..........';v.. .....Z... ......3... ...... .......,.x. .....v.._l.._... ,.,....:..V £12.... .....TA .. .. If. .... ..r ....x... .11.... ......z.. . ...... .... 3.5.... ...v..:.. . ....:....._. N: vmm Io D . . . . g. . R I. .. , ; ....N.._.._,m.. . , . . , .. . a...“ T. .... _ . . ._ N. ...-M M . , . ... m .W . . M . m.” 0-169 This is to certify that the thesis entitled TAXONOICY, l-TOI‘TENCLATURE, AND VARIATION WIT—TIN THE PINUS FLEXILIS COICPLEX presented by Raphael John Steinhoff has been accepted towards fulfillment of the requirements for L degree in My Major professor Datemyé %’ \MMWMM 4 L! B RA R Y Michigan State University ABSTRACT TAXONOMY, NOMENCLATURE, AND VARIATION WITHIN THE PINUS FLEXILIS COMPLEX by Raphael John Steinhoff The Pinus flexilis complex is composed of two populations of 5- leaved.pines of the subgenus Haploxylon of the genus Pinus. The northern population, usually known as Pinus flexilis James, has a range from southern Alberta to northern New Mexico. The southern population ranges from southern Colorado into northern Mexico. This population is referred to as Pinus strobiformis Engelmd, Pinus flexilis var. reflexa Engelma, Pinus reflexa Engelm., or Pinus ayacahuite var. brachyptera Shaw. The primary purposes of the study were to evaluate the extent of differences between the taxa and to evaluate the variation within each. The results were to be used to attempt clarification of the nomenclature and classification of the taxa. Materials for the study were collected in 1959 and 1960 from 61 native stands in the mountains of the western United States and.A1berta, Canada. Cones, seed, and a single branch of foliage were collected from each tree. The cones and foliage were measured and scored for several characteristics. The seed were planted in a replicated nursery test in 1961. Observations and measurements were made on the resulting seedlings during their first two years of growth. Distinct differences between the taxa were exhibited in the seedling test. Cotyledon number, length of secondary leaves, and height growth were the most satisfactory characters for distinguishing between the taxa. Traits measured on the cone and foliage specimens from.the parental trees exhibited less distinctive differences between taxa. Secondary leaf length was the most reliable parental character for separating the taxa. Other traits which served to distinguish the taxa were: (1) seed weight, (2) number of rows of stomata on the dorsal leaf surface, (3) length of cones, and (h) degree of cone scale reflexing. Raphael JOhn Steinhoff It was concluded from the results of the study that the taxa deserve separate specific rank. According to the rules of nomenclatural priority the proper name for the northern species is Einus flexilis James. The proper name for the southern species is Einug strobiformis Engelm. The patterns of variation in the regions where the species' ranges are contiguous or sympatric indicate that hybridization has occurred in the past and may still be taking place. Within the northern species, B. flexilis, the population structure had three principle characteristics. First, there was very little variation in either seedlings or parental specimens from that portion of the range extending from Alberta to central Colorado. Second, seedlings of southern origins grew faster and the cones from the parents were longer than those of northern origin. Third, seedlings from three restricted and isolated areas performed like those of the southernmost origins. These areas were in east-central Idaho, near Pine Bluff, wyoming, and in Douglas County, Colorado. Variation within 3. strobiformis was more random than in E. flexilis. Seedlings from the northernmost origins grew more slowly than those of more southern sources. The parental specimens from northern Arizona trees had shorter leaves, smaller cones, and less cone scale reflexing than those from central and southern Arizona. Both seedling and parental leaves of New Mexico and Texas sources were shorter and less serrulate than those from central and southern Arizona. Significant differences between progenies within one or more stands were found for all 13 seedling traits investigated. In some of the northern areas within—stand differences were almost as large as between-stand differences. TAXONOMX, NOMENCLATURE, AND VARIATION WITHIN THE PINUS FLEXILIS COMPLEX By Raphael John Steinhoff A THESIS Submitted to thhigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 196M ACKNOWLEDGMENTS Financial support for various phases of the study has been provided by the.American Museum of Natural History, Michigan State University, and the National Science Foundation. Special thanks are due to the many co-operators who have collected materials used in the study and to the staff of the Bogue Experimental Nursery for assistance in processing the specimens and conducting the nursery.test. Dr. W. B. Critchfield of the Pacific Southwest Forest and Range Experiment Station has generously supplied supplemental information for the species' range map. The author wishes to express his appreciation to the members of the guidance committee for their continuing assistance and encouragement throughout the study. ii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . PURPOSES . . . . . . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . . . Taxonomy and Synonymy" . . . . . Experimental Study of Variation and Speciation History of Provenance Testing in Forest Trees . . . Geographic Variation in Rocky Mountain Conifers . . . . Ponderosa pine (Pinus ponderosa Laws.) . . Douglas- fir [ Pseudotauga menziesii (Mirb. ) Franco] Lodgepole pine—(Pinus contorta Dougl. ) . . White fir [Abies concolor (Cord. and Glend.) Hoopes] Geographic variation in the White Pines . . . . . . Eastern white pine (Pinus strdbus L.) . . Western white pine (Pinus monticola Dougl.) Limber pine (Pinus flexilis James) MATERIALS AND METHODS . . . . . . . . . Materials . . . . . . . . . . . . . Acquisition . . . . . . . . . . Handling . . . . . . . . . . . thhods . . . Design and installation of the nursery test . Measurements of seedling traits . . . . . . Measurement of mature traits on herbarium.specimens . Statistical analysis . . . . . GENERAL NURSERY OBSERVATIONS . . . . . . . . . DIFFERENCES BETWEEN THE TAXA . . . . . . . . . . Seedling Differences . . . . . . . . . . . Individual characters . . . Simultaneous consideration of several characters . . Parental Differences . . . . . . . . . . . . . Individual characters . . Simultaneous consideration of several characters . DIFFERENCES BETWEEN STANDS WITHIN SPECIES . . . . . . . Pinus flexilis . . . . . . . . . . . . . . Seedling characters . . Parental characters . . . . . . . . . Pinus strdbiformis . . . . . . . . . . . Seedling characters . . . . . . . . . . . Parental characters . . . . DIFFERENCES BETWEEN PROGENIES WITHIN STANDS . iii Page vi ANALYSES OF CORRELATIONS BETWEEN CHARACTERS . . Correlations Between Seedling Characters . Correlations Between Parental Characters . Correlations Between Seedling and Parental Characters Correlations Between Seedling Characters and Origin Data . . . . . . . Correlations Between Parental Characters and Origin Data . . . . . . . . . DISCUSSION . . . . . . . . . . . The Question of Distinct Species . . . . Variation Within Pinus flexilis . . . . Variation Within Pinus strobiformis . . Correlations Among Characters . . . . . SUMMARY AND CONCLUSIONS . . . . . . . . . LITERATURE CITED . . . . . . . . . . . iv Geographic Geographic l’. Table l. 10. ll. 12. 13. 11+. 15. 16. 17. LIST OF TABLES Comparison of descriptions assigned to members of the Pinus flexilis complex by various authors. . . . . Two-year growth data for Pinus flexilis progenies, summarized by stand-progeny . . . . . . . . . . Two-year growth data for Pinus strobiformis progenies, summarized by stand-progeny . . . . . . . . . . Data for adult characteristics of Pinus flexilis trees, summarized by stands. . . . . . . . . . . Data for adult characteristics of Pinus strobiformis trees, summarized by stands. . . . . . . . . . . . . Examples of hypothetical, idealized, summation of differences tables . . . . . . . . . . . . . Results of analysis of variance tests of differences between the southernmost Pinus flexilis and northernmost Pinus strobiformis seedlings. . . . . . Summation of differences for seedling characteristics of Pinus flexilis and one Pinus strobiformis Stand’progenies o o o o o o o o o o o o o o Summation of differences for seedling characteristics of Pinus strobiformis and two Pinus flexilis stand-progenies . . . . . . . . . . . . . . Summation of differences for adult characteristics of Pinus flexilis and one Pinus strdbiformis stand-progenies . . . . . . . . . . . . . Summation of differences for adult characteristics of Pinus strObiformis and two Pinus flexilis stand-progenies . . . . . . . . . . . . . . Differences between progenies within stands for eight selected characters . . . . . . . . . . . . . Correlations between seedling characters of Pinus flexilis and Pinus strdbiformis . . . . . . . . . . . . Correlations between parental characters of Pinus flexilis and Pinus strObiformis . . . . . . . . . . . Correlations between parental and seedling characters of Pinus flexilis and Pinus strcbiformis . . . . . Correlations between seedling characters and geographic origin data for Pinus flexilis and Pinus strobiformis. . Correlations between parental characters and geographic origin data for Pinus flexilis and Pinus strobiformis. . Page 11 25 26 29 3o 35 1+2 1+3 1.6 A7 55 59 62 61+ 66 66 LIST OF FIGURES Figure Page 1. Distribution of members of the Pinus flexilis complex. . . . . . . . . . . . . . . . . 2 2. Examples of naturally occurring stands of members of the Pinus flexilis complex. . . . . . . . . . . . A 3. Location of stands sampled for this study . . . . . . 23 h. Portion of a nursery bed illustrating differences among seedling progenies . . . . . . . . . . . . . 38 5. Mean values for seedling characteristics of Pinus flexilis and Pinus strobiformis stands grouped by area of origin . . . . . . . . . . . . . . . . 39 6. Mean values for adult characteristics of Pinus flexilis and Pinus strdbiformis stands grouped.by area of origin . . . . ‘. . . . . . . . . . . . . A5 7. Boundaries of area of origin groupings of stand collections . . . . . . . . . . . . . . . 53 8. Hypothetical scatter diagram illustrating correlation patterns . . . . . . . . . . . . . . . . 60 vi INTRODUCTION This paper is concerned with members of the Pinus flexilis complex which occur in western North America. Taxa in the complex belong Within the group Flexiles, sensu Shaw (1914:2A-28), of subsection Cembra of the Haploxylon or soft pines. Members of this subgenus are characterized by non-decurrent leaf-fascicle bracts and a single fibro-vascular leaf-bundle. The subsection Cembra contains those members of the Haploxylon with terminal cone-scale umbos. Members of the group Flexiles of the subsection Cembra have Wingless seeds and dehiscent cones. The two taxa under consideration will be referred to as species and the specific epithets first publiShed for them will be used. The northern species, Pinus flexilis James, or limber pine, has been described as follows: Leaves in five-leaved clusters, thick, rigid, 35 to 75 millimeters in length with several rows of stomata on the dorsal surface, cones 75 to 250 millimeters long, scales rounded or pointed at the apex; tree 13 to 16 meters in height with a short, massive trunk 0.6 to 1.2 meters or more in diameter (Sargeant 1897, XI:35-37). Some other authors consider the maximum cone length to be considerably less,e.g. Engelmann (1863) 110 milli- meters. The species is found from southwestern.Alberta, south along the Rocky Mbuntains and related chains, to northern New Mexico and Arizona. It also occurs from Nevada westward to the Sierra Nevada Range with western outposts in southwestern California and the Wallowa Mountains in Oregon. At the eastern extreme it is found in isolated stands in western North and South Dakota (Figure l). The southern species, Pinus strobiformis Engelm., or Mexican border white pine, is found in the mountains of extreme southern Colorado, Arizona, New Mexico, Texas, and northern Mexico (Figure 1). Sargent (1897, XI:33-3h) gave the following description for the species: Leaves in clusters of five, slender, from.85 to 100 milli- meters in length, usually without stomata on the dorsal leaf surface, leaves sharply serrate or entire; cones 125 to 225 millimeters long, their scales thin and reflexed; tree 26 to 32 meters in height with diameters ranging to 0.6 meters. \\ . e . 1 . \‘\\\\\\W« M. \\ xx,” 00 I” ‘9”9’ :oMMOI ./ 9.. V/% mm : Km... I . E 1 R A m m a .1 F 3 In the days before railroad transportation, P. flexilis was the primary lumber source for the desert regions of Nevada (Jepson 1910:75). The tree also has value as protection for watershed lands and its windswept forms add beauty to the mountain scene. Its seeds are large and edible. They were gathered by Indians and trappers for food (James 1823;II:3A). The nutritious seeds are also an important item in the diet of many rodents (Hatt 191(3). Rodents are the prime agents in seed dispersal (Eggler l9h1). 'P. flexilis has been recommended for shelterbelt planting in wyoming and western Nebraska (U. S. Dept. of Agriculture l9h9z8h8). Pinus strobiformis, on the other hand, grows on less exposed sites than does P. flexilis and, under mesic conditions, develops to a size and quality that is quite satisfactory for producing lumber. A trial planting of 150,000 seedlings is currently being grown in central Colorado for use in reforestation efforts (Milodragovich, R. R. 1963. Personal communication to Dr. J. W..Andresen, Dept. of Forestry, Michigan State University). Examples of typical members of the two species and the asso- ciations in which they occur are presented in Figure 2.. .onmEoo mflaflxoam mdqfim map mo whopaoa mo m@dem mnflhhdooo aaadhdedm mo moagadxm .N madmflm .mpopoe oamm "coapw>oam .wpopofi mema ”soapn>oam .nnouflp< aoHHH>powchmm ado: mHEMOMHponpm mquL .wqopsoz aohz new: mHHflXoHM mdsflm PURPOSES Two primary purposes prompted the study: first, a desire to clarify the correct taxonomic rank of the two taxa through an analysis of the differences between the two as well as the variation within each; second, to recOmmend proper specific epithets which are consistent with good usage and accepted nomenclatural rules. The study also has long-range objectives. Among them are: (l) The establishment of plantations containing materials of known origin for continuing anatomical, morphological, and physiological studies of the species, (2) Correlation of juvenile with mature performance, (3) Production and evaluation of hybrids between members of the Pinus flexilis complex and other 5—leaved soft pines, and (4) Evaluation of the timber potentialities of the species in the Midwest. LITERATURE REVIEW Taxonomy and Synonymy. The discovery and naming of Pinus flexilis by James (1823,II:3M—35) initiated a series of nomenclatural controversies and confusions which have persisted until the present day. Unfortunately, James did not collect specimens and.his later description was based only on his field notes. Engelmann (1863) published a Latin description of P. flexilis and attributed the inconsistencies in the account of James to the inclusion of Observations on another 5-leaved species of pine, which was later named Pinus aristata Engelm..‘ This assumption might explain James's description of the cones as being erect and smaller than those of Pinus rigida Mill. However, the prominent arming of the cones of P. aristata and James's note that those of .P. flexilis are unarmed made such confusion seem unlikely. Nuttall (1853,III:107) attempted to elaborate on the earlier description by James, but his text and poor figure did little to clarify the situation. .A questionable nomenclatural addition placed in the synonymy'by some authors (e.g. Sudworth 1897:16) was made by Hooker (1838,II:161) in Flora Boreali-Americana. He listed a variety of sugar pine, Pinus lambertiana var. E Hooker, which may have referred to either P. flexilis or P. albicaulis Engelm. The collection upon which the varietal description was based was made by Drummond (1830) in Canada while portaging from the Red Deer River to the Columbia River at "Height of Land”. Drummond's description of the foliage would fit either P. flexilis or P. albicaulis but the cones which he observed had been attacked by rodents or birds. Because the cones are necessary for distinguishing between these species no positive conclusions can be drawn from his brief remarks about the species of pine represented. Endlicher (18A7zl50) questioned Hooker's interpretation and altered the varietal epithet to P. lambertiana var. .B. brevifolia in his synopsis of the conifers. As late as 1855, Carriere (1855:392), in Traite/ ge’ne’ral desConiferes, attributed authorship of P. flexilis to Wislizenus. This designation was based on a note by Engelmann (18A8) in his description of the collections of Wislizenus regarding a specimen sent to him.by A.Fend1er from Sante Fe, New Mexico. Because Fendler was unable to 6 7 collect at high elevations it is questionable if his specimen was _P. flexilis, but rather P. strobiformis. Although not pertinent to the question of similarity or difference between P. flexilis and P. strobiformis, Rydberg (1905) proposed that P. flexilis should be called.Apinus flexilis (James) Rydb. to conform to the nomenclature of the classification system of Necker (1790,III:269). The initial description of Pinus strobiformis was published.by Engelmann (18A8) when he described.material collected by Wislizenus in northern Mexico near Cosihuiriachi, Chihuahua. However, the specific epithet does not appear in Engelmann's later works (1878, 1880, and 1882). Shaw (1909 ll) attributes this ommision to the fact that Engelmann did not learn of Ehrenberg's (1838) description of Pinus ayacahuite until after l8h8 and that Engelmann then considered what he had named and described as P. strobiformis as synonymous with .P. ayacahuite. The range ascribed by Parlatore (1868, XVI,pt.II:H06-AO7) to P. ayacahuite Ehrenb. includes some areas where P. strdbiformis is found. He also considered the two species as synonymous. After abandoning P. strobiformis as the name for the trees found in northern Mexico, Engelmann (1878), in describing specimens col- lected by Wheeler's Expedition in Arizona, assigned varietal rank under the species Pinus flexilis to various forms which display some of the characteristics he attributed to P. strobiformis. These characteristics included serrulation of leaves, reduction in number or lack of stomatal rows on the dorsal leaf surface, elongation of cones, and elongation and reflexion of cone scales. The varieties were designated as: var.gZ -serru1ata- referring specifically to the serrulate leaves, var. S -macrocarpa- cones enlarged, and var. K -reflexa- cone scale apophyses elongated and reflexed. In the "Revision of the Genus Pinus” Engelmann (1880) did not mention the varietal forms of ‘P. flexilis. However, he soon (Engelmann 1882) proposed raising the variety reflexa to specific rank as Pinus reflexa. Based on their analysis of leaf anatomy, Coulter and Rose (1886) considered_P. flexilis and P. reflexa to be distinct at the species level. 