from the center of the macroplot. Within plants of each present microplot, were all species of herbaceous Approximate canopy coverage listed. each species was then recorded, according to the canopy coverage classes of Daubenmire (1959): Cover class canopy coverage % Results. trees and Midpoint % 2.5 15.0 37.5 62.5 85.0 97.5 0-5 5-25 25-50 50-75 1 2 3 4 5 6 75-95 95-100 sampling The shrubs are results for the associated given in Table 4. The results for the herbaceous vegetation are given in Table 5. The associated in the scattered, "savanna" plant at tree community with which P. capitata is Beatty Creek is dominated by Pinus jeffreyi Individuals layer. giving appearance. the community Notable of the the latter are well-documented shrubs include Ceanothus cuneatus and Rhus, diversiloba. Thirty-six herbaceous species were encountered in the 37 Table 4. Associated tree and shrub vegetation at Beatty Creek. Data were collected in five 1000 m2 circular plots, and are presented as percent canopy cover; for Pinus jeffreyi and Pseudotsuga menziesii, figures in parentheses indicate the number of stems greater than 3 m, and the number less than or equal to 3 m, in height. Study dates: 25-26 April 1984. * Plot 2 3 Slope position low low low Slope (%) 56 63 60 E SE W Aspect 4 5 high middle 50 49 WSW SW Trees Pinus jeffreyi 30(18;6) 15(18;19) Pseudotsuga menziesii 1.6(1;0) 2.0(1;0) 7.5(10;6) 7.5(18;7) 30(30;35) 0 0 0.15(0;1) Umbellularia californica 0 3.5 0.25 0 0.4 Arbutus menziesii 0 1.8 2.5 0 0 Ceanothus cuneatus 0.1 1.1 0 Holodiscus discolor 1.8 0 0 Shrubs Rhus diversiloba 0.15 0.6 0.05 0.05 0 0 *-Plots located within the Beatty Creek Research Natural Area. 0.05 0 0.15 38 Table 5. Herbaceous vegetation at Beatty Creek. Data are averages of percent cover (using canopy cover class midpoints), and percent frequency, in 40 0.25 m2 plots; study dates: 25-26 April 1984. Species Average % cover % Frequency Grasses and graminoids Bromus carinatus Luzula campestris Melica geyeri Poa scabrella Vulpia microstachys 1.0 6.5 0.1 16.3 1.5 15.0 40.0 5.0 45.0 47.5 0.9 37.5 42.5 5.0 37.5 5.0 10.0 5.0 Annual forbs Collinsia parviflora Cryptantha intermedia Crocidium multicaule Galium aparine Githopsis specularioides Plectritis congesta Thysanocarpus curvipes 1.1 0.1 0.9 0.1 0.6 0.1 Perennial forbs (including ferns) Achillea millefolium Aspidotis densa Brodiaea pulchella Castilleja pruinosa Cerastium arvense Delphinium menziesii Dodecatheon hendersonii Eriophyllum lanatum Eriogonum nudum Erythronium oregonum Hieracium scouleri Iris chrysophylla Lithophragma parviflora Lomatium nudicaule Lupinus albifrons Microseris sp. Monardella villosa Phacelia capitata Ranunculus occidentalis Saxifraga occidentalis Sedum sp. Silene hookeri Thlaspi montanum var. montanum Viola hallii 0.4 12.6 0.2 0.4 0.9 0.1 0.1 1.1 0.5 0.7 6.5 6.4 0.4 1.6 0.1 0.1 2.3 0.4 2.9 0.1 0.5 1.9 3.8 0.1 2.5 82.5 7.5 15.0 10.0 5.0 5.0 32.5 20.0 27.5 35.0 20.0 15.0 27.5 5.0 5.0 43.5 2.5 40.0 5.0 20.0 50.0 77.5 5.0 39 microplots. Aspidotis Prominent densa, species Luzula Iris chrysophylla, other important and include montanum Thlaspi Ranunculus are cabrella, Hieracium scouleri, campestris, species Poa var. montanum; occidentalis, Monardella villosa, and Silene hookeri. Within serpentine at successional the observed by White (1971), the vegetation sites Beatty Creek appears to be between the intermediate and climax two stages. species status, ieffreyt savanna grasses and herbaceous good cuneatus and Aspidotis densa, Ceanothus considered successional a pattern for lower elevation are the intermediate However, well-developed forbs flora. present. are is of indicative here, and perennial components major the Pinus of the Overall, the Beatty Creek community is example of Pinus a ieffrevi/grass savanna association. Many in communities with habitat features and vegetation fairly similar at of the other 21 populations of P. capitata occur Red Mountain highest Table the one studied above. to elevations and 3), different from quantitative partial of lists interest. include and Pinus where the the P. was capitata plant described were These represent the Gap. associated one studies of Goolaway Notable exceptions are conducted at studied (see communities Beatty Creek. are No at these sites, but associated species at each location are At the Red Mountain site, unique associates lambertiang, Garrya buxifolia, Amelanchier 40 pallida, and Silene, Goolaway Gap site, locations campanulata ssp. glandulosa. associated species not noted at other Whipplea modesta, and chrysophylla, Castanopsis Ceanothus viscida, Arctostaphylos include pumilus, At the Vicia californica. The Hill and I II sites appeared unique as occurring at higher elevations than Beatty Creek and well, in Boomer more mesic Boomer Hill site I Lathyrus ursinus, perforatum, thapsus, As perfoliata. appear to nature than capitata The formosa, americana, dubium. be species hastata, at the Rubus Bromus ;igidus, Hypericum polvohyllus, munitum, Verbascum Phacelia include Cvnosurus Polystichum Associated situations. echinatus, Torilis arvensis, sativa, Vicia Lonicera hispidula, Collomia beteroPhylla, and Montia noted previously, the soils at this site of different, a less stringent chemical the more typical serpentine soils of other P. sites. Boomer Hill II site includes Salix sp., Aauilegia Petasites frigidus, Chrysanthemum This site is Eauisetum arvense, Veronica leucanthemum, unusual in the and Trifolium high amount of seepage which occurs from the gentle slopes. There Pinus were jeffreyi Otherwise, the only was two Phacelia not noted: capitata sites where Boomer Hill I and Bridge. species is associated with P. ieffrevi (in varying amounts) throughout its range. 41 Vegetation history Due to topography, gradient from to factors have its Klamath of in mixed and and summarized by America; 2) forest North evergreen communities) occur vegetation The offered is a center of floristic diversity and is characteristics summary are to most closely resemble Arcto-Tertiary endemism. these The most striking 3) the region has an exceedingly rich flora for latitude, narrow vegetation patterns thought and These 1) the area has one of the most highly (1961): (redwood there; climatic very wide range of habitats. a to exist in the region. vegetation forests steep a had a major influence on the vegetation types come formations and the coast inland, the Klamath Mountains have characteristics complex geologic history, dissected diversity, contain have Whittaker complex edaphic evolved that long, a history giving rise to must have been extremely complex. A to aid in postulating about the here, evolutionary history of Phacelia capitata. The been vegetation considered region particular in According sources evolution to the of Detling (1968); the Pacific Northwest has discussed was that of the Klamath by Whittaker (1961). there have been three principal former, plant associations in the Pacific Northwest: 1) of in situ; 2) the northern Arcto-Tertiary Geoflora; and 3) the southwestern Geofloras by history the Madro-Tertiary United Geoflora of western Mexico and States. The movements of these began during the Oligocene epoch, as a drying and 42 cooling trend allowed the replacement of Eocene subtropical forests by more Metasequoia forests temperate much across comprised forests the Pacific Northwest. of Arcto-Tertiary the and Sequoia of These Geoflora, and their spread across During the late Miocene, continued cooling and increasingly dry to conditions affected the Arcto-Tertiary forests, leading a decrease forests A the region continued into the Miocene epoch. the more humid coastal strip (Whittaker, 1961). to general increase dry-adapted (Detling, in deciduous forms and restriction of these herbaceous 1968). vegetation probable centers southwest, was Northwest; these Geoflora. in the there thought the which of origin very these changes spread of these latter dry-adapted many of and Pinaceae the of accompanied types The types, members in moved scattered in from northward areas the of important to the flora of the Pacific comprised the Madro-Tertiary assemblages This flora reached southwestern Oregon by early Pliocene epoch, and reached its maximum development by the that Klamath middle of the epoch (Detling, 1968). during Pliocene epoch the vegetation of the Mountains It is took on essentially its present-day character (Whittaker, 1961). During culminated the late eventually epoch. At retreat occurred, higher Pliocene, least elevations in four the mixed cooling trend ice ages of the Pleistocene periods leaving further a of glacial northern with more relict advance and species xeric, at southern 43 vegetation Pacific types (i.e., the Pine-oak Forest) throughout the Northwest xerothermic period Hypsithermal northward More 1968). recently, a and warmer than the present, the dryer interval, occurred 4-8,000 years ago. The spread of the Madro-Tertiary Geoflora resumed and continued. the (Detling, This trend brought new genera and species over Klamath chaparral Mountains vegetation from northern California, and became established in the Rogue River basin (Detling, 1968). The events of last 4,000 years, under a cooler, the moister climatic trend, are summarized by Detling (1968). With possible temporary unrecorded reversals the Mesic Coniferous Forest has again become dominant over much of the western portion of the region. It has largely replaced the xeric Rogue flora except for numerous refugia where soil conditions have probably hindered its (the xeric flora's) competitors. Serpentine areas of southwestern Oregon are perhaps good examples of such soil-influenced refugia. The end results these of historical events are presented by Whittaker (1961). The central relation of the Klamath Region is regarded primarily not as one of a center of origin for forests of other parts of the West, but as a center toward which mesophytic forests of the past have shrunk, and as a center of accumulation of species of varied evolutionary history in the diverse habitats of ancient land surfaces. In summary, of 53 considered members of the out Pacific it is interesting to note that 28 genera, by Detling Madro-Tertiary Northwest. Phacelia (1968) to be important Geoflora, capitata extend into the is found in 44 association has geographic seems grass as these of 28 genera: Ceanothus, Umbellularia. and Rhps, ieffreyi It nine Arctostaphvlos, Arbutus, pinus, with In Garrya, Amelanchier, Holodiscus, the genus Pinus, P. its area of northernmost distribution the region to which Phacelia capitata is restricted. to postulate that the Pinus ieffreyi/ reasonable woodland, associated, may communities of with which Phacelia capitata is frequently be a good example of one of the diverse Madro-Tertiary derivation which entered the Klamath Region from the south. 45 BIOSYSTEMATIC STUDIES Cytology Materials microsporocytes, collecting These and methods. and young inflorescences fixing acetic chromosome counts from buds were obtained in the field by floral several For inflorescences from the plants. immediately placed in a killing were solution of 3 parts 95% ethanol: 1 part glacial which acid, was mixed in the field, for 24 hours. The material was then 70% ethanol, and stored in a freezer in a third change of washed in two overnight changes of the same. buds the In progression therefore carmine stain or from 45% acetic and the the alcoholic hydrochloric acidby buds hours Snow (1963), to ensure that Inflorescences were warmed in in a 55-60°C oven, were then in 70% ethanol in a freezer. Anthers were the buds and placed on a slide in a drop of acid. cells inflorescences were in distilled water and studied immediately, stored dissected several 24-48 briefly were using described for entire The buds were obtained. stain rinsed May). stained, suitable the helicoid cymes there is a distal meiosis throughout the early growing season of April-late (late the of A cover slip was placed on the material were squashed by applying pressure. from one inflorescence had Often, to be prepared before an appropriate meiotic stage was observed. Results. Twenty-eight chromosome counts were made in 46 the course for P. of corvmbosa, hastata. 