Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies on the probable origin of some European douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) plantations Berney, Jean Louis Ami 1972

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1973_A6 B47.pdf [ 7.28MB ]
Metadata
JSON: 831-1.0075390.json
JSON-LD: 831-1.0075390-ld.json
RDF/XML (Pretty): 831-1.0075390-rdf.xml
RDF/JSON: 831-1.0075390-rdf.json
Turtle: 831-1.0075390-turtle.txt
N-Triples: 831-1.0075390-rdf-ntriples.txt
Original Record: 831-1.0075390-source.json
Full Text
831-1.0075390-fulltext.txt
Citation
831-1.0075390.ris

Full Text

STUDIES ON THE PROBABLE ORIGIN OF SOME EUROPEAN DOUGLAS-FIR (PSEUDOTSUGA MENZIESII [MIRB.] FRANCO) PLANTATIONS by Jean Louis Ami Berney Forest Engineer, The Swiss Federal Institute of Technology, Zurich, Switzerland, 1970 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF FORESTRY in the Faculty of FORESTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December, 1972 ( i ) In p r e s e n t i n g t h i s t h e s i s in p a r t i a l f u l f i l m e n t o f the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I f u r t h e r agree that permission f o r extensive copying of t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head of my Department or by h i s r e p r e s e n t a t i v e s . I t i s understood that copying or p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed without my w r i t t e n permission. Department of The U n i v e r s i t y of B r i t i s h Columbia Vancouver 8, Canada Date ABSTRACT In Europe, the origin of older Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) plantations i s generally unknown. The tracing of the probable o r i g i n of one Swiss and twenty Polish Douglas-fir stands was attempted. Cone and seed c h a r a c t e r i s t i c s were measured on material collected from a l l twenty-one European stands, and compared with the displayed var i a t i o n of the same c h a r a c t e r i s t i c s from Douglasrfir growing in i t s natural range on the P a c i f i c Coast of North America. A seed morphology test,, as described by Allen (1960), was used to indicate the Coastal or Interior o r i g i n of the samples. Based on this t e s t , the stands from Boezingen/Switzerland and from Purda Lesna/ Poland proved both to be of Coastal o r i g i n . A d i r e c t comparison of averages of six cone c h a r a c t e r i s t i c s (width and length of cone scale, width of bract, length of f i r s t and second prong, and rating of bract) with the ranges of averages of these ch a r a c t e r i s t i c s as they vary in climatic seed c o l l e c t i o n zones of the natural range (Yao [1971]) did not allow an unequivocal assessment of o r i g i n . From these comparisons, i t could be concluded that cone scale and bract width are more influenced by the environment thannthe other c h a r a c t e r i s t i c s . The six c h a r a c t e r i s t i c s of cone morphology, one thousand-seed weight (measured or assessed on 124 samples from the natural range) and r e l a t i v e DNA content (measured on t h i r t y - s i x samples of the Coastal ( i i i ) part of the natural range) were used concurrently to perform discriminant analyses based on d i f f e r e n t subspecies, climatic and l a t i t u d i n a l groupings. The discriminant functions obtained were then used to c l a s s i f y the twenty-one provenances of unknown o r i g i n . Their o r i g i n was assessed at the intersection of the groups to which they were attributed in the d i f f e r e n t analyses. The low number of provenances, when compared with the immense natural range, and the low number of c h a r a c t e r i s t i c s taken into consideration, did not allow a complete separation of the groups of provenances. Therefore, the assessment of or i g i n based on cone and seed morphology i s s t i l l to be considered with caution. Pre-established c u r v i l i n e a r regressions of r e l a t i v e DNA content of embryo c e l l s on latitude in both Coastal and Interior Douglas-fir (El-Lakany and S z i k l a i £1971, 1972]), were used for inverse estimation of the la t i t u d e of o r i g i n . It could be established, by this method, that the stand from Boezingen/Switzerland originates between l a t . 44° and 47°N on the Coast. When samples are s u f f i c i e n t from the unknown stand, as was the case for provenance Boezingen, seed c h a r a c t e r i s t i c s and r e l a t i v e DNA content gave promising indication of the probable o r i g i n . (iv) TABLE OF CONTENTS Page L i s t of tables (vi) L i s t of figures (ix) Acknowledgements (x) 1. Introduction 1 11. The problem of determining the or i g i n of an unknown provenance 1 12. Douglas-fir and i t s importance in Europe 6 13. The present knowledge of the v a r i a b i l i t y of DouglaS'-fir and i t s potential use 13 14. Purpose of this work 22 2. Material: and methods 22 21. The Swiss Douglas-fir stand 22 22. The Polish Douglas-fir stands 23 23. Methods 25 231. Allen's seed tests 25 232. Yao's cones and seeds studies 26 233. El-Lakany and S z i k l a i ' s c l i n a l v a r i a t i o n in r e l a t i v e DNA content of embryo c e l l s 29 3. Results and discussion 30 31. The seed tests 30 32. Seed y i e l d 32 33. Cone scale morphology 35 331. The Swiss stand 35 332. The Polish stands 37 34. One thousand-seed weight 41 35. Relative DNA content of embryo c e l l s 41 (v) Page 351. The Swiss stand 41 352. The Polish stand at Purda Lesna 43 36. The determination of or i g i n 47 361. General 362. Display of the cone scale morphology and one thousand-seed weight variables, and the i r individual use 47 363. D i s t i n c t i o n between Coastal and Interior Douglas^fir, and i t s application to the Polish stands 61 364. Discriminant analyses using climatic zones, and th e i r application to Polish provenances 65 365. Discriminant analyses using l a t i t u d i n a l zones, and t h e i r application to Polish provenances 70 366. Discriminant analysis including the variable one thousand-seed weight, and i t s application to Polish provenance No. 6 Purda Lesna 76 367. Assessment of the or i g i n of Polish provenances No. 1-20 78 368. Discriminant analyses for the c l a s s i f i c a t i o n of the Swiss provenance 80 369. Relationship between r e l a t i v e DNA content and l a t i t u d e , and i t s application for the o estimation of o r i g i n of the Swiss provenance 83 4. Summary and Conclusion 86 Bibliography 94 (vi) LIST OF TABLES Table 1. Material available from the Polish stands 2. Relationship between biologi c a l and geographical variables for eight climatic regions (from Yao [1971]) 3. Average values of four seed c l a s s i f i c a t i o n c r i t e r i a 4. Seed y i e l d record for provenance Boezingen/ Swi tzerland 5. Seed y i e l d record f o r Polish provenances No. 1-20 6. Mean values and standard deviations of cone scale c h a r a c t e r i s t i c s of trees, in millimeters, f o r provenance Boezingen/Switzerland 7. Analysts of variance of cone scale widths for provenance Boezingen/Switzerland 8. Analysis of variance of cone scale lengths for provenance Boezingen/Switzerland 9. Mean values and standard deviations of cone scale c h a r a c t e r i s t i c s of provenances in millimeters, for the Polish provenances 10. Analysis of variance of cone scale widths, for Polish provenances No. 1-20 11. Analysis of variance of cone bract widths, for Polish provenances No. 1-20 12. Analysis of variance of cone scale lengths, for Polish provenances No. 1-20 13. Analysis of variance of f i r s t prong lengths, for Polish provenances No. 1-20 14. Analysis of variance of second prong lengths, for Polish provenances No. 1-20 15. Analysis of variance of rating of bract, f o r Polish provenances No. 1-20 ( v i i ) Table Page 16. Duncan new multiple range test of six cone scale measurements, f o r Polish provenances No. 1-20 40 17. Number of observations, mean and standard deviation of r e l a t i v e DNA content, in Feulgen Absorption Units, for provenance Boezingen/ Switzerland 42 18. Analysis of variance of r e l a t i v e DNA content, for provenance Boezingen/Switzerland 43 19. Number of observations, mean and standard deviation of r e l a t i v e DNA content, in Feulgen Absorption Units, at the embryo l e v e l , f o r provenance Purda Lesna/Poland 44 20. Analysis of variance of r e l a t i v e DNA content, for provenance Purda Lesna/Poland 45 21. Duncan new multiple range test of r e l a t i v e DNA content, f o r provenance Purda Lesna/Poland 46 22. Number of climatic zones from which each unknown provenance could originate 60 23. Discriminant analyses for the d i s t i n c t i o n between Coastal and Interior Douglas-fir 62 24. C l a s s i f i c a t i o n of.Polish provenances No. 1-20 into Coastal or Interior Douglas-fir 65 25. Discriminant analyses f o r the d i s t i n c t i o n between climatic zones 66 26. C l a s s i f i c a t i o n of Polish provenances No. 1-20 into climatic zones (CLIAMP) 68 27. C l a s s i f i c a t i o n of Polish provenances No. 3 and No. 15 into climatic zones (CLIAME) 69 28. Discriminant analyses for the d i s t i n c t i o n between l a t i t u d i n a l zones 71 29. C l a s s i f i c a t i o n of Polish provenances No. 1-20 into l a t i t u d i n a l zones (LATPC0) 74 ( v i i i ) Table Page 30. C l a s s i f i c a t i o n of Polish provenances No. 3 and No. 15 into l a t i t u d i n a l zones (LATICO.) 75 31. Discriminant analysis LATPWE, for the d i s t i n c t i o n between l a t i t u d i n a l zones 77 32. C l a s s i f i c a t i o n of Polish provenance No. 6 into l a t i t u d i n a l zones (LATPWE) 77 33. Assessment of the or i g i n of Polish provenances No. 1-20 79 34. Discriminant analyses for Coastal Douglas-fir 82 35. C l a s s i f i c a t i o n of the Swiss provenance Boezingen into the Coastal Douglas-fir zone 82 36. Analysis of variance for the regression of r e l a t i v e DNA content on la t i t u d e 83 (ix) LIST OF FIGURES Figure Page 1. Seed c l a s s i f i c a t i o n 31 2. Geographic d i s t r i b u t i o n of Douglas-fir seed sources and climatic zonation (from Yao [1971]) 48 3. Natural v a r i a t i o n of the cone scale width of Douglas-fir inside the seed c o l l e c t i o n zones 50 4. Natural variation of the bract width of Douglas-fir inside the seed c o l l e c t i o n zones 51 5. Natural variation of the cone scale length of Douglas-fir inside the seed c o l l e c t i o n zones 52 6. Natural v a r i a t i o n of the f i r s t prong length of Douglas-fir inside the seed c o l l e c t i o n zones 53 7. Natural v a r i a t i o n of the second prong length of Douglas-fir inside the seed c o l l e c t i o n zones 54 8. Natural v a r i a t i o n of the rating of bract of Douglas-fir inside the seed c o l l e c t i o n zones 55 9. Natural v a r i a t i o n of the 1000-seed weight of Douglas-fir inside the seed c o l l e c t i o n zones 56 10. Relationship:" between the number of provenances and the range of v a r i a t i o n of the f i r s t prong length of Douglas-fir inside the seed c o l l e c t i o n zones 58 11. Delimitation of Coastal and Interior Douglas-fir 63 12. Delimitation of l a t i t u d i n a l zones for analyses LATPCO and LATPWE 72 13. Delimitation of l a t i t u d i n a l zones for analysis LATICO 73 14. Delimitation of l a t i t u d i n a l zones for analysts' LATDNA 81 15. The relationship between the amount of DNA (in FAU) and the latitude of seed source in Coastal Douglas-fir 85 (x) Acknowledgements The author wishes to express his gratitude f o r the advice, help and encouragement given by his research supervisor, Dr. 0. S z i k l a i , throughout the course of the research and preparation of this thesis. The author also wishes to express his thanks to the other members of his thesis committee, Drs. P.G. Haddock and C.W. Roberts, f o r taking the time to read his t h e s i s , to Dr. M.H. El-Lakany for valuable advice, to Mrs. M.A. De Vescovi, Mrs. M. Home, and Mr. C. Yao for technical help, and to Mrs. K. Hejjas and Miss L. Cowdell for assistance in computing. Thanks are also due to Miss A. Adamovich, Mr. M.Feller and Mr. D. Plackett for time spent checking a manuscript written often in whimsical English, and to Miss L. Gould for the d i l i g e n t typing of the f i n a l copy. - 1 -I. INTRODUCTION II. The prob 1 em o f determf ning the o r i g i n of an unknown provenance Each individual plant or group of individuals (population, provenance) of a species can be assigned geographical co-ordinates. Consequently, the set of longitudinal and l a t i t u d i n a l pairs corresponding to a l l trees of a pa r t i c u l a r species f u l l y describes i t s precise range. The third geographical co-ordinate, the a l t i t u d e , does not add more information to the location. It can, however, replace the latitude or the longitude in certain cases. This might be the case, for example, i f the range is limited to a steep valley or to a regular mountainside, p a r a l l e l to one of the geographical co-ordinates. On the other hand, each individual plant can be described by an unfixed number of taxonomical characters, morphological or physiological; the set of a l l "multiplets" describes the total v a r i a b i l i t y of the species. In many cases, i t is possible to establish a r e l a t i o n explaining the v a r i a b i l i t y of the species as a function of i t s geographic d i s t r i -bution, d i r e c t l y or i n d i r e c t l y through the medium of some environmental factors. The problem of determining the or i g i n of a sample from an unknown population involves the solving of a system of relationships between the geographical co-ordinates and the taxonomical characters (which are measured on the sample). A c h a r a c t e r i s t i c , as i t i s meant here, i s a phenotypic character, q u a l i t a t i v e or quantitative. Its basic requirements are those of."good taxonomical characters", as described -2-by Davis and Heywood [1963): they (1) are not subject to wide variation within the samples being considered; (2) do not have a high i n t r i n s i c genetic v a r i a b i l i t y ; (3) are not e a s i l y susceptible to environmental modification; (4) show consistency j i . e . agree wtth the correlations of characters existing in a natural system of c l a s s i f i c a t i o n which was constructed without their use. (This l a s t requirement, however, does not hold in the case of i n t r a s p e c i f i c v a r i a t i o n ) . The point (3) is c e r t a i n l y one of the most important for our concern. Indeed, only characters with a high h e r i t a b i l i t y are suitable. For lack of such a knowledge, a gross estimation could be provided by the comparison of measurements taken on a population in i t s natural habitat with those taken on a sample from the same population growing in a foreign environment. Once introduced into a foreign environment, the sample of a population