8 After examining specimens collected by Pringle in 1887 in the same area Where Wislizenus collected the specimen described by Engelmann as P. strdbiformis, Sargent (1889) concluded that P. reflexa anqu. strobiformis were identical. He suggested that P. strobiformis was probably only a northern form Of.E' ayacahuite with short leaves and small cones. Shortly afterward, Lemmon (1892z3) used the epithet P. ayacahuite var. strdbiformis Sargent, to refer to the "Arizona white pine" even though Sargent had not proposed the varietal designation. Lemmon also specifically mentioned '3. reflexa as being synonymous with P. ayacahuite var. strdbiformis. As part of the synonymy for P. flexilis var. reflexa, Shaw (1909:12) listed, in Pines 9f Mexico, P. ayacahuite var. strdbiformis Lemmon and cited the above article as the reference. The nomenclatural confusion was further increased When Sargent (1897, XI:33-3h) in Silva of North America assigned.specific rank to Pinus strdbiformis and cited Engelmann as the authority. This was a complete reversal of his 1889 opinion that P._strdbiformis was only a form of P. ayacahuite. Sudworth (1897:17) accepted and concurred with Sargent's 1897 position and, in his Nomenclature 2f Arborescent Flora, listed P. strdbiformis Engelm., for which he listed as synonyms P. flexilis var. reflexa Engelma, P. reflexa Engelm., and P. ayacahuite var. strobiformis Lemmon. Two of Engelmann's varieties of P. flexilis were separated by Sudworth (1897:16) from P. strobiformis. Pinus flexilis var. serrulata Engelm. was placed as a synonym under P. flexilis; P. flexilis var. macrocarpa Engelm. was placed by itself as P. flexilis megalocarpa. The synonymy was further expanded when Voss (1907) considered that P. reflexa Engelm. should be designated as P. ayacahuite var. reflexa Voss. In his treatmenth. strobiformis Engelmd was placed as a synonym of P. ayacahuite Ehrenb. When Shaw (1909) authored The Pines of Mexico, he held that 'P. strdbiformis Engelm. was synonymous with the northern-form of _P. ayacahuite which he then designated as a new variety, namely Pinus 9 ayacahuite var. brachyptera. Pinus flexilis var. reflexa was retained as a variety Of.§° flexilis with P. ayacahuite var. strobiformis Lemmon and P. strdbiformis sensu Sudworth and Sargent as synonymy. Shaw evidently did not feel that the P. strObiformis of Engelmann, Sudworth, and Sargent was the same tree. By the time The Genus Pinus was published (Shaw lglu), the controversy had completed a full circle. In that publication Shaw considered P. strdbiformis Engelm. as synonymous with P. ayacahuite and all other types previously mentioned as being only forms of _P. flexilis. Onlqu. reflexa Engelm. and P. strobiformis Sargent are mentioned in the synonymy Of.§° flexilis. .After the publication of Shaw's The Genus Pinus in 191H, Sudworth (1917:12-13) pointed out thath. strobiformis, P. reflexa, and P. ayacahuite var. brachyptera all referred to the same tree. He called attention to the fact that the name strdbiformis was the oldest and thus implied the idea of priority in his preference for the continued use of P. strobiformis to refer to the species. However, Sudworth’s views seem to have found little favor until very recently, for P. reflexa or P. flexilis var. reflexa are the names most commonly seen in publications from the intervening years. Sargent (1922), in the second edition of the Manual 2f the Trees .gf North.America dropped Pinus strdbiformis Engelm. from the place he gave it in the first edition. As a synonym for P. flexilis he listed P. strobiformis Sarg., not Engelm. This change represents a complete reversal of Sargent's opinion of the status of the trees first described and named P. strdbiformis by Engelmann. In the Trees and Shrubs of Mexico, Standley (1920:5h-55) assigned separate specific rank to P. flexilis and P. reflexa. He placed P. strobiformis Engelm. in the synonymy of P. ayacahuite Ehrenb. The native student of the Mexican pine flora, Martinez (1948:10h-105) accepted Shaw's 1909 treatment Of.§' ayacahuite but separated P. reflexa Engelm. from P. flexilis James at the specific level. Pinus strobiformis Sudworth and P. ayacahuite var. strobiformis Lemmon were placed in the synonymy of P. reflexa. After a year of study in Mexico and British Honduras, Loock (1950), produced an English language treatise on the Mexican pines Which lO paralled and concurred with the taxonomic treatment of Martinez. A study of P. flexilis and P. reflexa samples led Douglas (1958) to conclude that the criteria used by previous authors for separation were not valid. She felt that a gradient of morphological characters connected P. flexilis var. flexilis, P. flexilis var. macrocarpa, and P. flexilis var. reflexa but that they were sufficiently distinct to warrant subspecific rank. The French taxonomist Gaussen (1960:202-205), however, listed both P. strobiformis Engelm. and P. reflexa Engelm. as separate species in addition to 3. flexilis. Mirov (1961:34-35), after analysis of the gum turpentines of P. flexilis and P. reflexa concluded that they should be regarded as separate species. A preliminary investigation of P. flexilis and P. reflexa disclosed that the species were attacked by different forms of dwarf mistletoe (Arceuthobium campylopodum). Pinus flexilis was attacked by A. campylopodum forma cyanocarpum and P. reflexa by A. campylopodum forma blumeri (Hawksworth, F. G. 1962. Personal communication to Dr. J. W. Andresen, Dept. of Forestry, Michigan State University). During the past decade the views of the U. S. Forest Service dendrologist, E. L. Little, Jr., have changed from the position that the complex should be treated as the species P. flexilis with varieties flexilis and reflexa (Little 1950:13-14, 1953:265-266) to the view that the more correct treatment would be to designate the two as separate species, namely P. flexilis James and P. strobiformis Engelm. (Keng and Little 1961, Little 1962 88). For comparative purposes the descriptions assigned to members of the complex by some of the more prominent authors are presented in Table l. A summary of the preceding literature review yields the following synonymy: Pinus flexilis James. Pinus flexilis James, Account of an expedition from Pittsburg to the Rocky Mountains. II:27 and 3h—35 (1823). Pinus lambertiana var. E Hooker, Flora Boreali—Americana. II:16l (1838). 11 .ovov mm.opfldnmomwm .M.Op moaammm moapmahomow n.3mnm m ooquOQOHm vooqsoqonm oomuOmH onoq opossum mmauom “mamav Nonappmz \ coonsqunm woundoqonm OOmIOON oqoq opossum moomuooa Aaamav swam shopmhgomap .> opfidnmomhm .m. poxoamoy moon no -- more www-mma noon cesarean ooa-om Asmwav anomaom -- -- mam -- unmade ms-mo Awawav sooaaomsm I. maauomapoapm .m coXonoh mzoa nos moss -- Am-avo -- ms-o: Acmmav oaseaq mom wok mmmumma oqon opmmhom maaumm Amfimav Nouflpnmz \ mom mom oomuooa ouoq inaam u: Amwwav andaaowmml. mxoawon .m n monon .> mHHHMoHM .m. nowpoadmom muonpsom coon osos oom-ooa egos m unmade mm-om Aommav nausea vomqoaoam ooxoawom mopmo so omoq mmmumm pmomonm moon omnom Azamav scam unmaan pemaan In so oqoq mmmnmw In no osoa mwuo: Ahmwav Pmowmmm -- -- maa-ms snowman unmaao 0m-0m Ammwav ssmsaomsm .I mHHonam .m mowpmadmom uponpsoz he use moapquoao wnHXonoh nwwqoq upmaopm soapmhhom upmooq mamhnmoga carom Homnom moapmflmopodnmgo oqoo moaemwhopomhdno urea .mnonpsc m50ahb>.hp xoagaoo mHHonHm msnflm one mo myopfima op donwflmmn mQOHPQHaomoU mo momflamgaoo .H magma 12 Pinus lambertiana var. P. brevifolia End1., Synopsis coniferarum. 15o (18h7). Apinus flexilis (James) Rydb., Bull. Torr. Bot. Club. 32:258 (1905)- Pinus strobiformis Engelmann. Pinus strdbiformis Engelmt, Sketch of the botany of Dr. A. Wislizenus's expedition. Sen. Misc. Doc. No. 26 (l8h8). Pinus flexilis var.g} serrulata Engelm. Coniferae of Wheeler's expedition. ‘in Report upon U.S. geographical surveys west of the one hundreth meridian. VI:258 (1878). Pinus flexilis var. Q macrocarpa Engelmt Coniferae of Wheeler's expedition. ig Report upon U.S. geographical surveys west of the one hundreth meridian. VI:258 (1878). Pinus flexilis var. { reflexa Engelm. Coniferae of Wheeler's expedition. in Report upon U.S. geographical surveys west of the one hundreth meridian. VI:258 (1878). Pinus flexilis var. megalocarpa Sudworth. Nomenclature of the arborescent flora of the United States. USDA, Div. of For. Bull. No. lu:l7 (1897). Pinus reflexa Engelm.. Bot. Gaz. 7zh (1882). Pinus ayacahuite Ehrenb.. Linnaea. l2zh92 (1838). Pinus ayacahuite var. strdbiformis Sargent ex Lemmon. Handbook of west-American cone-bearers. A (1892). Pinus ayacahuite var. reflexa Voss. Deut. Dendrol. Gesell. Mitt. 16 92 (1907). Pinus ayacahuite var. brachyptera Shaw. Pines of Mexico. Pubs. Arnold.Arb. No. 1:11 (1909). Experimental Study of Variation and Speciation The transition from the Observational to the experimental approach in taxonomic investigation permitted greater evaluation of variation in morphological characteristics. Experimentation allowed an estimate of the degree to which variation was due to the genetic potential of the organism.and the degree to Which the final expression was influenced by differences in the environment. Recent studies Which have examined the relative importance of heredity, environment, and their interactions received their prime impetus from.the extensive investigations of Turesson (1922a, 1922b, and 1930). This Swedish worker grew specimens of numerous herbaceous 13 species from a variety of habitats together in a common garden and compared the resultant plant forms with those found in native habitats. He recognized the differences between populations from different environments and the part played by the local conditions in altering the genetic composition of the population of a particular area. At the same time, Turesson called special attention to the variability present in all populations. The variability was often masked by the special environmental conditions in effect in the native habitat but was readily seen and compared with that of other populations in the common test site. Turesson (1922a) proposed the term "ecotype" to apply to an ecological unit to cover "the product arising as a result of the genotypical response of an ecospecies to a particular habitat”. He inferred discontinuity between ecotypes but did not stress the point very much. Extensive transplanting studies were conducted in California by Clausen, Keck, and Hiesey (1940). They utilized three trans- planting areas at elevations of 30, lAOO, and 3050 meters. Clonally propagated material from.each collection was transplanted to each area. More than 50 species were studied. Each species was represented.by several specimens from.a number of locations. Transplants were generally taken from an east-west transect from Mbntara on the Pacific Coast west of Palo Alto, thence through the Coast Ranges, San Januin Valley, and Sierra Nevada to Benton in the Great Basin. For a few species the collection area was ex- tended north to southern Oregon and south to southern California. Each species maintained its identity when transplanted into the new environments. Although extensive modification of some organs occurred the modified parts did not assume the form characteristic of other species. The results showed that species differences were strongly controlled by hereditary factors rather than being mere adaptations to envirnomental conditions. Maritime and Coast Range clones were usually reduced in size at the mid—altitude station, Mather. At the Timberline station they were unable to set ripe seed and soon died. The dormant period of mid-altitude plants was considerably shortened at the lu lowest station, Stanford. Mid-altitude forms generally made greatest growth at the mid-altitude station, with less at the low-altitude station, and a great reduction at Timberline. High-elevation clones generally made greater growth at the mid-elevation station than at Timberline but less at Stanford. Other characteristics which showed extensive modification included; number of stems per propagule, number of flowers per stem, size of leaves, time of flowering, and length of dormant period. In a review paper on the study of ecotypic variation, BOEher (1963) called attention to the idea that the most important goal of comparative cultivations Should be the study of the variation itself and the factors responsible for the variation. He proposed the separation of taxonomical and morphological motivations. History of Provenance Testing in Forest Trees Extensive general reviews of the provenance testing literature may be found in the works of Kalela (1937), Schfitt (1958), or wright (1962). The first provenancetestixfbe reported in the literature was initiated by a French nurseryman, Louis de Vilmorin during the years 1820 to 1850. He made several unreplicated plantings of Scotch pine (Pinus sylvestris L.) from different origins (cited from Wright 1962:142). Numerous investigations similar to those of de'Vilmorin during the remainder of the century led to the establishment of a several-nation co-operative study under the auspices of the International Union of Forest Research Organizations (IUFRO) in 1907. Wiedemann (1930) compiled and compared the results Obtained in the various plantations which survived the war years. An extensive study of the variability of foliar dry matter content and the relation of this measure to frost hardiness of Scotch pine seedlings was undertaken by Langlet (1936) in Sweden in the early 1930's. He collected seed from 582 areas in Sweden and grew the resultant seedlings for 2 years in an unreplicated nursery trial. In 1936 the IUFRO Congress initiated another series of studies on variability in Scotch pine, Norway spruce [Picea abies (L) Karst], and European larch (Larix decidua Miller). 15 In 1952 Veen (1952) visited.most of the test plantations and recommended measurement procedures. The Czechoslovakian, New Hampshire, and Michigan plantings of Scotch pine have been the subject of published reports covering variation in several growth, wood, and chemical properties. After 19h5 many geographic variation studies were initiated. In the United States two large co—operative tests were organized. The Southwide Pine Seed Source Study was initiated in 1951. Collections. were made and plantation established in 16 states (wakely 1961). These studies have been carried out with the four southern pines: .P. echinata Mill°:.E- elliottii Engelm., P. palustris Mill., anng. PEES§.P°° Approximately 50 collection areas and slightly over 100 plantations are included in the study. These plantations, in addition to providing a study of variation, can also provide materials for intra- and inter—specific hybridization.programs. In the north central states the NC—51 program was initiated in 1960. Co-operating agencies in 10 states are currently conducting geographic variation studies in 9 species. In several of the NC—51 tests, the progeny of individual trees have been kept separate to allow examination of the relative amounts of Within-and'between-stand variation. Currently, variation studies are being conducted on some three-score species throughout the world. During the past century many changes have occurred in the design, analysis, and philosophy of studies of geographic variation. For example, in Scotch pine, sampling intensity has varied from 12 widely scattered origins in the 1907 IUFRO test, to 582 origins from Sweden alone, in Langlet's study during the 1930's. At present, most studies attempt to get a broad, even sampling for the initial phase of a test and a more intensive sampling in follow-up studies in areas of special interest. In early variation studies little consideration was given to the test design. Consequently, the conclusions to be drawn from such tests were very limited. Then experiments began, about l9h0, to follow the work in other fields such as agriculture and horti- culture. Ideas and.practices on replication, plot size and shape, arnd selection of test areas were adapted to forestry prOblems. JTinally, experiments were conducted to devise experimental designs 16 which were specifically intended to provide the information desired. Prior to 1940 most tests utilized either short-term.nursery results or field.plantation measurements but not a combination of both. Analyses have progressed from the stage of recognition of differences between areas of origin, to comparison of growth of seedlings to that of the stands where the seeds were collected, to comparison of growth at early and later ages, and finally to the evaluation of differences between and within stands from.a particular region. The philosophy behind variation studies has changed as much as the methods employed. The early tests were set up to examine the existence and extent of possible differences between various origins of a species. Soon afterwards the practical approach of testing to find the most suitable source for making seed collections became prominent. This practical attitude continues at present but is tempered to allow or encourage the simultaneous investigation of theoretical aspects as well. Included in this latter category are the study of variability of anatomical features and physiological processes, the evaluation of taxonomic affinities through breeding tests, the study of ways to improve experimental design and analysis, and the determination of heritability estimates for use in future selection and breeding programs. The literature on geographic variation in trees is too voluminous to be reviewed completely here. However, work done with species which grow in areas where Pinus flexilis grows and with species related to P. flexilis will be reviewed. Geographic Variation in Rocky Mbuntain Conifers Ponderosa pine (Pinus ponderosa Laws.).--A 21—origin provenance test of ponderosa pine was started at Priest River, Idaho in 1911. The test contained a single plot of each origin. When examined at age A0 there was a good correlation between height at age 12 and at age #0 (Squillace and Silen 1962). In an earlier report on the same test Weidmann (1939) presented evidence for close correspondence 'between growth rate of the progeny and growth rate of trees in the 17 parental locality. Differences in several traits were evident. There was a 5 to 3 growth rate difference between the fastest growing trees from northern Idaho and the slowest growing ones from eastern and southern origins. South Dakota trees had a high proportion of two-leaf fascicles as compared to others which had mostly three. One California seedlot was completely eliminated by sudden freezing weather. A second provenance test was started in 1926 that included 10 origins and outplantings at six locations in washington and Oregon. Two-year nursery heights were strongly correlated with 30-year plantation heights (Squillace and Silen 1962). These same authors compared the results of these two studies and a third one conducted in New Zealand and found very close agreement in the relative heights of progenies from.Similar regions in all three tests. The effect of altitude of origin on seedling growth has been followed for a 20—year period in California (MirOV'EE'EE. 1952, Callaham and Metcalf 1959, and Callaham and Liddicoet 1961). Until age 12, mid—elevation origins outgrew low-and.high-elevation ones at all three planting sites (290, 830, and 1700 meters). .At 15 years, Callaham and Hazel (1961) found a significant correlation between the second year growth increment and 15-year height. They also found that within elevational zones, 39 percent of the variation in height growth was due to genetic causes. By age 20, differences due to elevation of origin had disappeared at the high—altitude but not at the other test sites. At all ages high-elevation origins performed poorly at low-and mid-elevation planting sites. In a 2-year nursery test of 60 origins grown in Michigan, Wells (1962) found a sharp break between sources from.Arizona and southern New Mexico and those from farther north. Seedlings of these southern interior origins grew taller, had longer leaves, and formed a greater number of secondary leaves in the first year than did others. There was approximately a 2 to 1 difference be- tween the greatest and the least development in these characters. Northern origin progenies of the coastal variety formed more terminal buds in the first year than southern ones and were less subject to winter injury. l8 Douglas-fir [Pseudotsuga menziesii (NHrb.) Franco].