20 for Phacelia capitata, five this study: These two for E. heterophvlla, and one for P. previous plus results, counts for E. capitata and P. corvmbosa, are shown in Table 6. The base chromosome Magellanicae is counted, capitata P. particular Callahan interest may reveal all six At 19 of 20 sites two populations, Elk is mixture a counts from of now are Of Creek and diploids The and the Callahan Creek site the presence of diploids there, as well. in group occurs at this chromosome level. are Further species diploids species the 1960). (Heckard, population tetraploids. in which contain tetraploid individuals. Creek, Creek Elk x=11 number Thus, Magellanicae that occur as basic the known to have morphological counterparts at the tetraploid level. Two in previous tetraploid counts (n=22) for E. corvmbosa Oregon additional E. Coast Sierra These, plus five counts obtained during this study, indicate that corvmbosa Diploids cited by Heckard (1960). were occurs of this Ranges of Nevada primarily species this level in Oregon. are known to occur in the North California, (Heckard, at and 1960). in one locality in the counts may yet Further reveal the existence of diploids of E. corymbosa in Oregon. Three Oregon, 1960). that all The this counts for P. n=11, had previously been recorded (Heckard, two heterophvlla in southwestern additional counts reported here indicate species (as ssp. viraata) is primarily 47 Table 6. Species Chromosome numbers for Phacelia species group Magellanicae in southwestern Oregon. Chromosome Collection Population number (n) P. capitata P. corymbosa P. hastata Shelly 419 Shelly & Nelson Shelly & Holmes Shelly & Holmes Kruckeberg 2703 Shelly 440 Shelly 405 Shelly 411 Shelly 681 Shelly 706 Shelly & Nelson Shelly & Nelson Heckard 2930; Shelly 433 Shelly & Nelson Shelly 399 Shelly 690* Shelly 414 Shelly 354* Shelly & Nelson Shelly & Nelson Shelly 689 Beatty Creek Bilger Creek Boomer Hill I Boomer Hill II Bridge Callahan Creek Cow Creek I Cow Creek II Cow Creek Turn-off Doe Creek The Drew Elk Creek Elk Creek 11 11 11 11 11 Goolaway Gap Lee Creek Little River Myrtle Creek North Peel Red Mountain Salt Creek Weaver Road 11 11 Eight Eight Grave Rough Waldo Wimer 22 22 22 22 22 22 Shelly 700 Shelly 702 Shelly 441 Shelly 698 Shelly 697 Constance & Bacigalupi 3394 22 Shelly & Holmes 767 11 11 Shelly 714 Shelly & Nelson 728 Dollar Mtn I Dollar Mtn II Creek & Ready Road Boomer Hill I P. heterophylla Cow Creek Elk Creek 22 11 11 11 11 11 11 22 11 11 11 11 11 11 724 766 781 734* 726 738 736 725* *-Representative voucher from the same population, the actual count being obtained from a separate, unvouchered individual. 48 diploid that in tetraploid although area, elsewhere level it its in does occur at the range, mainly as ssp. heterophylla (Heckard, 1960). The the reported count first here for E. hastata is apparently from southwestern Oregon. one Only tetraploids of this widespread species have been found (Heckard, 1960). Morphology of the species Taximetric The morphological characteristics studies. of Phacelia capitata were studied, to describe the degree of diploid variation within the species and the variation pattern populations the in comparative purposes, containing tetraploids. For studies of P. corvmbosa, P. hastata, and P. heterophvlla ssp. virgata were also made. Methods. studies, due material of corolla to capitata P. obscured capitata no were of and used in the numerical representative herbarium because the features of the pressed specimens. It was hoped could be conducted to augment the However, survival of transplants (8/29) of and E. overwatering. capitata) in studies studies. field plants paucity a greenhouse that P. are Living Of the flowered. pre-treatment, corymbosa was survivors, poor, possibly due to only one plant (P. Seed germination of P. capitata, with was good (80%), but none of these plants flowered. A was total studied. of 530 For individuals from natural populations Phacelia capitata, 420 plants from 20 49 populations were were examined. omitted, present Creek plants sites. plants) and measurements population, With number individuals of the exception of the Elk Bridge (27 plants) populations, 20 location each at small the to these at (33 due Boomer Hill II and Red Mountain measured. were were not taken because the plants Goolaway the for Corolla Gap were not flowering at the times the site was visited. comparison For studied of with P. capitata, 110 individuals were representing other species of the Maciellanicae: 80 P. corvmbosa (4 populations), 20 of P. beterophylla ssp. virgata population), (1 population). Site and of 10 hastata P. (1 locations and characteristics for these populations are listed in Table 7. From each morphological leaf (mm), corolla number individual data studied, length (mm), (mm), width of corolla 10 blade length characteristics were collected: width for opening (mm), calyx length (mm), stem height (cm), of leaf lobes, type of leaf lobing, number of floral branches below the terminal cymes, and glandulosity. The five leaf and flower characteristics were measured with a Stem height was measured with a ruler, from the base of the Zeiss 10x handlens which contains a 10 mm scale. flowering stem to the tip of the terminal cymes. Qualitative the degree manner: of assessments of the type of leaf lobing and glandulosity were made in the following Table 7. Study populations of Phacelia corymbosa, P. hastata, and P. heterophylla in southwestern Oregon. Site Population number T R Sec Location County Elev. Aspect Slope (m) (CLUSTER) P. corymbosa 0030 38S 8W 19 Siskiyou NF rd 4201 at Illinois River Josephine 390 SW 0-15 Eight Dollar Mtn II 0040 38S 8W 8 Siskiyou NF rd 4103, 5 km W of Selma Josephine 450 S 10 Grave Creek 33S 4W 21 Grave Ck Rd, 3.5 km N of Slate Ck Rd Jackson 850 SE 5 Rough E. Ready 40S 8W 18 Rough & Ready Botanical Wayside, 8 km S of Cave Junction Josephine 425 Co Rd 5560, 4.2 km E of O'Brien Josephine 475 Eight Dollar Mtn I 0 Waldo 0020 40S 8W 28 Wimer Road 0010 41S 9W 4 P. hastata H010 Siskiyou NF rd 4402 Josephine 490 SE (Wimer Rd), 6.5 km SW of O'Brien Same as P. capitate Boomer Hill I population (see Table 3) E010 30S 7W 35 N side of Cow Ck, 0.1 km W of Doe Ck Rd Douglas 300 S 0-30 30S 2W 34 St Hwy 227, 0.8 km E of Tiller Douglas 350 S 10 SW 0-5 0-20 P. -heterophylla Cow Creek Elk Creek 51 Tyve of leaf lobing 0-leaves entire. 1-leaves pinnatifid (lobes with a broad base at the midvein). 2-leaves pinnately compound (lobes reaching the midvein, with entire leaflets distinguishable). Glandulosity (when viewed dissecting microscope) with Zeiss 40x 0-eglandular. 1-a few scattered glandular trichomes among the longer tapering trichomes, mainly in and near the inflorescence. 2-densely and conspicuously glandular, especially on intermediate-length trichomes in and near the inflorescence, on the pedicels, and on calyces. Results. ten The means and standard deviations for the morphological characteristics were calculated for each population, and these are given in Table 8. basic The Magellanicae degrees which consists of axis bears flowering stem. consists of branches bearing 1960) plants in the a tap root, modified by varying of rosette a and gives rise to an stem pattern fundamentally The one to several helicoid cymes (Heckard, . characters Heckard The size and outline of leaves are fundamental importance in the Magellanicae. of (1960) emphasized the importance of the "climax type" (the mature rosette leaves produced just before formation due the an axis, usually foliate (or bracteate), with Leaves. leaf of branching, and a simple or branched caudex, of each annual organization to of the the flowering stem) in making comparisons, large amount of variability possible within a Table 8. Means and standard deviations for ten morphological characteristics measured on living plants from 26 Phacelia populations. Left figure in each population is the mean, right figure is the standard deviation. lki20 except as noted. Phacelia capitate Characteristic Beatty Creek Bilger Creek Bridge Boomer Hill I (n=27) Callahan Creek Cow Creek I Cow Creek II Cow Creek Turn-off 36.9 7.0 46.2 6.2 45.2 9.2 37.7 7.0 57.9 11.4 37.7 9.7 56.3 14.1 36.5 9.0 3.4 0.6 4.1 0.8 4.1 0.8 4.7 0.9 5.4 1.1 3.5 0.6 4.5 1.0 3.4 0.8 COrolla opening (mm) 5.3 0.9 4.2 0.5 3.4 0.6 4.9 0.7 5.8 0.8 3.9 0.7 4.4 0.9 3.8 0.4 Corolla length (mm) 5.8 0.5 5.7 0.4 5.8 0.4 5.3 0.4 6.6 0.7 5.6 0.5 5.6 0.5 5.6 0.4 Calyx length (mm) 5.0 0.6 5.3 0.6 5.4 0.4 4,0 0.5 5.3 0.6 5.2 0.7 5.2 0.9 5.1 0.5 29.6 6.5 35.1 11.8 25.6 5.4 28.6 6.2 42.4 10.5 29.9 5.8 30.7 8.0 34.7 7.1 No. of leaf lobes 0 0 0 0 0 0 0.04 0.2 0 0 0 0 0.1 0.3 0 Type of lobingl 0 0 0 0 0 0 0.04 0.19 0 0 0 0 0.1 0.3 0 0 0.4 0.8 2.0 1.6 2.2 2.1 0.9 1.0 2.7 2.0 1.4 1.4 1.6 1.1 1.1 1.1 1.0 0 1.0 0 1.0 0 1.0 0 1.0 0 1.0 0 1.1 0.2 1.1 0.2 Blade length (mm) Leaf width (mm) Stem height (cm) No. of floral branches Clandulosity2 1-Type of lobing: 0-leaves entire 1-leaves pinnatifid 2-leaves pinnately compound 2-Glandulosity: 0-eglandular 1-a few scattered glandular trichomes among the longer tapering trichomes, mainly in and near the inflorescence 2-densely and conspicuously glandular, especially on intermediate-length trichomes in and near the inflorescence, on the pedicels, and on calyces Table 8. Continued. Phacelia capitata (continued) Characteristic Blade length (mm) Doe Creek Doe/ Thompson Ridge The Drew Elk Creek Goolaway Gap* Lee Creek Little River Myrtle Creek North (n=33) Peel 44.9 10.6 50.2 9.3 47.4 12.4 60.7 15.0 33.4 9.9 40.4 7.9 40.6 8.7 48.0 12.3 38.2 10.4 3.8 0.7 4.4 0.8 4.2 0.9 6.8 2.1 3.3 0.8 3.7 0.6 3.9 0.7 4.5 1.3 4.0 1.0 Corolla opening (mm) 4.5 0.6 5.3 1.0 4.9 0.7 7.0 0.8 - - 4.1 0.7 4.6 0.5 3.9 0.4 4.7 0.8 Corolla length (mm) 6.1 0.4 6.6 0.6 6.1 0.4 6.8 0.5 - - 5.7 0.4 6.1 0.6 5.5 0.4 6.2 0.6 Calyx length (mm) 6.0 0.7 6.1 0.6 6.0 0.7 5.6 0.7 4.7 0.6 4.8 0.6 5.0 0.5 5.2 0.6 4.8 0.5 33.8 9.6 34.2 6.2 32.2 8.4 39.5 9.3 18.4 5.8 26.3 5.3 31.2 6.1 38.2 11.4 32.1 5.0 No. of leaf lobes 0.5 1.1 0.1 0.5 0.1 0.5 0.2 0.8 0 0 0 0 0 0 0 0 0.2 0.9 Type of lobing 0.3 0.6 0.1 0.2 0.1 0.2 0.2 0.6 0 0 0 0 0 0 0 0 0.1 0.5 No. of floral branches Glandulosity 3.2 2.5 3.2 2.5 1.8 1.4 3.0 1.9 1.1 1.7 1.7 2.4 1.3 1.0 3.3 4.5 1.0 0.7 0.9 0.3 1.0 0 1.0 0 1.0 0 1.0 0 1.0 0 1.0 0 1.0 0 1.0 0 Leaf width (mm) Stem height (cm) *-Data are pre-flowering. Table 8. Characteristic Blade length (mm) Continued. Phacelia capitata (continued) P. corymbosa Rice Creek Eight Dollar Mtn. I Salt Creek Weaver Road P. hastata Eight Dollar Mtn. II Waldo 8.7 47.8 7.4 44.1 14.2 37.8 6.3 36.1 5.8 29.8 5.6 3.5 0.6 3.9 0.8 4.1 1.1 10.0 2.1 8.9 1.2 7.9 1.7 10.0 Corolla opening (mm) 5.0 0.8 4.0 0.7 4.1 0.7 7.3 1.0 7.1 1.4 7.0 1.3 Corolla length (mm) 6.4 0.5 5.4 0.6 6.0 0.5 7.3 0.6 6.8 0.6 Calyx length (mm) 5.4 0.6 4.9 0.9 5.4 0.9 5.1 0.6 4.8 26.8 6.0 33.5 8.5 29.3 6.2 18.1 3.4 No. of leaf lobes 0 0 0 0 0.2 0.6 1.1 Type of lobing 0 0 0 0 0.1 0.3 1.3 1.7 2.6 3.1 1.9 1.0 0 1.0 0 1.0 Stem height (cm) No. of floral branche Glandulosity Cow Creek (n=10) 35.5 Leaf width (mm) Boomer Hill I Wiener Road P. heterophylla ssp. virgata 36.2 12.0 58.8 10.2 59.6 11.4 2.7 9.8 2.2 12.0 2.8 6.8 0.8 7.2 0.7 5.2 0.8 7.6 0.8 7.1 0.9 6.8 0.5 5.4 0.4 0.7 5.5 0.5 5.1 0.7 6.2 0.8 4.5 0.5 14.3 4.2 9.9 2.5 17.7 8.9 32.5 7.9 50.9 12.0 1.6 3.0 1.7 1.6 1.4 1.7 2.1 0.3 1.0 4.1 1.7 0.7 0.9 1.7 0.7 1.2 1.0 0.9 0.9 0.2 0.6 1.8 0.4 1.7 1.8 0.4 1.5 0.5 1.0 0.5 1.6 0.5 3.9 2.5 17.4 6.9 0 2.0 0 2.0 0 2.0 2.0 0.2 1.0 0 2.0 0 0 55 single plant. living plants these climax leaves were frequently withered anthesis. at living For this reason, measurements were made on from sterile rosettes, which appeared to be leaves representative as method was found in P. capitata that even on It provides It is believed that this possible. as an adequate comparison of the distinctive leaves of P. capitata with those of related species. The sterile rosette leaves of diploid E. capitata are generally were entire, although in six populations lobed leaves observed. lobed In these six cases combined, the number of leaves, the number populations two On these 12 leaves was only 12 out of 147 observed. sterile such rosette leaves (Peel, Doe leaves were compound pinnately Creek), Thus, lobing of observed. infrequent of is lobes varied from one