is subjected to a d i f f e r e n t natural selection than at the place of o r i g i n (Bouvarel [1958]). The selection pressure is quite strong and might s i g n i f i c a n t l y change the characters affected. The characters used to i d e n t i f y an unknown sample should, therefore, neither be d i r e c t l y subjected to natural s e l e c t i o n , nor be c l o s e l y correlated with characters which are under a strong selection pressure. Falconer (1960) noted that: "the characters with the lowest, h e r i t a b i l i t i e s are those most closely connected with reproductive -3-f i t n e s s , while the characters with the highest h e r i t a b i l i t i e s are those that might be judged on b i o l o g i c a l grounds to be the least important as determinants of natural f i t n e s s " . Thus, our requirement of high h e r i t a b i l i t y and independance of selection pressure are not mutually exclusive. Simak (1967) studied the vari a t i o n in thousand grain weight (TGW) for European ilarch (Larix deoidua M i l l e r ) autochthonous and introduced stands of known o r i g i n . He found that the TGW of introduced stands generally f i t s into the frame of the TGW of the seven main regions of the natural range. Therefore, TGW can be used as a diagnostic character for the i d e n t i f i c a t i o n of the o r i g i n of l a r c h . So that one can determine the origin of the seeds using s p e c i f i c taxonomical characters, t h e i r v a r i a b i l i t y and i t s r e l a t i o n to the geography must be known for the whole range of d i s t r i b u t i o n of the species. This i s best realized with, an intensive systematic survey, but an appropriate sampling might give a picture of the vari a t i o n which w i l l be precise enough. Ba s i c a l l y two s t a t i s t i c a l techniques w i l l establish relations between taxonomical characters at the i n t r a s p e c i f i c level and geographical parameters which, are suitable f o r th.e purpose of or i g i n reconstruction. The two methods which can be applied are regression analysis and discriminant functions analysis. Regression analysis links a taxonomical character, measured with a random error, together with one or several geographical parameters assessed without error, -4-Following the rules of algebra, as soon as more than one geographical variable i s included in one equation, other equations explaining other taxonomical characters w i l l be required to find the solutions. What is a r e l a t i v e l y easy problem in algebra is made more complicated in s t a t i s t i c s by the fa c t that tolerance l i m i t s are affixed to any actual value in the regression equation. The inaccuracy of the estimation w i l l thus increase with the number of equations to be used. On the other hand, the technique of discriminant functions analysis w i l l give the best possible separation between pre-established geographical groups for a set of taxonomical characters. It w i l l provide the basis f o r assignment of the unknown sample to one of the groups and.indicate the pr o b a b i l i t y of i t s correct c l a s s i f i c a t i o n . Qualitative characters, of course, do not require the use of s t a i t s t i c s , when they are considered alone. Their application w i l l help, to set up a rough process of elimination, helping to decide the ri g h t regression to be used,.for example, or to choose between two solutions. . The most common case in which genetic characters are used for or i g i n v e r i f i c a t i o n i s the determination of the blood type of the offspring and the suspected parents in helping to decide doubtful family r e lations in paternity cases. In the f i e l d of pharmacy, Jabbar and Brochmann-Hanssen (1961) remarked that existent differences in amino acid composition of opium samples from several geographic regions may be suitable as a means of origin determination. -5-Allen (1961) stated that his motive for the development of his Douglas-fir [Pseudotsuga menziesii [Mirb.] Franco) seed tests (gross seed morphology and germination behaviour) was the necessity of having a tool to v e r i f y the v a l i d i t y of data on orig i n quoted by the supplier. "Mistakes are possible and seed lots may be accidentally mislabelled or even mixed". In New Zealand, Sweet (1965) gave the same reason f o r r e l a t i n g early measurements and observations of seedlings to climate. Having established a series of regressions, he stated "two-year height and time lag between l a t e r a l and terminal bud burst of the provenances were both highly s i g n i f i c a n t l y correlated with temperature, climate and a l t i t u d e of the seed source. In each sample cor r e l a t i o n c o e f f i c i e n t exceeded 0.75. It is suggested that these rela t i o n s h i p s , i f used together, are s u f f i c i e n t l y strong to be useful in checking the data accompanying purchased seed of Douglas-fir..." In the U.S.A., Franklin and Greathouse (1968) found a l a t i t u d i n a l gradient in mean cotyledon number and an increase in seed weight with decreasing latitude f o r noble f i r [Abies proaera Rehd.), C a l i f o r n i a r.ed f i r {Abies magnified A. Murr.) and Shasta Red F i r [Abies magnified var.shastensis), They advocated the use of cotyledon counts from a representative sample for comparing with data from known populations to ve r i f y the seed source. Recently, Rosenstock (1972) suggested the use of A l l e n (1960); EULakany and S z i k l a t (1971); and Von Rudloff (1971) studies for the determination of orig i n of the old Douglas-firs growing in Germany. He advises one to attempt a h i s t o r i c a l analysis of the seed c o l l e c t i o n s -6-on the Pa c i f i c Coast, also. Rosenstock's argument was that provenance studies give indications only for a very limited area in the v i c i n i t y of the test s i t e . 12. Douglas-fir and i t s importance in Europe Douglas-fir, a forest tree species of North America, extends from l a t . 55°30' N. at Babine Lake, Takla Lake and at Tudyah Lake in B r i t i s h Columbia (Garman [1970]) into Mexico where isolated stands are found as far as l a t . 19° N (Fowells [1965]). In west-east d i s t r i b u t i o n i t grows from the islands of the P a c i f i c Coast east to long. n3°30' W in Canada, long. 105° W in the U.S.A., and long. 98° W in Mexico. In i t s enormous natural range, Douglas-fir varies greatly in i t s morphological as well as ecological behaviour. This i s well reflected in the many attempts to arrive at a taxonomical c l a s s i f i c a t i o n described in Allen (1960), Tusko (1963), and S z i k l a i (1964). E a r l i e r attempts by Flous (1936) subdivided North American^Pseudotsuga into eleven species and one variety. The next year (1937), she even extended her c l a s s i f i c a t i o n to 18 subspecies! Today, however, authorities do not universally agree beyond the separation into Coastal {pseudotsuga menziesii [Mirb,] Franco var. menziesii) and Interior Douglas-fir {Pseudotsuga menziesii var.glauca i B e i s s n . J Franco) (Fowells 11965]). In i t s natural range, Douglas-fir i s a very prominent species. Fowells (1965) ca l l e d i t "one of the world's most important and valuable timber trees" and Preston (1966) stated that "this tree produces more timber than any other species". -7-In 1828, Douglasrfir had already been planted in England and Scotland. The seeds had been collected in the P a c i f i c Northwest region of North America by David Douglas in 1826. Floehr (1958) reported that there was a rapid increase of i n t e r e s t for Douglas-fir in Europe, due to an unusually high' growth in comparison with the native species. At the present, Douglas-fir i s found in most European countries. In respect to Great B r i t a i n , MacDonald et al_. (1957) considered Douglas-fir as "one of the most important exotics grown in B r i t a i n " . By 1964, 32,000 ha had been planted with Douglas-fir with a yearly increase of 2,000 ha (Dawson [1964]). Vredenburch and La Bastide (1969) stated that the Netherlands have an area of more than 12,000 ha planted with Douglas-f i r . In other words 8% of the coniferous forest area of the country consists of Douglas-fir, Concerning France, Lanier (1966) noticed that "Coastal Douglas-fir i s a c t u a l l y one of the chief species f o r reforestation in the low and intermediate a l t i t u d i n a l zones [except in the south-west and in the Mediterranean zone). For some time, i t has occupied probably the second position (after Norway Spruce) in the reforestation programs with seedlings". (Trans!. from French). The reasons he gave are as follows; the growth rate, the ease of treatment, the q u a l i t i e s and the quantity of i t s timber, the r e l a t i v e resistance to winds and Its genetic p l a s i t i c i t y . Similar descriptions are used by authors of other countries when discussing this remarkable tree. Schober (1963a) compares the mean annual increment of Douglas-fir in six European countries with than of domestic species, which backed up -8-Galoux's [1951] statements: "In short, Douglas-fir, up to the present, seems to be the best purchase for the European forestry from the American f l o r a . It i s c h i e f l y interesting in low a l t i t u d i n a l zones where no other conifer surpasses i t as to y i e l d in volume of high-quality timber". (Transl. from French). Several European foresters have described the present and future potential use of Douglas-fir. In Germany, i t i s already used in place of Scots pine {Pinus silvestris L.) f o r the conversion of coppices (Viebig [1967]). It takes the f i r s t place on "mediocre s i t e s " (Volk [1969]) and gradually replaces Scots pine and s e s s i l e oak (Queraus petraea [Mattuschka] L i e b l . ) in Lower Saxony (Otto [1972]). Douglas-f i r should be used where "Norway spruce (Piaea abies [L.] Karst.) constantly suffers under the spruce sawfly (Prisiphora abietina Christ.) or Scots pine under wet snow, where hardwood composite f o r e s t and coppice p r e v a i l " (Transl. from German)(Abetz [1971]). In Bosnia and in Herzegovina, Douglas-fir with other conifers could be also introduced in standard fore s t where the proportion of mixture i s unsatisfactory, and the ecological ( p a r t i c u l a r l y edaphic) conditions are good (Pintaric [1967]). In It a l y , Merendi (1965) advocated growing the species in the Castanetum and Fagetum zones of the Appeninnes because of i t s rapid growth and large volume production and claims that the r i s k for disease is much less than anticipated. The introduction of Douglas-fir did not go on without some accidents and f a i l u r e s . Like any other exotic species, i t shows a dif f e r e n t i a t e d resistance to abiotic and b i o t i c diseases. In summing -9-up the experiences in Europe, Schober (1963b) declares that the most important problem in central and northern Europe is Douglas-fir's f r o s t s u s c e p t i b i l i t y at the seedling stage. Lacaze and Tomassone (1967) stated that "some blows were registered, e s p e c i a l l y on some shallow limestones, on poorly drained s o i l s and in f r o s t pockets, but c h i e f l y , the attention of many Douglas-fir users was attracted by evident differences in the behaviour or' in the quality of the stands according to the origin of the seeds used". (Trans!. from French). In 1930, the Douglas-fir needle b l i g h t (RhaMooline pseudotsugae Syd.) arose, followed fn 1935 by the even more dangerous Swiss rieedle cast (Phaeooryptus gaumannl [Rhode] Petr.). Koestler (1950) mentioned that not only young but also f i f t y to f i f t y - f i v e year old stands were k i l l e d . Lanier (1966) found that the degree of resistance to the Rha.bdocline-disease varied with the provenance. Simi l a r l y i t could be observed that the growth of Douglas-fir is not uniform, c a l l i n g for a careful selection of the best suitable provenance (Floehr, 1958, Schober, 1963a). Lacaze (1968) concluded his study on juvenile growth of some provenances by these words: "The economy of a reforestation with Douglas-fir depends therefore mainly on the choise of the provenance. It appears indispensable to popularize those, which, in the present state of our knowledge, are l i k e l y to guarantee the best y i e l d s " . (Transl. from French). Bouvarel.0969) recommended three steps before deciding on the choice of the provenance: 1. Search, in the natural range of the species, for a region of -10-provenance whose cli m a t i c c h a r a c t e r i s t i c s are analogous to those in the area to be reforested; 2. establish a provenance study in the neighbourhood or look at the results of e a r l i e r provenance studies showing any resemblances to the s p e c i f i c area, and 3. investigate the performances of the old stands having grown on similar s i t e s . The f i r s t path was taken at the time of the f i r s t introduction of North American trees in Europe (Irgens-Moller [1966]). It allows one to avoid important errors, but as Bouvarel (1969) mentioned: "If the analogy of climates guarantees good adaptation of the provenance and generally the survival of the plantation, i t is not enough to make sure^of getting the maximum f i n a n c i a l y i e l d in every case; nature did not always select l o c a l l y the growth and form c h a r a c t e r i s t i c s in the di r e c t i o n we wish", (Transl. from French). A provenance study allows one to follow very closely and precisely the performance of a group of provenances. The inferences which may. be drawn from the results are a function of the s t a t i s t i c a l design of the experiment and of the s i m i l a r i t y of the treatments with those applied in the commercial fo r e s t . A very important point is how far in time the results can be extrapolated. Haddock et_al_. (1967), in B r i t i s h Columbia, established strong correlations between the height of Douglas-f i r at two years and the height at f i v e , s i x , seven, and eight years. Nanson (1968), using the results of the Kaiserlautern (Germany) provenances comparison study established in 1910, found r e l a t i v e l y high correlations between the height at forty-two years and the height at -11-four and eight years. The generality of these results was unfortunately weakened by the observations made on the 1912 Douglas-fir heredity study in Western Oregon and Washington: Silen (1966) noted that "most of the results published when the study was 17 years old were reversed by age 50". Wakeley and Bercaw (1965) working on iloblo l l y pine (Pinus taeda L . ) , found d i f f e r e n t results at age t h i r t y -f i v e than at age twenty-two, which i s one more indication of the f r a g i l i t y of early t e s t s . Thus, provenance studies might give good preliminary indications in a case in which no previous information has been accumulated. On the other hand, there