-—A test of 120 single tree progenies from 13 coastal sources was started in 1912 in washington and Oregon (Munger and Merrie 1936 and 19h2). The first seedlings were outplanted in 1915 to six test areas. The individual progenies were planted in the same sequence in each plantation. A second, smaller replicate was planted at each site a year later. There were no significant relationships between the maternal parent's altitude of origin, age, or soil type and growth of the progeny. Two seed source progenies from Granite Falls and Darrington, washington exceeded the height-growth average for all stocks on every plantation. Time of bud-bursting was studied on three of the plantations by Merris.et'ai. (1957). The three earliest and latest origins to burst buds did so very consistently in all areas. Relative time of bud-bursting was related to the spring temperature pattern at the place of origin. Sources from areas with warm days and cold nights began growth later than those from areas where warm days and nights were prevalent. In a 24-year-old German test of Douglas-fir from Colorado, Oregon, washington, and British Columbia, the Colorado progenies grew very slowly in all test areas (SchOber 195A and Schober and Meyer 1955). They were also very susceptible to needle blight and frost damage. Coastal washington sources performed best in maritime planting areas and those from the Fraser River Valley of British Columbia performed best at intermediate elevations. Douglas-fir from a high-elevation New Mexican source showed the best survival and growth after 5 years in the field in a New Hampshire test (Baldwin and Murphy 1956). Among the remaining three origins, those from.Idaho were next best and were followed by Mentana and California progenies. In a 19—origin Christmas tree test in Pennsylvania, seedlings frOm the central and southern Rocky MOuntains were heavily damaged by late spring frosts but did not suffer from winter cold (Byrnes_et_al. 1958). Coastal and western interior sources were heavily damaged by cold winters but not by late spring frosts. In an Oregon test involving only origins of the coastal variety (of Douglas-fir, Irgens—MOller (1958) found that higher elevation l9 progenies grew less at Corvallis than lowland origins because they stopped growth earlier in the season. Two-year nursery results from another Oregon test of Oregon, washington, and British Columbia sources showed that most northern sources were faster growing than southern ones (Ching and Bever 1960). Mbst northern sources also had longer leaves. Southern sources began growth earlier and continued to grow longer than northern ones. First-year results from a 135 origin test in Michigan show origins from extreme northern Idaho and adjacent Mbntana to be the fastest growing ones from the Rocky Mountains (personal communication from.Dr. J. W. wright, Dept. of Forestry, Michigan State University). Lodgepole pine (Pinus contorta Dougl.).--Mbst early studies of the species were concerned.primarily with the differences between the coastal and inland forms and not with variation within regions. For an extensive review of these works see Edwards (1959 and 1955). However, Critchfield's (1957) study of P. contorta was quite similar to the present investigation. He utilized both seedlings and parental specimens in analysing variation patterns. However, in his study the emphasis was placed on the parental specimens. In contrast, the seedlings were considered of greater importance in the present study. His study also sought to clarify the complex nomenclatural and taxonomic treatment of the species. Samples of lodgepole pine were collected from.AO native stands throughout the species' range during the A-year period from 1952 to 1955. Leaves from adult trees growing in the Sierra Nevada were wider, on the average, than those from other geographic regions. An increase in leaf width with an increase in altitude was also observed. Mendocino Coastal and White Plains populations were distinct from all others in their lack of leaf resin canals. Leaves from interior stand collections were slightly longer than those from coastal collections. However, for seedling materials, leaves from interior sources were shorter than those from coastal sources. The angle between the cones and the branch on which they were borne was much more uniform for coastal P. contorta sources than for those from the interior. In particular, cones from the northern lRocky MOuntain collections had grown at a wide variety of angles. 20 Frequently the cone position overlapped the angular range characteristic of Pinus banksiana Lamb.. These results supported the earlier observations of Moss (l9h9) of hybridization and introgression between _P. contorta and P. banksiana in Alberta. With regard to the specific gravity of cones, those from the Rocky Mbuntains had the highest values and those from the Sierra Nevada the lowest. Cones from the coastal regions and the Cascade and Blue Mbuntains had intermediate values. The light cones from the Sierra Nevada usually shed their seed soon after maturity and did not persist for long on the tree. The denser cones from other regions were often indefinitely indehiscent and persistent. 0n the basis of his findings, Critchfield assigned subspecific rank to the four most distinct elements of the species. White fir [Abies concolor (Cord. and Glend.) Hoopes].-- The first preliminary results of a geographic origin study being conducted jointly by Michigan State University and the University of California indicated that among southern Rocky Mountain origins those seedlings from.Arizona and southern New Mexico grew fastest, had long and straight leaves, and were light in color (personal communication from Dr. J. W. Wright, Dept. of Forestry, Michigan State University). Seedlings from Utah were shortest, had shorter and curved leaves, and were darder colored. NOrthern New Mexico and Colorado origins were intermediate. Geographic Variation in the White Pines Eastern White pine (Pinus strObus L.).--In a test of 67 origins from the vicinity of Petersham, Massachusetts, Pauley et ai. (1955) found only random variation in growth rate. For another portion of the test involving sources from scattered locations throughout the species range, seedlings of Massachusetts origins grew fastest during the first two years while those from.more distant sources grew progressively slower. After lh and 15 years growth in field tests, the local origins were superior with regard to diameter growth. Results of a rangewide test Of.E' strObus initiated by the INortheastern Forest Experiment Station in 1957 have been reported from ‘three areas: (1) New Jersey (Santamour 1960), (2) Southern ftppalachians (Sluder 1963), and (3) southern Michigan (Wright_eE‘aP. 21 1963). In all areas seedlings from the more southern origins have made the fastest growth. Contrary to the expected reaction, seedlings of northern origins had the most 1ammas shoot growth in Michigan. western white pine (Pinus monticola Dougl.).--Differences in growth rate were found among progenies originating as little as one-half mile apart in Idaho by Squillace and Bingham (1958). Seedlings produced by moist site and low elevation sources grew faster on the lower and'better planting sites than did those from dry or high sources. At high elevation planting sites seedlings from.high sources did best after recovering from nursery and transplanting effects. At Placerville, California, 15-year height of P. monticola seedlings from Idaho is greater than that of California Sierra Nevada sources. Limber pine (Pinus flexilis James).--Although the botanical descriptions of members of the Pinus flexilis complex indicate considerable variation in morphological traits there has been very little systematic investigation of the variation. A study of the variability present in native P. flexilis stands in Colorado was begun by Douglas and Douglas (1955), but illness forced discontinuation of the work. MATERIALS AND METHODS Materials Acquisition.--The study was initiated in June of 1959 by Dr. J. W..Andresen of Michigan State University. Requests for seed from several widely scattered areas throughout the range of the complex were sent to selected co-operators. A portion of the seed received was planted in 1960 to determine the best methods of handling the larger scale test which was to follow. The majority of the collections for the study were made in 1960, To provide the widest possible sample of naturally occuring populations of the complex, U.S. Forest Service personnel, state foresters, and botanists were asked to co-operate in making the col- lections. The co-operators were asked to gather cones and a foliage specimen from up to 10 trees per stand and to keep the materials from.each tree separate. Edaphic, ecologic, and geographic data pertinent to the collection sites was also requested. .All collections were forwarded to East Lansing, Michigan for further processing. In addition, Dr. Andresen has spent the summers of 1960 to 1962 supplementing the collections and notes of the co-operators. The distribution of collections is presented in Figure 3. Handling.--When the collections arrived in East Lansing they were assigned accession numbers in accordance with the thhigan State Forest Genetics (MSFG) system of identifying new acquisitions. The cones were dried to facilitate seed extraction and the foliage specimens were dried and pressed in preparation for mounting on herbarium sheets. After the seeds had.been removed, the cOnes and foliage specimens were stored for later measurements and observations. The seeds were cleaned after extraction to eliminate foreign materials and then placed in 95% ethyl alcohol to float off any that were not completely filled. The alcohol flotation technique does not distinguish between seed which are filled and have a sound, living embryo and those which are filled but not sound. It also may eliminate some seeds which have germinative potential even though not «completely filled.' The filled seed were placed in cold storage at A5° F. until sowing izime some 180 days later. The seed were not stratified because of a 22 ‘.ll.ll l‘l {I 1" ll!!! Figure 3. o 2724 "3002 23 .2992 2825 o 2579 o O 990 D 967 o-28l9 *0‘42765 l030-o "I059 l006~ 3404: o 2900 o 2652 ’ 2877 .2929 'esn e 2964 930 2 F——;’-——— 25631 0-2919 o 2909 546 .D'2553 2987.. g; 2 9'6‘0 .4 S25632925 2509.. , "902 2978' ‘260! 2796-- ..Z \ 0932 0269 0905 .,2939 ‘2649 Location of stands sampled for this study. 2A desire to test the effect of place of origin on time of germination. In all, seed from 325 collections representing 61 stands, have produced seedlings for this study. thhods Design and installation of the nursery test.--A randomized complete block design with four replications was used in planning and establishing the experiment. Plots within replicates consisted of a single row which contained seed from a single tree with the exception of seven of the collections made in 1959. In each of these seven collections seed from several trees had been mixed together. Planting was done at the Bogue Experimental Nursery on the thhigan State University campus on May 20, 1961. The nursery soil was a sandy clay loam. It had been maintained at a high fertility level prior to the test. No fertilizers were applied during the course of the experiment. The nursery beds had been treated with ”MILON", a combination fungicide and herbicide, in the autumn of 1960. The seed were planted at A centimeter intervals in rows 100 centi- meters long. The rows were perpendicular to the length of the bed and were 15 centimeters apart. Seeding was started at the north end of each row to provide a common reference point for those cases Where sufficient seed was not available to fill the row. Accurate spacing was accomplished by using a steel tape measure and a wooden template which made a slight depression in the soil at each seed—spot. After sowing, a thin layer of fine sand was spread over the seed. The nursery beds were lined with A—inch boards which served to prevent disturbance of the edges. The boards were also used to support wire screen and lath to provide shade and protect against bird damage. Adequate soil moisture was maintained by sprinkler irrigation. Weeding was done by hand in conjunction with regular measurement activities. A one-half—inch sawdust mulch was applied in November of 1961 to reduce seedling damage by frost heaving. Measurement of seedling traits.--Traits chosen for analysis in this study are listed in Tables 2 and 3. The criteria for selecting a trait was either: (1) That it exhibited such pronounced row-to-row differences as to make the presence of between-progeny differences 25 Table 2. Two-year growth data for Pinus flexilis progenies, summarized by stand-progeny. : : E 3;: I0: .2. 5 E = = “I: .- .. s! 0 a a '- 58 E '5 E35: ‘ “ 5 33 3.2 :- 2 .. _ _ _ ‘ 0 ~ e é .. 2:. . s s 3: 2“ g s .. or- .2 .2 3L” '5. *5; . a ... 3 cu .- ‘u 3 .- and Ha . 0 9.- MC- I I I. h a: z a has a a u o: 08 c: on on no on L a at E- h» . . n“ 41-. 9.: A» a a u 0'0 ‘8 ~-' 08" 'l' "' ”8 a . . 2 :5 é-m :: z: z a: 3. :3 r. :2 !§ §.- :2 :2 a g. . . i5 4 a a: sis .2: :3 a: as as is as as a; ..85 is 2.3 si as a: .2 a e (I) (2) 413), (a) 5 0 7) 0 9) (IO) ll) (12) ()3) (14) (I5) no. “ ' ” ' feet day of year days no. ==; .g, g --'r.de.o— ... 'f.‘° no. ------.-,- ..... 2002 AJta. 1 51 00 115 10 0700 200 250 50 0.4 26 I.0 I0 2.4 3.0 30 0.0 1.0 25 43 5 3002 Alta. 7 5| 00 II5 10 4200 100 229 63 0.0 25 I.l IO I.7 3.I 25 0.0 1.0 25 43 3 2025 Alta. 9 49 20 114 20 5100 I72 230 04 0.0 25 I.0 I5 2.I 3.0 25 0.2 I.5 20 43 2 2040 lent. I0 47 50 II2 40 5000 2I2 243 3| 0.0 25 |.O 5 I.9 3.3 23 0.1 l.0 25 43 3 2050 lent. 0 47 50 112 40 5000 2I0 250 34 0.0 20 |.0 IO 1.0 3.0 25 0.I I.5 23 4| 3 2570 lent. I 40 50 _ III 43 _ 7000 _ I01 _ 229 _ 40 _ 0.4 _ 20 _ I.0 _ 45 _ 3.0 _ 3.2 _ 30 O 0 _ I.5 20 - 40 _ 0 2007 - loot. 4 _ 40 30 II2 20 7200 214 257 43 9.0 23 1.0 5 I.9 3.2 23 0 2 I.4 20 4| 3 977 lent. 0 40 I5 IIO I5 0500 205 230 3| 9.0 20 I.I I0 I.0 3.2 25 0.5 I.5 25 43 2 900 lent. 2 45 20 III 02 7200 2|7 250 33 9.0 23 1.0 I5 I.5 3.! 25 0.0 I.5 25 40 5 907 l. Dan 5 45 45 104 00 2500 211 204 53 9.0 27 |.2 IO 2.0 3.3 20 0.] l.0 30 40 5 2920 Iyo. IO 44 45 109 20 0300 100 243 55 0.2 20 1.3 20 I.0 3.4 30 0.3 1.4 30 53 0 2971 iyo. 3 44 29 109 49 ---- I75 250 75 9.0 24 I.0 30 2.0 3.0 30 |.0 2.0 30 50 ID 2904 'ye. 3 43 40 I09 35 6500 I00 229 49 0.4 20 [.0 25 |.4 3.2 20 0.1 |.0 30 5| 0 1005 Idaho 4 43 27 113 35 5000 209 250 41 9.2 20 I 2 IO I 0 3.2 30 I O 2. 30 53 0 907 Idaho -‘ 43 20 II3 30 5000 190 257 59 9.0 32 I 5 IO 5 3.2 30 0 2.5 30 53 0 2590 Idaho 2 43 I0 III 05 0000 217 250 33 0.0 33 I. I5 I 3.0 33 0 5 2.0 30 00 l3 2019 Idaho | 42 25 III 3| 0900 212 250 30 0.0 20 |.0 I5 2.0 3.0 30 |.0 I.5 30 50 I3 2009 Iyo. IO 42 33 I00 45 0300 I09 230 47 0.0 24 1.0 I5 |.0 3.2 25 0.I I.7 20 40 5 2765 Utah 3 42 I0 III 30 7200 I93 243 50 9.0 20 1.1 I5 2.1 3.3 30 0.0 1.0 25 53 IO I050 Utah I 41 35 Ill 20 0500 200 250 50 0.4 27 I.O 55 I.2 4.0 30 0.0 2.0 25 5| 5 I030 Utah 7 4| 22 II2 02 9300 200 230 30 9.2 20 I.2 I5 I.7 3.1 20 0.0 |.0 30 40 5 930 Utah -‘ 40 55 110 00 7000 I90 243 53 7.0 27 I.O 20 1.0 2.2 30 0.0 I.5 30 50 I3 1000 Utah 3 40 40 III 40 9700 202 250 40 9.0 30 l.4 IO I.0 3.2 30 I.0 1.0 30 50 I0 I04| Utah 2 40 3| III 4| 9000 I94 229 39 0.6 20 I.0 25 1.0 3.1 20 0.5 2.2 25 40 0 2553 Hebr. IO 4| 40 I04 02 5200 I00 27I 03 I0.0 27 I 7 5 2.3 4 2 30 0.2 2.I 30 6| IO 2563 Iyo. 5 41 00 104 04 5300 I79 204 05 9.6 32 I 0 0 2 43 O 4 2.0 4| 50 5 2724 CaIIt. I0 37 30 110 IO 10000 107 230 49 7.2 24 I 0 25 I 9 3 4 25 0 2 I 6 23 43 5 2540 Colo. 4 40 50 |06 50 0000 2|3 264 5| 0.2 29 1.1 20 I.7 3.2 33 0.2 2.0 20 40 0 2919 CoIo. 2 40 30 105 41 I0200 I71 250 79 9.2 24 1.0 I5 1.0 3.0 20 0.0 2.0 20 4| 5 2522 Colo. I 39 40 100 00 9900 I90 250 52 0.2 24 1.0 I5 I.0 3.0 20 0.0 2.0 25 4| 0 922 Cole. 7 39 32 100 00 9900 I71 230 65 0.0 26 I.I 25 1.0 3.4 20 0.I |.0 20 40 5 909 Colo. 7 39 40 105 30 10600 166 230 70 7.0 27 |.2 20 1.0 3.3 20 0.0 I.7 20 40 0 2523 _ Colo. _ _ 0 _ 39 36 _ 105 30 _ I0500 _ I90 _ 243 _ 53 _ 0.2 _ 25 - |.O _ IO _ I.7 _ 3.6 _ 25 _ 0.| _ 1.0 25 _ 40 - 5 2509 Colo. IO 39 I3 I00 05 10900 I73 236 63 0.2 25 I.I 30 I.5 3.6 30 0.3 2.0 20 40 5 260) Colo. Io 39 05 105 33 I0600 107 250 63 0.6 27 1.2 IO 1.9 3.0 20 0.1 1.0 20 40 0 2907 Colo. 2 39 55 I07 05 0300 I75 236 6| 9.6 21 1.4 IO 1.0 3.2 20 0.2 I.7 30 04 13 9I0 Colo. 5 39 36 107 I4 9300 209 251 40 0.4 29 1.2 I5 1.0 3.4 33 I.0 2.0 30 50 I0 2970 Cole. 0 39 05 I06 20 IIOOO I63 236 73 6.6 20 I.5 I5 2.0 3.5 33 0.2 1.9 33 50 lo 902 Colo. —- 39 20 I05 00 0500 I66 243 11 9 6 20 2 0 IO 2 2 4 0 30 0 0 2 0 30 04 I5 2900 Utah 9 39 00 |l2 50 9500 2I0 250 40 9 4 30 I.6 IO 2.4 3.5 33 0.4 I.7 33 6| 13 2652 Utah I0 30 31 112 3| 0500 200 250 50 9 0 29 1.6 15 2.3 3.4 30 0.3 1.0 4| 09 I3 2077 Utah 3 37 4O II2 40 6600 167 236 69 9 4 30 1.0 10 2.0 3.5 33 0.0 2.0 41 00 I3 2790 Colo. 5 30 30 106 30 9000 169 250 0| 9 4 32 I.0 I0 3.2 4.0 30 0.0 I.7 43 74 I0 2599 Colo. 2 30 00 I05 IO 9000 167 250 03 9 2 3| 2.0 O 2.6 4.I 30 0.0 2.0 4| 00 I0 2612 N. lax. 3 35 I2 I06 27 IO500 I67 264 97 l0.4 20 I 9 0 3.2 4.0 33 0.0 1.9 43 74 Standard devtallon of a progeny mean 4.0 IO.5 7.1 0.5 l 0 0 2 0 0.3 0.2 3.5 0.1 0.4 J J 5.0 1 150.05 bet-eon progeny meana |0.0 J7.| 24.0 1.6 3.5 0 0 17.5 0.9 0.0 10.7 0.4 1.3 I) 7 20.0 14 2 LSD 9' between erogeny leans 2I.I 40.2 31.2 2.0 4.4 0.7 32.0 I.| 1.0 I5.4 0.5 1.7 5 53.7 7 0 l Collected for the pIIot study -- progenleu vIthIn the stand grouped at tIao of collectIon. KEY TO GRADES Character (0) (9) (II) l-year Follage Color 2-yenr Pollage Color Degree of Lent Serrulntlon Grade 0 Sorrulnttona abaent or extremely nlnute and very acarce Grade I Yellow-green Yellow-green Sorrulattona fee In nuaber and minute but readily dtacernable Grade 2 LIght green Light green SerrulaIIona proatnent and numerou- Grade 3 Green Green Grade 4 Dark green Dark green 5 Grad Blue-green Blue-green neeuuuosam noeauneann n oueao aouuu shun amen» shun v Queue gecko gecko a Queue «echoes: use unenufloan uncuueflshuom ueouu can“: coca» «Judd a macho manauaooaav aaaeaou asp masque one Lopez: :0 ae« uncuaeasuaom coouunabaaor aeeuunaoHHor a wedge oouduu auo> end wanna! afloaeuuuo uo ascend unoduuazhuom o evaao nauaaasaaom «dog «0 oeauea uoHoo oun«~om usehsn aoHoo.wummw0h uaowwa maav Amy Aav uouoauano . mfln<¢c OF rfll .co«ueenuou «0 mafia an uneven» venue on» nasaua neuaououa :1 aosun pagan on» how neuoeuaoo a 26 m.u~ ~.n« n.v~ p._ n.o v.m~ o.a a.“ o.«« p.o v.v o.« «.a« u.ov 2.0“ undo: anwuoua couaaon ao.ama u.v~ c.o« 0.5“ a." «.o ~.o~ c.o o.o n.pa o.o o.n 0.“ ¢.v« H.5n n.wd unac- maououa coouaon no am; ~.v 0.0 n.n v.o a.o n.n «.o n.o o.n «.o o.a m.o 3.5 n.oa o.¢ ado: acououa a «o noauaa>ou unavuuum an aaa a» a.“ «.5 an o.n n.v o o.n an w.u~ and non op” caps an mod 0» an « magma hvou an ova we n.~ n.o mo o.n o.v o o.n an o.«~ 020 ohm «ma coho we no" an an « .xo: .2 mama an on" vo o.~ c." an o.n o.v o o.n pa ¢.ua do" was no” cope vv no” an an a .xoa .z mvoa as had v0 0.” O.“ 30 o.v m.v o m.« an n.2H moa new era coon «n no“ an an a- .xo: .2 mom on «Na on m.~ o.~ Ho m.v u.v o N.