to four; in of occurrence in diploid P. capitata. It should populations populations whether or from the of leaf of the 18 diploid studied, capitata E. 13 and in both containing tetraploids, it could be determined climax leaves of fertile rosettes were lobed lobing populations, total. However, in Percent of the 13 diploid populations ranged in (Bridge) to 85% (Boomer Hill I); in nine of these 0% degree that These results are shown in Table 9. not. climax 13 the noted be of lobed Thus, lobing climax the climax compared leaf leaves were 50% or less of climax lobing leaves do show a greater sterile to is rosette sporadic leaves. in occurrence 56 Table 9. Climax leaf lobing in Phacelia capitata. Population Number of stems observed with lobed climax leaves Beatty Creek 6/20 30 Bilger Creek 4/20 20 14/17 82 0/27 0 13/20 65 Cow Creek II 7/20 35 Cow Creek Turn-off 2/20 10 14/20 70 Doe/Thompson Ridge 6/16 38 The Drew 7/20 35 Elk Creek 18/33 55 Goolaway Gap 13/20 65 Rice Creek 10/20 50 Salt Creek 7/20 35 11/20 55 Boomer Hill I Bridge Callahan Creek Doe Creek Weaver Road 57 since and, these leaves withered frequently are at anthesis, is of limited use in comparative studies. The degree leaf of lobing is greater in both E. corymbosa and E. heterovhylla than in P. capitata. corvmbosa the number six, with entire leaves being infrequent. of but lobing ranges sometimes heterophylla, eight; of leaf (Heckard, 1960). In E. number of leaf lobes ranged from one to the the single a The nature pinnatifid to pinnately compound, from within lobes ranged from zero to leaf of In P. individuals observed, 16 had pinnately 20 compound leaves. Heckard entire being The lobes. leaves may leucophylla E. hastata describes the leaves of E. hastata as (1960) or occasionally tendency have with one for this species to possess entire led to some of the early confusion of E. (=E. hastata) with E. capitata. are to two pairs of The leaves of usually longer and wider than those of E. capitata. Leaf blade; width terminal diploid and lobed in was measured at the widest point of each leaves, lobe was measured. E. the the widest point on the large, The narrow width of leaves in capitata is distinctive. predominantly This characteristic, simple blades of sterile rosettes, are major distinguishing features of P. capitata. Blade lowermost the length lobes (as measured from the base of the in lobed leaves, and from the junction of blade with the petiole in entire leaves) appears quite 58 variable useful diploid in to It is taxonomically capitata. P. some extent, but not to the same degree as leaf outline and width. Throughout Flowers. biserial are circinate and crowded peduncle, Maaellanicae, the flowers the along forming the upper helicoid a cyme side of a (Heckard, 1960). Corolla in is significant for species comparisons Magellanicae. the distinct P. shape a broad, open-campanulate corolla (i.e., types: egena); margins of extremes in shape are of two The and a the lobes narrowly (i.e., P. imbricata). types, though not cylindrical corolla with the incurved and embracing the stamens Between these extremes, the corolla sharply defined, are stable enough to delimit certain taxa (Heckard, 1960). this In (measured length of the lobes incurved. spreading face of the corolla opening expanded flowers) and the of corolla (measured from the base to the tip used capitata, P. width the the were lobe) diploid the across of a study, being to quantify corolla shape. In corollas are mostly cylindric, the neither nor spreading-campanulate Rarely, a few individuals showed some degree of in the lobes (Boomer Hill I, Doe Creek, Doe/ Thompson Ridge). The correlated Creek size of with populations the corolla ploidy level. opening appears to be The Elk Creek and Callahan of P. capitata, containing tetraploids, 59 have the highest largest mean openings on average, of The corolla corymbosa and P. Jastata also are larger, P. populations The values for diploid populations. of the average corolla openings, exceeding the than those of diploid P. capitata. The the former two species are all tetraploid. corolla openings of diploid E. heterophvlla are within the range of diploid P. capitata, however. Differences width for in corolla length show the same trends as the corolla opening. of tetraploid the Again, there is a tendency populations have longer corollas to than diploids. The result increased of the individuals increased effects, of cell corolla "gigas" effect (Stebbins, 1971), in which higher and of tetraploids may be a size ploidy organ size. levels show a tendency for Because of such apparent corolla shape should be considered carefully when comparing these species. The were the at calyx length full anthesis. species was also measured in flowers which The range of variation among all and populations overlapped considerably; this characteristic thus is probably of little use taxonomically. Stem height. Magellanicae, species dm) show overlap as In considering height of stems in the Heckard P. (1960) noted that, while such corymbosa (1-3 dm) and E. nemoralis (10-15 striking differences, the ranges of most species considerably, "...limiting the utility of height 60 as taxonomic a general, He also noted that, "...in criterion." there was difference in height significant no between diploid and tetraploid forms of the same species." The results difference consistent E. capitata (9.9-18.1 ieterophylla plants in mean stem height between diploid (25.6-38.2 cm). The studied, cm) tall also is study indicate that there is a this of tetraploid E. corvmbosa and nature similar are heights biennial E. In the few P. hastata distinctive. stem the of to those encountered in diploid P. capitata. There is populations from a tendency for P. capitata plants from the containing tetraploids to be larger than those diploid populations, but individual observations from these two groups do overlap. Floral majority could The considered be (1960) stated that the Notable paniculate. exceptions those species having virgate inflorescences (i.e., heterophylla floral Heckard of inflorescence patterns within the Magellanicae include E. branching. virgata), ssp. in which numerous short branches overlap to form a densely branched system. term condition "capitate" in which is defined several Heckard by congested as "...the cymes are more or less separate from other cymes and form a head." his In states that scorpioid that is description of P. capitata, Kruckeberg (1956) the cymes usually inflorescence consists of two to three in a congested, subcapitate erect cluster borne singly and terminally. In this 61 study, number the cymes was counted, feature. The branching is variable allows from variability results (Table 10) indicate terminal in highly Even if one that many of the capitata show branching P. Exceptional individuals capitulum. populations diploid apparent still observed stems is "capitate" terminology used by the is such a single cyme below the terminal of fit populations. of this that and capitata P. in between and it the plants with 13 and 17 included Creek North), 14 (Salt Creek), 10 (Lee Creek), and (Myrtle 9 the to Kruckeberg, below assess presence inflorescence flowering to frequent within the floral branches below the terminal of Hill (Boomer Doe/Thompson Ridge) cyme branches and I below the terminal cluster. the Of zero, one, stems had stems 80 two or two studied branched the virgate to corvmbosa studied, all had below the terminal cluster; 38 cymes. Thus, the tetraploid for is not same extent as diploid P. capitata. The inflorescence distinctive. P. cymes such populations of in of Further Oregon, this species P. heterophvlla ssp. virciata is studies of P. hastata would be necessary to fully assess cyme branching in that species. is interesting to speculate on the extent to which It the variable, branching habitat whose in sometimes P. high, degree of inflorescence capitata may be a phenotypic response to disturbance. Of the three populations studied habitats are largely undisturbed (Rice Creek, Beatty 62 Table 10. Floral branching below the terminal cymes in Phacelia capitata. Population Occurrence of floral brinches below the terminal cymes Occurrence of more than one floral branch below the terminal cymes Number Number Beatty Creek 6/20 30 1/20 5 Bilger Creek 17/20 85 12/20 60 Boomer Hill I 16/20 80 14/20 70 Bridge 15/27 56 6/27 22 Callahan Creek 19/20 95 12/20 60 15/20 75 7/20 35 Cow Creek II 17/20 85 10/20 50 Cow Creek Turn-off 12/20 60 7/20 35 Cow Creek I Doe Creek 18/20 90 14/20 70 Doe/Thompson Ridge 20/20 100 14/20 70 The Drew 16/20 80 10/20 50 Elk Creek 32/33 97 28/33 85 Goolaway Gap* 11/20 55 4/20 20 Lee Creek 16/20 80 5/20 25 Little River 15/20 75 9/20 45 Myrtle Creek North 17/20 85 9/20 45 Peel 15/20 75 5/20 25 Salt Creek 19/20 95 11/20 55 Rice Creek 12/20 60 5/20 25 Weaver Road 15/20 75 11/20 55 *-Data from pre - flowering individuals. 63 Creek, Bridge), the latter two contained the smallest and proportions below terminal the diploid individuals with two or more cyme branches of The majority of other stems of ones. P. capitata came from highly disturbed areas, such Further studies would exposed roadbanks and rock pits. as needed be determine to relationship and define to nature the possible this of what factors, e.g., improved soil texture, might be influencing it. Vestiture the two common tapering wide trichome and the glandular trichome. trichome may be erect or appressed. The It exhibits range of sizes, the extremes of which may give the appearance large Heckard (1960) described basic types of trichomes in the Magellanicae, the tapering a and glandulosity. of two trichomes smaller distinct reaching trichomes types of tapering trichomes: lengths of four to five mm, and which often form velutinous a undersurface beneath the large ones. The part of They are septate the because when present, are trichomes, covering shorter trichomes. of most prevalent on upper parts of the stem, calyx leaves. the velutinous pedicels, lobes, glandular The peduncles, glandular and petioles, trichomes are the veins of often obscured of their small size and relative infrequency among tapering small, observable on trichomes. They are readily the upper portions of the stem, etc., using a 40x dissecting microscope. Kruckeberg (1956) stressed the eglandulosity of P. 64 capitata. that The species the throughout individuals the visible stem one individual (Doe Cow Creek Turn-off) appeared II, glandular. It is important to emphasize trichomes of P. microscopically. the Only discernible glandular trichomes; two Creek glandular near study indicate, however, does possess short, glandular trichomes no (Cow conspicuously my of populations. its possessed Creek) that results capitata are only are most abundant on the They inflorescence, on the pedicels, and on the calyx lobes; they are less frequent on the leaves. Phacelia capitata the by corymbosa is readily distinguishable from P. its short- to medium-length trichomes are all gland-tipped SEM (see section). glandulosity hastata shows patterns of Phacelia similar heterophvlla P. conspicuous dense glandulosity, in which to Phacelia capitata. E. was seen to be quite glandular, and resembled corymbosa in this respect, at least on the stems in the inflorescences. The vestiture leaf of E. capitata is illustrated in the SEM section. Species Phacelia capitata, (1956), is cespitose stems erect, presented. arising sterile unbranched A modified perennials each numerous description. with from leaf description revised from that Plants are 1-many thin, of of Kruckeberg deeply taprooted wiry flowering a rosulate tuft of leaves, with rosettes often present; stems are or occasionally branched, (11)20-45(63) 65 cm finely tall, consisting sericeous, herbage silvery-gray the a hyaline-shiny pubescence made up of three of trichome types: longer, shorter, matted trichomes, and with scattered, very short glandular trichomes, especially near the inflorescence, on pedicels, the the on basal calyx lobes, and leaf veins, rare or absent lower pairs are basal of long; climax basal lobe helicoid or borne more pedicels lobes in are 0.5-0.8 surface singly 2-3 capitate or subcapitate branches below the terminal cymes; fruit are with 1-3 mm