are now a r e l a t i v e l y large number of pure o r mixed Douglas-fir stands in many European countries. They are the remaining stands from plantations made at the end of the nineteenth and at the beginning of the twentieth century, when foresters showed a great interest for exotic trees. The Douglas-fir has adapted i t s e l f to the new environment and, in most cases, surpasses the surrounding native trees. Several authors are convinced that these pioneer-stands have a great value and recommend the use of their seeds as being the safest method for establishing Douglas-fir plantations. Thus, Koestler (1950) stated that " p a r t i c u l a r l y in the case of Douglas-fir and eastern white pine, the proved old trees deserve consideration with regard to th e i r properties and the associated h e r i t a b i l i t i e s ; for these old trees have already passed d i f f i c u l t t e s t s " . (Trans!. from German). After Rowe (1964): "Once an exotic i s established there may be more gain in reproducing i t from i t s own seeds as rapidly as possible (...) than in -12-continuing the search for better provenances". Jahn (1955), having compared German and American Douglas-fir provenance studies, recommended a more intensive use of seeds from old German stands and the inclusion of them in new provenance studies. Expatiating on the evolution of introduced species, Bouvarel (1958) gave the following practical recommendation: "The local plantations which have given good results are seed sources of the greatest interest f o r new reforestations in the same region". (Transl. from French). Recently, Zimmerman (1972) placed the "acknowledged German stands" in the f i r s t position of a l i s t of provenances to be planted in Hesse (Germany). Douglas-fir produces variable quantities of seeds from year to year in America as well as in Europe, but too often i t is not enough to cover the regional needs. Allen and Owens (1972) wrote "... some seeds are produced annually in a region, except for about one year in any four- or five-year period. Abundant or medium cone crops occur every two to seven years, and commonly about every f i v e years in a region, but the cycle i s unpredictable". The situa t i o n in Europe i s si m i l a r , or even worse. In Holland, good cone crops occur only once every eight years on the average and the more frequent l i g h t to moderate crops are generally attacked by the Douglas-fir seed f l y (Megastigmus spermotvophus Wachtl) (Vredenburch and La Bastide [1969]). In Belgium, Beckers (1972) observed that "since 1959, there has been only one good crop,.in 1960, and one moderate crop, in 1964". Therefore Wheat (1966) considered the lack of available seeds from the proper source as the main reason for the seed movement problem in the Douglas-fir region. In Belgium, Cornet D'Elzius 0963) admitted that the seeds gathered from the -13-best stands would only cover a t i n y part of the needs of the country. The state of a f f a i r s has not changed within the l a s t nine years: "The rarety of good crop-years constitutes a strong handicap for practice..." (Beckers [1972]). In addition to the i r r e g u l a r i t y of good seed-years, numerous prominent European Douglas-fir stands are l i k e l y to disappear as Bouvarel (1969) quoted i t in the case of France. Thus, in many areas, there w i l l be a d e f i n i t e need f o r seeds imported from the natural range because of both the s c a r c i t y of seed-years and the decreasing number of good stands. From the preceding considerations, we can conclude that one way to obtain "sound seeds" f o r reforestation purposes is to attempt to find out the origin of the prominent European stands and to buy seeds from these l o c a l i t i e s on the North American West Coast. 13, The present knowledge of the v a r i a b i l i t y of Douglas-fir and  i t s potential use Allen (1960) had made a study of the morphological differences between Coastal and Interior Douglas-fir seeds. He based his i d e n t i f i c a t i o n on differences in seed coat markings and seed body shape. His basis is the r e s u l t of having examined over 160 l o t s of seeds from Bella Coola, B.C., to C a l i f o r n i a on the Coast and from Babine Lake and Prince George, B.C., to Wyoming and Colorado in the Interior regions. No proof was given for the genetical or i g i n of the differences in seed morphology. However, in a more extensive paper on the same work, Allen (1961) observed the constancy of the seed characters inside the -14-subspecies ranges: seeds from the dry Qulf Islands in the wet coastal range show Coastal characters and seeds from the Interior wet belt have the same appearance as the seeds from the generally dry Interior range. Such a fa c t supports the supposition that the t r a i t s involved are strongly genetical. Tusko's (1963) studies on seeds from twenty-eight provenances across B r i t i s h Columbia substantiated Allen's findings with the exception of a population on Revelstoke Mountain (Interior) above the 3.400 foot level (1036 m). The quickness of the test and i t s r e l a t i v e goodness: "Mixtures of Coast and Interior seed can be separated with an error not greater than 10% (Allen [I960]), makes i t most suitable f o r the primary purpose of discrimination between the subspecies. In his second a r t i c l e on seed morphology t e s t i n g , Allen (1961) found that "a low-temperature germination test w i l l a s s i s t in sorting these (seeds from t r a n s i t i o n a l areas) into easterly and westerly provenances". Moreoever, i t was said that these germination tests are subject to the effe c t s of other factors such as the nu t r i t i o n of the parent, the cone and seed maturity, the cone storage conditions, the processing, and the seed storage. Allen concluded that much more study i s needed before the practical p o s s i b i l i t i e s of the germination tests could be evaluated. The other meticulous investigations of Tusko (1963) did not enable him to go beyond the subspecies l e v e l . He found s i g n i f i c a n t differences between Coastal and Interior populations in six quantitative characters: needle thickness, cone diameter, cone scale length, cone bract length, and cone s p e c i f i c gravity, but no differences i n needle length and cone scale angle. S z i k l a i (1969) used ninety-one provenances collected by IUFRO Section -15-22 tn 1966 tn British. Columbia, Washington and Oregon. From the measurements of cone length, seed length, seed width, seed wing length and seed wing width, i t was possible to detect l a t i t u d i n a l clines for the length of the cone and the length of the seed. However, only trends were noted and no mathematical relationships calculated. Further treatment of the data w i l l c e r t a i n l y give u t i l i z a b l e r e s u l t s , but in th e i r present state they can only be of very limited use for the purpose of origin determination. Bouvarel et al. (1969) applied multivariate s t a t i s t i c a l techniques on the same provenances used by S z i k l a i (1969). Canonical variables analysis was performed on the same characters plus the weight of one thousand f i l l e d seeds, and weight of one thousand empty seeds, A l l canonical variables were s i g n i f i c a n t , suggesting that no variate could be eliminated. The calculated Mahalonobis distances allows the grouping of the provenances on the map: "There appears a quite homogenous area in the center, whereas along the l i m i t s of the sampled area, groups between neighbouring provenances appear. Evidently, a more detailed analysis should be done..." (Transl. from French). These results suggest that more variables should be used when beginning the study. Considering only the inter-provenances v a r i a b i l i t y of the number of seeds per cone, the standard deviation of the number of seeds per cone, the weight of one thousand f i l l e d seeds, and the characters used by S z i k l a i (1969) such as cone length, seed length, seed width, seed wing length, and seed wing width, the same authors established canonical correlations with the a l t i t u d e , the la t i t u d e and the longitude. A l l -16-btological variables are again s i g n i f i c a n t , mainly with the longitude or the l a t i t u d e , the importance of the a l t i t u d e being much les s . The map drawn from the results of the canonical correlations i s d i f f e r e n t from the map drawn from the canonical variables analysis. Principal components analysts was carried out using the same variables. The r e s u l t provided a d i f f e r e n t grouping of provenances. On the whole, the results obtained with the use of multivariate s t a t i s t i c a l techniques suggest that more variables be added before similar.groupings can be obtained from the d i f f e r e n t approaches and u t i l i z e d for the purpose of o r i g i n determination. The most recent comprehensive study of the geographic v a r i a b i l i t y of Douglas-fir i s a master's thesis written by Yao [1971). Beside the previously mentioned ninety-one provenances, he also included in his study thirty-three more provenances collected by IUFRO Section 22 in 1968. Even with this extended provenance c o l l e c t i o n , the Interior Douglas<-fir regions east of longitude 117° W in the upper part of the range and east of longitude 121° W below latitude 47° N were s t i l l not sampled. He investigated one thousand seed weight, cone scale width, bract width, cone scale length, f i r s t prong length and second prong length. He also studied the position of the bract in r e l a t i o n to the scale and the f i e l d germination inside eight c l i m a t i c regions. He was able to establish relationships with latitude and a l t i t u d e within the regions. The c l i m a t i c regions are those established for the purpose of seed c o l l e c t i o n by Haddock and S z i k l a i (1965) for Canada, by Western Forest Seed Council (1966) for Washington and Oregon, and by Buck et a l . -17-0 970) for C a l i f o r n i a . It was not possible to set an equation with la t i t u d e or a l t i t u d e in each region for any character. Furthermore, no general relationship could be established, which would l i n k the climatic zones. Three main reasons might be responsible f o r these r e s u l t s : the number of samples, the grouping, and the technique of analysis. As the characters used are e a s i l y measured on the unknown samples, i t w i l l be worthwhile to question the grouping and the technique and try to adapt the results of the analysis to the purpose of or i g i n determination. The 1966 and 1968 IUFRO Section 22 co l l e c t i o n s u t i l i z e d for the precedent studies were extended in 1969 to the Douglas-fir zones in the states Utah, Colorado, Arizona and New Mexico, U.S.A. Twenty-one provenances from the total c o l l e c t i o n were selected by El-Lakany and S z i k l a i (1971) to explore the v a r i a b i l i t y in nuclear volume and r e l a t i v e DNA content from the protoderm of embryo c e l l s . For both characters strong c u r v i l i n e a r relationships were established with the l a t i t u d e , separated but p a r e l l e l for both subspecies. Furthermore, a highly s i g n i f i c a n t correlation was found between nuclear volume and r e l a t i v e DNA content. Although'they sample size might seem i n s u f f i c i e n t when compared with the magnitude of the natural range of Douglas-fir, the strength of the relationships i s astonishing and their a p p l i c a b i l i t y encourages the investigation of further correlations. Thirty-one provenances were selected in a follow-up study by E l -Lakany and S z i k l a i (1972). The results substantiated the relationships between the latitude and the r e l a t i v e DNA content of the protoderm of -18-embryo c e l l s found in the previous study. At the same time, a s l i g h t influence of the a l t i t u d e of the seed source was discovered. Although the technique was s l i g h t l y d i f f e r e n t (the embryos were squashed on the microscope s l i d e for the measurement and not sectioned and mounted in Canada balsm as i t had been done previously), there i s no evidence to suggest that both studies of the r e l a t i v e amount of DNA cannot be used conjointly to supply more confident regression equations. The essential o i l s or terpene compounds analysis i s a r e l a t i v e l y recent but promising device for the study of v a r i a b i l i t y and hybridization in natural populations. Working with western white pine {Pinus montioola DouglJ), Hanover [1971] and preceding papers) found that monoterpene compounds in the c o r t i c a l oleoresin are under strong, predictable genetic c o n t r o l , probably involving one or several l o c i . Terpene concentrations in woody c o r t i c a l or leaf tissues lend themselves quite well to quantitative analyses. Wilkinson e_t al_. (1971) working on white spruce (Piaea glauca [Moench] Voss), and Flake et a l . (1969) on eastern redcedar (Juniperus virginiana L . ) , f o r example, were able to display c l i n a l v a r i a t i o n . The f i r s t quantitative examination of the geographic differences in monoterpenes concentrations of wood oleoresin of Douglas-fir was carried out by Hanover and Furniss (1966) on three samples from Idaho. Signifcant differences were found in certain compounds, but no proof for lack of environmental influence on the t r a i t s was given. An extensive study of the v o l a t i l e leaf o i l s of Douglas-fir was reported by von Rudloff (1972). His measurements on forty-two samples collected -19-from Vancouver Island up to the foothills- of Alberta in Canada showed clear-cut q u a l i t a t i v e and quantitative d i s t i n c t i o n s between Interior and Coastal Douglas-fir and suggest a quantitative description of Interior and Coastal intermediates. However, some of the samples used came from a forest nursery whereas others came from natural stands growing along roads. Even i f Hanover (1966) said that monoterpene concentration of white pine cortex oleoresin i s quite stable with respect to many environmental f a c t o r s , i t does not give Von Rudloff the l i b e r t y to assume that this conclusion could be applied to his study of leaf o i l samples of Douglas-fir, also. Thus, more study i s needed before a coherent description of the chemical composition of v o l a t i l e o i l s can be made and the results u t i l i z e d for o r i g i n determination. Numerous provenance comparison studies of Douglas-fir have been designed inside the natural range and in the regions of introduction. Their results supply invaluable information about the v a r i a b i l i t y of the species. However, provenance studies are generally established for very p r a c t i c a l purposes. The number of provenances enclosed in a study is limited because of i t s high cost. The choice of the provenances is guided by the knowledge acquired from previous experiments and i s neither random nor systematic. Thus, provenance studies generally bestow information which i s v a l i d for only the included provenances. Interpolation for other provenances should be done cautiously. There are two ways of u t i l i z i n g provenance t r i a l s for the assessment -20-of the o r i g i n of an unknown stand. The f i r s t method would be to grow unknown seeds together with seeds of the natural range in a new t r i a l and make a d i r e c t comparison. This i s done occasionally. Due to the natural selection on the old introduced stand, i t s survival and growth performances are often as good or even superior to those of the best provenances t r i e d for the f i r s t time. In the case of Douglas-fir, this can be found in Guinier (1948() and Jahn (1955). Although such a test would give the most unequivocal indication of the o r i g i n of the unknown stand when using "good" taxonomical characters, as defined e a r l i e r , i t s establishment i s unlikely because i t would be expensive. It could be done as a secondary task in a more general study. Another method would be to grow the unknown seeds under conditions identical or at least very similar to those of a pre-established provenance t r i a l and measure the same characters. The obtaining of a comparable environment can only be realized for studies in v i t r o . These studies generally include a r e l a t i v e l y low number of provenances. The number of characters measured is l i m i t e d , as the seedlings cannot remain very long in the growth chamber due to the lack of space. On the other hand, f i e l d tests o f f e r less r e l i a b i l i t y . P a r t i c u l a r l y in respect to short term studies, the yearly climatic variations would have a marked e f f e c t on phenotypic characters. When discussing the provenance studies with Douglas-fir, which have a potential use, the work of Wright vet-al.