“ on 0.55 mm ku an” ace» ov mod on on a- .xo: .2 «mm ma on“ am N.o o.u on o.n o.n o m.n on o.aa and «as mod cons ma oaa vs an H .ufiu< noun on ~52 om o.~ 5.2 vs ¢.v v.n o o.u ov o.v~ ma“ was so” econ ma «OH ov an n .nau< Hahn ma «a up n.a 0.2 an o.n o.v o o.~ «a v.«a cog aha oh" cone Mn oHH av an m .n«u< anus n5 mod 35 o.a v.“ we p.v 2.0 o n.« an o.n~ so" now 052 com» ma mod an an m .uau< nvhn mv and flu m.o «.2 an o.n 0.0 o o.« on «.aH ooa mom and coon o¢ oaa an an 9 .uau< amen as oma vm «.5 o.” mo o.n a.v o o.n an «.na mo“ 30" no” cove on mofi ov an a .u3u< «nun an and mu o.“ m.“ on o.n u.v o m.u an «.53 mod man as” oops on ed” a” an m .n«u< «mod an and mo o.o e.” we o.n «.v o v.« an w.oa nod gun cog oovu ”a Add an «a u .Nau< moan on «v2 mu «.5 m.o an v.v m.n o m.n on «.25 mm «on nod can» av HHH on on m .nau< menu nu mm mm m.” o.” an o.v o.n n u.« an o.- Hod a»« on“ ooam ov AHA ca mm a- .N3u< vow pm bag vo o.m o.o mm m.v «.n o N.“ am «.23 mm ku and coco ov ”an on an «1 .nfiu< mom -----.EEIIIIII 4mm ouauu hum --uouuuunu IN .5: .aa .o: muse ago» no age noon . o . o .02 away Avav ”may Aua~ mafia 2°20 “m0 any 590 A00 any Ava any Auv “av mw 4 4 mm «m_ wm m« w« awn mm mm mm mm mm mm m mm mm new mm m o o K K o n J 8 o u 1.5 I.A o a m u 3 u 2 o u P 3 m 7 a u 9 3 J 0 m o 9 J an a 9 mm. MM 03 09 00 ...—09 M 3.4 WK AB 9 o A 33 1.1. 83“. A3 9 .40 0 0 0 0.4 J. ...—0 .100 a 1.1 I .11 S tr 0 I. 1". 36 TO u.3 J J .4... .19 Du. J 1. JD. 03. “.3 13 uu. 90 no 1 1. n u... u N 0 0 0 901. 39 D. D. 3 13 03 I. n D. “t. 00 n 1.0 H H 10 10 30 J J 130 .AJ 0 o .0 3. 0 D. a 0.90 OJ w HI. 9 a ..o .11: .AI. 0 O on“ .1 u u 910 '1' u a are a 1 I W o I I A o a o a W 48 3 3 um. M. .... I. 3%.... I 0 us 1 a a U u a W e a e u s u u o 3 3 s 1 0 0 S 3 J I 3 o 90 s ..o A a a A u u u e a 9 10. I s B .msemoamndcdpm an GMNHHQSSSm nmmflqmwomm mflahowflpoapm macaw pom epww npzoam awmmuose .m mapde 27 likely, or (2) That considerable variation in the trait had been reported in the species' descriptions. The validity of the assumption of differences was tested by analysing data from the pilot study. Measurements were always begun at the start of replicate one and continued in sequence to the end of replicate four. To reduce bias, the measurement for each plot was Obtained before the identity of the row was checked. The plot mean is the measure used in all compu- tations. l. Germination date was recorded for all seed in each plot. Seedlings were recorded When the hypocotyl became visible. Germination counts were made twice a week for 5 weeks after the first seedlings appeared. Further observations were recorded at weekly intervals for 4 weeks and then at 3-week intervals until the ground froze in NOvember. Mean germination date was calculated as follows: i = faxa + fbxb + fcxc‘° . . . . . + f x nn f + f + f . . . . . . + f a b c n where i = day of the year mean germination was reached. Xa’ Xb’ xc, etc. = day of the year for the mid-point of each observation period. fa, fb’ fc, etc. = number of seeds germinated since the previous examination. 2. Date of bud set, or terminal bud formation was recorded at weekly intervals. Bud set was reported to have occurred when one-fifth (l/S) of the seedlings in a plot had visible terminal buds. 3. Length of growing season was obtained by subtracting date of germination from date of bud set. M. Cotyledon number was Obtained by counts of five seedlings per plot. 5. Cotyledon length was measured on the longest cotyledon on each of the same five seedlings used to determine number. 6. Diameter of hypocotyl was estimated to the nearest l millimeter for the plot as a whole. 7. The percentage of seedlings forming secondary leaves in the first year was determined by counting. 8,9. First— and second-year foliage color were recorded 28 according to a series of color grades which were established at the time of observation in November 1961 and September 1962. The seedlings with leaves eXhibiting the most yellow coloring were always scored as Grade 1. Seedlings with blue-green foliage were at the other extreme and were scored as Grade 5. 10. Length of secondary leaves was measured to the nearest millimeter on leaves collected at the end of the second growing season. For each plot the sample consisted of one fascicle of leaves from each of five seedlings. ll. Degree of leaf serrulation was scored under a dis- secting microscope on leaves from one replicate that had been used for length measurements. The grades used were 0 (no serrulation), l, and 2 (serrulations prominent and numerous). 12. The number of dorsal leaf surface stomatal rows were counted with the aid of the microscope at the time that the leaf serrulation estimates were made. Incomplete rows were counted as half—rows. l3. First-year height was determined by measuring the tallest and shortest seedlings on each plot. (Only epicotyl growth was measured, i.e. the distance from the upper surface of the cotyledons at the point of insertion into the stem to the tip of the terminal bud. The validity of means based on the tal- lest and shortest seedlings had.previouSly been established by O. 0. Wells on ponderosa pine, by J. B. Genys on Japanese larch [Lariz leptolepis (Sieb. and Zucc.) Gord.], and by J. W. wright on Scotch pine in the same nursery (personal communications). 14. Second:year height was determined by measuring five seedlings on each plot to an accuracy of l millimeter. 15. The amount of the second growth increment in the second year was measured from a point 5 millimeters above the last-formed secondary leaves to the tip of the new shoot. The measurements were made on the same seedlings used for second-year height measurements. Measurement of mature traits on herbarium.specimens.--Traits chosen for measurement are listed in Tables h and 5. Leaf length, 29 Table h. Data for adult characteristics of Pinus flexilis trees, summarized by stands. Degree o! Cone Beale - 7 letlext ‘ E :. u n “ g 3 35 33 e . h 2 x 8 fl ‘4 04 b N N d N ' 40 0! I F4 '2 a a o o e o o o-u o a o e e o a . rs ‘\ a o o v a a g a I d an 3 v1 I - a i E H o z I: a u a 1- n I. a A A D a )1 e on 2: a v4 0 g 8 3 E .4 u: g 0 o; a «I: z a E g g I- “ E 0 5 j I O E E aE as 3 a 38 53 a a 3 a: o i: (10) (17) (10) (10) (20) (21) (22) (23) (24) (25) (20) ° ' ° ' teet no. u. no. grade -. -. n. -. ° ° ° 2002 Alta. 51 00 115 10 4700 11 53 2.3 0.1 0 04 51 5 0 30 40 3002 Alta. 51 00 115 10 4200 13 50 2.4 0.0 0 04 40 3 O 30 50 2025 Alta. 40 20 114 20 5100 11 53 2.0 0.1 0 01 53 4 O 10 30 2040 lent. 47 50 112 40 5000 11 40 2.4 0.0 0 70 40 4 0 20 40 2050 lent. 47 50 112 40 5000 10 53 2.0 0.3 1 01 51 4 10 50 00 2570 lent. 40 50 111 43 7000 12 51 2.3 0.0 2 01 51 4 0 0 10 2001 lent. 40 30 112 20 7200 12 40 2.0 0.3 0 71 53 4 0 10 10 077 lent. 40 15 110 15 0500 11 43 2.0 0.1 3 00 53 5 0 10 40 000 lent. 45 20 111 00 1200 12 51 2.7 0.0 3 01 40 4 0 0 20 901 I. flak. 45 45 104 00 2500 0 53 2.4 0.0 4 04 40 4 0 10 30 2020 Iyo. 44 45 100 20 0300 11 51 2.4 0.1 1 70 40 5 0 30 50 2071 lye. 44 29 100 40 ---- 10 53 2.4 0.5 1 01 51 0 10 30 50 2904 Iyo. 43 40 100 35 0500 11 51 2.4 0.0 I 00 53 0 0 30 50 1005 Idaho 43 27 113 35 5000 9 51 2.2 0 0 00 53 7 0 20 50 007 Idaho 43 20 113 30 5000 0 -— --- --- - --- -- - -- -- -- 2500 Idaho 43 10 111 05 0000 7 40 2.2 1 0 00 51 0 0 50 00 2019 Idaho 42 25 111 31 0000 12 51 2.5 0.4 2 00 53 5 0 20 40 2009 Iyo. 42 33 100 45 0300 15 40 2.3 0.1 1 00 43 0 0 20 50 2705 Utah 42 10 111 30 7200 13 40 2.0 0.2 5 70 40 0 0 10 50 1050 Utah 41 35 111 20 0500 10 40 2.5 0.1 5 00 53 5 0 0 10 1030 Utah 41 22 112 02 0300 0 50 2.0 0.2 2 00 50 0 0 0 40 no can so as no on 1000 10 --‘ --— --- - --- -- - -- -— -- 1000 Utah 40 40 111 40 9700 9 50 2.0 0.1 5 91 50 7 10 40 50 1041 Utah 40 31 111 41 0000 10 50 2.5 0.0 5 01 50 0 10 20 50 2553 lebr. 41 10 104 02 0200 9 50 2.2 0.7 4 01 53 0 0 20 40 2503 Iyo. 41 00 104 05 5300 0 04 2.7 0.0 5 102 50 0 20 50 00 2724 Calll. 37 30 110 10 10000 14 50 2.9 0.1 0 74 51 5 0 0 30 2540 Colo. 4O 50 100 50 0000 10 43 2.4 0.0 2 91 50 0 0 50 00 2010 Colo. 40 30 105 41 10200 11 53 2.9 0.0 0 71 53 5 0 0 30 2522 Cole. 39 40 100 00 0000 15 40 2.5 0.0 3 70 51 3 0 0 0 022 Cole. 30 32 100 00 0000 12 --‘ --- ~-- - --- -- - -- -- -- 000 Cole. 30 40 105 30 10000 11 --‘ --— —-- - -—- -- - -- -- -- 2523 Colo. 39 30 105 30 10500 11 40 2.0 0.0 1 74 51 5 0 10 20 2500 Cole. 30 13 100 05 10900 12 53 3.2 0.0 1 01 53 4 0 0 10 2001 Cole. 30 05 105 33 10000 0 53 2.3 0.2 1 00 53 5 0 0 40 2007 Cole. 30 55 107 05 0300 12 53. 2.0 0 2 3 11 40 6 0 10 40 010 Cole. 30 30 107 14 0300 10 -- --- --- - --- -- - -- -- -- 2070 Colo. 30 05 100 20 11000 0 53 2.5 0.3 0 70 53 4 0 20 40 002 Cole. 39 20 105 00 0500 7 --‘ --- --- - --- -- - -- -- -- 2900 Utah 39 00 112 50 9500 0 50 2.0 0 5 I 90 50 0 10 4O 50 2052 Utah 30 31 112 31 0500 0 -- --- --- 2 01 50 5 0 30 50 2077 Utah 37 40 112 40 0000 9 50 2.2 0 1 4 00 50 0 0 10 20 2790 Cole. 30 30 100 30 9000 0 50 2.2 0 I 2 102 01 7 0 30 40 2500 Colo. 30 00 105 10 0000 7 53 2.0 0 1 0 90 50 0 0 10 40 2012 N. lex. 35 12 100 27 10500 7 40 1.0 0 0 00 51 0 0 20 50 Standard devlatlon of a progeny aean 0.73 2.04 0.10 0.22 1.00 3.70 1.27 0.70 5.01 0.52 7.40 L00.05 between progeny aeana 2.55 10.3 0.50 0.17 3.70 13.0 4.45 2.04 17.5 33.3 20.2 L50.01 between progeny aeana 3.21 12.9 0.10 0.97 4.74 10.7 5 50 3.40 22.1 41.0 32.0 ‘ Collected for the pllot etudy -- no parental aatertale coIIected. [BY TO DIORII OP LIA! BIRINEAWION OIADIS Grade 0 - Serrulatlona absent or extreaely a1nute and very acarce Grade 1 - Serrulatlona tee 1n nuaher and a1nute but readily dleeernable Grade 2 - Berrulatlona proalnent and nuaeroua 2.6.333: can unocfiaouq mnogadfiuom u N 0095 oananuoomav mauuooa pan manque cad nonanq :3 Bow mucuunfishuom I H ovauc oouuom huo> use caucus afloaouuxo ho uaomna mucuueazuuom I o ovouo mug zanfiggum a mo gown Oh. E 3O .vOuooaaoo maauuoaaa aauzoaaa on an mosum vegan on» now uopooaaoo a o.~o o.Ho ~.Nu mo.” on.m >.on vu.v uo.o os.o o.n~ ~u.n mcama anououo awesomn Ho own u.on n.no n.u~ vo.m nv.v o.nH ou.n uu.o oo.o o.oH on.u magma anemone cowaoon no no; ov.u «o.o Ho.m os.o no.3 ou.o no.5 mo.o oH.o «o.m op.o cams acououo a mo soaoaa>mo ouaocaom ooH oo oo m on «55 o o.o o.H on o oops on won Ho Ho maxop soon oo on on m on and o o.H o.H Ho n oooo we mod on no .xo: .2 moon on on o o no boa o u.o o.~ oo o oouo ov mod on no .xm: .z ovom --- --- u- I- -- --- -- --- --- a-- v oooo on noH on no .xo: .2 moo --- --- -- -- -- --- -- --- --- a-- h ooou ov mod on on .xo: .z «no oon ooH on mg on «an «H o.H o.o oo o oono o3 odd om Hm .uwu< moon odd om on oH on «HA o «.3 o.o oo o oooo o5 mod me an .n3t< anon on“ own op on on own 5 o.H >.o oo o oono Hm oHH Ho Ho .Nfin< mono ooa ooH on «H on on“ «H v.H p.o no a comp n5 ooH on do .uou< ovum oo oo oo o no and u o.o o.o on o oooo ov odd om no .Nflu< Home oo oo oo o Ho and o o.H m.H o» v oovo on ooH ov no .Nnu< moon oo oh on o no pad 53 o.H o.o A» o oops om oHH on no .unu< moon on on on o oo hofl o o.H 3.3 as u oovn an 555 on on .nflu< noon on o o n on oo o o.o o.H oo o oooo no HAM on no .nfln< ooom --- --- u- u- .. --- .. --- --- mu- o ooHo ov HHH on on .Nau< ooo --- --- -- -- I- nu. -- --- --- an- m oooo ov HHH om no .~«»< mom 0 o o .EE .5. .55 .EE ovdum .0: .EE .0: you“ . o . o Ammo Ammo Avuo Ammo Amoo Aflno homo Aoao Aodo A530 Aogo wm wm mu Wm m w am am mm an mm m mm mm mm m s u .J on u u Pu 43 cm on 1m 8 us 1... 08 d 38 33 3m .03 a a 03 M... NO. 03 a. A 31 .11. A3 9 or. OJ OI. 0.1 u3 a 09 01. Da 8 I. 1a. In. u no uu .Au. M w. on. .19 .41 0.0. J; 1 1 n u N OI 6.1 88 S I. a T. U B a B I. n D. 00 n /. I .10 D. u 90 30 .10 :0 mo 0 D. a 91 m 8 T. 5...... .4 3 I. 1.1. I ..Art I. u a O. / .n u m m w 1 s m 8 8 m. mm a a 8 SJ 8 e I. B A D. B a T. S .wmdxoauom oaoom onoo no ooumon .monoum an omufisdaadm .mmopp mflahomHQOMpm mdqflm mo moflpmflsopowhoso pHdow Mom wpwm .m oHQwE 31 serrulation, and stomatal patterning were chosen to correspond to similar measurements made on the seedlings. Cone characteristics were chosen to investigate points of controversy among the early descriptions. A single branch about 18 inches long had.been collected from each tree. Leaves to be measured were removed from the main stem.of the branch. They were taken from the central portion of the growth produced in the year prior to collection of the specimen. An average of 10 to 15 cones had been collected from each tree. 16. The number of sound seed per gram was calculated at the time the seed lots were weighed in preparation for sowing. 17. Leaf length was measured to the nearest millimeter. Five fascicles of leaves were measured for each specimen. 18,l9. Number of dorsal surface stomatal rows and degree of leaf serrulation were observed under the dissecting micro- scope using the same procedures employed for the seedling leaves. The leaves used for length measurements were used for these observations. 20. Length of peduncle was measured to the nearest 5 millimeters on five cones taken at random from.the collection. Mere accurate measurement was not warranted because of the difficulty of determining the point of attachment without destroying the cone base. 21,22. Cone length and cone width were measured to the nearest millimeter on each of five cones. Cone width was measured at the widest point on each cone. 23. Length of cone scale apophyses was measured on five scales from the central portion of each of five cones to an accuracy of l millimeter. 2h, 25, and 26. Cone scale reflexing was estimated as the angle between a line extending parallel to the adaxial surface of the cone scale and the adaxial surface of the apophysis. The estimates were based on scales in the center of the terminal, central, and basal one-third of each of five cones. 32 Statistical analysis.-—Two basic types of statistical analyses were applied to the data: viz. analysis of variance and correlation. Most of the computations were performed.by an electronic computer (MISTIC). In addition, the results of the analysis of variance tests for several characters were combined by using the ”Summation of Differences" approach of wright and Bull (1962). The combined analyses were used to evaluate the patterns of variation for any evidence of discontinuity. Analyses of variance were performed on the data for each nursery character. Plot means of the 278 progenies represented in all four replicates were used as items. The form of the analyses was as follows: Source of variation Degrees of Freedom. Parameters estimated Stands 60 0’2 + r 7T2 + rt 0’ 82 Trees within stands 217 0’2 + r 0132 Replication (error) 83h CT'2 Total llll Where: r the number of replicates = h and t I! the harmonic mean of the number of trees per stand = n.16 The appropriate value for testing the differences between individual trees, the standard error of a progeny mean (sit) is equal to the square root of the error mean square divided by the number of replications (r), viz. sit== Vfihror mean square . An r approximate value for testing the differences between stands, the standard error of a stand-progeny mean (sis), is equal to the square root of the mean square for trees within stands divided by the harmonic mean (t) of the number of trees per stand.multiplied.by the number of replicates (r), viz. Sis = [Mean square for trees within stands \I lt) x (r) A standard error of a stand-progeny mean is strictly valid only if there were no significant differences between trees within stands. ————f ' 33 Such differences were present in some stands but the amount of within-stand variation was generally small enough for the above formulas to be valid. Data from measurement of herbarium.specimens collected from.Sl stands in 1960 were also subjected to analyses of variance. There was no replication of single trees so only between stand comparisons are possible. To maximize sample size, observations from all trees sampled in a stand were included in the analyses even though some did not produce seedlings. The form of the analyses was as follows: Source of variation Degrees of freedom Parameters estimated Stands 50 0‘2 + t0;2 Trees within stands 3A2 0’ 2 Total 3‘95 The appropriate value for testing the differences among stands (sis) is equal to the square root of the mean square for trees divided by the harmonic mean of the number of trees per stand (t), viz. Sis = \/Mean square for trees (t) Differences among stands were tested by the methods of Duncan (1955). The appropriate standard error of the mean was multiplied by a factor from Duncan's tables to Obtain a "Least Significant Difference" (L.S.D.). A single multiplying factor was chosen to represent a rank difference of 20 for the complete experiment. This value was chosen to facilitate separation of stands which occur in or near the area where the ranges of the two taxa overlap for placement with the correct taxon. At the same time the value is not so large as to obscure differentiation within each taxon. Choosing a single multiplier tends to underestimate the significance of differences between similar means. In actual practice, however, there were very few instances in which the use of the single multiplier caused any loss of precision. In order to combine data from several characters the "Summation of differences" technique of Wright and Bull (l962) was employed. 3A These summations made it possible to compare the progenies as entities rather than trait by trait. Differences for the analyses were calculated by subtracting the least significant difference (L.S.D. 05) from the actual difference between two stand.means for each character. This process eliminated non-significant differences from further consideration. The remainder was multiplied by four and divided by the L.S.D. The factor "A" was an arbitrary one to eliminate the .05' need for decimals. The process was repeated for each character. Finally, the resulting values were summed to give a single value applicable to a stand-pair. The procedure described is represented by the following formula taken from wright and Bull (1962:36): ) A (Difference-— L.S.D. Z .05 L.S.D..O5 Two hypothetical examples may serve to illustrate the use and interpretation of the analyses. Consider a series of stands sampled along a given transect, e.g. latitudinal or altitudinal. In Example A, Table 6, each stands differs only slightly from.the adjacent one but by continuously greater amounts from more distant ones. In Example B, some widely separated stands, a and f were similar whereas the neighboring stands f and g were different. The pattern presented in Example B would usually be interpreted as discontinuous variation. Stand means were used as items for computing all possible simple correlations among seedling traits, parental traits, and geographic origin data. Only the 50 stands which were represented by both seedling and parental materials were included in the analyses. 