long, hispid; calyx mm (3.0)4.2-6.6(7.7) long-hispid corollas long, margins, the abaxial are white, rotate, mm long, (2.3)3.0-6.0(8.4) mm broad, entire, obtuse-rounded lobes; appendages are attached barely the of or occasionally slightly campanulate-spreading, (4.1)5.0-6.9(8.0) with consist or terminally, or frequently with 2 cyme early inflorescences congested, a short-hirsute; cylindric, long, cauline leaves are gradually but not in wide, cm to petioles 1-2 cm tapering linear-oblong, mm with 1-2 (1.7)2.5-6.5(8.8) wide, upwards; lateral compound are entire, or occasionally have 1-2 pairs; cymes cluster mm leaves reduced pinnately or lobes, (2.3)2.9-6.0(13.0) wholly linear-lanceolate, simple and entire, pinnatifid rarely sterile leaves of the portion of the stem; rosettes very bristles underlain by appressed a free distally; millimeter portions stamens above forming and style the base of the corolla tube, a long (3 mm) and narrow "V" are exserted 5-7 mm, the 66 filaments glabrous mid-length along ovoid, mm long, densely clothed with stout bristles 2 mm 2-3 the mature long; specimens), with or few scattered hairs about a immature filament; seeds are are capsules mm long (in diploid 1.9-3.0 brown, with a reticulate-pitted surface; n=11, 22. Cluster analysis. taximetric studies, morphological were cluster to CLUSTER. capitata mean the values for the 10 attributes measured for 25 study populations subjected program summarize the results of the To was analysis using the computer The Goolaway Gap population of Phacelia omitted the analysis because data for from this site were collected prior to flowering. The Methods. agglomerative, dendrograms (Keniston, many different This introduces options i.e., until Nonetheless, degrees of fusion analysis, of 1978). the strategies subjectivity one nature; in thus, process are CLUSTER is interactive, as exist data hierarchical, is clustering the possibilities standardization CLUSTER combinatorial and summarizing obtainable well; program set, and into for and transformation or for selection of dissimilarity measures. the process of data must actually choose between various the most satisfactory results are obtained. CLUSTER allows for a useful summary of the difference between the species and populations studied. In this analysis, the Bray-Curtis dissimilarity index 67 been has utilized. attributes the with measure, relatively whereas values largely determine because four for (Boesch, leaf the number and this of attributes index glandulosity) low is that of scores Thus, in are this stem height would and Due to this, and results. (type cyme of with 1977). length cluster attributes lobes, cluster, feature high scores largely determine the value of unimportant instance, leaf A leaf lobing, number of branches below the terminal were assessed in a qualitative or non-metric way, the data require standardization (Dr. C. D. McIntire, pers. comm.). was division for a the by the attribute maximum, in which all values given maximum morphological value "high The standardization method used observed score attribute for are divided by the This removes that attribute. bias," and allows use of quantitative and qualitative values together. The fusion individual thus as the group result, mean In the strategy, inter-group resemblance is defined of resemblances all members to of another. between members of one Thus, group average has no tendency toward "space distortion." it introduces relationships measure the way in which group/ final clusters which are formed. the average clustering defines and group/group resemblances are calculated, and defines group strategy produces only moderately sharp clustering, but relatively originally (Boesch, As a 1977). little expressed distortion by the to the dissimilarity The group average fusion strategy 68 best summarizes exist between the which differences the Phacelia species and populations studied. The Results. shown results in Figure 3. seen at cluster analysis are the of The four major clusters of interest are They a dissimilarity level of approximately 0.12. include, from left populations), tetraploids 2) two and P. 1) diploid P. capitata (17 right, P. populations containing capitata hastata populations), (four corymbosa. to population), (one 3) E. and 4) P. Jieterophvlla ssp. virciata (one population). The mean results values, variation at least on the basis of that, populations diploid the uniform fairly indicate morphologically. within the species, of P. capitata are The actual degree of especially with respect to size and outline, and floral branching, were discussed leaf above. similarity The containing tetraploid population is these are the Creek and Callahan to capitata interest, of intermediate these actual E. between intergrading Creek). but and the the populations E. hastata it does not "prove" that taxa at the former sites (Elk The similar tendencies toward entire leaves, larger corolla openings and lengths, longer, and expressed three glandulosity have most likely contributed populations being clustered together. The source of morphological variation in the E. capitata populations containing tetraploids, is considered in the final section. which is problematic, 69 Figure 3. Cluster species. analysis of Phacelia 010-Little River; 020 (phacelia capitata=A: Peel; 030-Lee Creek; 050-Bilger Creek; 060 Myrtle Creek North; 06A-Weaver Road; 070-Rice Creek; 080-Boomer Hill I; 090-Beatty Creek; 120-Cow Creek I; 130-Cow Creek II; 140-Salt Creek; 150-Doe/Thompson Ridge; 15A-Doe Creek; 160-Cow Creek Turn-off; 170-Elk Creek; 180 Callahan Creek; 190-The Drew; 210-Bridge; E. corymbosa=0: Road; 020-Waldo; 030 010 -Wiener Eight Dollar Mtn. I; 040-Eight Dollar Mtn. II; E. hastata=H: 010-same site as Boomer Hill I above; P. heterophvlla ssp. viruata=E: 010-Cow Creek. See Tables 3 and 7 for population locations.). r-{ 0 OL r 1-4 zo orn x 1-1 01- mDD -0-1 2> 73-0)0 m3)-0 z (r) -1-103> z 1-41_, 0 cn cul rsm. im .2 1-1 1-4 rri C) X rn-irrip 71 Z 3> CO 2> o <73--1 0 C rn C-<71U1 OD-1C 7:173-1 r ocoDo L,J F-6 °C) D ___ E010 A180 0010 0030 0020 0040 H010 A.170 A15A A190 11. A080 A130 A210 A150 AOGA AOGO A020 A070 A090 A120 A160 A030 A050 A140 A010 N -4 LT1 01 .1 CO DISSIMILARITY 71 Cluster the major analysis appears to be useful in illustrating differences Magellanicae, corymbosa, which exist the species group in at least when well-defined entities (i.e., P. heterophvlla P. virgata, ssp. and diploid E. capitata) are compared. The major capitata, morphological corymbosa, P. differences between Phacelia Jastata, P. and E. heterophvlla ssp. virgata are summarized in Table 11. Scanning Phacelia leaf E. electron capitata vestiture capitata electron Seeds of Methods. E. corvmbosa, pollen morphology and and of E. capitata, and upper stem vestiture of and corymbosa were studied using scanning E. microphotographs. specimens. taken microscopy. All pollen, The seeds used were dry, mature stem segments, and leaves were from herbarium specimens and hydrated; they were then critical-point dried Al. All (1968). studs following the procedure of Cohen, et specimens were then mounted on aluminum using Duco adhesive (Microstick adhesive for pollen), coated with approximately gold-paladium alloy photographed. University in angstroms 180 of 60-40 a vacuum evaporator, examined, and The photography was done by the Oregon State Scanning Electron Micrography Lab, Department of Botany and Plant Pathology. Results. is similar (Figure 5A); Magellanicae The reticulately ridged and pitted seed coat in P. this capitata (Figure 4A) and E. corymbosa feature is shared by all species of the (Heckard, 1960). Of particular interest is Table Species capitata n 11,22 II. Comparison of Phacelia capitata, P. corymbosa, P. hastata, and P. heterophylla ssp. virgata in southwestern Oregon. Habit Stem height (cm) perennial, mat-forming 20-45 7.5-25 corymbosa 22 perennial, mat-forming hastata 22 perennial, with 1-many stems, rarely mat-forming 20-50 biennial 30-75 heterophylla "P. virgata 11 Leaf outline Leaf size length(cm) width(mm) Inflorescence Corolla shape Clandulosity linear-lanceolate, 2.5-6.5 entire; rarely, with 1-2 pairs of leaflets at base of lamina 2.9-6.0 Cymes often terminal, or sub-terminal,or with I-many lateral inflorescences below the terminal cymes linear to lanceoblong, entire, or often with 1-2 pairs of leaflets at base of lamina 2.4-5.2 6.0-14.0 cymes terminal, or cylindricoften with 1-2 subcampanulate terminal cyme branches conspicuosly glandular overall, with short-medium glandular trichomes lanceolate, entire, or rarely with 1-3 leaflets 3.5-7.5 7.0-13.0 cymes terminal or sub-terminal, often with 1-many lateral inflorescences below the terminal cymes cylindric scattered, short glandular trichomes, especially near the inflorescences, on the pedicels, calyx lobes, and leaf veins pinnately dissected, with 1-4 pairs of lobes 3.8-8.0 virgate, with 10many overlapping lateral cymes cylindricslightly flaring 8.0-17.0 cylindric (campanulate in tetraploida) campanu late scattered, very short glandular trichomes, especially near the inflorescences, on the pedicels, calyx lobes, and leaf veins medium-length glandular trichomes frequent in the upper portions of the inflorescence 73 Figure 4. Seed of Phacelia capitata. A Dorsal surface (50X) B Portion of dorsal surface (150X) C Dorsal surface pit detail (340X) D Detail of a double "pore" complex (700X) 74 75 Figure 5. Seed of Phacelia corymbosa. A Dorsal surface (40X) B Portion of dorsal surface (160X) C Dorsal surface pit detail (280X) D Detail of "pore" (1600X) Figure 5. 77 the presence, distributed junctions or and more coat. the of they not the in previously of genus Phacelia, documented. concerning supplying seed of or study for improved shows cell-wall (Figure 5D) channels through the seed been include regulation These presence of such seed "pores" in other species possibilities "pores" hardened by the with may provide seemingly open (Figure 4D) Mactellanicae, apparently bordered be to constricted The associated ("pores") pit margins (Figures 4B-D, 5B-D). the appear material, species, of a set of distinctively structures of "pores" both in has Intriguing the function of these oxygen to the embryos, dormancy (i.e., in P. corvmbosa, which germination after cold stratification seed (Quick, 1947)), and/or facilitation of water imbibition. Another prominent circular This is this area gone, a second do (Figures is material (Figures 5C, D). frequently cracked, and when one layer deeper lining is revealed in the pit bottoms; surface often has a small crack near its "cracking membrane" system may also provide study in determining seed germination ecology. to appear 4B, morphology is interesting not coat In the bottom of each pit is a thin-looking of This relationship seed of P. corvmbosa. in "membrane" center. an peculiarity to C), be its possible These membranes as clearly developed in E. capitata seeds of which germinated well in the greenhouse with no cold stratification treatment. An SEM study of pollen morphology in the 78 Hydrophyllaceae, species two The imbricata. described uniformly baculate; grains extending pole, the equally pseudocolpi pollen these of exine the foveo-reticulate; colpi The pollen of nearly capitata P. but toward and almost are isopolar, the grains the colpi the finely is are species tricolpate-tripseudocolpate, as conspicuously P. heterophylla and P. Magellanicae, the in and Chuang (1982), included Constance by reaching either not long as the true colpi. as fits this description closely (Figure 6A, B). A surface illustrates and of of the long, tapering trichomes, trichomes shorter undersurface the leaf vestiture in E. capitata of presence the the presence view (Figure form velutinous a reveals Magnification 6C). glandular small which the hairs, in this case near the leaf margin, in the species (Figure 6D). The P. conspicuous, corvmbosa (Figure distinguishing E. capitata possess very medium-length glandular trichomes of from it is not small 7C) are very important P. capitata (Figure 7A). in However, eglandular (Figure 7B), since it does glandular trichomes similar to those which are found in P. corvmbosa (Figure 7D). Reproductive bioloav Flowering and Phacelia capitata elevation sites pollination begins in Flowering biology. early May at the in lower- and continues until early to mid-June. At 79 Figure 6. Pollen and leaf vestiture of Phacelia capitata. A Pollen grain (4500X) B Detail of pollen grain colpus (18000X) C Leaf vestiture (30X) D Detail of vestiture at leaf margin (300X) 80 81 Figure 7. Stem vestiture of Phacelia, capitata and P. corymbosa; stem segments taken from just below the terminal cyme branches. A Phacelia capitata stem vestiture (150X) B Short glandular trichome of E. capitata (600X) C Phacelia corymbosa stem vestiture (40X) D Short glandular trichome of P. corymbosa (500X) Figure 7. 