(1971) on 128 provenances of the Interior subspecies should be mentioned. A l l parts of the Interior range were well represented with the exception of B r i t i s h -21-Columbia. Fiye characters (date of f i r s t - y e a r bud set in the nursery; height; winter damage; f r o s t damage; and colour in the plantation) measured on s i x t y - s i x provenances were used to build a summation-of-difference analysis. This analysis allows one to build up more or less homogenous groups and to separate them. Morris et al_. [1957) found the time of bud bursting to be under a strong genetic control. This was confirmed by Silen (1962), who estimated the genetic component of the total variation of the bud bursting t r a i t at 94% and 96% over two years of observation. The other t r a i t s are subject to the yearly climatic variations. Consequently, the use of these measurements in another decade on an unknown sample would c e r t a i n l y give unreliable results, The German Douglas-fir provenance experiments comprised r e l a t i v e l y few seed sources, the most extensive being the one at Chorin (Eberswalde near Berlin) with nineteen provenances (Guillebaud [1956]). Three of the seed sources were from Colorado, one from New Mexico, three from Montana, f i v e from Idaho, four from Washington and three from C a l i f o r n i a . These sources are very few when compared with the extent of the natural range. Ferrel and Woodard (1966) subjected Douglas-fir seedlings to drought conditions in the greenhouse and laboratory. They showed s i g n i -f i c a n t differences in drought resistance between Interior and Coastal origins and within each of these groupings. The study, however, included few provenances and f a i l e d to give averages v a l i d for regions. Therefore, i t can hardly be used for the determination of the or i g i n -22-of a sample. Baker (1971) measured the epicotyl length of the majority of provenances collected by IUFRO Section 22 at the end of the f i r s t growing season in the nursery. He was able to develop regression equations between epicotyl length and a combination of environmental variables for zonal groups. Unfortunately, the best equation obtained does not explain more than 48% of the variation and the others explain less than 20%. This f a c t , added to the considerations on the use of provenance t r i a l s lead to the rejection of Baker's study for o r i g i n determination. 14. Purpose of this work This thesis deals with the establishment of the o r i g i n of one stand of unknown o r i g i n , growing in Switzerland and of twenty stands in Poland, by means of comparing cone and seed character!sties to samples growing in the natural area of d i s t r i b u t i o n of Douglas-fir on the-Pacific Coast of North America. 2. MATERIAL AND METHODS 21. The Swiss Douglas-fir stand In August 1971, the Swiss Federal Forest Research Station at Birmensdorf collected cones from a Douglas-fir stand at Boezingen near Biel on the southern side of the Jura Mountains. The sixty to eighty year old plantation i s situated at an a l t i t u d e of 670 m and grows on a limestone substratum. The Douglas-firs are mixed with Norway spruce, -23-European l a r c h , black pine (Pinus nigra Arn.) and s i l v e r f i r (Abies alba M i l l e r ) . This stand belongs to a group of about ten stands of ages sixty to ninety years, each with an area of less than 0.5 ha disseminated over the lowland. There are no indications as to their o r i g i n . Fifteen cones were gathered from ten trees. The seeds were extracted and the cones dried, before being sent by a i r to Vancouver on November 29, 1972, where they have been stored at 0-2°C, in the G.S. Allen Forest Genetics and Seed Laboratory. Most of the cone bracts and seed wings were damaged, due to processing and transportation. 22. The Polish stands In the years 1968/69 and 1969/70, cones were collected by a crew from the Arboretum Kornickie (Polish Academy of Sciences) in Poznan on twenty Douglas-fir stands, seventy to eighty year old. Chylarecki (1970) indicates that Douglas-fir i s cultivated in Poland under various fores t conditions of climate and s o i l s , but did not supply a description of the precise location and of the state of the twenty stands. The experimental stands generally occupy between 0.25 and 0.50 ha. No indication seems to exi s t as to their o r i g i n . Table 1 indicates the number of trees sampled and the number of cones available on each c o l l e c t i o n s i t e . The seeds were extracted and the cones dried. The material was sprayed with a fungicide before being mailed by a i r to Vancouver on October 7, 1970. An appreciable number of cones were damaged, p a r t i c u l a r l y -24-the larger ones. No d i s t i n c t i o n had been made between the trees from which the c o l l e c t i o n had been taken. The material was stored in the cold room, at 0-2° C, at the Faculty of Forestry ? University of B r i t i s h Columbia. Table 1. Material available frora the Polish stands Provenance Number of -sampled on trees Number of cones available 1, WOLrNSKI 15 27 2. MIEDZYZDROJE 14 24 3. DOBRZANY 17 24 4. BIALY BOR 15 21 5. S0SN0 15 26 6. PURDA LESNA 15 29 7. RYN 15 30 8. MYRADZ 16 27 9. JAROCIN 15 26 10. KACZORY 17 25 11. GOLABKI 15 27 12. SKRZECIN 15 26 13. DOLICE 16 26 14. SWIEBODZINN 16 32 15. KROSNO 15 26 16. UJSOLY 15 19 17. POKRZYWNO 15 24 18. DUSZNIKI 13 15 19. KOWARY 15 12 20. BROJCE 13 22 -25-23. Methods Due to the damaged condition of the material, the freedom of choice for the methods was considerably reduced. Four studies on the morphology were selected as representative for naturally growing Douglas-fir; seed tests (Allen £1960]), nuclear c h a r a c t e r i s t i c s (El-Lakany and S z i k l a i £1971, 1972]), and cone morphology and seed weight observations (Yao £1971]). 231. Allen's seed tests Allen's observations can be described in the following way: the seed t i p of the Coast seed i s longer, more pointed and is pinched in jus t behind the extreme t i p while Interior seed has a shorter, more blunt and less pinched t i p . The body shape of Coast seed is generally longer and narrower than the broader and more d e f i n i t e l y triangular Interior seed. The dark side of Coast seed has a rough outer surface and appears to be darker and d u l l e r than the d i s t i n c t i v e sheen of the Interior seed surface. The body contour of the Coast seed is characterized by a pronounced ridge on the dark side as well as by the wrinkled appearance of the seed coat ju s t behind the t i p . The Interior seed has a much less pronounced ridge and has a f u l l and smooth body contour. The Interior seeds often appear to be brighter and are commonly marked by d e f i n i t e stripes on the outer dark surface. These stripes may be wavy and continue into the wings or wing bases. A preliminary test with seeds of known origin showed that the colour c l a s s i f i c a t i o n was d i f f i c u l t to observe. Consequently, the -26-colour test was dismissed afterwards. As i t i s currently done for undergraduate training in Forest Genetics at the University of B r i t i s h Columbia, the four remaining characters were rated in the following ways: Seed t i p length: 1. long - more than 1.0 mm 2. medium - between 0.5-1.0 mm 3. short - less than 0.5 mm Body shape: 1. not d i s t i n c t l y triangular 2. intermediate 3. triangular Body contour: .1. d i s t i n c t (ridge) 2. intermediate 3. not d i s t i n c t Body surface: 1. rough 2. medium 3. smooth From the above c l a s s i f i c a t i o n s , number one (1) characterizes Coastal seeds and number three (3) Interior seeds, number two (2) standing for seeds of intermediate o r i g i n . For the Swiss provenance, eleven seeds (minimum number of seeds in tree No. 7) of each of nine trees were subjected to the test and averages were calculated. From the Polish provenances, f i f t y seeds from provenance No. 6 Purda Lesna (the only provenance with more than t h i r t y seeds available) were tested. 232. Yao's cones and seeds studies Yao (1971) established relationships between b i o l o g i c a l variables Table 2. Relationship between biological and geographical variables  for eight climatic regions (from Yao 11971]) VARIABLE LATITUDE I 1 2 3 4 3 5 6 3 7 8 3 ALTITUDE 1 2 3 4 3 5 6 3 7 8 3 SCALE WIDTH BRACT WIDTH SCALE LENGTH 1st PRONG LENGTH 2nd PRONG LENGTH RATING OF BRACT 1000-SEED WEIGHT 0.772 0.52 0.70 0.72 0.50 0.64 0.71 0.64 0.60 0.73 0.82 0.72 0.70 0.87 0.54 0.73 0.93 0.77 0.80 0.63 0.79 0.81 0.34 0.83 0.80 0.40 0.37 0.54 0.87 0.56 0.64 0.88 0.44 0.48 1 cli m a t i c region (Section 13), from 1 to 8 correlation coefficients 3 No v a l i d correlation could be calculated in Regions 4, 6 and 8 due to the limited number of samples —I. c u 3 3 C L —1 P J —t C T P J — 1 r l -co i . c + ro £Z Q . CD P J 3 C L , O 0 3 — i . c + sz Q -fl> .-s CD t o -a CD rt c+ — J . < CD • << • 1. in -s CD t o c ; — i r + C/l p j -s n> t o c 3 3 CO -s i . t o f D C L ro -28-The cor r e l a t i o n c o e f f i c i e n t s are r e l a t i v e l y high, but the great number of nonsignificant or absent regressions (empty squares in Table 2) suggests the use of another technique. For the Swiss provenance, i t was only possible to measure the width and length of the scale and the one thousand-seed weight. Two cones were randomly selected for each of the ten trees and f i v e scales were taken in the middle of each cone. This gives one hundred observations, while Yao used three scales of two cones f o r each of f i f t e e n trees or ninety scales per provenance. A nested analysis of variance was performed on each character. For the Polish provenances scale width, scale length, bract width, f i r s t prong length, second prong length and rating of bract were appraised. The one thousand-seed weight was calculated for the provenance No. 6, which had 134 seeds available. The number of observations varied from thirty-two in provenance No. 19 to as much as ninety-six in provenance No. 14. A one-way analysis of variance and a Duncan new multiple range test were performed on each variable (except thousand-seed weight) f o r a l l stands. Yao's results were used in two ways. F i r s t , a d i r e c t comparison of the averages of the unknown samples with the ranges of averages for the climatic zones was done. This gave some indications concerning the influence of the new environment on the characters chosen. Secondly, Yao's dats were used to calculate discriminant functions f o r the climatic zones and for l a t i t u d i n a l zones (see, for example, Freese [1964]). The discriminant functions were then used to class the unknown samples. -29-The technique of measurement of the cone and seed c h a r a c t e r i s t i c s was similar to that applied by Yao. 233. El-Lakany and S z i k l a i ' s c l i n a l variation in r e l a t i v e DNA  content of embryo c e l l s The seeds from a l l unknown sources were separated into f i l l e d and empty ones using X-ray fluoroscopy. F i l l e d seeds are defined as those which have a well-developed endosperm and an embryo that f i l l e d 75-100% of the embryo cavity. A more refined c l a s s i f i c a t i o n of the empty seeds, as was done by S z i k l a i (1964) was not attempted. The sc a r c i t y of f i l l e d seeds led me to l i m i t the measurement of the r e l a t i v e DNA content to the Swiss provenance Boezingen and to the Polish provenance No. 6 Purda Lesna. Ten seeds from four trees, or fo r t y seeds, were used from the Swiss sample and f i f t y seeds from the Polish one. The technique applied was similar to that used by El-Lakany and S z i k l a i in the i r second paper (1972). Nested analyses of variance were performed f o r both stands and a Duncan new multiple range test was further conducted for the Polish stand. In t h e i r 1971 paper, El-Lakany and S z i k l a i gave as the best f i t t i n g regressions f o r the scatter diagrams, two c u r v i l i n e a r l i n e s : Log (DNA) = 14.61-10.87 log (lat)+0.12 l a t , R = 0.967, for the Coastal provenances; and log (DNA) = 9.37-6.61 log (lat)+0.08 l a t , R = 0.919. for the Interior provenances. In the 1972 paper, the scatter diagrams were best explained by two d i f f e r e n t equations of analogous shape to those of 1971. However the authors suggested the use of the following ones, p a r t i c u l a r l y f o r the purpose -30-of inverse estimation: Coastal provenances: DNA = 679.5540-29.9788 l a t + 0.3939 l a t 2 , R = 0.980; Interior provenances: DNA = 1100.4700-48.0090 l a t + 0.5791 l a t 2 , R = 0.949. The sets of points from El-Lakany and S z i k l a i (1971, 1972) were used together to calculate new regression equations, on the assumption that neither of the sets alone could explain f u l l y the true relationship between r e l a t i v e DNA content of embryo c e l l s and the l a t i t u d e of o r i g i n . The c u r v i l i n e a r relationship obtained was then used for the estimation of latitude from the DNA measurements and tolerance l i m i t s were affixed to i t . The r e l a t i v e DNA content variable was used thereafter conjointly with Yao's variables for the c l a s s i f i c a t i o n with discriminant functions. 