35 Table 6. Examples of hypothetical, idealized, summation of differences tables. Example.A. Continuous variation pattern. Differences between stands listed below and on the left. Stands a E b l g c 2 l .E d 3 2 l d e A 3 2 l .E f 5 u 3 2 l 3 g 6 5 u 3 2 l _g_ n 7 6 5 u 3 2 l n Example B. Discontinuous variation pattern. Differences between stands listed.below and on the left. Stands a .8; b J. b c 2 l E d 3 2 l ‘d e l O l 2 .e f 3 2 l O 2 f g 12 ll 10 9 ll 8 .5 GENERAL NURSERY OBSERVATIONS Germination began on June 6, 1961 and continued during the summer and fall until the ground froze in Nevember. Some germination also occurred during May l962, especially among seed lots which had late germination during the first season. Second year germination was concentrated during a 10 day period. Several seedlots contained less than the desired lOO seed. A small percentage of the seed failed to germinate. Rodents consumed many seed and emerging seedlings even though poisoning and trapping were employed to reduce such losses. The combination of these factors reduced the stocking of nursery beds well below the planned maximum. Approximately one-fifth of the rows had fewer than 6 seedlings, one- fifth had 6 to lO, one-fifth had ll to 15, one-fifth had 16 to 20, and one-fifth had 21 to 25 seedlings. Fumigation of the nursery beds in the fall preceding planting with "MILON", a combination fungicide and herbicide reduced fungus and weed problems to a minimumt Application of “CAPTAN” as germination progressed further reduced damping-off losses. Minor weeding was done by hand in conjunction with regular observations and measurements. Because a test on some extra seedlings indicated that the "Stoddard solvent" normally used to control weeds in the nursery had a detrimental effect on the seedlings, at least during the first year of growth, it was not applied to the experimental materials. A sawdust mulch was applied to control frost heaving during the 1961-62 winter. The treatment was successful except in low spots where water was occasionally trapped on the surface. The spacing employed in planting appeared adequate during the two year period. There was no competition between rows for light or space. There was competition between seedlings within rows of the fastest growing progenies. The roots of most seedlings were long enough to interlace with those from seedlings in other rows. However, the resultant competition did not noticeably restrict or favor seedling growth. Seedling color and growth rate served to indicate that soil nutrients and moisture were maintained at adequate levels. 36 DIFFERENCES BETWEEN THE TAXA In discussing the differences between the taxa the author prefers to use the specific names originally proposed, Pinus flexilis for the northern taxon and E. strObiformis for the southern. It was felt that the use of specific names would help clarify the discussion by reducing the terminology. Seedling Differences Form and size differences among seedlings in the nursery that indicated the presence of different taxa were apparent within two months after germination began. .A portion of a seedbed showing differences at the end of the second growing season is illustrated in Figure A. Individual characters.--Data resulting from the measurement of seedling characters are presented as stand averages in Tables 2 and 3 for E. flexilis and E. strObiformis, respectively. Figure 5 summarizes the data of Tables 2 and 3 by areas of origin. In Figure 5, the tWo taxa appear to be distinct with regard to ’most characteristics. Stand-progeny means were used as items in analyses of variance to test the difference between the species' means for each character. All differences between the species' means were significant at the 0.1 percent level. F values ranged from a low of lu for the number of dorsal surface stomatal rows, to a high of 370 for l-year height. The analyses are available on request. A more critical series of analyses was performed to further test the differences between the species. In these analyses seedling progenies from the southern extreme of the range of E. flexilis were compared with those from.the northern extreme of the range of E. strobiformis. Six stand-progenies were chosen to represent E. flexilis; the three most southern ones from.Utah, the two most southern ones from Colorado, and the one from.northern New Mexico. .Four standrprogenies were chosen to represent 3. strObiformi53 the one from northern New Mexico and the three most northern ones from Arizona. The results of these analyses are presented in Table 7. There was very little overlap between species in the characters of cotyledon number, length of secondary leaves, and height growth. 37 38 .mvfldomong wcflaevvm macaw mooqohommfiw wnflpdppmdaafi cop anomhdo a mo Goappom .: vsdmflm (i) lean germination date 220‘ I- a o >“200., b-t o >n 8 1.0fi II I I I 160,, I JIILII A B C D I F 0 R l J K Area of origin (4) Cotyledon number 12‘ 10 _ ML! Are- ot origin Number @ w I )vllllllllll. (7) Trees forming secondary leaves in first year is Per cent in '1] [Jill I _.. n A B C D E P G H i J K Area of origin (10) Length of uocondnry lcnvca 80 60 - 40 4 2. I] i II I , A B C D E P G H l J K Area of origin (13) l—yvur hvluhl 80 1 60_ s E 10 - I 20 I I In, I I I A D C D E F G H i J K Arvn of origin Arena of origin SW Alberta. A B. 8" Wyoming, C N h C Colorado -u Eflsl Cunirnl iduho 1M 5 Uinh, 5 Colorado. Figure 5. Montana, ). Eusl Cuntrnl California SW Nebrasku E 52 Wyoming 0_ SW Douglas Cn., Culnrudn I; N NC" W'lltll Day of year Grade Grade 39 (2) Date oi bud sci (J) Length of growing season 1204 290 g 1004 II a 250 60. 1 I I I I 230 I I 7 40 I n I n A B D E I G H l J A B C D E I O i J I Area of origin Area of origin (5) Cotyledon length (6) Diameter of hypocotyl 40 35 . 30 q 3.0 _ 25 d I I I E 2.0 q I H n H 20 I I 7 , , 1,0 LL A I I I A B C D E F G a I J K A D C D E I G a i J K Aron oi origin Area of origin (8) l-)car loilugu inlur (9) 2-year foliage color 5 q d 3_ 54 3 2 . g 4 _ I- I I] ° L 1 I, . 3 l I I A] _ A B C D E F G N l J K A B C D E F G H I J K Area of origin Area of origin (11) Dvgrcc of leaf serrulation (12) Number of dorsal stomatal rots 1.5 a 1.0 4 3 . - 2.1 a 0 5 4 g g 1 ( I I I o J 1 l. 1 o , A B C D E F G H I J K A B C D E F 0 a i J K Area of origin Arel of origin (l4) Z-yvnr height (l5) Amount of second growth incrumunt I101 120 100 < I‘D-1 ‘IOfi on . l s 20 . H w ill , 0 . III I ,,, A n C 0 H I J K A n C D E F G H i J K W North Dakota. & NW Wynming SE Idaho, E NE Uiuh Area of origin KEY Mil—H strobiformis =— Aruan of origin New Mexico L N Arizona N . C 8 SE Arizona S New Mexico & NW Texas XL.— Mean values for seedling characteristics Area of origin of Pinus flexilis and Pinus strobiformis stands grouped by area of origin. Table 7. #0 Results of analysis of variance tests of differences between the southernmost Pinus flexilis and northernmost Pinus strdbiformis seedlings. Character Value of F resulting from the test of differences ‘between species (a) Percent of total variance attributable to differences between species 2 Number CrD 2 2 x loo .1. CfIB C7NJ . Mean germination date 6.7* 54 Date of bud set ll.l* 68 Length of growing season 5.7* #9 h. Cotyledon number h7.9*** 91 5. Cotyledon length 1.1 2 6. Diameter of hypocotyl l5.H** 75 7. Seedlings forming secondary leaves in the first year 3.6 35 8. l-year foliage color l3.l** 72 2-year foliage color 9.h* 6h 10. Length of secondary leaves 110 .8*** 96 ll. Degree of leaf serrulation 9.h* 6h 12. Number of dorsal stomatal rows 0.0 O 13. l-year height 1.011396% 96 ill. 2-year height 1+8.Ll*** 91 15. Amount of second growth increment 37.0*** 88 (a) 96* For each analysis degrees of freedom.were l and 8 for between- and -within-species variation respectively. Greater than 5.32 needed for significance at the 5 percent level. Greater than 11.26 needed for significance at the 1 percent level. Greater than 25.4 needed for significance at the 0.1 percent level. 111 With respect to these traits both stand-and single—tree-progenies could usually be assigned definitely to one species or the other. Date of mean germination, date of bud set, length of growing season, diameter of hypocotyl, first and second year foliage color, and degree of leaf serrulation were less satisfactory as diagnostic characters. There was little overlap between species if the means applicable to regions of origin were considered. However, there was considerable overlap if stand-progeny and single-tree-progeny means were considered. Cotyledon length, number of dorsal stomatal rows, and secondary leaf formation during the first year were of little value in dif- ferentiating the species. Variation was almost as great within as between species. Simultaneous consideration of several characters.--The summation- of—differences technique was used to combine the data from eleven characters: Date of bud set Length of growing season Cotyledon number Cotyledon length Diameter of hypocotyl l-year foliage color Length of secondary leaves Degree of leaf serrulation 1-year height 2-year height Amount of second growth increment ...: [.4 ...—I l—1 HAAAAAA U1 47w l-' O OO O\\n 4:00 R) VVVVVVVVVVV AAA/\A The summations for E. flexilis are presented in Table 8. The bottom line of that table contains the summations for progeny 903, the most flexilis-like of the E. strobiformis seedlings. Study of the summations in that line shows that, all traits considered, progeny 903 is more different from all E. flexilis than is almost any 3. flexilis progeny from any other in the same species. In other words, there is almost no overlap between the species. Species distinctness is also indicated in Table 9, which includes the summations for E. strObiformis and the two southernmost E. flexilis progenies (Nos. 2599 and 2612). Considering all traits, almost all Summation of differences for seedling characteristics of Pinus flexilis and one Pinus strobiformis stand—progenies. Table 8. Totei ditterence (in eu-etion-unite). between progeny lined belo- enti progeny on left. Prowl! . state . or province n :OO-‘OOOO fl glOO—tHOOON IOOOONOOOO Dal 2222 A1 ta. 2242 bat . 2252 but . 2572 but . 2227 bat . 277 but . 290 loat. 227 I. 2002 Alta. 2222 Alta. 0 2929 "0 2271 '10. 2924 Iyo. 0 2 1025 Idaho ldaho 2520 Idaho 207 21512 2 11 O 2212 Idaho 2902 Iyo. L12 2202 0 2 5 2 7 2 1059 Utah 10” Utah no Utah Utah 2125 Utah 1002 Utah 1041 2 2 lebr . 2222 2222 1 14 12 14 12 11 2522 Iyo . 2124 12 12 14 4 1 2724 Cal 11. 0 C010. 2542 12 12 12 21214 2219 Colo. 2522 Colo. 222 Cole. 202 Colo. 2522 2010. 2502 C010. 0 2201 Colo. 3927 Colo. Colo. Colo. 212 2272 11 202 Colo. 2 2200 Utah 12 2222 Utah 12 2217 Utah 10 11 10 2 10 12 10 41212 11 10 14 10 12 1 12 14 10 12 11 C010. 2 12 10 11 2222 C010. 10 10 2722 110 12 12 2 10 10 12 14 2212 0 11 12 10 12 12 12 12 11 11 2 12 10 10 2 21210101012121012 2212 2.2.8. 22 12 12 202 12 12 14 10 12 15 22242224 228222127211225 1212 22 2.2442222224342227222112 21272432432725” 903' Arie e. lu-etien-unit - Z“ Actual difference - bl 110.05 ) / 1.83.05 for 11 ditrerent charactere. um etend included tor coneretive purpeeee. #3 .mnopowhmno poonoMMHd AH 90% mo own \ Amo own |.eoqmneoene Heepoev: w n pens-eonpeeesm e H m o o o o o o o o o o o o o _ e e om mm waxes seem mmmm o m m w o e o o e o o o m 0 am mm .xez..z mmmm mom 0 o o o o o o o o o o H : mm mm .xez .z mom meow o o H m o o o o o o a e om mm .me: .2 meow mmm m o o o o e o o o o 0 me am .xmz .z mmm e m a H o m o o a A an we mm mm .enea. mmem Anew o o o a o o o m m em om .Nen< Hmem mmem m m m o o a o e we we .Neea mmem m mm o o o o o o o om em .Nnha. mesm Hmmm a o o o m 0 mm om .Nen< ammo mmmm o m 0 me on mm m: .Nen< mmsm mmoa o o m : mm mm .eeea mmoa owm o o 0 an em .Neea mowm mmmm m H mm mm .Net< mmmm how 0 me me .Neea. eom mom me me .Nee< mom mflauowflpompm mdqflm mamm o .xmz .z mamm mmmm .oaoo mmmm mnaexman eaten .pwoa map oomfl>onm no anowonm can 30Hmp_©opmfla hoomonm moozpop oAmPHQSIQOHPmaESm QHV mononowMHo Hopoa no opmpm «hqowonm .moflnowonmtoqopm mHHonHm mdqflm 039 and mHaHOMHponpm msnflm mo weapmflaopoonogo mnflaooom pom mooqonommwo mo qoapdaasm .m canoe Ah the E. strobiformis progenies are more similar to each other than to lg. flexilis. Parental Differences Individual characters.--Data on parental characters are presented by stands in Tables A and 5 and by broad areas of origin in Figure 6. The sampling was less complete than in the case of the progeny characters. This was especially true of the northern E. strdbiformis area (Area "I") which was represented by parental material from only a single stand (No. 2569). The length of secondary leaves is the best diagnostic character for separating the species. Stand means range from A3 to 6A millimeters for E. flexilis and from.69 to 96 millimeters for E. strObiformis. Other characters for separating the species were: (1) The number of sound seed per gram, (2) The number of dorsal surface stomatal rows, and (3) The length of cones. The degree to which cone scales were reflexed provided distinct separation between species for all materials except those from.stand 2569. There was considerable overlap between stand means of both species for four characters: (1) Leaf serrulation, (2) Length of peduncle, (3) Cone width, and (4) Length of apophysis. Simultaneous consideration of several characters.--Values for all parental characters were included in the summation-of—differences analyses which yielded the results presented in Tables 10 and 11. The arrangement of these tables is similar to that of Tables 8 and 9 with one or two stand values for the opposite species included for comparison. When all parental characters were considered the distinction between species was smaller than for seedling characters. Materials from two 3. strobiformis stands, No. 2569 from north-central Arizona and No. 2649 from central New Mexico were unusual. They were more different from.those of most other E. strobiformis stands than from materials of some 3. flexilis stands. These stands were flexilis—like in their parental traits and strdbiformis-like in their progeny traits. That might be explained in either of two ways. First, the sites on which the stands were growing were so atypical as to result in abnormal H5 (16) Number of seed per gram (1?) Length of secondary leaves (18) Number of dorsal stomatal rows 14 13 1 3 - 10 80 n h I 3 S 2 . E 2 a 3 . . 5 = EGO ~ '3 l . 6 . I I u I I I I H 4 , 11 n 40 | I o , , 7 W ‘ A B C D I F (1‘ H 1 J K A B C D B F H 1 J K A D C D I P H i J K Area of origin Area of origin Area of origin (19) Degree of leaf serrulation (20) Length of peduncle (2]) Cone length lJO 110 . 10 ~ 2 a ‘ E . 6 ‘ 0 E 90 -I g l. I] 4 - In °° I n H 2 ‘ I l I I o I l A A L l 0 I l a l 7 70 l 7 _ A B C D 8 F H I J K B C D E F H l J K A D C D B F H 1 J Area of origin Area of origin Area of origin (22) Cone width (23) Length of apophysis (24) Degree of cone scale rationing Terminal l/3 of cone 59 12 45 I1 10 J 3 I- 55. 8 1 3 30 q . o E 6 n c .4 50‘ 4 -I 0 15— 2 -I u I 5 l ‘15 . .. , O 7 7 0 2 L L A A a - A B C D B F N I J K A B C D B F H I J K A B C D B P H I J E Area of origin Area of origin Area of origin (25) Degree of cone scale reflexing (26) Degree of cone scale reflexing Central l/3 a! cone Basal l/3 of cone 90 105 75 d 90 _ . 60 75 3 “ g _ 2 E w M £3, 45 .. 8 60_ c a .e -4 o . e ... J0 —< 45 w 4 go - S a 15 _. I 30‘ I 0 I A , u 15 , A B C D E F H l J K A B C D I F B I J K Area of origin Area of origin KEY Pinus flexilis —- Pinus strobiformis = Areas of origin Areas of origin 8' Alberta. Montana. W North Dakota. & NW Wyoming 1. N New Mexico a N Arizona 5' Wyoming, SE Idaho, & NE Utah J. C & SE Arizona N h C Colorado K. S New Moxieo & NW Texas A. B C D. East Central California E East Central Idaho F SW Nebraska & SE Wyoming G SI Douglas Co., Colorado M. S Utah. 8 Colorado, & N New Mexico a Single stand collection. NSFG l 902, collected {or the pilot study; no parental materials other than seed available. Figure 6. Mean values for adult characteristics of Pinus flexilis and Pinus strobiformis stands grouped by area of origin. 1+6 .nauoahan o>Auaaanlou now ewe—:2: vulva ...—ch-hanu uaOLoAAAAv AA Low 00.93 \ 3:339:- 35E .0 A 00.93 i aoauaoAAAv A133 :N n uAaauaoAualaam .- @691. n I n A AA OA VA N CA nA 0A 0A NA NA AA A” o o o 0 0A NA 0A 1A 0 0A AA NA a MA AA oA 0A 0A NA NA NA 0A .NA.-< AQUON NAON n N A N n A. n 6 n vA I n A. h A A o A n 0 NA o n A n v N O Q n NA A. a a Q m n .KOli NAoN gnN o N A N a o o v AA r. v A. 0 A n N o 1 a 0A a A. o N N v A v N A. n a v N n N .voo mmnN oahN o o H o 0 AA r. «A A. A NA v o o A o n o a n o o A N n N r. n n h v v n n v .vou wth NNDN v N n A n n o o 1 a a O N A o A N O A AA 0 o N A o n o m A N N N v N "Adan hth cooN n o v a 0 0A a A 0A N A A N N c 0 0A N n A A A n n m N AA 0 N v v r. M £1:— 8mN INQN A o o O 0 O N n NA A N n o v v v A v o o o O A A o n n A o c N o .vou aha haaN N N o v 0 n 0 NA n N n N N A A A AA N A o o A o A A A n o O o o .voo NDMN ASN v A a N v n AA 0 N n o n v 0 A A. o o o o o A o A. N n o o N o .vou ACUN §N o n A OA 0 ON w w a 1 n A. n v 6A a v o n n o N N N u n A v m .voo §N nNNN N O o A DA n o ... A n v n 0 AA A A N o o o o o o n o o o o .vou NNnN NNnN 6 VA A 0N h a BA A. n n v a 0A a 0 a o 0 A v A A 0A v v n m .voo NNNN a A DA n n n A v v n A NA 6 N N A A o A o A n o o c A .vou aAmN 0 AA 1 N o o N A. A. n A n o o o O a h n 0A a A v o A. A .015 OvnN VNhN NN a A. 0A v n v A 0 VA n N n a v o A o A h A A o n .AAAIU QNNN nonN r. n N 5 NA NA NA 0A 6 v n v A. AA 0A NA AN 0A 0 wA mA AA AA .0». nwnN HNNN o o o N n o A n o o o O o A A 0 H o N A n A ...—nun nnmN AvoA o o n N A. ,o 0 o A A N o N o h n N n N v N flaw: AvoA 08A A V v a A n o N N N A n o o o n V v u a lay: 86A 80A N n u A A. o o o o o A o n n n o o n o in: onoA 30A n o 5 NA N n m n o o A v o m n N n n fie: 36A nahN N v n N A N N N N A v N A. N n N N Ada: nwhN 83 n 2 o v a u n n v n a v u u _ n .9? Sou NAQN b A o o o A o o n N A o o c o 05:: QADN oonN N n n a NA a NA NA A. A. a o h ofivA ommN nqu o o o o A o w n A o A n o oaavA awoA vomN o o o o o v N o o o o o .9». $NN Con o o n o n n o o o A o .3- :mn onNLA, A o N N A o o o c .0». mNmN boo A O N A v A A n A .MIA .2 hum O8 0 o o w 0 o o A .uncl com :5 N A N o o A O 5:0. :6 thN o A. A o N N Jan, houN mth A. A o N N ...—.01 menN wnON A n N A .uEOl unuN ova o o o Jae: ovnN nNnN O o .auA< nNON NOON O .IAA< Noon NQON .auz NOQN aunt's“. .5 .0uaum £qu :0 AdONOhN and {Son tau-AA manual:— aaaauon aAnuAE—uquuaalau a: 00:29::- Aauah ..Aawuoam .moAnomoag-wqopm mASMOMAQoapm msqflm ago and mAAAxOAm mdqfim mo mOApmAhopowMono pAduo pom mooqmawmon mo QOAPGBESw .OA oAQmB .mnmpoondno pnmsmmwflo AA Mom mo.mmq \ Amo.amq I.moqoamw%A© Addpoov: N." PAQSIGOAPaSSSm m memo o m m o o m o o o 0 ma ma em mmxma szwm mmmm o m o m m o o o o m NA om .xmz .z mmmm mzmm em AA om :m m m m m m a e .xmz .z memo mmsmlnbu. m o o A o m mm mm m: .Nfln< mmsm Amsm o o o A A a ma mm :m .Naaa Hmsm mm m o A s m w mm :m a: .Nfla< mmsm mdsm o m m a mm mm s: .Nfiaa. m4sm Hmmm o o 0 ma mm mm .Nfla< Hmmm mmHm o o m :A NN .NHH< mmsm mNOA m ma NA mm .Nfla< mmoa momm 0A ma :m .Nfla< mowm mw mmmm m w .Nfla< mmmm mASfiOAAnOapm msnflm mamm m .xmz .z mamm mmmm .oaoo mmmm mflaflxmam manam .pmmA map no knowoam monA>oad and 30Am3.©opmAA hammoam mmmepmn.dAmPAQSJQOApaaSSm QAV mommamwmflo Aopoa no apnea «hammoam .mmAQmwoamndqopm mAAAxmAw msawm .Aa.mapma 03p one mfiahomfinonpm mdqflm mo mOApmAampoomogo PAsoo Mom moomoamaeflo mo HOApnaadm #8 development of the parental trees. If so, their phenotypic appearance might not be a reliable indicator of their genetic potentialities. Second, the parents might be members of a hybrid swarm Which was surrounded bva. strobiformis pollinators. If so, the offspring would be expected to be much more strobiformis-like than the parents. However, they would not be expected to be indistinguishable from.pure .3' strObiformis as was actually the case. The primary question which the study sought to answer was: “Are there two distinguishable taxa within the Pinus flexilis complex?” When based on seedlings grown together in a nursery the answer is an unqualified "Yes". On the basis of parental performance under field conditions the answer is ”Yes" but with reservations about materials from stands 2569 and 26kg. The second question to be answered was: "If the taxa are distinct, what taxonomic rank does each merit?” The author feels that the differences are of sufficient magnitude that the taxa merit equal rank as species. This ranking is in accord with the general level of species distinctiveness recognized within the genus EiEEE' According to the rules of nomenclatural priority in the International Code 22 Botanical Nomenclature, Article ll, (1961) the proper name to be applied to the northern species is Pinus flexilis awmes. The proper name for the southern species is Pinus strobiformis Engelm.. DIFFERENCES BETWEEN STANDS WITHIN SPECIES The analysis of the materials for evidence of distinctness of species also revealed the presence of variation within each. The next logical step was to analyze the differences between the stand- collections for each species. Pinus flexilis Seedling characters.