83 higher-elevation the until May late first flowers the lower portions that the flowering is delayed initial early June and lasts until early July. or The noted sites, open within an inflorescence are on to the cyme branches, although it was of lowermost buds are not always the first to bloom. As flowering proceeds, the initially helicoid cyme branches uncoil until, at the time of seed maturation, they are nearly or completely ascending. flowers In the stamens 5-7 mm with of and beyond the capitata which are fully expanded, P. are well exserted, extending from styles the corolla tube. ends style the of The anthers are subequal branches. The stigmatic surfaces are very small, the stigmas being confined to the tips the style of branches and consisting of the ends of only a few cells (Heckard, 1960). The of quite capitata are congested, forming a P. many-flowered tight, be flowers inflorescence attractive which, as a whole, may pollinators. to In cases where the in an inflorescence were counted, the number ranges flowers 53-303; these inflorescences varied from approximately from 2.5 to 5.0 cm in diameter. Throughout protandrous, stigma of the MaQellanicae, flowers are the pollen being ripe and available before the is receptive (Heckard, 1960). dichogamy, the or temporal This is one mechanism separation of sex functions in flowers, whereby Protandry is only effective, however, if it is total, i.e., self-pollination may be inhibited. 84 if all of stigma is Another receptive as situation seen protracted and may likelihood be (Faegri may which factor protandry ages pollen is removed from a flower before the the Pijl, 1971). effectiveness of self-pollination is the the Magellanicae, in which flowering is in many adjacent flowers of slightly differing open at one time on a single individual. of having ripe simultaneously within one single within of der the decrease inhibitor an van and pollen and inflorescence The receptive stigmas opposed (as to is undoubtedly increased in such a flowers) The congested arrangement of the flowers may make system. pollen transfer to such receptive stigmas likely. Owing in P. capitata, pollen must apparently be transferred occur from the anthers to geitonogamy. equivalent to flower), flower to the stigma of another same individual. that on as one of flower one factors, for self-pollination to above the to This process is referred In the genetic sense, geitonogamy is autogamy except (self-pollination occurring within that the services of a pollinator are required (Faegri and van der Pijl, 1971). The process of geitonogamy in Phacelias with congested inflorescences is well-described by Willis (1895), in a discussion of pollination in E. tanacetifolia. dichogamy of the flower gives it some chance especially as it is cross-fertilization, largely visited for honey and pollen; but it seems probable that self-fertilization is more common, even when it is visited by insects. The are closely packed together...Insects flowers crawl over the whole mass of flowers, touching stamens and styles indiscriminately, and probably often knock pollen on to the stigmas from the The of 85 surrounding anthers...Every flower on the plants examined set a full complement of seed. Heckard concurred (1960) observing that, common the in with the observations of Willis, although self-pollination probably is Magellanicae, "...it does not seem to occur to any extent when bees are lacking." During visitors insect behavior was landing the on Macior, just "floral this opportunity for the same moving (1895). Upon It could be seen stamens styles are inter-plant visitation observed. In as many of giving the Even autogamy is possible if Insects were often observed mature. plants cross-pollination and undoubtedly about, another to close often so the same by way, is if pollen same of together opportunity an P. that was also pollinator there well, the on populations dense particularly this (Dr. L. the flowers. inflorescence In nectar on anthers long enough for stigmas of to one the Willis by collecting geitonogamy. flower individual. the pushed remains from capitata from process are pollen surface," comm.) flowers some described as inflorescence, the insects would crawl over pers. in the action of studies, field of E. capitata plants was observed. Their on an entire that course the for is retained on the body of the insect. Cross-fertilization evidenced by interspecific observed by the does instances intergradation Heckard occur (1960). of at in hybrid the these plants, as swarms and by tetraploid level, as The suspected allopolyploid 86 origin tetraploid species, such as E. hastata, certain of would require cross-pollination events as well. general, the pollination syndrome of E. capitata is In described best geitonogamy facultatively as probably is cross-pollination Though xenogamous. frequent, opportunities for most No evidence of apomixis in the do exist. Magellanicae has been observed (Heckard, 1960). visiting Insects and by potential most and of Entomology, Oregon State University. pollinators frequently spp.) J. Lattin, A. Moldenke, and W. Stephen Drs. Department the These of E. capitata corvmbosa were collected during May, 1984, and were P. identified of populations several listed in Table 12. The visitors were bumblebees (Bombus observed honey-bee, the are Avis mellifera. The variety of this limited sample indicates a insect visitors found rather general, "open" pollination system for E. capitata. In of pollination ecology in California, Moldenke study a both that Phacelia species are heavily visited by stated (1976) "specialist" and visiting P. species while generalists, "generalist" one ordinarily members the family The Bombus E. corymbosa instance, an individual genus a vectors. and capitata in Melissodes, of in of specialists are of visiting Asteraceae (Dr. A. Moldenke, pers. comm.), was found collecting nectar from P. capitata. Seed which production. Phacelia pollinators Methods. To compare the extent to capitata sets seed with and without natural available, five exclosures were established. Table 12. Insect pollinators of Phacelia capitata and P. corymbosa. Phacelia species Order Family Genus and species P. capitata Hymenoptera Apidae Apis mellifera L. Bombus californicus Smith B. caliginosus Frison B. edwardsii Cresson B. huntii Greene B. vosnesenskii Rad. P. corymbosa Anthophoridae Melissodes sp. Megachilidae Osmia sp. Lepidoptera Lycaenidae Everes comyntas Godart Hymenoptera Apidae Apis mellifera L. Bombus vosnesenskii Rad. 88 These exclosures Area Natural set up in the Beatty Creek Research were on April fruiting The 1984. 25, inflorescences were retrieved on June 30, 1984. The which exclosures consisted of small milk carton sections were placed in a short segment of nylon netting. The below the netting tied was inflorescence; supported tied was it entire the stem the around off the top to a stake, which on were exclosures All exciosure. established before the first flowers had opened. retrieving Upon from natural and same the the exclosures, paired inflorescences The number of flowers pollination, were collected. produced seeds by been left open to had which individuals, five inflorescence pairs were the then counted. Results. enclosed was were inflorescences significant no excluded agreement noted production Seed with that is open-pollinated in summarized in Table 13. and There seed production in inflorescences that from pollinators. the observations These of results Heckard are in (1960), who caged plants of P. californica produced only an occasional seed per cyme. Although for the data indicate that pollinators are needed appreciable provide information self-pollinated however, seed or set in P. capitata, do not on the extent to which the flowers are out-crossed; as selfing is probably prevalent. self-pollination they previously stated, Careful artificial would be needed to determine the extent of Table 13. Inflorescence* 1-1 open Seed production in open-pollinated and enclosed inflorescences of Phacelia capitata. Number of flowers 158 Number of flowers producing seed Percent flowers producing seed 61 38.6 Number of flowers with evidence of seed predation Number of flowers producing: 1 seed 2 seeds 3 seeds 18 55 6 0 0 0 0 0 1-1 closed 130 0 0.0 1-2 open 108 40 37.0 4 40 0 0 1-2 closed 136 0 0.0 0 0 0 0 2-1 open 247 175 70.8 20 85 87 3 2-1 closed 219 1 0 1 0 0 2-2 open 154 82 53.2 30 55 27 0 2-2 closed 154 5 3.3 0 4 1 0 3-1 open 59 49 83.1 0 49 0 0 3-1 closed 53 0 0.0 0 0 0 0 0.46 *-First number indicates individual plant; second number indicates inflorescence pair from each plant. 90 self-compatibility. in the stigmatic small breaking the of difficulties complex. a owing field, hand pollinations are difficult to the congested inflorescences and surfaces; several attempts stems. Heckard (1960) ended in the alludes to the in artificially self-pollinating plants in the With few exceptions, artificial pollinations gave percentage low Such of seed set compared to that obtained by natural pollinations. Some capitata no seeds occurs inflorescences of E. in In the cases where this was observed, (Table 13). holes small predation seed were present in the sides of the capsules, and had No insects were found that could developed. be implicated in this seed predation. Pollen viability. A Methods. survey of pollen viability in corvmbosa, P. hastata, and P. heterophvlla ssp. viraata was populations Phacelia of capitata, P. The pollen grains were sampled from dried herbarium made. specimens. One single flower, cotton blue in to three were used. dehisced The anthers, all from a pollen was stained with and counts were made using a lactophenol, Zeiss light microscope. Resul s. The pollen viability counts are listed in of pollen viability is often useful as a Table 14. The degree conservative For example, sterile indicator reduced hybrids of the pollen resulting fertility of individuals. viabilities may characterize from interspecific crosses. 91 Table 14. Pollen viability in Phacelia species group Magellanicae. Number of pollen grains examined Species Population n P. capitata Bilger Creek Boomer Hill I Boomer Hill II Bridge Callahan Creek Cow Creek I Cow Creek II Cow Creek Turn-off Doe Creek Doe/Thompson Ridge The Drew Elk Creek Goolaway Gap Little River Myrtle Creek North Rice Creek Salt Creek Weaver Road 11 P. corymbosa P. hastata Eight Eight Rough Waldo Wimer 11* 11 11* 22* 11* 11* 11* 11 ? 11* 11 11* 11* 11* ? 11* 11* Dollar Mtn I 22 Dollar Mtn II 22 & Ready 22 Road Boomer Hill I P. heterophylla Elk Creek 22 22* 22 11* Percent viable pollen Collection 65 Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly Shelly & Nelson & Holmes & Holmes & Nelson & Nelson 709 412 719 706 & Holmes & Nelson & Nelson & Nelson 349 721 & Holmes & Nelson 720 447 579 552 418 568 45 96 97 81 36 Shelly Shelly Shelly Shelly Shelly 700 702 698 697 691 1703 49 Shelly & Holmes 767 538 96 Shelly & Nelson 727 480 758 572 593 583 645 710 802 817 526 516 710 865 198 761 813 223 687 95 93 58 95 98 97 92 85 86 99 98 71 98 93 96 80 71 *-Pollen viability obtained from a specimen other than the chromosome count voucher. 