3. RESULTS AND DISCUSSION 31. The seed tests The seed test was conducted as shown on Figure 1 for Boezingen tree No. 12. The results are summarized in Table 3. A l l but one averages are between 1.00 and 1.73. It can therefore be assumed that both unknown stands are of Coastal o r i g i n . The homogeneity of the tree averages for the Boezingen stand i s remarkable and might be an indication of a close parentage between the trees from which seeds had o r i g i n a l l y been gathered in North America. S e e d C l a s s i f i c a t i o n 1 Douglas -fir: Boz j nqe n (Switzerland), Tree Number 12 Morphological Chdrdcteristics Seed Number 1 2 3 4 5 6 7 8 9 10 11 Total Mean Seed tip i I I > 1.0 mm 1 X X X X X X X 7 o-5-io mm 2 X X X X 8 <o-5 mm 3 — Length Z 15 1.36 not tri&ngular 1 X X X X X X 6 Body intermediate 2 X X X X X 10 Shape triangular 3 — 16 145 Body Contour Cridge) distinct 1 X X X X X X X 7 intermediate 2 X X X X 8 not distinct 3 — Z 15 1-36 Body rouqh 1 X X X X X X X X 8 medium 2 X X X 6 Surface smooth 3 — Z 14 1-27 Overall Mean 1-56 l:Allen,G-S-(l960) A method of distinguishing coastal from interior Douglds-fir seed BC Lumberman--32-T a b l e 3. Average v a l U e s o f f o u r seed c l a s s i f i c a t i o n c r i t e r i a TREE jHO.OF SEED TIP BODY BODY BODY AVERAGE NO. SEEDS . LENGTH . SHAPE. CONTOUR SURFACE B0EZINGEN/SW1ITZERLAND: 6 11 1.45 1.73 1.00 1.09 1.32 7 11 1.27 1.00 1.27 1.64 1.29 8 11 1.18 1.27 1.18 1.18 1.20 10 11 1.36 1.36 1.09 1.18 1.25 11 11 1.27 1.18 1.00 1.18 1.16 12 11 1.36 1.45 1.36 1.27 1.36 13 11 1.73 1.45 1.18 1.27 1.41 14 11 2.55 1.09 1.09 1.55 1.57 15 11 1.55 1.18 1.00 1.27 1.25 MEAN 99 1.52 1.30 1.13 1.29 1.31 PURDA LESNA/POLAND: 50 1.26 1.48 1.60 1.52 1.47 32. Seed y i e l d T a b l e 4 r e v e a l s the seed y i e l d f o r t he Swiss provenance Boenzingen. The p e r c e n t o f f i l l e d seeds i s r e l a t i v e l y low. When compared w i t h d a t a g i v e n by Orr-Ewing (1965), i t s u g g e s t s a c l o s e p a r e n t a g e between the p a r e n t s and, a c c o r d i n g l y , between the t r e e s i n the s t a n d . However, such an a s s e r t i o n must be made p r u d e n t l y as G r i f f i t h (1968) i n d i c a t e d a v a r i a t i o n o f 3.47% to 57.72% f i l l e d seed p er cone f o r n o n - f e r t i l i z e d .< w i n d - p o l l i n a t e d t r e e s i n d i f f e r e n t y e a r s . The average number o f e x t r a c t e d seeds per cone i s lower than Orr-Ewing's r e c o r d s . S i n c e the -33-Table 4. Seed y i e l d record for provenance Boenzirigeri/Switzerland TREE NO. NO. OF CONES NO. OF SEEDS EXTRACTED AVERAGE NO. OF SEEDS/ CONE NO. Of FILLED SEEDS AVERAGE NO. OF FILLED SEEDS/CONE PERCENT OF FILLED SEEDS 6 15 611 40.73 24 1.60 3.92 7 15 1005 67.00 183 12.20 18.20 8 13 83 6.38 13 1.00 15.66 9 15 - - -10 15 1018 67.87 41 2.73 4.02 11 15 876 58.40 - -12 15 487 32.47 95 6.33 19.50 13 15 135 9.00 - - -14 15 19 1.27 7 0.47 36.84 15 11 11 1.00 1 0.09 .; 9.09 TOTAL 144 4245 364 AVERAGE 424.5 29.48 36.4 2.53 8.57 cones come from the only seed bearing trees in year 1971, i t can be assumed that 1971 was not a good seed-year. Or i f 1971 was a good seed year, a substantially strong i s o l a t i o n exists among the Douglas-fir trees at the time of p o l l i n a t i o n to avoid inbreeding within population. Adverse climatic conditions could have been also the cause of the low percentage of f i l l e d seeds. Table 5 gives the seed y i e l d for the 20 Polish stands. Only the f i l l e d seeds were counted and, therefore, the percent of f i l l e d seeds is not indicated. The number of f i l l e d seeds per cone is extremely low. Here again, three main causes can be considered: bad cone crop, -34-inbreeding, and adverse conditions of adaptation. Table 5. Seed y i e l d record for Polish provenances No. 1-20 PROVENANCE NO. OF CONES NO. OF FILLED SEEDS AVERAGE NO. OF FILLED SEEDS/CONE 1. WOLINSKI 27 -2. MIEDZYZDROJE 24 1 0.04 3. DOBRZANY 24 4. BIALY BOR 21 5. SOSNO 26 6. PURDA LESNA 29 157 5.41 7. RYN 30 20 0.66 8. MYRADZ 27 23 0.85 9. JAROCIN 26 4 0.15 10. KACZORY 25 29 1.16 11. GOLABKI 27 1 0.04 12. SKRZECIN 26 1 0.04 13. DOLICE 26 - -14. SWIEBODZIN 32 22 0.69 15. KROSNO 26 1 0.04 16. UJSOLY 19 - -17. POKRZYWNO 24 18. DUSZNIKI 15 6 0.40 19. KOWARY 12 - -20. BROJCE 22 3 0.14 AVERAGE 23.1 13.4 0.58 -35-33. Cone scale morphology 331. The Swiss stand The results of the cone scale measurements are given in Table 6. A nested analysis of variance was then performed on each character, as shown in Table 7 and Table 8. The presence of s i g n i f i c a n t differences between trees for both characters is in accordance with Yao's findings. The s i g n i f i c a n t differences between cones within trees i s probably due to an i n s u f f i c i e n t sampling. Since no information exists as to the sampling, this source of error was neglected. The averages used la t e r on w i l l be the provenance averages. Table 6. Mean values and standard deviations of cone scale c h a r a c t e r i s t i c s of trees in 'millimeters, for provenance  Boezingen/Switzerland TREE NO. NO. OBS. CONE SCALE WIDTH MEAN S.D. XONE SCALE LENGTH MEAN S.D. 6 10 18.00 0.87 13.30 0.75 7 10 20.89 0.37 15.79 0.90 8 10 19.33 0.66 18.42 0.61 9 10 17.48 1.36 14.90 0.58 10 10 19.24 0.76 14.79 0.48 11 10 17.90 0.56 15.78 0.52 12 10 20.36 2.11 17.94 1.52 13 10 21.63 0.44 16.46 0.52 14 10 22.09 0.75 15.68 0.65 15 10 22.99 .1.06. 16.92 0.83 MEAN 100 20.00 2.06 16.00 1.64 -36-Table 7. Analysis of variance of cone scale widths for provenance  Boezingen/Switzerland SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUES TREES 9 326.85 36.317 5.93** CONES/TREES 10 61.247 6.1247 14.82*** ERROR 80 33.069 0.41336 TOTAL 99 421.17 Note: In a l l ANOVA tables *** F value s i g n i f i c a n t at 0.1% level ** F value s i g n i f i c a n t at 1.0% level * F value s i g n i f i c a n t at 5.0% level NS F value non s i g n i f i c a n t Table 8. Analysis of variance of cone scale lengths for provenance  Boezi ngen/Switzerland SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUES TREES 9 208.20 23.134 8.53** CONES/TREES 10 27.134 2.7134 7.40*** ERROR 80 29.329 0.36661 TOTAL 99 264.67 T a b l e 9. Mean v a l u e s and s t a n d a r d d e v i a t i o n s o f cone s c a l e c h a r a c t e r i s t i c s o f  provenances i n m i l l i m e t e r s , f o r the P o l i s h provenances PR. NO. NO, OB, CONE SCALE WIDTH MEAN S.D. CONE BRACT WIDTH MEAN S.D. CONE SCALE LENGTH MEAN S.D. 1 s t PRONG LENGTH MEAN S.D. 2nd PRONG LENGTH MEAN S.D. RATING OF BRACT MEAN S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 80 72 72 63 77 78 80 80 78 80 80 72 75 96 62 56 64 45 32 56 21.24 20.72 21.52 20.31 19.38 20.13 19.87 20.98 18.89 19.83 21.57 15.90 19.06 19.90 23.97 22.65 23.08 20.89 17.96 20.77 1.37 2.01 2.50 1.53 1.62 1.43 1.63 2.19 1.44 2.35 2.04 0.90 1.53 1.81 3.22 1.21 3.37 2.25 1.68 1.39 4.89 4.95 4.92 4.77 4.66 4.63 5.35 4.49 4.67 4.73 5.08 4.10 4.33 5.05 4.88 3.93 4.84 4.09 5.78 5.26 0.84 0.59 0.46 0.89 0.50 0.42 0.60 0.61 0.87 0.66 0.43 0.31 0.48 0.50 0.61 0.63 0.57 0.40 0.75 0.48 17.04 16.60 17.78 16.62 16.23 18.08 17.72 16.67 16.40 17.03 16.92 12.06 15.64 16.26 20.11 19,19 18.91 17.18 14.92 17.22 1.53 1.68 1.53 1.47 1.40 1.18 1.80 1.53 1.92 1.77 1.32 0.63 1.32 1.22 2.73 1,20 2.78 1.32 1.32 1.46 7.05 7.95 7.17 6.98 6.73 6.28 8.29 7.46 6.87 7.17 8.02 5.6.9 6.56 9.80 7.60 6,47 7.17 6.73 8.05 7.56 1.22 1.15 1.16 0.88 1.22 0.70 1.26 1.54 1.26 1.06 0.95 0.59 0.85 1.38 1.40 1.15 1.07 1.00 1.03 0.85 3.50 3.74 3.14 3.01 2.94 3.11 2.85 3.26 2.96 3.28 3 t l 9 1.88 2.94 2.24 3.81 2.59 3.17 2.90 3.04 3.22 0.69 0.63 0.50 0.45 0.54 0.37 0.52 0.92 0.62 0.81 0.58 0.28 0.48 0.64 0.69 0.59 0.67 0.38 0.75 0.44 2.50 2.04 2.40 2.19 2.52 2.88 2.27 2.39 2.62 2.34 2.86 1.43 2.33 1.00 2.66 1.75 2.20 2.49 1.50 2.71 0.69 0.76 0.69 0.72 0.60 0.32 0.66 0.49 0.52 0.67 0.35 0.50 0.66 0.00 0.51 0.55 0.65 0.66 0.57 0.49 AVERAGE 20.41 2.57 4.76 0.71 16.91 2.25 7.32 1.45 3.03 0.74 2.26 0.75 -38-Analyses of variance were performed on each variable to study the s i m i l a r i t y of the Polish stands. Tables 10 to 15 show the results of the one-way analysis of variance. Table 10. Analysis of variance of cone scale widths, for Polish  provenances No. 1-20. SOURCE D.F. SUM OF SQUARES MEAN SQUARE r VALUES PROVENANCES ERROR 19 1378 3956.8 5281.8 208.25 3.8329 54.33*** TOTAL 1397 9238.6 Table 11. Analysis of variance of cone bract widths, for Polish provenances No. 1-20 SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUES PROVENANCES ERROR 19 1378 211.05 491.27 11.108 0.35651 31.16*** TOTAL 1397 702.32 Table 12. Analysis of variance of cone scale lengths, for Polish provenances No. 1-20 SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUES PROVENANCES ERROR 19 1378 3467.4 3612.5 182.50 2.6216 69.61*** TOTAL 1397 7079.9 -39-Table 13. Analysis of variance of f i r s t prong lengths, for Polish provenances No. 1-20 SOURCE D.F. SUM OF SQUARES MEAN SQUARE . F VALUES PROVENANCES ERROR 19 1378 1196.8 1740.3 62.992 1.2629 49.88*** TOTAL 1397 2937.2 Table 14. Analysis of variance of second prong lengths, for Polish provenances No. 1-20 SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUES PROVENANCES ERROR 19 1378 279.28 494.95 14.699 0.35918 40.92*** TOTAL 1397 774.23 Table 15. Analysis of variance of rating of bracts, for  Polish provenances No. 1-20 SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUES PROVENANCES ERROR 19 1378 345.62 449.60 18.191 0.32627 55.75*** TOTAL 1397 795.22 -40-For each character, the differences between the twenty stands are highly s i g n i f i c a n t , so that i t would not be rat i o n a l to c l a s s i f y them by studying one and drawing similar conclusions on the others. The next step was to try to group the stands by means of the Duncan new multiple range t e s t . This was only performed for a pro b a b i l i t y level of 0.05. The six groupings obtained are pictured in Table 16. Table 16. Duncan new multiple range test of six cone scale measurements, for Polish provenances Mo. 1-20 SCALE WIDTH GROUPS 15 17 16 11 3 1 8 18 20 2 4 6 14 7 10 5 13 9 19 12 BRACT WIDTH GROUPS 1st PRONG SCALE LENGTH LENGTH 2nd PRONG . RATING OF LENGTH BRACT 19 7 20 11 14 2 3 1 15 17 4 10 9 5 6 8 13 12 18 16 GROUPS GROUPS GROUPS GROUPS 15 16 17 6 3 7 20 18 1 10 11 8 4 2 9 14 5 13 19 12 14 7 19 11 2 15 20 8 17 10 3 1 4 9 5 18 13 16 6 12 15 2 1 10 8 20 11 17 3 6 19 4 9 13 5 18 7 16 14 12 6 11 20 15 9 5 1 18 3 8 10 13 7 17 4 2 16 19 12 14 -41-The ordering, as well as the grouping, i s d i f f e r e n t for each character. Consequently, as none of the variables are a p r i o r i more important than the others, no general grouping was attempted and each provenance is considered separately. 34. One thousand-seed weight Seven samples from six trees provided 321 seeds and were used to calculate the one thousand-seed weight of provenance Boezingen/ Switzerland. This gave an average of 10.1167 g. For provenance Purda Lesna/Poland, three samples, t o t a l i n g 134 seeds, were weighed. The average one thousand-seed weight obtained is 13.7567 g. It is not possible to.judge the value of both averages at this stage. However, as they are based on few samples, when compared with the cone scale morphology variables, less confidence can be attributed to them. Therefore, the one thousand-seed weight was neglected in the attempts of c l a s s i f i c a t i o n based on the individual variables. It was used only in the multivariate analyses, where i t s influence is weighed by the presence of other variables. 35. Relative DNA content of embryo c e l l s 351. The Swiss stand Table 17 l i s t s the number of observations, the mean value (in Feulgen Absorption Units) for each embryo and tree in the four -42-sampled t r e e s . E l - L a k a n y and S z i k l a i (1972), u s i n a e x a c t l y t he same t e c h n i q u e , o b t a i n e d s t a n d a r d d e v i a t i o n s not h i g h e r than 11.19 f o r t h e provenance a v e r a g e s , but i n t h i s s t u d y s t a n d a r d d e v i a t i o n s up to 33.3 Feu l g e n A b s o r p t i o n U n i t s were o b t a i n e d . T a b l e 17. Number o f o b s e r v a t i o n s , mean and s t a n d a r d d e v i a t i o n o f  r e l a t i v e DNA c o n t e n t , i n Fe u l g e n A b s o r p t i o n U n i t s , f o r  provenance B o e z i n g e n / S w i t z e r l a n d TREE NO. 6 TREE NO. 7 TREE NO. 10 TREE NO. 12 NO. MEAN S.D. NO. MEAN S.D. NO. MEAN S.D. NO. MEAN S.D. OBS. OBS. OBS. OBS. 20 122.0 18.1 20 104.0 13.0 20 133.4 17.9 • 20 137.9 23.7 19 123.0 21.0 20 105.5 9.9 20 170.1 15.6 20 144.1 21.3 20 165.4 33.3 19 125.4 27.2 20 117.8 12.9 20 108.4 17.0 18 100.2 30.5 20 144.5 18.0 20 112.8 18.5 20 = 129.0 16.9 20 124.7 18.9 20 142.0 22.1 20 115.2 19.6 20 136.0 21.8 20 123.3 13.8 20 120.9 16.9 20 107.1 16.5 20 122.9 19.2 20 128.8 18.3 20 122.4 13.1 20 128.0 21.0 20 147.2 14.9 20 137.5 17.6 20 129.6 13.2 20 140.1 19.2 20 135.8 26.7 20 122.1 17.9 20 137.4 19.3 .20 107.2 10.0 20 118.1 17.3 20 117.9 19.1 20 136.1 17.0 20 109.7 13.2 20 99.2 18.3 197 126.8 26.2 199 126.8 21.8 200 124.1 24.9 200 127.9 24.5 A n e s t e d a n a l y s i s o f v a r i a n c e was performed on the d a t a , the r e s u l t s o f which a re shown i n T a b l e 18. The a n a l y s i s o f v a r i a n c e r e v e a l s no d i f f e r e n c e s between the t r e e s , but shows s t a t i s t i c a l l y s i g n i f i c a n t d i f f e r e n c e s between the embryos w i t h i n the t r e e s . I t seems r e a l i s t i c to c o n s i d e r t h e s e d i f f e r e n c e s as due m a i n l y to a l a c k o f s k i l l i n p r o c e s s i n g ( m a i n l y i n the s q u a s h i n g o f the embryos on the m i c r o s c o p e s l i d e ) and i n the s e a r c h f o r w e l l - f o r m e d n u c l e i -43-surrounded by a clear cytoplasm. High values are probably due to a very dark cytoplasm v/ith impurities and very low values to c e l l s which partly l o s t their cytoplasm during the squashing. These values could be reduced with improved technique and experience. Table 18. Analysis of variance of r e l a t i v e DNA content, for  provenance Boezingen/Switzerland SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUES TREES 3 1510.861 503.6201 0.0918NS EMBRYOS/TREES 36 197489.7 5485.824 15.1052*** ERROR 756 274560.5 363.1750 TOTAL 795 473561.1 As only a provenance average was required for the c l a s s i f i c a t i o n task, the average of a l l observations was calculated for further use: Number of observations: 796 Minimum value : 60.95 Feulgen Absorption Units Maximum value : 232.47 Feulgen Absorption Units Mean value : 126.38 Feulgen Absorption Units Standard deviation : 24.46 Feulgen Absorption Units 352. The Polish stand at Purda Lesna The results of the determination of the r e l a t i v e DNA content in the Polish provenance No. 6 (Purda Lesna) show a high v a r i a b i l i t y , -44-even after elimination and substitution of doubtful values: Number of observations Minimum value Maximum value Mean value Standard deviation 985 89.65 Feulgen Absorption Units 201.41 Feulgen Absorption Units 147.94 Feulgen Absorption Units 21.99 Feulgen Absorption units The re s u l t i n g number of observations per embryo, averages and standard deviations are given in Table 19. Table 19. Number of observations, mean and standard deviation of  re l a t i v e DNA content in Feulgen Absorption Units at the  embryo l e v e l , for provenance Purda Lesna/Poland NO. MEAN NO. MEAN NO. MEAN NO. MEAN OBS • S.D. OBS. S.D. OBS. S.D. OBS. S.D. 20 139.3 15.7 20 161.6 18.1 20 130.1 15.9 20 155.3 12.9 20 148.2 12.2 20 155.2 17.6 20 151.0 8.0 18 181.1 13.8 20 146.1 17.5 20 151.3 16.5 20 154.2 13.8 19 134.9 22.3 20 137.0 20.5 19 160.4 17.2 20 160.9 11.9 20 133.8 12.2 20 133.6 15.9 20 146.2 15.4 20 134.3 10.9 20 160.4 10.1 20 124.3 12.6 19 113.6 11.1 19 169.5 14.5 20 163.2 10.2 20 148.2 16.9 19 173.9 10.5 20 135.1 16.2 20 156.8 8.7 20 138.6 18.3 18 179.9 10.4 20 134.4 14.4 20 153.2 14.0 20 119.3 15.3 20 145.7 13.3 20 168.5 13.9 20 155.3 13.3 20 125.7 12.0 20 132.2 16.6 20 124.7 19.1 19 172.1 14.9 19 132.9 15.4 20 117.6 13.3 20 141.9 12.9 20 152.6 27.7 19 162.6 15.9 20 147.7 12.3 20 142.8 11.9 17 175.9 15.6 20 157.4 8.1 20 169.5 13.2 -45-The data were then submitted to an analysis of variance to test the value of the provenance average. The analysis of variance is shown in Table 20. The s t a t i s t i c a l l y s i g n i f i c a n t F-value requires the rejection of the overall-mean. As i t was said in Section 22, no d i s t i n c t i o n had been made between the trees, so that this source of variation could not have been studied. Table 20. Analysis of variance of r e l a t i v e DNA content, f o r  provenance Purda Lesna/Poland SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUE EMBRYOS 49 266690. 