--The outstanding feature of the geographic variation pattern is the performance of progenies 2553 and 2563 from a small area near Pine Bluff, in southeastern wyoming and the adjacent part of southwestern Nebraska. These seedlings had the most cotyledons and the longest secondary leaves of any of the limber pine. They also set first-year buds latest, formed the fewest secondary leaves during the first growing season, and had the greenest leaves. They were among the fastest growing origins (see Table 2 and Figure 5). Progenies 907, 1085, and 2590 from east-central Idaho had the longest cotyledons. They were nearly equal to seedlings from the Pine Bluff with regard to cotyledon number, length of secondary leaves, and lateness of bud set. Also outstanding was progeny 902 from Douglas County, Colorado, It was from the easternmost collecting area in the state. Although not extreme in any character, it had considerably more and longer cotyledons, greater height, longer leaves, and darker foliage than other seedlings from the same latitude. There was only one collection from the western part of the species' range. Stand-progeny 272M from the Sierra Nevada of east- central California had the fewest cotyledons recorded. It also had short cotyledons, short leaves, and a slow growth rate. The main body of the species occupies high elevations in the Rocky Mbuntains from southern Alberta to northern New Mexico. Within this broad area, there were no trends evident in that portion of the range extending from southern Alberta to central Colorado. Progenies from this area had the shortest cotyledons and secondary leaves, were the yellowest, and grew the slowest. Southern Colorado, southern Utah, and northern New Mexico progenies were the tallest within the species. They also had long l+9 5O cotyledons and secondary leaves. When several traits are considered simultaneously by the "Summation-of-Differences" method (Table 8), stand progenies from three areas show differentiation from those of the main.portion of the range of the taxon. These areas are in5(l) east-central Idaho, (2) the Pine Bluff region of Nebraska and wyoming, and (3) Douglas County, Colorado. Seedlings from the more highly differentiated stands in these areas very closely resemble seedlings from the southern portion of the species range. Progeny of the single California collection are very similar to those from the main body of the species. Within the main range from Alberta to central Colorado, the differences between stands are small and mostly of random nature. Southern Utah and especially southern Colorado and New Mexico progenies show considerable differentiation from.the more northern ones. Parental characters.--The collections from the Pine Bluff area were again outstanding (see Table 4 and Figure 6). These specimens had the longest leaves, the most extensive leaf serrulation, the longest cones, and the longest peduncles. The cone scales were among the most reflexed. Cones from east-central Idaho had the heaviest seeds, as reflected in the low number per gram sample, and the most reflexed cone scales. However, the leaves from adult trees were among the shortest for the species in contrast to the long leaves of the progenies. The Douglas County, Colorado collection was represented only by seed. These were among the heaviest obtained for the species. Materials from the California collection were distinctive in several ways. The seeds were the lightest, the cones and their peduncles the shortest, and the cone scales the least reflexed. The leaves had the most rows of stomata on the dorsal surface and the least serrulation. Within the main portion of the species' range the specimens were remarkably uniform. Cones from this area were narrowest and their cone scales had the shortest apophyses. In general, materials from northeastern Utah and adjacent areas showed slightly greater 51 development than those from.Alberta, Montana, and northern wyoming or northern and central Colorado. When all characters are cOnsidered together the entire species shows little consistent differentiation (Table l0). However, the results indicate that the samples from within any given area were more variable than were their progeny. Qne collection from the Pine Bluff area and one from east-central Idaho are seen to be quite different from.all others. In general, it was not as easy to differentiate between the southernmoSt stands and those from the northern and central areas on the basis of parental specimens as it was by observing their progeny. Pinus strobiformis Seedling characters.--Within the portion of the range of the taxon represented in this study, variation was not extensive (see Table 3 and Figure 5). Seedlings from northern.Arizona and New Mexico had fewer and shorter cotyledons, thinner hypocotyls, shorter leaves, and more rows of stomata on the dorsal leaf surface than more southern origins. They also had the shortest growing season and the least serrulate leaves, grew the slowest, and were the lightest colored. Central and southern.Arizona progenies had the longest secondary leaves. The leaves from these progenies also had the most pronounced serrulations and fewest dorsal surface stomata. Progenies of southern New Mexico and northern Texas stands had the longest cotyledons, the thickest hypocotyls, and the fastest growth rates. They were about equal to Arizona progenies in length of growing season, number of cotyledons, and foliage color. The "Summation—of—Differences" analysis indicates that the northern stand-progenies have differentiated slightly from the southern ones (Table 9). Within all three areas the variability is nearly random. Parental characters.--The single northern Arizona stand from which parental materials were available was very different from.more southern stands in Arizona, New Mexico, and Texas (Table 5 and Figure 6). The specimens from this stand had the shortest leaves of any. The leaves also had the most rows of dorsal surface stomata. This collection had 52 the shortest cones, the shortest apophyses, and the least cone scale reflexing. Among other Arizona collections there were slight north to south trends. Seed weight, leaf length, length of apophysis, and cone scale reflexion increased from north to south but cone length and width decreased. Central and southern.Arizona trees had the longest leaves with the fewest dorsal stomatal rows and the most pronounced serrulation. Their cones were longest and widest, had the longest apophyses, and showed the greatest amount of scale reflexing. Cone: scales with apophyses 25 millimeters long and reflexed into a full curl were Observed. Specimens from.New Mexico and Texas stands showed slightly less development than those from central and southern.Arizona in almost all characters. When all characters were considered together the northern Arizona stand appeared to be very different from other P. strdbiformis stands and almost intermediate between P. strobiformis anqu. flexilis (Table ll). However, when judged by seedling characters it was definitely associated with P. strObiformis. A slight amount of differentiation between the Arizona and New Mexico - Texas populations appears to have taken place. Area—of—origin groupings used in preparing Figures 5 and 6 were rather arbitrarily made by utilizing a combination of geographic features and the results of the "Summation-of—Differences” analyses (Tables 8, 9, lO, and ll). The outlines of these areas are illustrated in Figure 7 superimposed on the map of the collection locations. 53 ‘s ~~-~ Q ~‘ ‘s “~’ . 2579 - 967/ w \ " 0 990 ,I .2929 x' .‘597l ’1’ :‘ \ o 2964 ,o’ I '9073 .--------’ ’/ M5" Jams 0 2909/" ‘ #2765 1" __ IOBO-O "I059 ..2553 I005 l'—I‘§:, .2546 ‘ ‘2553 3‘1” . ~ ~':---- “04' O 29l9 ‘ ______________. § 95 L.-.“ I, 313 1 o 2900.. “"~' '1' __ o 2652 .' . ----.. k W : o 2724 ' l ---; ---‘li KEY emu-S MUS. ------ Piflgfi firgbifgtmis -------- 2763’ Figure 7. Boundaries of area of origin groupings of stand collections. DIFFERENCES BETWEEN PROGENIES WITHIN STANDS Preceding evidence has shown that the taxa and stands within taxa differed. The next step is the examination of the stands for differences among the individual trees. Sampling of parental materials involved the collection of only one herbarium.specimen from each tree. With one specimen per tree it is possible to derive a figure showing the amount of variation within a stand but it is not possible to state whether the variation is due to genetic or environmental factors. That can'be done only with a replicated test, as when the progenies from each parent are represented several times in the nursery. For that reason all conclusions as to within-stand variability are based on progeny performance. Only stands represented by four or more individual tree progenies were considered in the analyses. Significant differences between progenies within stands were found for all 13 traits analyzed. Eight of the traits were selected for presentation and further discussion in this report (see Table 2). The characters selected to illustrate the variability are:(l) Length of growing season, (2) Cotyledon number, (3) Cotyledon length,_(h) Trees forming secondary leaves in the first year, (5) l-year color, (6) Secondary leaf length, (7) l-year height, and (8) 2-year height. Dates of germination and'bud set were not presented because they are functions of length of growing season. Height was considered to be a better measure of l-year growth than stem diameter. The patterns of within-stand variation in first- and second-year color were similar. No test of within-stand differences was possible for leaf serrulation and number of dorsal stomatal rows because observations were made on only one replicate. Two-year total height includes the amount of growth during the second increment so the latter measure was not presented. Differences between progenies within stands are numerous in comparison with the small number of differences between stands from most regions. Significant within-stand differences in length of growing season were found in approximately half of the stands. The stands with differences were uniformly distributed throughout both species. In slightly over half of the P. flexilis stands there were 51+ . ‘ . e . . > . , . . u - u ( . . . . , .. _ a . ' . u a .. _ w .L l‘ .. v . . - . . \- . . .. J -. . a u . 4m ‘ , . . .. 1. ... I. . s. u a“! - g .. u v . . ... . . . .. . a ,- . .. i. . ' n. p. . . . . . . . . . I a . . . , .. x a . - - ». . . L \. a a. . , ,. \. ...... u . . u c- ,- . - u c a . . u u . t. .,. . I \ . . ‘ . . _ ‘ . . t . . I . ‘ . .— vL . . 'u - _" u A . _ . .- . .. I 'i. ' . .. "-n' .... ‘ ' 9 ‘ > - .. . . . ‘ - . .. . . . . . . a... . N Table 12. 55 Differences between progenies within stands for eight selected character S. E3 E2 number Length of growing season Cotyledon number Cotyledon length secondary leaves Secondary l-year l-year color leaf length l-year height 2-year height Pinus flexilis 2992 3002 2825 28h6 2856 2697 977 967 2929 1085 2909 1030 2553 2563 272M 922 909 2523 2509 2601 916 2978 2900 2652 2796 a + + I + + + + l+| + + + + + + + + ++III + + + + I I + + I + + + + I Pinus strobiformis 2569 2805 1023 2621 27h3 + I+++ l I I + + I I I + + + + I + + + + + + + + + + + + I + + + I + + I I + + + + + I + I I I + ' + + + + + + + + + I + l l + _ + + + + I I I I + + + + + + + + l + + + + + I I + + + + + + + + + + + + + + I + + + + + + + + + + + + a. . . + indicates stand. - indicates a significant difference between progenies within the no significant difference between progenies within the stand. 56 significant differences in the number of cotyledons per seedling. However, only within one of five P. str0biformis stands were differences evident. Within-stand differences in cotyledon length were present in almost all stands of both taxa. One might expect the number of cotyledons per seedling to be genetically controlled. However, cotyledon length is pr0bably strongly related to seed size or weight because cotyledon growth is nearly complete by the time the reserve food supply of the seed is expended. Correlation coefficients presented in the following chapter indicate that seed weight is related to'both measures. However, the use of plot means in the analyses may have masked the extent of the relationships. Pinus flexilis progenies within half of the standsoH vacuums A an uaaoauficmwm a «o o=Ha> moo. ohm. mum. so». vms. man. man. A * no ~a>aH unmouma m »a ucaoauH:Mam a mo o=Ha> m NH m a b nu an mammnaca aw mmwcmmouq «o amnsaz wauxwmwou mammnaoma *snmm.+ ..uom.+ *wnm.+ swam.+ «nu.+ osm.+ **mmm.+ oHaom once no omummn Ammo mo nausea Ammo amv.' woo.+ ons.+ oun.+ oom.+ oom.+ *samm.+ mamsaaona mo sumac; Ammo nun“; ocoo Ammo wawxmaumu owm.+ mov.+ vsm.+ «om.+ uno.+ nn~.+ **o~v.+ «Haom «:00 no omuwwn Ammo mnm.+ usv.+ mmh.+ uov.+ onv.+ mvm.+ **anm.+ mfimsnaoaa no sauna; Ammo camcofl omm.+ .ovo.+ «Hm.+ *wms.+ *.omm.+ om~.+ §*nms.+ can“: «coo Ammo ocoo AHNV ovv.+ *vmn.+ osv.- om~.+ vmm.+ vma.+ .mmn.+ mdmsnnoaa no season Ammo maocsumq vwo.+ ovm.+ som.- mmo.+ omm.+ o~¢.+ **mnv.+ numamd ocoo AHNV mo sumac; Aomv wcwxoamou Hmo.+ nmo.+ om¢.+ «do.- vom.c mvv.+ **vmv.+ ofiaom «coo no omuuon Away mfio.u mv~.+ ann.+ «mo.- van.- {*nmw.+ *nov.+ mumsnaona Ho cameo; Ammo cofiumasuuom man.+ *vnm.+ omm.u mo~.- moo.u «mow.+ mac.+ can“: mcoo Away Mama mmv.+ *hum.+ moH.+ «vo.+ ~mo.- om~.+ *ovn.+ nausea waoo name we woumon Amao wafixwaumu mach *mmn.- *wmo.- nHv.+ ~mv.- mnH.+ ofiu.- ma~.- maaom ocou “o mmummo Ammo Hauasopm *vmn.- **mws.u Ham.+ mom.- umm.+ mov.- an.- mamaaaoaa Ho sumac; Ammo Ho tonssz Amfio mafiXOAHmu gm~u.+ *omo.+ omm.+ vmm.- mmm.+ vvm.+ va.+ oamom onoo no omummo Ammo *oao.+ «mmo.+ vmm.+ mon.- snu.+ mv~.u non.+ mfimssaoaa «o spasm; Ammo *mmm.u osu.- *omm.+ msn.u *vsm.+ Hm¢.u *«Hv.+ can“; once Ammo mm>amH «mo.- mpo.+ vnH.+ Hou.u *nnm.+ nun.+ **«mv.+ nuwnmd anoo Afimv sumocoomm am«.- *osm.- Ham.+ mvm.+ .mmp.+ wma.- mnm.+ mach Hauasoum no umnasz Amflv co spasms AsHV mnuxofimou Hum.- omH.- «mo.+ mmv.- a~o.+ mma.u *mflv.- mflaom mcoo no autumn Ammo mom.u mon.n nno.- v~n.- vmm.n mHv.n .*Hmo.n mumsaqona mo sumac; Ammo Hoo.+ «no.- can.. van.- ..Hva.- mvm.- saoas.- spasms acoo AHNV sap» mvm.+ oeo.+ oHv.- oov.u oHH.+ omm.u **osv.u :ofiaaazguom mama no magmas “may you comm mmv.n man.- mmo.u m-.- sono.a w-.u *ovn.- mm>aoH suaocooom Ho sumac; ANHV mo umnasz floss .xo: .z an»: 2 .os: z mowoonm an»: m .oHoo o .omt mm .uzo: moaoQO .nfiu< masses .oHoo m .oHoo z ogauu 3m .ap~< magnum mfieuomanonpm macam mwafixmau mscwm mowmqaa :OwHaHouuoo scan; 0» mucuoaumno .mHEQOMHQoan msqfim cod mHHonHm mfinflm go myopowsdgo Hmpqohmm mooSPmp mQOprHohhoo .aa magma 63 within the Arizona population, the wider cones had shorter apophyses. The longer apophyses of E. strobiformis were reflexed to the extent that they did not proportionately increase cone width. Comparison of correlations between length of leaves and number of dorsal stomatal rows for seedlings (characters lO-12 in Table 13) and parental specimens (characters 17-18 in Table lh) reveals some striking reversals. Seedlings from southern Idaho-southern wyoming-northern Utah and southern Colorado-southern Utah-northern New Mexico with longer leaves had fewer dorsal stomata. For parental materials the longer leaves had more stomata. Correlations Between Seedling and Parental Characters Seed weight, as measured by the number of seed per gram, was strongly correlated with almost all seedling characters of E. flexilis if the entire species range was considered (Table 15). Within smaller areas, however, seed weight was significantly correlated only with cotyledon length and secondary leaf length. There were no significant correlations involving seed weight for E. strObiformis. No significant effect of seed weight was found on either first- or second-year height of seedlings from.within subdivisions of the ranges of either;E. flexilis or E. strobiformis. A seed weight—height correlation has commonly been found in early growth results for other species. The use of average seed weights and plot means for heights might have obscured such a relationship if it existed. Length of leaves of parents was not consistently related to any seedling characters Of.§' flexilis for limited areas. There was a surprising absence of correlation between parental and seedling leaf length for E. strobiformis. The number of rows of stomata on the dorsal surface of parental leaves was significantly correlated with the same measure on seedling leaves only for E. strObiformis. A correlation between the degree of serrulation of parental and seedling leaves was found for the species as a whole and for the most northern and southern populations Of.§° flexilis. Length of peduncle, cone length, and cone width were significantly correlated with seedling characters only when the entire ranges of the species were considered. 61L Table l5. Correlations between parental and seedling characters of Pinus flexilis and 3.2.13.5. strobiformis. Characters to which correlation aeglies Pirfl flexilis W Parental Seedling lntire Alta. 8' Idaho N Colo. S Cole. Entire Aria. characters characters species lent. SS Iyo. C Colo. 5 Utah species N Iyo. u um: s. Iex. (16) Nunber of (3) Length of growing season -.4l4‘ -.352 -.036 +.400 -.452 +.143 -.344 seed per (4) Cotyledon number -.645‘* -.474 -.613 -.237 -.336 -.217 -.567 gran (5) Cotyledon length -.76400 -.466 -.762' -.655 +.130 +.040 -.334 (6) Diameter of hypocotyl -.138*# -.l74 -.577 -.349 -.478 +.033 -.279 (7) l-year secondary leaves -.336* +.07l +.091 -.069 -.791 -.186 -.550 (8) l-year foliage color -.378* +.07l +.l54 -.309 -.439 +.043 -.332 (10) Length of secondary leaves -.67400 -.260 -.573 -.806* +.596 +.113 -.120 (13) l—year height -.682t* -.369 -.372 -.l79 +.062 +.065 -.472 (14) Z-year height -.665¢# -.319 -.278 -.234 -.063 -.131 -.183 (15) Second growth increment -.548" -.271 -.026 -.550 +.539 -.033 +.l7l (17) Length of (3) Length of growing season +.354’ «537. -.3as +.021 -.160 +532. +.sasu secondary (4) Cotyledon number +.260 -.35l +.461 +.634 -.176 +.576‘ +.766‘ leaves (6) Dianeter of hypocotyl +.3Sl* +.078 +.521 +.519 -.665 -.340 -.251 (8) l-year foliage color +.356'I +.134 +.078 -.024 -.355 -.325 -.232 (10) Length of secondary leaves +.503*t +.388 +.306 +2261 +.423 +.327 +.382 (13) l-year height +391: +399 +.4ss +.179 -.391 -.198 -.114 (15) Second growth increlent +.266 +.204 +.063 +.446 +.464 -.604‘ -.626 (18) Number of (3) Length of growing season -.301 -.506 -.046 -.331 -.744 -.600¢ -.683* stomatal (4) Cotyledon nunber -.353# +.113 +.l72 +.086 -.708 +.096 +.058 rows (10) Length of secondary leaves -.222 -.70l$‘ +.155 +.002 -.358 -.658‘ -.732‘ (11) Degree of leaf serrulation -.122 -.316 +.404 +.355 —.313 -.601* -.7220 (12) Nulher of Ito-Ital rows -.078 -.417 -.090 +.177 -.288 +.747‘* +.702‘ (13) l-year height -.381‘ -.551 +.389 -.314 -.781 +.110 +.061 (14) 3-year height -.423*‘ -.450 +.04O -.218 -.839 +.395 +.335 (15) Second growth increnent -.393# -.810* -.l79 -.195 -.306 «.325 +.269 (19) Degree of (4) Cotyledon nulber M423“l +.308 +.275 +.284 +.650 -.001 +.092 leaf (5) Cotyledon length +.473#’ +.16l +.078 +.568 -.660 -.013 +.084 serrulation (6) Dianeter of hypocotyl +.483‘* -.311 -.086 +.933'* -.431 -.022 +.311 (10) Length of secondary leaves +.517" +.150 +.119 +.358 .000 +.384 +.236 (11) Degree of leaf serrulation +.394‘ +.789*O +.286 +.357 +.918# +.109 .000 (13) 1-year height +.424‘* +.138 +.185 +.757‘ -.554 -.081 +.077 (14) z-year height +.501#* +.325 +.604 +.158* -.126 -.032 +.010 (15) Second growth increlent +.424¢O +.423 +.585 +.732* -.167 -.212 -.188 (20) Length of (4) Cotyledon nulber +.427** +.303 +.408 -.486 +.840 +.03O -.057 peduncle (10) Length of secondary leaves +.410‘ +.158 +.706 -.338 +.263 +.6750 +.443 (13) l-yelr height +.338‘ +.317 +.060 +.322 +.340 +.376 +.451 (l4) 2-year height +.357’ +.323 +.336 +.216 +.497 +.077 +.059 (21) Cone (a) length of growing season «412. +.019 .000 -.102 +.319 +.251 +.osa length (4) Cotyledon number +.493*' -.372 +.421 -.383 +.047 -.091 -.116 (5) Cotyledon length +.665*t +.34l +.547 +.612 +.435 —.24s -.260 (6) Din-eter of hypocotyl +.626#* +.072 +.343 +.009 +.019 -.175 -.070 (B) l-year foliage color +.479¢‘ -.187 -.139 +.303 +.825 +.358 +.533 (10) Length of secondary leaves +.684** +.396 +.523 +.454 +.395 +.456 +.309 (l3) l-year height +.660i‘ +.434 +.270 +.155 +.l53 +.010 +.O73 - (l4) Z-yeer height +.590*‘ +.242 +.281 -.079 +.438 -.054 -.038 (15) Second growth increment +.450#* +.232 +.l48 +.043 +.536 -.163 -.080 (22) Cone (4) Cotyledon number +.334‘ +.199 +.634 -.131 -.748 -.429 —.509 width (5) Cotyledon length +.503*‘ -.189 +.748 +.366 +.867 -.365 -.306 (a) Dianeter of hypocotyl +400 -.509 +.608 -.213 -.