724 769 781 754 733 783 734 726 737 792 725b 92 viability Thus, suspected degree may be of use in studying known or counts intergradient interfertility of In this project, be used to in assessing the and populations species or populations. between it was hoped that pollen viability could study capitata populations Callahan Creek), hybrid nature of the P. potential the containing tetraploids (Elk Creek and and to describe the range of variation in pollen viability among diploid populations of the species. The results percentage of Although plants from of E. capitata generally show the high populations diploid inconclusive. are pollen expected in nonhybrids, there viable are four exceptions (Boomer Hill II, Elk Creek, Salt Creek, and Weaver severely high lowered and low tetraploid pollen viabilities were found in plants from of E. diploid viability pollen Due corymbosa. nonhybrid in populations, In addition, both (i.e., less than 75%). localities variability an in which pollen viability of diploids is Road) and to this tetraploid is of limited usefulness as indicator of hybrid origin in the tetraploid E. capitata populations. Environmental Magellanicae, the observed, be but attributed involving found may strongly influence pollen Heckard (1960) found that in nonhybrid species viability. of factors pollen viability was always that "...a variation of about 10 percent can to studies that high the of pollen environment." E. In an extreme case californica and E. imbricate, he viability varied within the same 93 populations concluded through depending that decrease observed abortion period was "...largely a result of a blooming the pollen increasing the Heckard when they were sampled. on in available water and an accompanying decrease of Such vigor." temperature environmental variations, perhaps including may fluctuations, to contributed have the variability in pollen viability observed in this study. Competitive ability and ruderal response The hypothesis serpentine soils, physiological because by restrictions of several as result a possible of but habitats such to ability competitive poor of communities only not are limited to species certain that more in also closed non-serpentine substrates, has been offered botanists. (1951), Kruckeberg in found studies and off comparing growth serpentine soils, concluded that the superior growth of all the tested species serpentine uniquely merely serpentine non-serpentine aside endemics of restricted In later a to requirement specific some serpentine." plants than chemical from serpentine by soil suggests that not are on report 1954), he again suggested that, because growth (Kruckeberg, of endemics because provided species non-serpentine on serpentine "...the plant of unable found to be better on on serpentine soils, some other factor tolerance restriction. are was to must be sought to explain He postulated that the serpentine maintain themselves against the 94 competition rigorous Stebbins Major and ability grow to that from extinction by preserved are stating agreed, (1965) endemics "...serpentine the non-serpentine plant communities. of on serpentine where more competitive plants cannot grow." Phacelia example of A and grow However, a on E. possibly the Boomer species the apparently provide poor with endemic excellent an competitive the species has the ability to germinate non-ultramafic soils (80% germination rate). capitata is not found off serpentine soils (or metamorphic derivative thereof, i.e., at close Hill observations to greenhouse germination test on non-serpentine that showed appears serpentine a ability. soil capitata indicate may poor be the that result, a The following field in nature. site) I competitive serpentine restriction of in part, of the species' evident even in its ability, own native, undisturbed habitat. often serpentine communities, P. capitata is undisturbed In found in serpentine slopes frequently along channels in very specific types of microsites. at the Peel intermittent, which cover of site, individuals vernally other On the scoured species occur stream is low. The plants which occur in these shallow channels are often very large, as well. of annual In the surrounding vegetation, where cover grasses and other forbs is higher, the frequency of E. capitata is noticeably less. The apparently poor competitive ability of the 95 species, is again within its own undisturbed native community, also evidenced Creek. Here, infrequent throughout outcrops interesting are However, layer instances in of the Pinus large where E. capitata plants bedrock occur, In addition, an the cracks of the rocks. in and scattered widely herbaceous the serpentine of frequent are individuals savanna. jeffreyi serpentine slopes along Beatty the on example which suggests the need for a very open seed germination was observed here. "safe site" In location where a dead Pinus jeffreyi trunk had fallen a over, root hole large a capitata left in the soil by the upturned was hole contained seven large, floriferous This mass. Phacelia good for individuals, while the surrounding in layer, very few plants, and no other such concentrated herb "clumps," were found. Owing to apparent the competition, perhaps in establishment phases, one invade the grow and case--at serpentine abundant rock quarries, might all sites 18 has on for lowered germination seed the expect well in disturbed areas. soil more requirement P. capitata and to Indeed, such is where some disturbance of the occurred, P. capitata is noticeably the open, barren areas (i.e., roadbanks, etc.) than in the surrounding, undisturbed serpentine habitat. Quantitative were collected traversed data at Lee to illustrate this ruderal response Creek, by a logging road. on a serpentine hillside Seven 4x4 m square plots were 96 set three up: plots adjacent bedrock The of compared however, the on cover of fact that to adjacent undisturbed areas where cover The plants are most abundant, serpentine raw species is lower. other the cotyledons an still on the exposed a higher percent cover species is higher. other on The results are given in Table outcrop. have and plot one and show that P. capitata individuals are more data numerous, outcrop vegetation, undisturbed in 15. meters above the uphill edges of the roadbank five located plots on the barren roadbank, three plots only Also of interest is the observed seedlings were apparent) where percent roadbanks in with (i.e., of the roadbank two plots. would It be naive to conclude, on the basis of these observations alone, that competitive exclusion is the field or even the most important, cause of the serpentine single, restriction The situation may be much more of E. capitata. and more controlled experiments would be necessary complex, to fully understand the autecology of the species. As example, an which growth rates of E. capitata on and normal soils could be compared, to determine serpentine whether the the have species grows ordinarily toxic elements, such as magnesium. evidence indicating associated with serpentine soils optimally that, of concentrations certain Also, Grime (1979) summarizes in instances nutrient-deficient were on ultramafic soils where habitats plants such as supplied with high rates of mineral Table Plot 15. Ruderal response of Phacelia capitata at Lee Creek. Data collected from seven 4x4 m plots. Number of plants (non-flowering; flowering) Percent cover Number of seedlings Percent cover of other vegetation Herbs Shrubs Trees 0.0 0.0 A. Roadbank plots 8.0 5 1.5 14; 31 5.0 1 2.0 1.0 10.0 3; 24 25.0 0 1.0 0.0 0.0 1 0; 28 2 3 B. Undisturbed plots 4 4; 4 1.5 0 55.0 0.0 15.0 5 0; 0 0.0 0 95.0 0.0 75.0 6 1; 0 0.1 0 90.0 0.0 70.0 5; 9 2.0 0 8.0 0.0 0.0 C. Rock outcrop plot 7 98 nutrients (i.e., in quantities detrimental to the growth of such accumulated plants. which these nutrients appeared to be calcium), P. capitata provides such an example, in Perhaps there a "requirement" for the low concentrations is of phosphorous and calcium which probably characterize many of its sites. and requirements case on capitata example of naturally If a soil generally in serpentine a may endemic lower the such as P. Lastly, E. capitata may be a good habitats characterized by high stress status) and light disturbance. nutrient characteristic competitive the substrates, the intrinsically slower growth rate which is then so, operate "stress-tolerator" (Grime, 1979), in that it exists poor (i.e., of such soils. on may Lack of required soil properties, in non-serpentine ability competitive ability competitive poor also exists. together this Of course, the possibility that special soil of such species may also decrease ability of this and other serpentine endemics. Thus, (1975) as state, hypothesis, ecology unknown. Kruckeberg and Proctor and Woodell the full nature of the competitive exclusion which and (1954) may be so important when considering the evolution of serpentine endemics, remains Clearly this is a fertile area for research, and Phacelia capitata may provide a very revealing subject. 99 DISCUSSION AND CONCLUSIONS $peciation and evolutionary relationships provides 4agellanicae) both the morphological and complexes, such a mature polyploid of example an group (species alliance magellanica Phacelia The complex. In ecological extremes species. However, the diploids are often less common than are usually represented by the diploid polyploids. the restricted, and amounts the populations their geographic Their more are genetic variation within of usually is ranges other, each with the addition, In less. or diploids are usually sympatric only in restricted areas, so that the likelihood hybridization of of doubling and different involving allopatric diploids much is number chromosome reduced (Stebbins, 1971) . the Of species six exclusively almost considered in "morphological in and ecological attributes occur is its the is in P. capitata and P. characteristics the form serpentine soils. of example good a extreme," ecological morphological to study: this these of distinct restriction Maaellanicae that occur the or partially as diploids, two have been detail Each corymbosa. of having and a very special of virtually complete Of all six species which as diploids in the Magellanicae, however, P. capitata most restricted morphological in its geographical range. distinctiveness, restriction Thus, to 100 distribution pattern, diploid the (or a close derivative), narrowly endemic soils serpentine level almost and make exclusive occurrence at an capitata E. unusual, unique entity in the Mapellanicae. It of difficult to elucidate the evolutionary history is such diploid a speculation of in a mature complex, but In considering possible modes interesting. is polyploid speciation for E. capitata, it is convenient to consider types two origins: insular endemics are diverged from ancestral those upon depleted and which their in Insular common, but which were never differ endemics. following the were more once of these establishment a small, isolated area. which those which species, a small group of individuals of a widespread species, individuals are endemic of Depleted species common, widespread, and richer in biotypes (genetically different populations), but whose present variability 1942a) two endemic to depletion genetic of of restricted species may be further types one another by two criteria. from insular living Secondly, specialized the form continental species; form, if First, if closely related to widespread species which is surrounding in strictly other due subsequent loss of biotypes (Stebbins, through distinguished occur is . These an rarity it if areas, it is probably a it is closely related to no is more likely a depleted species. endemic is morphologically a highly compared to its relatives, it is probably 101 an insular descendent if it is less it may be a depleted ancestor (Stebbins, 1971). specialized Based competitive distribution, and morphological characteristics of Phacelia theories regarding the origins of modern-day and vegetation geographic restriction, edaphic ability, cytology, the regarding observations on capitata, while these, of Klamath the in can one Mountains, support, much difficulty, an insular mode of origin for this without species. polyploid In with complexes lowest the number in nature, the members found chromosomes arc the primitive of while the polyploids are derived from them (Stebbins, ones, 1942b). addition, In irreversible 1971). at least such as the divergence and the can Magellanicae an be evolution in known, and the diploid level will have at mode primary of direction the speciation initial, predominantly is lower to higher levels (Stebbins, from