5442.6 24.64*** ERROR 935 206490. 220.85 TOTAL 984 473180. The stand could have originated from a mixture of seeds from d i f f e r e n t o r i g i n s . To test this hypothesis, a Duncan new multiple range test was performed at the 0.05 level of probability on the averages of the embryos. This resulted in a subdivision of the f i f t y embryos in twenty homogeneous subsets, which a r e . l i s t e d in Table 21. None of these subsets are c l e a r l y isolated from the others, but assuming representative sampling of the stand, mixed seed may be the answer. However, i t i s d i f f i c u l t to assert that the sample is representative of the stand, because i t s extent i s not known. -46-Therefore, the hypothesis of mixed seeds cannot be accepted. Table 21. Duncan new multiple range 'test of r e l a t i v e DNA content, for provenance Purda Lesna/Poland^ 0 9 , 24, 9) (24, 9, 6, 36, 10) (6, 36, 10, 27, 23, 11, 5, 42, 31, 34, 41, 33) (27, 23, 11, 5, 42, 31, 34, 41, 33, 4, 8, 1) (23, 11, 5, 42, 31, 34, 41, 33, 4, 8, 1, 37, 38) (41, 33, 4, 8, 1, 37, 38, 25, 22) (4, 8, 1, 37, 38, 25, 22, 3, 18) (8, 1, 37, 38, 25, 22, 3, 18, 7, 2) (37, 38, 25, 22, 3, 18, 7, 2, 28, 16, 49) (38, 25, 22, 3, 18, 7, 2, 28, 16, 49,.46) (25, 22, 3, 18, 7, 2, 28, 16, 49, 46, 29, 15, 39, 47, 45) (18, 7, 2, 28, 16, 49, 46, 29, 15, 39, 47, 45, 13) (28, 16, 49, 46, 29, 15, 39, 47, 45, 13, 17, 43, 30, 14) 0 6 , 49, 46, 29, 15, 39, 47, 45, 13, 17, 43, 30, 14, 12) (49, 46, 29, 15, 39, 47, 45, 13, 17, 43, 30, 14, 12, 44) (17, 43, 30, 14, 12, 44, 35, 26, 32) (14, 12, 44, 35, 26, 32, 48) C35, 26, 32, 48, 20, 50) (26, 32, 48, 20, 50, 21) (48, 20, 50, 21, 40) Consequently, whatever the cause was for the lack of homogeneity between the embryos, the provenance average has a limited value. The variable " r e l a t i v e DNA content" was therefore abandoned. Based on f i f t y embryos studied. The parentheses enclose the embryo numbers which are not s i g n i f i c a n t l y d i f f e r e n t from each other.. -47-36. The determination of origin 361. General For the purpose of determining the o r i g i n of the unknown samples, i t was necessary to establish discriminations and groupings as well as to combine regression l i n e s together. Their v a l i d i t y is r e l a t i v e to the practical'.purpose and no general conclusions should be drawn from them. 362. Display of the cone scale morphology and one thousand-seed  weight variables, and t h e i r individual use Figure 2 shows the portion of the natural range of Douglas-fir used by Yao (1971). The area i s subdivided into twenty-one climatic subzones used for seed c o l l e c t i o n purposes. Due to the p a r t i c u l a r geography of the P a c i f i c Coast, the c l i m a t i c regions are elongated p a r a l l e l to the coast. -48-Figure 2. Geographic d i s t r i b u t i o n climatic zonation (from of Douglas-fir seed sources and Yao [1971]) -49-For each of the seven characters of Yao (1971) (Fig. 3-9), a graph was drawn, displaying the range of the provenance averages inside the zones. The seed c o l l e c t i o n zones were arranged from north to south and from the Coast to the Interior of the continent. For each zone, the number of provenances sampled was indicated. F i g - 3- N a t u r a l v a r i a t i o n o f t h e c o n e s c a l e w i d t h o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s - 1 mm; 26 -25 24 231 22| 21 201 II 13 8 15 6 4 5a 7a 8 2a 3a la 3b 2b laa 3c 5b 7b 2c 2d lb 3d 2e Ic 3e zone 1 Data from Yao [1971] F i g - 4 - N a t u r a l v a r i a t i o n o f t h e b r a c t w i d t h o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s - 1 mm 6 0 5-8 13 8 6 15 1 6 4 5a 7a 8 2a 3a la 3b 2b laa 3c 5b 7b 2c 2d lb 3d 2e Ic 3e zone Data from Yao 11971] F i g - 5 - N a t u r a l v a r i a t i o n o f t h e c o n e s c a l e l e n g t h o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s 3 1 mm 22 21 20 19 -18 4 II 4 13 8 8 15 6 4 5a 7a 8 2a 3a la 3b 2b laa 3c 5b 7b 2c 2d lb 3d 2e Ic 3e Data from Yao £1971] F i g - 6 - N a t u r a l v a r i a t i o n o f t h e f i r s t p r o n g l e n g t h o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s - 1 mmj 11-0 IO-0 9 0 8-0 -7-0 6-0 2 6 15 13 8 6 4 5a 7a 8 2a 3a la 3b 2b laa 3c 5b 7b 2c 2d lb 3d 2e Ic 3e zone Data from Yao £1971] F i g - 7 - N a t u r a l v a r i a t i o n o f t h e s e c o n d p r o n g l e n g t h o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s - 1 15 13 6 8 6 4 5a 7a 8 2a 3a la 3b 2b laa 3c 5b 7b 2c 2d lb 3d 2e Ic 3e zone 1 Data from Yao [1971] F i g * 8- N a t u r a l v a r i a t i o n o f t h e r a t i n g o f b r a c t o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s -3-5 3 0 2-5 2 0 1-5 8 ® 15 13 4 6 4 5a 7a 8 2a 3a la 3b 2b laa 3c 5b 7b 2c 2d lb 3d 2e Ic 3e ]Data from Yao 119713 1 F i g - 9 - N a t u r a l v a r i a t i o n o f t h e l O O O - s e e d w e i g h t o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s l 15 10 9 8 7 -H 3 5 II 1 3 8 6 4 6 4 5a 7a 8 2a 3a la 3b 2b laa 3c 5b 7b 2c 2d lb 3d 2c Ic 3e Data from Yao 11971] zone c n i -57-With the exception of the variable "rating of bract", there seems to be a s l i g h t increase from north to south in the average values. Nevertheless, the ranges are hardly distinguishable except for the regression c o e f f i c i e n t s of t h e i r relationship with the l a t i t u d e . Also, in each case there exists quite a strong relationship.between the range of variation and the number of provenances included in the climatic zones. This is exemplified by the variable " f i r s t prong length" in Figure 10. -58-F i g l O - R e l a t i o n s h i p b e t w e e n t h e n u m b e r o f p r o v e n a n c e s a n d t h e r a n g e o f v a r i a t i o n o f t h e f i r s t p r o n g l e n g t h o f D o u g l a s - f i r i n s i d e t h e s e e d c o l l e c t i o n z o n e s - 1 mm o2d o3b 2-0 1-0-0 0 o3d o2b. > l a ,o5c b5b o4 olb o[a o2e °2c o3a o7a 8 o7b 0 6 ?3e, I M 1 o3c oic ' ' ' ' I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 number of provenances 1 Data from Yao [1971], -59-Consequently, higher numbers of provenances could have displayed wider ranges. Each unknown sample was then compared with Figures 3-8 and the number of c l i m a t i c zones where i t t h e o r e t i c a l l y could have come from was recorded in Table 22. In eight provenances, the average of at least one character was outside the presently known range of the natural variation of the species. The character for which the highest number of unknown provenances f a l l outside the range i s the cone scale width, followed by the bract width. As indicated before, the complete variation inside the seed zones has probably not been described. Therefore, no v a l i d conclusion can be made from Table 22 concerning the influence of the environment on the phenotypic characters involved. On the other hand, i f we consider the displayed variation as representative for the total v a r i a t i o n , i t can be concluded that the environment has a greater influence on cone scale width and on bract width than on the other characters. It can be concluded from the previous comparisons, that the single variable approach does not allow- an unequivocal c l a s s i f i c a t i o n . This approach has to be abandoned, unless a stronger c h a r a c t e r i s t i c can be found which may switch the search to the correct climatic zone. The variable " r e l a t i v e DNA content" and i t s strong relationship with the latitude for a great portion of the Douglas-rfiT range, wtll be used.to narrow the search, to a l i t t l e group of climatic zones. This w i l l be done in Section 369_, after the multivariate approach is considered. -60-Table 22, Number of cl i m a t i c zones from which each unknown provenance coUId Noriginate CHARACTERISTICS1 PROVENANCE Wl W2 LI L2 L3 R SWITZERLAND: 1. BOEZINGEN 3 2 POLAND: 1. WOLINSKI 8 8 7 8 9 11 2. MIEDZYZDROJE 4 10 5 14 9 13 3. DOBRZANY 8 9 9 8 9 12 4. BIALY BOR 6 7 5 9 7 14 5. SOSNO - 5 3 . 7 8 11 6. PURDA LESNA 4 5 11 2 8 5 7. RYN 3 13 9 13 4 16 8. MYRADZ 5 2 5 10 10 12 9. JAROCIN - 5 4 9 6 14 10.KACZ0RY 3 7 7 8 11 12 ll.GOLABKI 8 16 7 ' 14 11 5 12.SKRZECIN — r- - 2 13.DOLICE - T 1 4 8 2 14.SWIEB0ZIN 3 14 3 4 -15.KR0SN0 8 8 4 11, 8 16.UJS0LY 13 14 3 1 10 17.P0KRZYWN0 10 7 14 8 10 14 18.DUSZNIKI 4 - 8 7 6 12 19.KOWARY - 4 - 14 8 3 20.BR0JCE 4 16 9 11 10 . 8 1 Wl = cone scale width W2 = bract width LI = cone scale length L2 = f i r s t prong length L2 = second prong length R = rating of bract -61-363. D i s t i n c t i o n between Coastal and Interior Douglas-fir, and  i t s application to the Polish stands S z i k l a i (1969) indicated the climatic zones l a , l a a , l b , 2a, 2c, 3a, 3b, 3c, 3d and 4 as enclosing the Coastal subspecies of Douglas-fir and the zones 5a, 5b, 6, 7a, 7b and 8 the Interior subspecies (see Figure 2). This corroborates Haddock and S z i k l a i ' s 0966) delimitation of seed zones and is in agreement with Isaac and Dimock II 0958) range of Douglas-fir (vary mehzires%i)• °n the contrary, El-rLakany and S z i k l a i 0971, 1972) considered sampled provenances No. 44 from cl i m a t i c region 3b; 59, and 76 from 3c; 86, 97 and 100 from 3d; and 112 and 113 from 3e as belonging to the Interior subspecies. Stepwise discriminant analysis (Seagraves [1972]) was applied to Yao's one hundred and twenty-four provenances, including the six cone morphology- variables on the provenance l e v e l . Climatic regions 1, 2, 3 (without 3d and 3e), and 4 were considered as Coastal, and 3d, 3e, 5, 6, 7 and 8 as I n t e r i o r , in a f i r s t attempt. The discriminant functions obtained for the problem subspecies allowed to c l a s s i f y 77.42% of the cases c o r r e c t l y (Table 23). Redistribution of marginal provenances in several steps produced the discriminant functions NEWSUB (Figure 11), which c l a s s i f y 95.97% of the cases in the correct group. In this new grouping, zones 1, 2, 3a, 3b, 3c and 3e are e n t i r e l y Coastal, zones 5a, 6, 7 and 8 e n t i r e l y Interior, and 3d, 4 and 5b partly Coastal and partly Interior. S t i l l m i s c l a s s i f i e d are provenances No. 13, 22 and 30 on Vancouver Island, provenance No. 27 situated at an a l t i t u d e of 915 m in the Coast Mountains of B r i t i s h o Columbia, and provenance No. 38 in the Puget Trough (Washington). -62-Table 23. Discriminant analyses for the d i s t i n c t i o n between  Coastal and Interior Douglas-fir TITLE SUBSPE NEWSUB NO. OF PROVENANCES 124 124 NO. OF GROUPS 2 2 NO. OF VAR. INCLUDED 6 6 NO. OF VAR. USED 2 5 % CASES CORRECTLY CLASSIFIED 77.42 95.97 APPROXIMATE F STAT. 14.34705*** 39.05072*** D.F. FOR APPROX. F 2; 121.00 5; 118.00 F VALUES FOR EACH VARIABLE: WITH D.F. 1; 121 1; 118 CONE SCALE WIDTH 8.1483*** BRACT WIDTH CONE SCALE LENGTH 24.1737*** 6.0267* FIRST PRONG LENGTH 32.3206*** SECOND PRONG LENGTH 51.6860*** RATING OF BRACT 13.2862*** 26.4102*** Figure 11. D e l i m i t a t i o n o f C o a s t a l and I n t e r i o r D o u g l a s - f i r -64-It i s well possible that this discrimination is biased. Indeed, only one provenance, No. 98, t y p i f i e s Interior Douglas-fir below latitude 47°N. Consequently, the Interior subspecies is heavily influenced by the provenances of upper latitudes. Subsequent analyses based on l a t i t u d i n a l groupings w i l l take account of the r e s u l t of NIWSUB in spite of i t s imperfection. Only the Swiss provenance and the Polish stand No. 6 could be y c l a s s i f i e d with the help of Allen's seed t e s t s . The twenty Polish stands were therefore c l a s s i f i e d into Coast or Interior using the discriminant functions. The method consists in computing the p r o b a b i l i t y that a subject (provenance) came from each group (Coast, I n t e r i o r ) . The subject is then classed according to the grOupito which i t has the highest probability of membership (Seagraves [1971]). Only four provenances (Table 24) are classed as Coastal Douglas-f i r . Surprisingly, provenance No. 6 i s classed as belonging to the Interior group in contradiction with the results of Allen's seed t e s t s . As enough evidence exists for the goodness of the seed tests (Section 13), the contradictory c l a s s i f i c a t i o n of provenance No. 6 is an indication of the low r e l i a b i l i t y of NEWSUB for o r i g i n determination. -65-Table 24. C l a s s i f i c a t i o n of Polish provenances No. 1-20 into  Coastal or Interior Douglas-fir PROV. NO. CLASSED WITH PROBABILITY 1. COAST 0.613 2. INTERIOR 0.549 3. INTERIOR 0.795 4. INTERIOR 0.975 5. INTERIOR 0.999 6. INTERIOR 0.899 7. INTERIOR 1.000 8. INTERIOR 0.970 9. INTERIOR 1.000 10. INTERIOR 0.947 11. INTERIOR 1.000 12. INTERIOR 1.000 13. INTERIOR 0.998 14. INTERIOR 1.000 15. COAST 1.000 16. COAST 0.963 17. COAST 0.969 18. INTERIOR 0.979 19. INTERIOR 1.000 20. INTERIOR 0.996 