«3 -.302 -.036 (10) Length of secondary leaves +.504*‘ +.062 +.623 +.573 +.381 +.ll3 -.292 (13) l-yenr height +.445" -.034 +.507 -.195 -.253 -.223 -.l52 (14) 3-year height +.359‘ +.024 +.364 -.390 -.349 -.004 +.160 (15) Second growth increnent +.355’ -.l4l +.l35 -.136 +.508 +.316 +.571 (23) Length of (3) Length of growing season +.397 +.321 .000 -.701 -.384 +.734#t +.828¢¢ apophysis (4) Cotyledon nulber +.469*‘ +.530 +.618 -.044 -.198 +.246 +.291 (5) Cotyledon length +3706” +.265 +.520 +.103 +.4o1 -.os7 -.001 (6) Diameter of hypocotyl +.562** +.017 +.826* +.098 -.648 -.604‘ -.373 (10) Length of secondary leaves +.732fi‘ +.585# +.506 +.5SS +.477 +.542 +.446 (11) Degree of leaf serrulation +.442“ +.870** +.491 +.363 +.075 +.531 +.517 (13) l-year height +.629‘* +.629* +.747 +.100 -.479 +.045 +.027 (14) 2-year height +.652‘* +.733*‘ +.190 +.353 -.140 -.504 -.648 (15) Second growth increment +.638*‘ +.763*# +.168 +.541 +.508 -.752** -.768‘ (25) Degree of (3) Lbngth of growing season +.100 +.206 +.l48 -.378 -.451 +.70l* +.572* cone scale (5) Cotyledon length +.502#* +.536 +.O42 +.754* +.l73 +.058 +.004 reflexing (10) Length of secondary leaves +.493#* +.220 +.344 +.609 +.568 +.487 +.423 (11) Degree of leaf serrulation +.368* +.273 +.743 +.302 +.233 +.484 +.406 (13) l-year height +.337* +.124 +.723 +.217 -.554 +.163 -.018 (14) 2-year height +.267 +.035 +.276 +.212 -.l56 -.412 -.669* (15) Second growth increment +.259 +.097 +.419 +.337 +.538 +.682t -.672¢ Number of erogenies in analysis 36 13 7 8 5 12 9 Value of r significant at 5 percent level (= t ) .325 .553 .754 .707 .878 .576 .666 Value of r significant at 1 percent level (= ** ) .416 .684 .875 .834 .959 .708 .798 65 A surprising result of the analysis was the finding of significant positive correlations between apophysis length and several seedling characters for the northern portion of the range of P. flexilis. Correlations Between Seedling Characters and Geographic Origin Data Latitude of origin was significantly related to most seedling characters of Pinus flexilis when the entire range was considered but only occasionally for smaller areas (Table 16). Although not significantly correlated over the whole range, there was a significant negative correlation between latitude and cotyledon number for two segments of the range of P. flexilis, the northernmost and southermost. For E. strObiformis, latitude was negatively correlated with length of growing season and cotyledon number and positively correlated with the . amount of growth during the second increment. Negative correlations between longitude and seedling characters of_P. flexilis when the overall range is considered, reflect a strong correlation between latitude and longitude. The latitudinal effect is prdbably the most important component of the correlation coefficient. . Fewer dorsal stomata on leaves of seedlings of.Arizona - as compared to Nevaexico and Texas — origins gave rise to the correlation with longitude. A decrease in latitude appears to compensate for increasing altitude of origin for more southeran. flexilis collections so that no significant correlations between elevation and seedling characters were detected. Elevational effects appear more consistent among P. strobiformis progenies. The relative lack of origin—progeny correlations within areas means that it did not pay to get detailed origin data within regions. Except for 2-year height (inversely correlated with latitude of origin for the Alberta—Montana-northern wyoming area of P. flexilis) it was impossible to forecast progeny performance from origin data. Correlations Between Parental Characters and Geographic Origin Data Correlations between both latitude, elevation, and parental characters of P. flexilis are very erratic (Table 17). However, for .P. strobiformis there were several consistent and highly significant 66 eon. see. one. van. see. «no. use. A e. no Ao>eA «nausea A e. eeaoeuemumw u do ease> ace. Oh». one. son. var. man. man. A e no ~e>o~ «sconce n as unuouuenuam u «0 o=~e> a «a n m!‘ b ”A on edemaecs :« uoacmuoum no heels: spbh.n eeanh.n v~v.+ nav.n oov.+ AoA.n ~0~.n umeonuon meson once we oouusn Anuv sobh.l eev0¢.n Hon.+ ouo.n ovo.+ «H«.I ooH.n nanhnnons no nausea Away ooo.+ .Aoe.+ nav.a oo«.+ esaoo.+ «An.+ th.+ uses Asa-scum uo tonssz Away soees>oAm esona.n coca..- nuv.+ abn.+ onu.u oo~.+ ~o~.+ asexuauou canoe once «0 acumen Anny coca..- oceae.n ooo.+ «ne.+ Ane.u evoe.- eeeev.u unmanned. «0 nausea Anne «in.n 06“.- eno.a ah~.+ ove.u «no.3 sewn.) cAocsooa no nausea Aomv sshuo.u eenvh.u ebua.+ oaa.u s~n¢.u hhn.+ e«~.n mo>eo~ Audenooee «0 nuance Ahuv oneness; .uea .2 he»: 2 .os- z seduce. noun a .o~oo o .0». an .ucOI seduces ease .fl«h< Ghana" OHOO Q .OAOO z O‘CUH .m .UUH< Ohauflfl ahfi¢0dhdno AIHCOhIQ OunnthOOO suluouunouue conga Idnwxeau Isaak nouamms ecuusuoquO nodes 0» muouueuseu now send sflwfleo oflflmwpmoom one myopoohozo finenessm soospon wooepoaopsoo .mHEMOMHQOMpm macaw one mHHonHm macaw .se magma «as. no». woe. van. use. ens. .Av. A e: no Ao>oA “seated A e. encodes «a t «0 o=Ae> one. one. one. see. van. non. n«n. A e no Ao>oA «season n e. unsuauenudu t we oaAu> O NH 0 Q 6 DA on nuhhudfid flu nvwfiquhm.HO BODIDZ onn.+ esae.+ ovn.+ mwm“- auo.- mmm.- oAn.+ ego-atone assoc vacuum Anne nvv.- .ooa.u ooh.+ uno.- aeo.+ v~o.- uno.- soeeessutoo «soA «o autumn AAAV enn.- enoo.- onu.+ po«.+ noA.n nao.+ voo.+ someo- sssuotu no sauces Ame eoeea>esu «so . oeo.+ Aeo.- noo.+ ecA.+ omv.u .eon.- season tees-" Ansc euA.- e0en.n evA.n «ou.u ooo.u Amo.- e«on.- each essences «o tosses A«_o has - oe«.r ooh.- onm.+ «on.+ vAn.- enem.u Aseooonss «o nose-see Ame Ava - nuA.+ voo.u op«.- eeA.- a««.u eenon.- some». season» we season Any oeaeeweos echo + eeao.+ nes.+ neo.- oAA.+ eeea.- .eoon.- sequences senate enouom AnAc Aev.+ eee.+ ooe.- mn«.- oeo.+ eeeee.- ..onn.- unseen tees-“ Avso nnA - opA.- no».. one.. pa«.- «sn.- .eoav.- ensue: teoA-A Anne onA.+ oeo.+ Hoe.- eon.+ evo.- Ano.- .«oe.- upon Aces-oee co tones: Anne «pv.- vvn.- noo.+ ae~.+ oov.- so«.- .enn.- ou>eos synagogue co nausea AoAc opA.+ oeo.+ voo.n 5.5.- «av.u hAa.- eeeon.- Aseooonsn uo nonessee on vua.u Aa«.u e~o.+ ouo.+ aav.u noo.u eenuv.- nuusoA :oeoAAeoo Anv noo.u .osn.- enao.- oa«.+ v~n.u eooo.u va.- ton-s: :oeeAAeoo Ave nan . eave.- ves.- AeA.n noo.+ «eo.- .evuv.u cameo. seasons do season Any assesses .mum..z .mu»: z .o». z unecon- aa»: a .voo o .ohs um .eeoa noncon- seen .ueu< one»:- .voo a .voo z on¢e~ an .eeA< unseen whose-gene seeAeoom oaeeetsooo sumhouwnouua esnwa eenaueAe eased OOanN‘ flOflHCAOflhOO Buwfl’ OH OHOHOIkISU pom spec sflmflpo ofismosmoow one myopoohozo msflaooom somepon muoflpoaossoo .maammwammmmm mmqmm was mHAmeac mamam .wa mamas 67 relationships. Length of secondary leaves, apophysis length, and degree of cone scale reflexing all decreased with increased latitude of origin. Higher elevation of origin was related to more dorsal stomata and shorter and less reflexed cone scale apophyses. These parental characters are the ones usually employed to separate 1P. flexilis and P, strobiformis, yet, all show gradation from the typical P. strobiformis condition toward.that of P. flexilis as the boundary between their ranges is approached. One would logically have expected.more origin-parent than origin-progeny relationships. Their relative absence indicates that either the right origin data were not measured, or the origin data chosen had little selective value. l DISCUSSION The Question of Distinct Species The taxonomy and nomenclature of the Pinus flexilis complex have been an almost constant source of controversy since the discovery and naming of P. flexilis by James (1823). Most of the disagreement is centered about the status of that portion of the complex which occurs from extreme southern Colorado into northern Mexico. The main question asked is whether there is a continuous gradation between this southern portion of the complex and that from.more northern areas. If the gradation is continuous, or nearly so, does it result from.hybridization between two separate taxa or is it merely the transition between the extremes of variation in only one? Seedling materials from a major portion of the range of the complex When grown together in a common nursery indicate that two distinct entities are involved. Although the number of collections from areas where intermediacy might be expected is small, the seedlings are sharply aligned with one group or the other rather than being intermediate. This distinction is maintained also for seedlings from New Mexico sources where the ranges clearly overlap. Although retaining their distinctness, progenies of the two taxa from the zone where the ranges meet and overlap exhibit some variational trends which indicate that hybridization has occurred or is presently occurring. Parental materials, in contrast, exhibit almost a continuous gradation of variation in several characters. How can the two views be reconciled? The seedling materials represent the expression of heritable differences under a very limited range of environmental conditions. The parental materials reflect the interaction of heredity and environment under widely differing conditions. Most of the intergradation in parental materials was found in traits of the cone. The seedlings do not yet have cones to allow a comparison of the magnitude of genetic and environmental effects on these traits. Variation Within Pinus flexilis Within the range of P. flexilis, the northern member of the complex, three main expressions of variation were observed. The first was 68 69 the unexpected uniformity of both seedlings and.parental materials from collections made in the area from.Alberta to central Colorado. The common finding in other studies has been that over a comparable latitudinal range there was significant variation in several traits. The low density and scattered distribution of trees would be expected to restrict gene flow. With restricted gene flow the action of natural selection to effect adaptation to the environment would.be very localized. As a result, considerable variation would be expected from place to place throughout the range. The absence of differences between stands might be attributed to several factors. Two of the more apparent possibilities are: (1) lack of genetic diversity upon which selection could act, and (2) uniformity of selective forces throughout the area. The first proposal, however, is opposed by the finding of significant within-stand differences for several traits. The second seems hardly plausible because of the broad elevational and latitudinal ranges involved. These two variables tend to be Somewhat compensating, for at higher latitudes the trees grow at lower elevations. Numerous other environmental factors such as exposure, annual precipitation, and soil type vary throughout the area. All of these contribute to the selective pressure exerted in each part of the area. With so many factors involved it seems unlikely that their composite effects could be equal throughout the area. The second feature of the observed variation concerned the extent of development of seedling and parental materials from isolated areas. The northernmost area was located in southwestern North Dakota. The stand is separated from other areas Of.§° flexilis by a distance of approximately 200 miles. This collection was made at a lower elevation than any other. Materials from this collection were very similar to those from Montana sources. The collections from the Pine Bluff region of wyoming and Nebraska were considerably different from others made at nearly the same latitude. The area is separated from the nearest stands of .P. flexilis in the Rocky Mbuntains by a treeless plain nearly 60 miles wide (Goodding 1923). In most traits, the materials from this area were most like those from.the southernmost portion of the species' range. Leaves of both parental and seedling materials were longer than 70 others from similar or more northern sources. Seedlings from the Pine Bluff area also grew faster than all but the most southern progenies. The areas where collections were made in east-central Idaho are separated from other collection areas to the east and south by the broad Snake River valley. Seedlings from these areas were fast growing. They were similar to seedlings of southern origin in most ways. Both the Idaho and Pine Bluff stands occur at lower elevations. than other stands of similar latitude. Also, in both areas materials from one stand showed less extensive development of traits than those from the others. Progeny of the easternmost stand collection (902) in central Colorado were taller than those from areas slightly farther west. They also had longer leaves and cotyledons. This collection was obtained from the eastern slope of the Front Range whereas the other Colorado collections were from the western slope or farther west. In his study of variation in lodgepole pine, Critchfield (1957:6510 also found that materials collected On the eastern and western slopes of the Front Range differed in several characteristics. The western portion of the species' range was represented by a single collection from the Sierra Nevada in eastern California. Both seedling and parental materials from this collection were very similar to those from high elevations in central Colorado. The third feature of the variation found in P. flexilis was that progenies of southern Colorado, southern Utah, and northern New Mexico trees grew considerably faster than those from northern Utah and Colorado. In Colorado, the Arkansas River appeared to be the dividing line between slower growing progenies to the north and east and faster growing ones to the south and west. The cones from the southern trees were longer and wider than those from northern ones. They also had increased cone-scale reflexing, a trait usually associated with P. strobiformis. Variation Within Pinus strobiformis Variation between the northern and southern, and eastern and western populations was found in P. strobiformis. Progenies of northern Arizona and New Mexico sources grew slightly slower than more southern ones. They also had shorter . .Jl'J. . v 1 -" .v ‘_ .. ' It - . L. ~ ' V 0 w L .— -. J V I . . . k. , - . .. x t' - ..., M. ...J u .. . i \ -.. .- ' .lu' . \el .. .. '- u u c -. b . _ l ' _ _ \J.- . - \I' c__ _. . I. - . .u .3. --. a ~ v . . l . - ~- ' ...- 1. .. _-.1 . . - ..-.tli- -..- . . . .\n. _ . _ . . u - - l u .. LII... ' .. . _ U ._ ,1 . . l r I L. l ' . s 1.. . 71 cotyledons and secondary leaves. Parental materials were available for only one northern stand in Arizona. Leaves from this collection were shorter and less serrulate than those from more southern sources. The cones were shorter, and their apophyses shorter and much less reflexed than more southern ones. Seedlings from New Mexico and Texas sources were slightly faster growing than those from.Arizona. However, Arizona progenies had longer and.more serrate leaves. When the adult materials are compared, the Arizona collections exhibit greater development of both leaf and cone traits than do those from New Mexico and Texas. It appears that the New Mexico materials show evidence of a slight amount of past or present hybridization with P. flexilis. This is reflected in their shorter leaves and cones, less serrulate leaves with more dorsal stomata, and less reflexed cone scales. In.Arizona, the valleys of the Colorado and Little Colorado Rivers separate areas where P. strobiformis occurs from.areas to the north and east Where P. flexilis or suspected hybrids between P. flexilis and P. strobiformis occur. A possible exception to the general separation may occur in the San Francisco Mountains near Flagstaff, Arizona. A specimen (Busby no. 831) seen at the University of Michigan Herbarium, collected in 1883, appears to bear typical P. flexilis foliage. There were no cones present with the collection. Three collections from the San Francisco Mountains were included in the present study. Seedlings from all three collections were distinctly of the P; strobiformis type. Parental materials were available for only one of these three collections. Foliage from trees in this collection was somewhat intermediate betweean. flexilis and P. strobiformis. The cones resembled those of P. flexilis in many' features. Thus it appears likely that P. flexilis has occurred in that area in the past if it is not presently growing there. Several recent attempts to finle. flexilis in the area have not yielded any likely specimens (personal communication, Dr. J. W. Andresen, Michigan State University). Differences between progenies within stands were more apparent for P. strdbiformis than P. flexilis. Height differences at the end 72 of the second year were especially noticeable among P. strobiformis progenies. Correlations Among Characters Most correlations between characters for P. flexilis were significant when the entire species range was considered. However, within smaller portions of the range only a few correlations were significant and their occurence was erratic. Thus it appears that few, if any, of the traits measured are causally related. For P. strobiformis, the correlations were much more consistent. Seed size has commonly been found to influence the amount of growth of seedlings during the first few years. No significant effect was found on either first or second year height within subdivisions of the ranges of either P. flexilis or P. strobiformis. The use of average seed weights and plot means for heights might have obscured such a relationship if it existed. However, observations on albino seedlings indicated that the seed provided nutrition to the seedling for only the first 10 to 15 days following germination. Most growth occurs after this time. Longer leaves had fewer dorsal stomata and more pronounced serrations in P. strobiformis collections but there were no consistent relationships for similar P. flexilis materials. Parental collections of P. strobiformis that had long leaves with few dorsal stomata also had cones with long and reflexed apophyses. There was a surprising lack of correlation between parental and seedling leaf length. The number of dorsal stomata on seedling and parental