Thus, complexes formed trend polyploidy of evolution for such groups. Because P. capitata occurs almost exclusively as a diploid, it leading to chromosome can from be concluded that the evolutionary events can the origin level. described be the This as ancestral occurred at this species the of being the case, then P. capitata one of the evolutionary "offshoots" stock diploid which formed the foundation of the Magellanicae. The Tertiary northward Geoflora migrations probably of brought elements with of the Madrothem, sometime 102 since beginning the Phacelia species correlates which from mature Pleistocene or 500,000 10,000,000 and Pliocene and selection for between approximately Upon the ancestral substrates serpentine the perhaps few, a may capitata ago. the in southwestern Oregon, the process of natural serpentine-tolerant divergence epochs, years with during originated probably contact California E. This polyploid complexes that may presently be of considered species' evolved. capitata E. well with the conclusion of Stebbins (1971) that majority the the Pliocene epoch, the ancestral of partially preadapted, could have begun. individuals Further ended with the "insular" isolation of have the serpentine outcrops which it presently on occupies. The ability fact helps only provides one possible explanation for the not species' present explain species. If serpentine reproduce this rise to capitata thus may imbalanced to serpentine soils, it also restriction the hypothetical evolutionary history of the the few tolerance individuals only were which were developing able to survive and in the more open communities on serpentine soils, could tolerance, E. capitata possesses poor competitive that reinforce selection the for serpentine and further divergence of the individuals giving the species through time, could occur. despite The persistence of E. poor competitive ability, be owed to its ability to tolerate the chemically soils, and to its consequent "escape" of the 103 of competition. The morphological distinctness of Z. rigors capitata from other widespread members of the Magellanicae, well as its restricted geographic distribution, provide as further evidence that the species is an insular endemic. The cytogeography the origin of species in Oregon. because both It logical is corymbosa Two problems make previously. location for location species First, Z. P. of diploid corvmbosa is in Siskiyou County, 175 km south of the southernmost capitata (Heckard, 1960). Z. both are known northernmost The approximately California, substantiate. to hypothesis not known to occur as a diploid in Oregon, as is discussed speculate that, are largely restricted to serpentine species difficult 8 pattern of these to they may share a common ancestor. a Figure species. distribution allopatric the this latter the for illustrates soils, a related is also of interest when considering a possible south, mode corymbosa, P. capitata on ultramafic soils to replaces which species phacelia of distinct they morphological very Second, the two entities in the different from one Magellanicae, and another. a common diploid ancestor was in fact shared If are two species, it must surely have been long extinct. by the No evidence for any present genetic interaction between Z. capitata and Z. corymbosa was found. Even restricted the the study of a morphologically distinct, narrowly endemic difficulties such as Z. capitata has not been spared of polyploidy and interspecific 104 capitata 2n * capitata 411 0 capitata ?n N * corymbosa 4n corymbosa ?n CALIFORNIA Figure 8. Distribution of Phacelia capitata and P. corymbosa in southwestern Oregon. Ultramafic formations outlined (Wells and Peck, 1961). Phacelia corymbosa locations are plotted from herbarium specimens, except for the tetraploid location at Sexton Mountain, Josephine County (Heckard, 1960). 105 with Further, diploids factors ...confounding as due introgression to Stebbins tetraploids." due to fact the taxonomists origin limitations necessary is in other or (1971) concluded that this may be guides tetraploid such diploids local from inadequate many of caution diploids." in the morphological criteria used by that are encountered that its later and variation of tetraploids in one place the "...interpreting taxon a the states, (1976) other polyploids may cause and not states he by polyploidy of "...attainment interaction Jackson As hybridization. evolutionary the to With these populations. mind, the interrelationships of E. capitata in within the Magellanicae may be considered. Fhacelias the early collections of Magellanicae -type of Certain in County, Oregon (see which Kruckeberg (1956) likened Relationships), capitata, are P. at Although genomes P. genetic between These collections may these two species, but most this time best referred to the widespread, variable The Heckard of which P. specimen collected by Constance and DS) is the most intermediate in appearance. (2956, conspicuous, latter hastata. E. intergrades hastata. Rollins to of the species with other members of the especially represent former of suggestive are interrelationships complex, Taxonomic Douglas (pers. comm.) felt this plant may combine capitata and P. corymbosa, it lacks the intermediate-length glandular trichomes of the would be expected in such a hybrid. On this 106 basis, because and was specimen easily found Nebo Mt. where this not on serpentine soil, it is most is to E. hastata, with a note as to possible referable with affinities near area the capitata. P. found the former have I species to be extant in the Mt. Nebo area. A location only with sympatry examples 768, their possession buds capitata and interest at P. perhaps predominantly chromosome counts on On the basis of other features, between P. at least be postulated. Of hybridization of can site is the less stringent nature of the between intermediate is habitats to which E. capitata the disturbed the factors nature which of the highly is largely and a more normal soil chemical profile. restricted, the hastata this which serpentinized plus occurrence the however, soil, meiosis. past of specimens were unsuccessful, as all the intermediate were possibility narrow, Attempted leaves. the be intermediate, however, in to relatively of as a tetraploid, a Some collections (Shelly & species. appear 771) silvery entire, limits ordinarily Holmes latter the "good" that the former species showed and between hybridization representing individuals species which situation capitata was found in biotic E. of diploid a as This is member of the complex (E. hastata). another each of occurs where counts Chromosome these is the Boomer Hill I site. intergradation species the which is potentially revealing with respect to site This, roadside habitat, are have allowed P. hastata and E. 107 capitata to become sympatric here. populations The origins determining population at Creek population, only a tetraploid count It may be best, in seeking to describe the obtained. was significance in a mixture of diploids and tetraploids, while is Callahan the km presence of tetraploid The other. approximately at both of these sites may therefore be related (i.e., hybridization event polyploidization single a each within occurring sympatry, of individuals to two sites, to realize that they are these of neighboring 0.8-1.0 in The Elk Creek individuals. such of involved problems the of examples further are of P. capitata containing tetraploids followed by chromosome doubling) in the past. the On basis diploid other appears This heterophvlla. P. with species is E. capitata, occurring along highway both north and south of the serpentine outcrops In surveys throughout the vicinity, no Creek. above Elk other morphologically were found. individuals undisturbed suggests Highly of E. at possibly variable, capitata are this not tetraploid restricted location, but to occur areas above the road as well. serpentine the in This that the possible hybridization event which led to polyploidization habitat distinct members of the Magellanicae roadbank disturbed the these intergradient populations for sympatric neighboringly the parent be to field observations, the most likely of disturbance of in E. the capitata area, occurred prior to the tetraploids having 108 since colonized the roadbank. With respect E, capitata of species conclusions have from hybridization in the resulted either a species similar to P. egena or with the with Long-term introgression sympatric E. beterophvlla. involving have It is feasible that the lines. and leaf lobing patterns of E. capitata at could Creek locally these along corollas larger past its morphological characteristics are too poorly known at this time to allow any strong here, Elk The occurrence of this egena. P. and Oregon, in population, Heckard have interacted with a Californian member may Magellanicae, the Creek Elk collection (2930, JEPS) the possibility that his on noted the to tetraploids the and diploids at this site may so obscured the hybrid nature of the plants that exact determination lengthy field origin their of and artificial is not possible. Only hybridization studies could perhaps offer answers to the origin of these populations. Figure P. capitata complications the 9 summarizes the possible interrelationships of within patterns of cytological Magellanicae. In the face of the discussed above, it is appealing to interpret occurrence "evolutionary the of limited noise," when interspecific gene exchange as one considers the distinctive morphological, ecological, and almost complete isolation southwestern Oregon. of E. capitata within the complex in 109 Figure 9. TETRAPLOID Possible interrelationships of Phacelia capitata within the species group Magellanicae. 1 egena 1 (n=22) DIPLOID (n=11) heterophylla ssp. virgata I-- 110 The endemic status o Although and discussions of area which taxa which Neoendemics evolved in a particular have Paleoendemics Major, 1965; Stott, 1981). former range (Stebbins and third described, category, which is above. extremes. area, (Richardson, the is holoendemic, has also been intermediate in character between the Such species, even when covering its a still narrowly restricted, most often by physical of or physiological (ecological) barriers characteristics The 1978). categories endemic formed the basis for have not yet spread. they types, extreme confined to a very limited portion of their now are virtue two which once possessed much wider distributions but taxa maximum have species has endemic endemism since the late 1800s. new A of complex, paleoendemics, represent which very been neoendemics are classification the traditionally from the species with of three the respect to various attributes are summarized by Richardson as follows: Endemic status Neo- Holo- Paleo- Attribute taxonomically isolated geographically isolated polymorphic derived characters environment stable ploidy level high potential to expand area age (recent +, old -) As he makes attribute clear, + +/+/- +/+/+/are - - +? + + + there E. capitata + + - + + + - many (+)/- (+)/- exceptions + - +? + to the states, and there is no clear distinction between the three types. 111 is geographically and ecologically Since E. capitata restricted to a small area, and surely did not once occupy a much larger a neo- former its seems factors which might prevent other or south of its present areas unless there are unknown soil logical, climatic, The seems to most closely fit the it serpentine the to distribution this. It is, The possibility of E. capitata expanding category. range a paleoendemic. not is holoendemic; or chemical, it clear if the species is best characterized as not however, area, soils occupied by E. capitata and E. analyzed corymbosa are quite similar in chemical nature. According localized generalized must specialized in North western be Stebbins to America, distribution recent of habitats This, origin. such as vegetation, surrounding the in regions in are close relatives of more have and which endemics (1971), too, supports the classification of E. capitata as a neoendemic. Thus, the most reasonable interpretation of Phacelia capitata may be to classify it as an insular neoendemic. Recommendations for preservation Endemic taxa such as Phacelia capitata are of great As a consequence of their scientific interest restricted geographical distributions and often specialized ecological requirements, become endangered destruction. and as a value. they are also most likely to result of habitat modification or It is imperative that such species be studied 112 and documented before they are lost to extinction. 