364. Discriminant analyses using c l i m a t i c zones, and their application  to Polish provenances Two analyses were performed, in which the provenances were grouped -66-into the climatic zones (Figure 2). One was done using the provenance averages, ca l l e d CLIAMP. As a group cannot consist of less than two observations, zone 8 had to be merged with zone 7A. The other analysis was done on the tree averages, and was.called CLIAME (Table 25). No analysis was attempted on the basic observations as i t would have meant handling 131,328 data! Table 25. Discriminant analyses for the d i s t i n c t i o n between climatic zones TITLE CLIAMP CLIAME NO. OF PROVENANCES 124 124 NO. OF.GROUPS 20 21 NO. OF VAR. INCLUDED 6 6 NO. OF VAR. USED 6 6 % CASES CORRECTLY CLASSIFIED 24.19 14.25 APPROXIMATE F STAT. 4.53732*** 17.57088*** D.F. FOR APPROX. F 114;577.27 120;10398.81 F VALUES FOR EACH VARIABLErWITH D.F. 19; 99 20; 1798 CONE SCALE WIDTH 3.3182*** 11.0289*** BRACT WIDTH 2.0988** 8.1620*** CONE SCALE LENGTH 1.9775* 6.0492*** FIRST PRONG LENGTH 3.0442*** 8.8502*** SECOND PRONG LENGTH 6.7738*** 31.8118*** RATING OF BRACT 4.4438*** 14.3137*** CLIAMP gives a better discrimination than CLIAME, which can be seen in the number of cases (observations) classed in the i r own group. -67-This is not surprising, since Yao (1971) found s i g n i f i c a n t differences between the trees for a l l six variables. CLIAMP discriminant functions were used to class the twenty Polish provenances (Table 26). Each unknown stand was characterized by i t s averages. The c l a s s i f i c a t i o n of provenance 2, 3, 6, 8 and 16 is in contradiction with the c l a s s i f i c a t i o n obtained in NEWSUB. However, provenances 2, 3, and 8 are classed into rather scattered groups, and provenance No. 16 could j u s t as well be classed into group 3b, which would be in accordance with NEWSUB. Concerning provenance No. 6, as the one thousand-seed weight is known, a special discriminant analysis w i l l be done in which i t w i l l be classed (see Section 366). Provenance No. V3 (classed in CLIAMP d i f f e r e n t l y than in NEWSUB) and provenance No. 15 (classed s i m i l a r l y in both analyses) were then classed in CLIAME (Table 27). Each unknown stand was characterized by i t s basic observations. -68-Table 26. C l a s s i f i c a t i o n of Polish provenances No. 1-20 into climatic zones (CLIAMP) PROV. CLASSED WITH PRO- GROUP OF WITH PRO-NO. TO GROUP BABILITY NEXT LOWER BABILITY PROBABILITY 1 2a ( 0%)T 0.514 la ( 0%) 0.147 2 2a ( 0%) 0.501 l a ( 0%) 0.171 3 2a ( 0%) 0.204 l c (60%) 0.192 4 5a {20%) 0.458 6 ( 0%) 0.196 5 5a (20%) 0.598 6 ( 0%) 0.238 6 3a (33%) 0.814 2a ( 0%) 0.047 7 7b (.0%) 0.814 7a+8 (20%) 0.159 8 2a .( 0%) 0.420 3b (18%) 0.124 9 5a (20%) 0.466', 6 ( 0%) 0.401 10 6 ( 0%) 0.396 2a ( 0%) 0.227 11 5b (13%) 0.323 7a+8 (20%) 0.293 12 6 ( 0%) 0.894 7a+8 (20%) 0.079 13 6 ( 0%) 0.717 5a (20%) 0.191 14 7b ( 0%) 0.594 7a+8 (20%) 0.404 15 l c (60%) 0.626 3e ( 0%) 0.208 16 6 ( 0%) 0.405 3b (18%) 0.392 17 3b (18%) 0.194 4 ( 0%) 0.148 18 6 ( 0%) 0.299 4 (;o%) 0.257 19 7b ( 0%) 0.683 7a+8(20%) 0.316 20 5a (20%) 0.330 7a+8 (20%) 0.233 The percentage in ^ parentheses indicates the frequency of cases corr e c t l y c l a s s i f i e d in CLIAMP -69-Table 27. C l a s s i f i c a t i o n of Polish provenances No. 3 and No. 15  into climatic zones (CLIAME) GROUP PROV. NO. 3 DOBRZANY PROV. NO. 15 KROSNO la ( o%)r 1 aa (17%) 4.2% 1.6% lb (34%) 1.4% 1.6% Ic (18%) 2.8% 4.8% 2a ( 4%) 2.8% 2b (25%) 6.9%; 1.6% 2c (33%) 6.9% 9.7% 2d ( 4%) 2e (31%) 4.8% 3a (20%) 9.7% 12.9% 3b ( o%) 3c ( 1%) 3d ( 8%) 1.4% 1.6% 3e (41%) 32.3%j 4 (27%) 22.2% 16.1% 5a ( 4%) 1.4% 1.6% 5b ( 7%) 6.9% 1.6% 6 (52%) 12.5% 4.8% 7a (23%) 2.8% 4.8% 7b (42%) 9.7% 8 (20%) 8.3% The percentage in parentheses indicates the frequency of cases corr e c t l y c l a s s i f i e d in CLIAME -70-For both provenances the c l a s s i f i c a t i o n i s very d i f f e r e n t from CLIAMp. Only at the subspecies level is the c l a s s i f i c a t i o n s i m i l a r . The s i g n i f i c a n t differences between observations in each variable are c e r t a i n l y responsible for this s i t u a t i o n . 365. Discriminant analyses using l a t i t u d i n a l zones, and t h e i r application  to Polish provenances The c l a s s i f i c a t i o n in the climatic zones gives a quite narrow indication of the distance to the coast, but remains more indeterminate in respect to the l a t i t u d e , most zones being at least 2° long. Therefore, two analyses using the same variables as in Section 364 were performed, based on l a t i t u d i n a l groupings. One, called LATPCO, was.done using the provenance averages, and the other, c a l l e d LATICO, using the tree averages (Table 28). The l a t i t u d i n a l zones have a width of 1°, but contain at least two observations. The grouping for LATPCO and LATICO are given in Figure 12 and Figure 13, respectively. Both discriminations give a higher percentage of cases correctly c l a s s i f i e d than the corresponding analyses based on climatic zones. This indicates that the variables included are more cl o s e l y connected with latitude than with longitude and l a t i t u d e together. LATPCO discriminant functions were used to class the twenty Polish provenances (Table 29). Each unknown stand was characterized by i t s averages. -71-Table 28. Discriminant analyses for the d i s t i n c t i o n between l a t i t u d i n a l zones TITLE LATPCO LATICO NO. OF PROVENANCES 124 124 NO. OF GROUPS 17 22 NO. OF VARIABLES INCLUDED 6 6 NO. OF VARIABLES USED 5 6 % CASES CORRECTLY CLASSIFIED 30.65 15.95 APPROXIMATE F STATISTIC 6.77875*** 17.13774*** D.F. FOR APPROXIMATE F 80;500.16 126;10428.19 F VALUES FOR EACH VARIABLE: WITH D.F. 16; 103 21; 1797 CONE SCALE WIDTH 6.5544*** 11.5536*** BRACT WIDTH 4.2951*** 9.8994*** CONE SCALE LENGTH ' 3.6681*** FIRST PRONG LENGTH 4.8996*** 8.8711*** SECOND PRONG LENGTH 12.7429*** 34.2631*** RATING OF BRACT 2.8516*** 10.0040*** Figure 12 Delimitation of l a t i t u d i n a l zones for analyses LATPCO and LATPWE Figure 13. Delimitation of l a t i t u d i n a l zones for analysis LATICO -74-Table 29. C l a s s i f i c a t i o n of Polish provenances No. 1-20 into  l a t i t u d i n a l zones (LATPCO) PROV. NO. CLASSED TO GROUP WITH PRO-BABILITY GROUP OF NEXT LOWER PROBABILITY WITH PRO-BABILITY 1 M (25%) ] 0.447 L (11%) 0.333 2 L (11%) 0.526 M (25%) 0.250 3 L (11%) 0.369 U (60%) 0.273 4 U (60%) 0.585 T (40%) 0.163 5 U (60%) 0.618 T (40%) 0.224 6 M (25%) 0.431 U (60%) 0.299 7 R (60%) 0.834 Q (68%) 0.127 8 M (25%) 0.750; U (60%) 0.141 9 U (60%) 0.498 T (40%) 0.286 10 U (60%) 0.453 0 (33%) 0.168 11 0 (33%) 0.646 U (60%) 0.155 12 T (40%) 0.776 R (60%) 0.124 13 U (60%) 0.802 M (25%) 0.093 14 R (60%) 0.999 Q (68%) 0.001 15 A ( 0%) 0.326 R (60%) 0.276 16 M (25%) 0.982 J (31%) 0.010 17 M (25%) 0.345 J (31%) 0.304 18 M (25%) 0.894 U (60%) 0.098 19 Q (68%) 0.904 R (60%) 0.093 20 T (40%) 0.311 0 (33%) 0.282 The percentage in parentheses indicates the frequency of cases corr e c t l y c l a s s i f i e d in LATPCO -75-Provenance No. 3 (classed in LATPCO d i f f e r e n t l y than in NEWSUB) and provenance No 15 (classed s i m i l a r l y in both cases) were then classed in LATICO (Table 30). Each unknown stand was characterized by i t s basic observations. Table 30. C l a s s i f i c a t i o n of Polish provenances No. 3 and No. 15  into l a t i t u d i n a l zones (LATICO) GROUP PROV. NO. 3 PROV. NO. 15 DOBRZANY KROSNO A (13%) ] 9.7% 21.0% B (14%) 14.5% C (23%) 3.2% D (33%) 1.4% 16.1% E (29%) 1.4% 4.8% F (21%) G (15%) 1.4% H (11%) 4.2% 8.1% I (16%) 1.6% J (10%) 4.2% K (13%) 6.9% L ( 6%) 4.2% M (12%) 2.8% 4.8% N ( 0%) 11.1% 0 (67%) 20.8% 9.7% P ( 3%) 4.2% 1.6% Q (30%) 2.8% R (43%) 6.9% S (21%) 1.4% 4.8% T (12%) 4.2% U ( 8%) 8.1% V (13%) 12.5% 1.6% The percentage in parentheses indicates the frequency of cases c o r r e c t l y c l a s s i f i e d .in LATICO -76-The classification is very different from LATPCO, even in the case of provenance No. 15. The separation between the groups is such that only a slight difference between two observations is enough to have them classed into different groups. The results of both CLIAME and LATICO are unreliable, as they only render the differences between the observations, as well in the natural range as in the unknown Polish stands. 366. Discriminant analysis, including the variable one thousand-seed  weight, and its application to Polish provenance No. 6 Purda Lesna On the assumption that with increased number of variables the discriminant functions will be better, the one thousand-seed weight was included in an analysis called LATPWE, LATPWE (Table 31) uses latitudinal groupings of provenances, as this supplies an analysis, which results in a better discrimination than climatic groupings. The groupings are the same as for analysis LATPCO (Figure 12). The discriminant functions classify 33.87% of the cases correctly. That is indeed better than LATPCO. LATPWE discriminant functions were used to class Polish provenance No. 6 (Table 32), Provenance Purda Lesna is classed in an Interior zone as was the case with NEWSUB, However, this is in disagreement with the seed tests,with the classing into CLIAMP and into LATPCO. -77-Table 31. Discriminant analysis LATPWE, for the d i s t i n c t i o n between l a t i t u d i n a l zones TITLE LATPWE NO. OF PROVENANCES 124 NO. OF GROUPS 17 NO. OF VARIABLES INCLUDED 7 NO. OF VARIABLES USED 6 % OF CASES CORRECTLY CLASSIFIED 33.87 APPROXIMATE F STATISTIC 7.28962*** D.-f. FOR APPROXIMATE F 96;584.70 F VALUES FOR EACH VARIABLE:WITH D.F. 16; 102 CONE SCALE WIDTH 6.5043*** BRACT WIDTH 4.0834*** CONE SCALE LENGTH FIRST PRONG LENGTH 4.8424*** SECOND PRONG LENGTH 11.6943*** RATING OF BRACT 2.1345* lOOO-SEED WEIGHT 9.6776*** Table 32. C l a s s i f i c a t i o n of Polish provenance No. 6 into  l a t i t u d i n a l zones (LATPWE) PROV. CLASSED TO WITH PRO- GROUP OF WITH PRO-NO. GROUP BABILITY NEXT LOWER BABILITY PROBABILITY 6 0 (33%) ] 0.827 U (60%) 0.135 The percentage in parentheses indicates the frequency of cases correctly c l a s s i f i e d in LATPWE -78-Since the one thousand-seed weight i s based on limited number of available f i l l e d seeds (Section 34), i t s value as i t was used in c l a s s i f i c a t i o n of LATPWE has to be treated with reservation. 367. Assessment of the origin of Polish provenances No. 1-20 For assessment of the origin of the twenty Polish provenances, the three c l a s s i f i c a t i o n s into NEWSUB, CLIAMP and LATPCO have been synthesized (Table 33). A provenance for which the c l a s s i f i c a t i o n in a l l three discriminations i s in agreement can be considered as corre c t l y classed. When the agreement-triangle (Table 33) i s open, some reservations have to be made. As the twenty stands are classed into LATPCO groups with an average prob a b i l i t y of 0.624 and into CLIAMP groups 0.512, more reliance can be granted to the l a t i t u d i n a l c l a s s i f i c a t i o n . It can be seen (Table 33) that fourteen provenances show the closed agreement-triangle. They might be considered as more r e l i a b l y classed as origin is concerned than the remaining six provenances not showfng agreement among the three tests (NEWSUB-CLIAMP-LATPCO). Out of the fourteen provenances, ten were classed as Interior and four as Coastal. Although Coastal provenances of Douglas-fir are generally found in Europe, i t is possible to find Interior provenances among the early introductions, Poland being partly suboceanic and partly subcontinental (Troll and Paffen £1965]). - 7 9 -Table 33. Assessment of the or i g i n of Polish provenances No. 1-20 PROVENANCE AGREEMENT1 NEWSUB / \ CLIAMP-LATPCO ASSESSMENT SUBSPECIES OF THE CLIMATIC ZONE ORIGIN LATITUDE 1. WOLINSKI Z\ COASTAL 2a 50-51 2 . MIEDZDROJE _ COASTAL 2a ' 4 9 - 5 0 3 . DOBRZANY ^_ COASTAL 2a 49-50 4 . BIALY BOR Zi INTERIOR 5a 52-54 5 . SOSNO Zi INTERIOR 5a 52-54 6. PURDA LESNA COASTAL 3a 50-51 7 . RYN Z\ INTERIOR 7b 49-50 8 . MYRADZ Zi COASTAL 2a 50-51 9 . JAROCIN INTERIOR 5a 52-54 10. KACZORY z\ INTERIOR 6 52-54 11 . GOLABKI zv INTERIOR 5b 46-48 12. SKRZECIN / \ INTERIOR 6 51-53 13. DOLICE z\ INTERIOR 6 52-54 14. SWIEBODZIN A INTERIOR 7b 49-50 15. KROSNO z\ COASTAL 1c 38-40 16. UJSOLY z\ COASTAL 3b 50-52 17. POKRZYWNO COASTAL 3b 50r<52 18. DUZNIKI COASTAL 4 50-52 19. KOWARY INTERIOR 7b 48-49 20. BROJCE Zi INTERIOR 5a 51^52 A straight l i n e t i e s two analyses, when the c l a s s i f i c a t i o n of the concerned provenance i s in agreement -80-368. Discriminant analyses for the c l a s s i f i c a t i o n of the Swiss  provenance The Swiss stand has been classed unequivocally as Coastal with the help of Allen's seed te s t s . Due to the bad conditions of the cones, only four metric characters remained for discriminant analyses: cone scale width, cone scale length, r e l a t i v e DNA content and one thousand-seed weight. Measurements of DNA ex i s t for forty-eight provenances, mainly Coastal ones. Therefore, discriminant analyses were performed only with provenances from the Coastal subspecies. Two groupings were made, one climatic (CLIDNA) and one l a t i t u d i n a l (Figure 14). For CLIDNA, i t was only necessary to group the climatic zones l a and l a a . The analyses are given in Table 34. LATDNA is based only on the variable r e l a t i v e DNA content. This is remarkable and w i l l allow a comparison with the regression of r e l a t i v e DNA content on latitude (Section 369). Both analyses produce a discrimination of a similar value, which can be seen from the percentage of cases c o r r e c t l y c l a s s i f i e d . CLIDNA and LATDNA were used to class the Swiss provenance Boezingen (Table 35). Both c l a s s i f i c a t i o n s are in agreement. The low prob-a b i l i t i e s are due to the high number of groups f o r so few observations. -81-Figure 14. Delimitation of l a t i t u d i n a l zones for analysis LATDNA -82-Table 34. Discriminant analyses for Coastal Douglas-fir TITLE CLIDNA LATDNA NO. OF PROVENANCES 36 35 NO. OF GROUPS 14 10 NO. OF VARIABLES INCLUDED 4 4 NO. OF VARIABLES USED 2 1 % CASES CORRECTLY CLASSIFIED 30.56 31.43 APPROXIMATE F STATISTIC 5.04798*** 17.33556*** D.F. FOR APPROXIMATE F 26;42.00 9;25.00 F VALUES FOR EACH VARIABLE: WITH D.F. 13; 21 9; 25 CONE SCALE WIDTH CONE SCALE LENGTH 1000-SEED WEIGHT 2.5235* RELATIVE DNA CONTENT 4.4910*** 17.3356*** Table 35. C l a s s i f i c a t i o n of the Swiss provenance Boezingen into  the Coastal Douglas-fir zone ANALYSIS CLASSED WITH PRO- GROUP OF WITH PRO-TO GROUP BABILITY NEXT LOWER BABILITY PROBABILITY CLIDNA 2d(33%) 1 0.217 2c (100%) . 