leaves was correlated only for P. strObiformis. A portion of the confusion encountered regarding the separation of the two taxa may be related to the effects that elevational differences have upon the cones and leaves of P. strobiformis. The leaves of higher elevation sources had more dorsal stomata and the cones had shorter apophyses. The cone scales were also less reflexed. Elevational effects in P. flexilis were significant in only one area and for one character. SUMMARY AND CONCLUSIONS The Pinus flexilis complex is composed of two populations of 5-leaved pines of the subgenus Haploxylon. Both members have cones that open at maturity and the seeds of both lack effective wings. The northern population has a range from southern Alberta and British Columbia south to north-central New Mexico. This population is almost universally known as Pinus flexilis James. Members of the southern population occur in northern Mexico, Texas, Arizona, New Mexico, and southern Colorado. The most frequently used names associated with the southern taxon are: Pinus strObiformis Engelm., P. flexilis var. reflexa Engelm., P. reflexa Engelm., and P. ayacahuite var. brachyptera Shaw. The primary purposes of the study were to evaluate the extent of differences between and within the two taxa of the complex. The results were to be used to attempt clarification of the names and ranks of the taxa. A review of the literature showed that most authors recognized differences between the two taxa, but disagreed about classification and nomenclature. The suggestion gathered from the literature on experimental studies of variation and Speciation in plants and trees was that samples of natural populations from various origins should be grown together in a common test area. This procedure would eliminate most of the differences due to environmental effects and expose the heritable differences. Collection of materials for the study began in 1959 and continued through 1960. The study utilized two sources of information. The first was drawn from observation of morphological traits of cone and foliage specimens. These specimens were collected from several trees in each of 61 native stands. The second source was obtained from seedlings grown for two years in a nursery at Michigan State University. Seed for this phase of the study was gathered from the same trees sampled for cones and foliage. The nursery test was established in 1961. A randomized block design with four replications was used in the test. Each plot contained seed from a single tree. Distinct differences between the two taxa were exhibited in the seedling test. Cotyledon number, length of secondary leaves, and 73 71+ height growth were the most satisfactory characters for distinguishing between the taxa. Both stand- and single-tree progenies could usually be assigned definitely to one taxon or the other on the basis of these traits. Diagnostic characters of less value were germination date, date of bud set, length of growing season, diameter of hypocotyl, first and second year foliage color, and degree of leaf serrulation. Characters of little or no value in differentiating the taxa were: cotyledon length, number of dorsal stomatal rows, and amount of secondary leaf formation during the first year. Traits measured on the cone and foliage specimens from.the parental trees exhibited less distinctive differences between the taxa. Secondary leaf length was the most reliable parental character for separating the taxa. Other traits which served to separate the taxa were: seed weight, number of rows of dorsal stomata, length of cones, and degree of cone scale reflexing. There was considerable overlap with regard to: leaf serrulation, length of cone peduncle, cone width, and length of cone-scale apophyses. It was concluded from the preceding results that the two taxa under consideration deserve separate specific rank. .According to the rules of nomenclatural priority in the International Code of Botanical Nomenclature, Article 11, (1961) the proper name to be applied to the northern species is Pinus flexilis James. The proper_ name for the southern species is Pinus strobiformis Engelm. I The patterns of variation in the regions where the species' ranges are contiguous or sympatric indicate that hybridization has occurred in the past and may still be taking place. Both seedling and parental materials of Pinus flexilis from southern Utah, southern Colorado, and northern New Mexico indicate the presence of some genes from P. strObiformis. There is also an indication that P. flexilis genes are present in the northern P. strobiformis populations. The cones from the parental collections eXhibited the most evidence of hybridity. Controlled hybridization studies to test these conclusions will be possible when the outplanted seedlings reach breeding age. Special field studies might detect the presence of hybrid swarms if hybridization is occurring naturally at present. Within the northern species, Pinus flexilis, the population 75 structure had three principle characteristics. First, there was very little variation in either seedlings or herbarium specimens from.their parents in that portion of the range extending from southern.Alberta to central Colorado. This uniformity was unexpected because of the common finding of considerable variation in other plants and trees from a similar latitudinal range. A portion of the uniformity may be attributed to the fact that the trees grow at increasing altitudes in more southern areas. The second characteristic was an increased growth of seedlings from southern origins. Cones from the parent trees were also longer than those of northern origins. These materials appeared to show evidence of immigration of genes from Pinus strobiformis. The third characteristic was the extreme performance of seedlings from three areas: (1) East-central Idaho, (2) The Pine Bluff area of Nebraska and wyoming, and (3) Douglas County, Colorado. Seedlings from.these areas equalled or exceeded.those from the southernmost portion of the species range in height growth and foliar development. The parental specimens also resembled those from the southern collections. The stands in these areas were isolated.from the main body of the species and were restricted in size. Selective forces acting within these sub-populations might more easily change the genetic composition than in extensive areas where gene migration was less limited. Variation between stands within Pinus strdbiformis was much more random than in P. flexilis. This randomness can.prdbably be attributed to the fact thath. strobiformis occurs only on small, widely separated mountain ranges, or individual peaks. Seedlings from the northernmost origins grew more slowly than those of more southern sources. They also had shorter leaves, fewer cotyledons, and more rows of stomata on the dorsal leaf surface than southern ones. These traits suggest the influence of P. flexilis genes. The parental specimens from northern Arizona differed from those of central and southern Arizona in several ways. They had shorter leaves, smaller cones, and less cone scale reflexing. Central and southern Arizona progenies had the longest leaves. The leaves from these progenies also had the most pronounced 76 serrulations and the fewest dorsal surface stomata. Cones from parental trees in that area were the largest of any. They also had the most reflexed scales. Seedlings from eastern New Mexico and northwestern Texas grew the fastest. Both seedling and.parental leaves of these sources were shorter and less serrate than those from central Arizona sources. Variation between trees within stands of either species could be satisfactorily investigated only from.the seedlings. Significant differences between progenies within one or more stands were found for all 13 seedling traits studied. Within some portions of the range of either species, differences between progenies within stands were almost as large as differences between stands. The wide spacing found in many stands would be expected to lead to considerable self-pollination or close inbreeding. If that has in fact happened a large amount of random within-stand variation could be expected. Mbst correlations between seedling and parental characters were significant when the entire species' range was considered as a unit. However, within smaller portions of the range, the corresponding values often were not significant. The number of observations in any particular area was too small for a completely satisfactory test. More correlations between seedling traits and geographic origin data were significant than those between parental traits and origin data. This finding suggests that either (1) the parental traits studied were not adaptive, or (2) the origin data chosen had little selective value. The study revealed the need for continuing research on several aspects of the problem, First is the need for more detailed study of samples from the area where the species are contiguous or sympatric. Second, materials from areas which were not sampled should.be examined. Among those areas are: (1) Northern Mexico, (2) Southwestern California, (3) Nevada, (4) The wallowa Mountains of Oregon, and (5) The Black Hills of South Dakota. Third, several areas, e.g., the Pine Bluffs, should be reexamined to determine the extent of ecotypic differentiation. LITERATURE CITED Baldwin, H. I., and D. Murphy. 1956. Rocky Mountain Douglas-fir succeeds in New Hampshire. Fox Forest Note No. 67. 2pp. 68 Bocher, T. W. 1963. The study of ecotypical variation in relation to experimental morphology. Regnum Vegetabile. 27:10-16. Byrnes, W} R., H. D. Gerhold, and W. C. Bramble. 1958. Douglas-fir varietal tests for Christmas tree plantations in Pennsylvania. Progr. Rep. Penn. Agric. Exp. Sta. No. 198. 6 pp. Callaham, R. z., and A. A. Hazel. 1961. Height growth of wind- pollinated progenies. Silvae Genetica. 10:33-42. Callaham, R. Z., and A. R. Liddicoet. 1961. Altitudinal variation at 20 years in ponderosa and Jeffrey pines. J. For. 59:81H—820 Callaham, R. Z., and W. Metcalf. 1959. Pinus ponderosa - altitudinal races confirmed. J. For. 57:500-507. .\ I .’/l .\ . Carriere, E. A. 1855. Traite general des Coniferes. Paris. XV + 656 pp. Ching, K. K., and D. Bever, 1960. Provenance study of Douglas-fir in the Pacific Northwest region. I. Nursery performance. Silvae Genetica. 9:11-17. Clausen, J., D. D. Keck and W. M. Hiesey. 1940. Experimental studies on the nature of plant species. I. The effect of varied environ- ments on western North American plants. Carnegie Inst. Wash. Pub. 520. vii + #52 pp. Coulter, J. M., and J. N. Rose. 1886. Synopsis of North American pines, based upon leaf-anatomy. Bot. Gazette. 11:256-262. Critchfield, W. B. 1957. Geographic variation in Pinus contorta. Maria Moors Cabot Foundation Pub. No. 3. viii + 118 pp. Douglas, M. M. 1958. Intraspecific variation in Pinus flexilis. J. Colo.-Wyo. Acad. Sci. IV, No. 10:30—31. Douglas, M. M., and J. R. Douglas. 1955. The distribution and growth of the limber pine in Colorado. J. ColorWyo. Acad. Sci. IV, No. 7 #6—u7. Drummond, T. 1830. Sketch of a journey to the Rocky Mountains and to the Columbia River in North America. in Hooker's Botanical Miscellany. London, John Murray. Vol. 1:178-219. Duncan, D. B. 1955. Multiple range and multiple F tests. Biometrics 11:1—42. Edwards, M. V. 195k. A summary of information on Pinus contorta. Part I. For. Abstr. 15 389—396. Edwards, M. V. 1955. A summary of information on Pinus contorta. Part II. For. Abstr. 16:3-13. Eggler, W} A. l9hl. Primary succession on volcanic deposits in southern Idaho. Ecol. Monographs. 11:277—298. Ehrenberg, C. 1838. Vorlaufige Nachricht uber die mexicanischen Coniferen, vom Herausgeber. Linnaea. 12zh86-h96. Endlicher, S. 18h7. Synopsis coniferarum. Sangalli, Apud Scheitlin & Zollikofer. iv + 368 pp. 77 78 Engelmann, G. 18h8. Sketch of the botany of Dr. A. Wislizenus's expedition. Senate Misc. Doc. No. 26. 1h1 pp. + maps and charts. Engelmann, G. 1863. On Pinus aristata, a new species of pine, discovered by Dr. C. C. Parry in the alpine regions of Colorado Territory, and on some other pines of the Rocky Mountains. Transactions St. Louis.Acad. Sci. Vol. 11:205-21h. Engelmann, G. 1878. Coniferae of Wheeler's expedition. ‘in Report upon U. S. geographical surveys west of the one hundreth meridian in charge of G. M. Wheeler. IV. Botany 255-26A. 1 Engelmann, G. 1880. Revision of the genus Pinus, and description of Pinus elliottii. Transactions St. Louis Acad. Sci. Vol. IV:161-190. Engelmann, G. 1882. Notes on western conifers. Bot. Gazette. 7zh. Gaussen, H. 1960. Les Gymnospermes. .Actuelles et fossiles. Fascicule VI, XI. Généralites, genre Pinus. Travaux du Laboratoire Forestier du Toulouse. Toulouse, France. Sec. 1. Vol. II, Pt. 2, 272 pp. Goodding, L. N. 1923. An interesting area of limber pine extending into southern Nebraska. J. For. 21:175—176. Hatt, R. T. l9h3. The pine squirrel in Colorado. J. Mammal. 2hz3ll-3u5. Hooker, Sir. W. J. 1838. Flora Boreali-Americana. Vol. II, Pt. 9:97-11h. International Code of Botanical Nomenclature. 1961. Adopted by the Ninth International Botanical Congress - Montreal, August 1959. Editors, Lanjouw, J. EE.§£° Regnum.Vegetabi1e. Vol. 23. 372 pp. Irgens-MBller, H. 1958. Genotypic variation in the time of cessation of height growth in Douglas-fir. For. Sci. 4:325—330. James, E. 1823. .Account of an expedition from Pittsburg to the Rocky Mountains. Philadelphia, H. C. Carey and I. Lee. Vol. II. #22 pp. + xcviii Appendix. Jepson, W. L. 1910. The silva of California. Mbmoirs Univ. Calif. Berkeley, The University Press. Vol. II. 480 pp. Kalela, A. 1937. Zur Synthese der experimentellen untersuchungen fiber Klimarassen der Holzarten. Communicationes Institute Forestalis Fenniae. 26:1—Ah5. Keng, H., and E. L. Little, Jr. 1961. Needle characteristics of hybrid pines. Silvae Genetica. 10:131—146. Langlet, o. 1936. Studier over tallens fysiologiska variabilitet och dess samband med klimatet. Medd. Statens Skogsf5rs5ksanstalt. 29:219-A70. Lemmon, J. G. 1892. Handbook of West-American cone-bearers. Pacific Press Publishing Co. Oakland, Calif. 2A pp. Little, E. L., Jr. 1950. Southwestern trees. Agric. Handbook No. 9. U. S. Gov. Print. Off., washington. ii + 109 pp. Little, E. L., Jr. 1953. Check list of native and naturalized trees of the United States (including Alaska). Agric. Handbook No. Ml. U. S. Gov. Print. Off., washington. iii + #72 pp. 79 Little, E. L., Jr. 1962. Variation and evolution in Mexican pines. 3g Seminar and study tour of Latin-American conifers. FAO. English Ed. No. 1. x + 209 pp. Loock, E. E. M. 1950. Pines of Mexico and British Honduras. Dept. Forestry, Pretoria, Union of South Africa. Bull. 35. x + 244 pp. Martinez, M. 1948. Los pinos Mexicanos. Ed. 2. Mexico. 361 pp. Mirov, N. T. 1961. Composition of the gum turpentines of pines. U. S. For. Service Tech. Bull. No. 1239. v + 158 pp. Mirov, N. T., J. W. Duffield, and A. R. Liddicoet. 1952. Altitudinal races of Pinus ponderosa - a l2—year progress report. J. For.‘ 50:825-831. Morris, W. G., R. R. Silen, and H. Irgens—Moller. 1957. Consistency of bud bursting in Douglas—fir. J. For. 55:208-210. Moss, E. H. 1949. Natural pine hybrids in Alberta. Canadian J. Research. Section C. Botanical Sciences. 27:218-229. NMnger, T. T., and W. G. Morris. 1936. Growth of Douglas-fir trees of known seed source. USDA Tech. Bull. 537. 40 pp. Monger, T. T., and W. G. Morris. 1942. Further data on the growth of Douglas-fir trees of known seed source. Pacific NW. For. Exp. Sta. Mimeo. 12 pp. Necker, N. J. de. 1790. Elementa botanica. Neowedae ad Rhenum. Apud Societatum Typographycam. 456 pp. Nuttall, T. 1853. The North American sylva. Philadelphia, Robert P. Smith. Vol. III. 152 pp. + 40 plates. Parlatore, P. 1868. Gymnospermae. E2 De Candolle. Prodromus systematics naturalis regni vegetabilis. Paris, V. Masson. Vol. 16. Pt. II:345-52l. Pauley, S. S., S. H. Spurr, and F. W. Whitmore. 1955. Seed source trials of eastern white pine. For. Sci. 1:244-256. Rydberg, P. A. 1905. Studies on the Rocky Mountain Flora. XV. Torrey Bot. Club. Bull. 32:598. Santamour, F. 3., Jr. 1960. Seasonal growth in white pine seedlings from different provenances. Northeastern For. Exp. Sta. For. Res. Note 105. 4 pp. Sargent, C. S. 1889. Notes upon some North American trees. XIV. No. 352 Pinus refleanEngelm. Garden and Forest. 2:496. Sargent, C. S. 1897. Silva of North America. Vol. XI. Coniferae (Pinus). 163 pp. + 55 plates. Sargent, C. S. 1922. Manual of the trees of North America (exclusive of Mexico). Ed. 2. Houghton Mifflin Co. Boston and New York. ‘ xxvi + 910 pp. Schober, R. 1954. Douglasien-Provenienzversuche. I. Allg. Forst-u. Jagdztg. 125:160-179. Schober, R., and H. Meyer. 1955. Douglasien—Provenienzversuche. II. Allg. Forst-u. Jagdztg. 126:221-243. 8O Schfitt, R. 1958. Zfichtung mit Kiefern. Teil I. Individualunterschiede und Provenienzversuche. Reinbek bei Hamburg. Bundesanstalt fflr Forst—u. Holzwirtschaft. Mitteilungen 40. 65 pp. Shaw, G. R. 1909. The pines of Mexico. Pubs. of Arnold Arb. No. 1. iv + 30 pp. Shaw, G. R. 1914. The genus Pinus. Pubs. of Arnold Arb. No. 5. 96 pp. Sluder, E. R. 1963. A white pine provenance study in the southern Appalachians. USFS Res. Paper SE-2. 16 pp. Squillace, A. E., and R. T. Bingham. 1958. Localized ecotypic variation in western white pine. For. Sci. 4:20-34. Squillace, A. E., and R. R. Silen. 1962. Racial variation in ponderosa pine. For. Sci. Monograph No. 2. 27 pp. Stanley, P. C. 1920. Trees and shrubs of Mexico. Contributions from the U. S. National Herbarium. Vol. 23. 1312 pp. Sudworth, G. B. 1897. Nomenclature of the arborescent flora of the United States. USDA, Div. of For. Bull. No. 14. vii + 419 pp. Sudworth, G. B. 1917. The pine trees of the Rocky Mountain region. USDA Bull. No. 460. 47 pp. Turesson, G. 1922a. The species and the variety as ecological units. Hereditas. 3:100-113. Turesson, G. 1922b. The genotypical response of the plant species to the habitat. Hereditas. 3:211—350. Turesson, G. 1930. The selective effect of climate upon the plant species. Hereditas. 14:99-152. U. S. Department of Agriculture. 1949. Trees — The year book of agriculture. U. S. Gov. Print. Off., washington. 944 pp. + xiv Appendix. Veen, B. 1952. Report of a tour of inspection of international provenance trials with larch, pine, and spruce of 1938/39 and 1944/45 and suggestions for future treatment and assessments. Sec. 22, IUFRO, Zurich. mimeo. 44 pp. Voss, A. 1907. Coniferen - Nomenklatur - Tabelle. Deut. Dendrol. Gesell. Mitt. 16 92. Wakeley, P. C. 1961. Results of the southwide pine seed source study through 1960—61. Proc. Sixth Southern Conf. on Forest Tree Improvement. Univ. of Florida, Gainsville. pp. 10—25. Weidman, R. H. 1939. Evidences of racial influence in a 25-year test of ponderosa pine. J. Agric. Res. 59:855-887. Wells, 0. O. 1962. Geographic variation in ponderosa pine (Pinus ponderosa Doug. ex Laws.). Unpublished PhD Thesis. Michigan State University. vi + 112 pp. Wiedemann, E. 1930. Die Versuche fiber den Einfluss der Herkenft , des Kiefernsamens aus der preubischen forstlichen Versuchsanstalt. Zeitschrift fur Forst —u. Jagdwesen. 62:498-522, 809—836. J. W. Wright translation. [Fl 8i Wright, J. W. 1962. Genetics of forest tree improvement. FAO Forestry and Forest Products Studies. No. 16. FAQ of the UN. Rome. xvi + 399 pp. . wright, J. W., and W. I. Bull. 1962. Geographic variation in European black pine - two-year results. For. Sci. 8:32-42. Wright, J. W., W. L. Lemmien, and J. Bright. 1963. Geographic variation in eastern white pine - 6-year results. Quart. Bull. Mich. Exp. Sta. 45:691—697. ROOM USE ONLY £29,044? {131? my! .v