'hacelia serpentine increase at time it is not appropriate to consider the species with was placed and endangered of Natural been Thus, at least abundance of the species. the "endangered" list paradoxical case for a a endemic, since habitat disturbance has led to an in this presents capitata Most recently, the species extinction. on the "watch list" of Oregon rare, threatened, taxa plant threatened Heritage after being removed from a species, Data throughout Base, range (Oregon their In addition, it has 1985). recommended for removal from candidacy as a threatened species under the Federal Endangered Species Act of 1973. The the geographically species capitata can on provide does a narrow a "watch list." restriction of serpentine basis for maintaining E. In this way, its populations be monitored to detect any decline in numbers which may occur with increased further mining tendency in disturbed for still important, population the serpentine substrates. of allowed populations destruction of its habitat, such as by areas has, for the Its weedy time the persistence of this unique species. of such ecology, however, species, to so protect that being, It is undisturbed features of their habitat requirements, and evolutionary uniqueness may be studied in natural settings. 113 BIBLIOGRAPHY Alt, D. and Hyndman, D. W. 1978. Roadside geology of D. Oregon. Mountain Press, Missoula, Montana. 268 p. Anderson, E. 1949. Introgressive hybridization. John Wiley and Sons, New York. 109 p. Atzet, T. and D. L. Wheeler. 1984. Preliminary plant Siskiyou Mountain Province. associations of the U.S.D.A. Forest Service, Pacific Northwest Region, Siskiyou National Forest. 315 p. Baillon, H. E. 1890. 10: 397-402. Hydrophylleae. Histoire des plantes W. H. 1956. Plants of Iron Mountain, Rogue River Range, Oregon. Amer. Midl. Nat. 56: 1-53. Baker, Baldwin, E. M. 1976. Geology of Oregon. Publishing Co., Dubuque, Iowa. 147 p. Kendall/Hunt Bentham, G. 1837. Review of the order of Hydrophylleae. Trans. Linn. Soc. 17: 267-282. Berg, and E. H. Gardner. 1978. Methods of soil M. G. analysis used in the Soil Testing Laboratory at Oregon State University. Agricultural Experiment Station, O.S.U., Corvallis. Special Report No. 321. D. F. 1977. Application of numerical classification ecological investigations of water pollution. U.S. Environmental Protection Agency, Corvallis, Oregon. Ecological Research Series Report No. EPA- 600/3 -77033. 115 p. Boesch, in Brand, A. 1913. Hydrophyllaceae. Pflanzenreich IV, 25: 1-210. In: Engler, A. Das Brown, R. 1818. Observations, systematical and geoSmith's Christian graphical, on Professor collection of plants from the vicinity of the river Tuckey, J. K. Narrative of an expedition Congo. In: to explore the river Zaire, p. 420-485. Candolle, A. P. de. 1845. Hydrophyllaceae. In: 9: 287-301, 564-565. Cave, M. S. DC., Prod. and L. Constance. 1942. Chromosome numbers in Publ. Bot. 18: Univ. Calif. the Hydrophyllaceae. 205-216. 114 Cave, Cave, M. S. and L. Constance. 1944. Chromosome numbers in II. Univ. Calif. Publ. Bot. 18: the Hydrophyllaceae: 293-298. M. S. and L. Constance. 1947. Chromosome numbers in III. Univ. Calif. Publ. Bot. 18: the Hydrophyllaceae: 449-465. Cave, M. S. and L. Constance. 1950. Chromosome numbers in Univ. Calif. Publ. Bot. 23: IV. the Hydrophyllaceae: 363-382. Cave, and L. Constance. 1959. Chromosome numbers in M. S. the Hydrophyllaceae: V. Univ. Calif. Publ. Bot. 30: 233-258. Choisy, J. D. 1833. Description des Hydroleacees. Mem. Soc. Phys. Geneve 6: 95-122. Cohen, A. L., D. P. Marlow, and G. E. Garner. 1968. A rapid critical-point method using fluorocarbons (Freons) as intermediate and transitional fluids. Journ. Microsc. 7: 331-342. Constance, L. 1963. Chromosome numbers and classification in Hydrophyllaceae. Brittonia 15: 273-285. Constance, L. and T. I. Chuang. 1982. SEM survey of pollen Hydrophyllaceae in morphology and classification (Waterleaf Family). Amer. J. Bot. 69: 40-53. Cronquist, A. 1981. An integrated system of classification of flowering plants. Columbia Univ. Press, New York. 1262 p. J. L. Cronquist, A., A. H. Holmgren, N. H. Holmgren, and Reveal P. K. Holmgren. 1984. Intermountain flora--vascular plants of the intermountain West, U.S.A. Volume 4. The New York Botanical Garden, Bronx. 573 p. method Daubenmire, R. canopy-coverage 1959. A vegetational analysis. Northwest Sci. 33: 43-64. of Detling, L. E. 1968. Historical background of the flora of the Pacific Northwest. Bull. No. 13, Museum of Nat. Hist., Univ. of Oregon, Eugene. 57 p. Faegri, K. and L. van der Pijl. 1971. The principles of pollination ecology. Pergamon Press, Oxford. 291 p. Gray, conspectus of the North American Hydrophyllaceae. Proc. Amer. Acad. 10: 312-332. A. 1875. A 115 Grime, J. 1979. Plant strategies and vegetation P. processes. John Wiley and Sons, Chichester. 222 p. 1956. The subspecies of Phacelia nemoralis Heckard, R. L. Greene. Leafl. West. Bot. 8: 29-32. Heckard, L. R. magellanica 32: 1-126. Taxonomic studies in the Phacelia polyploid complex. Univ. Calif. Publ. Bot. 1960. Hitchcock, C. L. and A. Cronquist. 1973. Flora of the Pacific Northwest. Univ. Washington Press, Seattle. 730 p. C. 1976. Evolution and systematic significance Jackson, R. of polyploidy. Ann. Rev. Ecol. Syst. 7: 209-234. Jussieu, A. L. de. 1789. Genera plantarum 129. Reniston, J. A. 1978. Program CLUSTER: an aid to numerical Marine University, State classification. Oregon Science Center, Newport (unpublished). 32 p. Intraspecific variability in the Rruckeberg, A. R. 1951. response of certain native plant species to serpentine soil. Amer. J. Bot. 38: 408-419. A. R. 1954. The ecology of serpentine soils. Plant species in relation to serpentine soils. Ecology 35: 267-274. Kruckeberg, III. 1956. Notes on the phacelia magellanica Kruckeberg, A. R. complex in the Pacific Northwest. Madrono 13: 209-221. Kruckeberg, A. R. 1969. Plant life on serpentinite and rocks in northwestern North ferromagnesian other America. Syesis 2: 15-114. McKee, B. 1972. Cascadia--the geologic evolution of the Pacific Northwest. McGraw-Hill Book Company, New York. 394 p. Moldenke, A. R. 1976. California pollination ecology and vegetation types. Phytologia 34: 305-361. Munz, A California Press, Berkeley. 1681 p. P. A. 1959. flora. Univ. California Oceanic and Atmospheric Administration. 1982. National Monthly normals of temperature, precipitation, and heating and cooling degree days 1951-80, Oregon. Climatography of the United States, Publ. No. 81. 18 p. 116 Oregon Natural Heritage Data Base. 1985. Rare, threatened and endangered plants and animals of Oregon. The Nature Conservancy, Portland. 31 p. Proctor, J. and S. R. J. Woodell. 1975. The ecology of serpentine soils. Adv. Ecol. Res. 9: 255-366. Quick, C. R. 9: 17-20. 1947. Germination of phacelia seeds. Madrono I. B. K. 1978. Endemic taxa and the taxonomist. Essays in plant taxonomy. Academic Press, London. p. 245-262. Richardson, In: Snow, 1963. Alcoholic hydrochloric acid-carmine as a stain for chromosomes in squash preparations. Stain R. Tech. 38: 9-13. Stebbins, G. L. 1942a. The genetic approach to problems of rare and endemic species. Madrono 6: 241-258. Stebbins, G. L. ecology and 36-45. 1942b. Polyploid complexes in relation to the history of floras. Amer. Nat. 76: Stebbins, G. L. 1971. Chromosomal evolution plants. Edward Arnold Ltd., London. 216 p. in higher Stebbins, G. L. and J. Major. 1965. Endemism and speciation in the California flora. Ecol. Monogr. 35: 1-35. P. 1981. Interpreting endemic distribution patterns. Historical plant geography. George Allen and Unwin, London. 151 p. Stott, In: Thorne, R. F. 1976. A phylogenetic classification of the angiosperms. Evol. Biol. 9: 35-106. N. 1982. Nutrient limitations to plant growth S. a California serpentine grassland. Amer. Midl. Nat. 107: 95-99. Turitzin, in U.S. Dept. of Agriculture, Weather Bureau. 1930. Climatic summary of the United States, from establishment of stations to 1930. Walker, R. B. 1954. The ecology of serpentine soils. II. Factors affecting plant growth on serpentine soils. Ecology 35: 259-266. Wells, F. G. and D. L. Peck. 1961. Geologic map of Oregon west of the 121st meridian. U.S. Geol. Survey Misc. Geol. Inv. Map 1-325. 117 1971. Vegetation-soil chemistry correlations D. serpentine ecosystems. Ph.D. thesis. University of Oregon, Eugene. 274 p. White, C. in Whittaker, R. H. 1954. The ecology of serpentine soils. IV. The vegetational response to serpentine soils. Ecology 35: 275-288. R. Whittaker, Mountains, 279-338. 1960. Oregon and H. Vegetation California. the Siskiyou Ecol. Monogr. 30: of Whittaker, R. H. 1961. Vegetation history of the Pacific Coast States and the "central" significance of the Klamath region. Madrono 16: 5-22. Willis, J. C. 1895. Fertilization of Phacelia. Jour. Linn. Soc. Bot. 30: 53-60. APPENDIX 118 APPENDIX cited Specimens addition (in to those discussed in the Taxonomic Relationships section): Hwy. isotype), m, OREGON: Coos Co.: N of St. km E of Bridge, 100 m, Kruckeberq 2702 (WTU, 3.2 42, Douglas Kruckeberg. capitata Fhacelia $undberg 1031 (OSC), Shelly k Nelson 754 (OSC). Beatty Creek, 0.4 km N of Cow Creek Road, 320 Co.: Shelly 419 (OSC); BLM rd. 29-5-11.0, 2.3 km NW of Bilger Creek Road, Road (BLM m, Shelly 0.5 km from 8.1 km SW of Interstate Hwy. 5, k Holmes 766, al (OSC); BLM rd. 30-6-4.1, BLM rd. 29-6-34.0, 850 m, Shelly £ Holmes 781 Creek, Umpqua NF rd. 3230, 0.5 km W of St. Callahan 227, Hwy. 30-6-4.2), rd. 760 (OSC); m, $helly k Nelson 724 (OSC); Boomer Hill 490 490 Fosback m, s.n,. (OSC), Shelly 440 (OSC), & Nelson 733 (OSC); Cow Creek Road, 0.5 km W of Doe $helly Road, Creek 300 m, Shelly 405, 70n (OSC); NW side of Cow km SW of Doe Creek Road, 300 m, $helly 411, 412 Creek, 1.2 (OSC); Cow Creek Road, 4.8 km W of Glenbrook Loop Road, 230 m, $helly of Doe Creek Road, 640 m, Shelly k Holmes 783 (OSC); NW The Drew, 2930 m, $helly (JEPS), Creek, (OSC), 270 810 Shelly k Nelson 734 (OSC); Elk Creek, St. Hwy. 227, 1.6-4.0 km SE of Tiller, 375 m, Beckard along Lee 719 (OSC); Doe Creek Road, 1.9 km N of Cow Road, 400 m, Shelly 706 (OSC); BLM rd. 30-7-14.0, 3.2 Creek km 681, m, BLM Shelly parker rd. 433 (OSC), Shelly k Nelson /21 (OSC); 28-4-15.0, 415 m, Fosback It al. s.n. 399 (OSC); Little River, 5.9 km SE of Glide, k Parker s.n. (OSC), Shelly 111, 1Q. (OSC); 119 14, 1.6 km NW of Myrtle Creek, 245 m, Howell 28807 Rt. Co. 721 (OSC); Nickel Mountain, ca. 365 m, Shelly 414, Detling 6339 (ORE); 295 Knouse (CAS), m, flank 736 Shelly (OSC), Shelly 354 (OSC); SW p.n. BLM rd. 29-7-36.0, 710 m, Creek, Rice (OSC); 792 (OSC); Salt Creek, BLM rd. 30-7-36.0, Holmes m, Shelly & Nelson 725, 725b (OSC); Weaver Road, 2.9 km 340 SW k Moore Red Mountain, 930 m, Detling 5314 (ORE), Shelly k of Nelson rd. 26-3-34.2, 0.5 km E of Peel, BLM of Interstate Hwy. 5, 210 m, Ornduff s.n. (JEPS), Shelly Goolaway Gap, 930 m, Shelly Jackson Co.: (OSC). 720 689, & Nelson 737, 738 (OSC). P. Road, Jackson Co.: OREGON: Creek Slate of Co.: Grave Creek Road, 850 m, Shelly 441, Eight Dollar Mountain, Siskiyou NF at Illinois River, 390 m, Shelly 700 (OSC); Eight 4201 Mountain, Deer Creek, 5 km W of Selma, 450 m, Shelly Rough (OSC); Cave N Josephine Dollar 702 km 3.5 (OSC). rd. Jepson. corvmbosa and Ready Botanical Wayside, 8 km S of 425 m, Shelly 698 (OSC); Waldo, 4.2 km E of Junction, O'Brien, 475 m, Shelly 697 (OSC); Wimer Road, 6.5 km SW of O'Brien, 490 m, Constance & Bacicialupi 3394 (WTU), Shelly 691 (OSC). E. Douglas hastata Boomer Hill Interstate (OSC). Hwy. Road 5, ex Lehmann. (BLM 760 rd. m, Douglas Co.: OREGON: 30-6-4.2), 8.1 km SW of Shelly & Holmes 767, 768, 771 120 P. heterophylla Douglas OREGON: Road, 227, (OSC). 300 0.8 Co.: Shelly m, km Pursh E ssp. yin:rata (Greene) Heckard. Cow Creek, 0.1 km W of Doe Creek 714 (OSC); Elk Creek, along St. Hwy. of Tiller, 350 m, Shelly k Nelson 727, 728