0.199 LATDNA G( 0%) 0.229 H( 0%) 0.219 The percentage in parentheses indicates the frequency of cases c o r r e c t l y c l a s s i f i e d in CLIDNA and LATDNA, respectively. -83-369. Relationship between r e l a t i v e DNA content and l a t i t u d e , and  i t s application to the estimation of or i g i n of the Swiss  provenance The discriminant analysis NEWSUB defines a new boundary between Coastal and Interior Douglas-fir. El-Lakany and S z i k l a i ' s (1971, 1972) provenances f a l l i n g on the Coastal side were used to calculate the following c u r v i l i n e a r regression, on the tree averages: DNA = 721.82739-32.16846 LAT+0.41966 LAT 2. The analysis of variance i s given in Table 36. Table 36. Analysis of variance for the regression of r e l a t i v e DNA content on latitude SOURCE D.F. SUM OF SQUARES MEAN SQUARE F VALUE DUE TO REGRESSION ABOUT REGRESSION 2 177 91553.43750 8362.56250 45776.71875 47.24611 968.8992*** TOTAL 179 99916.00000 The multiple c o r r e l a t i o n c o e f f i c i e n t , R=0.9572, is s i g n i f i c a n t at the 1% l e v e l . This equation, i f not i d e n t i c a l , i s very similar to that found by El-lakany and S z i k l a i (1972). Tolerance l i m i t s for the average of four observations were calculated with the following formula (Draper and Smith [1966]): Y£-T(V, 1-1/2PC) /s2/G+ V(Y) -84-A where: Y = estimated value T = student T-value V = degrees of freedom ©C= probability level 2 S •= residual mean square G = number of observations on which the X-value is based „ V(Y) = variance of Y Figure 15 displays the estimated equation with 95% and 99% tolerance l i m i t s . The graph was used for the inverse estimation of l a t i t u d e for the Swiss provenance Boezingen. Entering the graph from the DNA axis, the latitude is read on the abscissa. This gives a latitude of 45°25'N. The 95% inverse tolerance l i m i t s extend from 44?08' to 46°29'N and the 99% inverse tolerance l i m i t s from 43°42' to 46°47'N. With the help of the discriminant functions of LATDNA, i t was possible to place the Swiss stand between latitude 44° and 45°N. However, the probability that i t had come from latitude 45° to 46°N was only s l i g h t l y lower. Indeed, the discriminant functions based on one variable stretches the groups on a straight l i n e . S i m i l a r l y , a straight l i n e regression would have indicated a lower l a t i t u d e . While tolerance l i m i t s are narrower f o r the average value of the regression, inverse tolerance l i m i t s decrease when the slope of the curve increases. Consequently, a higher l a t i t u d e would have been indicated with much narrower inverse tolerance l i m i t s . For the case of the provenance Boezingen, they give a better indication of the goodness? of the average, than a true range. 210 200 — Fig-15- The relationship between the amount of D N A (in FA U) and the latitude of seed source in Coastal Douglas-fir* Legend Estimated regression 95% tolerance limits 99% tolerance limits 2-124 Provenance numbers as used by Yao(l97l) Tree averages * Based on El-Lakany and Sziklai (1971 and 1972) i CO cn 0 " 39 40 41 42 45 46 47 Latitude CN) 53 -86-A regression based on the embryo averages might have given narrower tolerance l i m i t s (smaller T-value, but higher G [=796] [from Table 17J, instead of =4). However, the residual mean square would have been substantially higher. From the discriminant analyses and from the regression analyses, i t can be concluded that the Swiss provenance Boezingen has i t s or i g i n in the Coast Ranges of Oregon between latitude 44°N and the estuary of the Columbia River. The estimation of the l a t i t u d e of o r i g i n from the r e l a t i v e DNA content can be applied to enter into the probable c l i m a t i c zone (Figure 2). Five zones can be considered: l b , 2c, 2d, 3c, and 3d. From Figure 3 and Figure 5, i t can be seen that the average cone scale width (20 mm) as well as the average cone scale length (16 mm) f a l l completely outside the range of the concerned climatic zones. This confirms the doubts expressed in Section 362 concerning the r e l i a b i l i t y of the cone scale measurements. 4. SUMMARY AND CONCLUSION The determination of the o r i g i n of one Swiss and twenty Polish Douglas-fir stands has been attempted. Literature on the geographical var i a t i o n of Douglas-fir was reviewed and four studies were selected to compare the unknown stands. Cones were collected in 1971 in Switzerland and in 1969-70 in Poland. Seeds extracted from cone samples were separated by hand in -87-" f i l led" and "empty" using X-ray fluoroscopy. In provenance Boezingen (Switzerland) and Purda Lesna (Poland), seeds were submitted to Allen's seed test, thousand-seed weight was determined and the relative DNA content of embryo cells was measured. In provenance Boezingen, cone scale width and cone scale length were measured. In the twenty Polish provenances, cone scale width, cone scale length, bract width, f irst and second prong length were measured, and the rating of bract was scored. The following results were obtained: 1 . Seed tests: Provenance Boezingen (Switzerland) and provenance Purda Lesna (Poland) are both of Coastal origin. 2. Seed yield: There are few fi l led seeds in all provenances, suggesting a close parentage between the trees inside the stands. 3. Cone scale morphology: Significant differences exist between trees and between cones within trees for cone scale width and cone scale length in provenance Boezingen. The information available does not allow one to discover the cause of the presence of differences. There are statistically significant differences between the twenty Polish stands in the six characters studied. Based on cone scale morphology, no satisfactory groupings of provenances were obtained. 4. Thousand seed weight: Thousand seed weight for provenances Boezingen and Purda Lesna -88-i s based on few seeds and was therefore only used with reservation. 5. Relative DNA content of embryo c e l l s : For the Swiss provenance, a r e l a t i v e l y high variation was found in the r e l a t i v e DNA content. S i g n i f i c a n t differences are present between embryos within the trees. The tree averages, however, are homogeneous. For the Polish provenance Purda Lesna, s i g n i f i c a n t differences were found between the embryos and no s a t i s f a c t o r y grouping could be obtained. Therefore, r e l a t i v e DNA content was not used for the determination of origin of provenance Purda Lesna. The high v a r i a b i l i t y seems to be due mainly to the technique involved. 6. Determination of o r i g i n of the Polish and Swiss plantations: A. "Indiividual use of the cone scale morphology variables: Each unknown provenance can be classed in several c l i m a t i c zones, for each variable. It is not possible to class a provenance unequivocally. B. Concurrent use of the cone scale morphology variables f o r the separation into Coast and Interior Douglas-fir: Discriminant analysis was performed on the six characters for the 124 provenances from the natural range for the separation into Coastal and Interior Douglas-fir. A remodelling of S z i k l a i ' s (1969) boundary between both subspecies produced a correct c l a s s i f i c a t i o n of 96% of the provenances, on the basis of this analysis. The discriminant functions of the analysis NEWSUB, based on cone scale width, cone scale length, f i r s t and second prong length, and rating of bract were used to class the twenty Polish provenances; four -89-seem to be of Coastal and sixteen of Interior o r i g i n . For provenance Purda Lesna, the c l a s s i f i c a t i o n i s in contradiction with the r e s u l t of the seed te s t s . C. Concurrent use of the cone scale morphology variables for the separation into c l i m a t i c zones and c l a s s i f i c a t i o n of the twenty Polish provenances: Discriminant analyses were performed on the six cone scale characters for 124 provenances at the provenance level as well as at the tree l e v e l . The discrimination is better at the provenance than at the tree l e v e l . The twenty Polish provenances were classed with the discriminant functions of the analysis CLIAMP at the provenance l e v e l , based on a l l six cone scale characters. In some cases the c l a s s i f i c a t i o n into c l i m a t i c zones contradicts the c l a s s i f i c a t i o n into Interior and Coastal groupings. The c l a s s i f i c a t i o n using the discriminant functions of the analysis CLIAME at the tree l e v e l , based on a l l six cone.scale characters gives inconsistent r e s u l t s , which are thought to be due to the lack of homogeneity among the data. D. Concurrent use of the cone scale morphology variables for the separation into l a t i t u d i n a l zones and c l a s s i f i c a t i o n of the twenty Polish provenances: Discriminant analyses using a grouping into l a t i t u d i n a l zones gave s l i g h t l y better r e s u l t s than the analyses using climatic zones. The same discrepancy as before was found between the use of the analysis at the provenance l e v e l , LATPCO, based on cone scale width, -90-bract width, f i r s t and second prong length, and rating of bract, and the use of the analysis at the tree l e v e l , LATICO, based on a l l six cone scale characters. The c l a s s i f i c a t i o n into l a t i t u d i n a l groups is similar to that into c l i m a t i c groups. E. C l a s s i f i c a t i o n of provenance Purda Lesna into l a t i t u d i n a l zones: A discriminant analysis including the cone scale variables and the thousand-seed weight was performed and used for the c l a s s i f i c a t i o n of provenance Purda Lesna (Poland). The c l a s s i f i c a t i o n with the discriminant functions of the analysis LATPWE, based on cone scale width, bract width, f i r s t and second prong length, rating of bract, and thousand-seed weight, i s in accord with that into the subspecies, but i n disagreement with a l l other c l a s s i f i c a t i o n s . It was therefore abandoned. F. Assessment of origin of the twenty Polish provenances: The c l a s s i f i c a t i o n into subspecies, into climatic and l a t i t u d i n a l zones were viewed together. When they did not contradict each other, the concerned provenance was classed at the intersection of the three zones (Table 33). G. Discriminant analyses and c l a s s i f i c a t i o n of the Swiss provenance: Discriminant analyses including two cone scale variables, thousand-seed weight and r e l a t i v e DNA content were established using c l i m a t i c , as well as l a t i t u d i n a l groupings. Provenance Boezingen was classed into both groupings (CLIDNA, based on thousand-seed weight; LATDNA, based on -91-r e l a t i v e DNA content) and the results are in agreement (Table 35). H. Determination of or i g i n of provenance Boeziingen using regression analysis: A regression of r e l a t i v e DNA content on l a t i t u d e in Coastal Douglas-fir was calculated and tolerance l i m i t s affixed to i t . The lat i t u d e of origin of the Swiss stand was determined from the graph. It agrees with the previous c l a s s i f i c a t i o n . Provenance Boezingen originates from the Coast Ranges of Oregon between l a t i t u d e 44°N and the estuary of the Columbia River. 7. Conclusion: Four methods have been used for the purpose of determination of the o r i g i n of unknown Douglas-fir populations: A. Seed tests: the c h a r a c t e r i s t i c s used are q u a l i t a t i v e and allow a s p l i t t i n g into two categories - the two subspecies of Douglas-fir. As long as the samples to be analyzed are not of intermediate o r i g i n , the tests y i e l d conclusive r e s u l t s . Indeed, although they are used to define the subspecies, the tests do not produce a clear-cut boundary between both Coastal and Interior Douglas-fir. From a practical standpoint, the seed tests are very good. They are e a s i l y learned and quickly carried out. They require the use of only a strong microscope. B. Direct comparison of the range of values of the individual variables: this method is best used to determine o r i g i n of stands when the intergroup v a r i a b i l i t y i s high. Unfortunately, however, the ch a r a c t e r i s t i c s used in this study show high intragroup v a r i a b i l i t y . The main use of the ch a r a c t e r i s t i c s was to give some indication of the influence of the environment. -92-C. Discriminant analyses: when the individual d i r e c t comparisons f a i l e d , l i n e a r combinations of the variables for a multidimensional display of groups of provenances proved useful. Once the discrimination i s obtained, the c l a s s i f i c a t i o n of an unknown sample is a s t r a i g h t -forward geometric operation. However, the c l a s s i f y i n g i s only as good as the discrimination i t s e l f . The c h a r a c t e r i s t i c s used in this study did not y i e l d a perfect separation. Then, the degree of separation in the discrimination should be considered before p o s i t i v e l y assigning an unknown sample to a certain group on the basis of the calculated p r o b a b i l i t i e s . An important aspect of the use of discriminant analysis is the grouping of the data. The climatic regions which were u t i l i z e d were only poorly related to the b i o l o g i c a l variables studied. However, the c l a s s i f i c a t i o n into climatic zones was generally in agreement with the better c l a s s i f i c a t i o n into l a t i t u d i n a l zones. The theory behind this method renders i t somewhat sophisticated for p r a c t i c a l use. Furthermore, i t requires the use of an electronic computer, as well as a r e l a t i v e l y high number of provenances. D. Regression analysis: when a simple relationship between a single environmental and a b i o l o g i c a l variable i s a v a i l a b l e , as i s the case between latitude and r e l a t i v e DNA content, this method can be applied for quite accurate determination of o r i g i n . The number of samples can be s i g n i f i c a n t l y lower than in the case of discriminant analysis, but the relationship obtained between the variables may be less accurate than i f a greater number of provenances had been used. In practice, the r e l a t i v e DNA content of embryo c e l l s is d i f f i c u l t -93-to measure. It requires a microscope equipped with a photometer, and furthermore, s k i l l , which cannot be acquired immediately. Concerning the unknown tree stands, the following points should be noted: as well as the cone and seed samples, information pertaining to the extent of the stand, i t s quality and sanitary condition, the treatments applied throughout i t s l i f e t i m e , and the environmental conditions should also be sent to aid c l a s s i f i c a t i o n . A better knowledge of the natural variation of the species w i l l , of course, increase the confidence with which the ori g i n of unknown stands can be found. At the present time, the above l i s t e d methods allow a reasonably r e l i a b l e determination of the twenty-one European Douglas-fir stands. -94-BIBLIOGRAPHY Abetz, P. 1971. Douglasien-Standraumversuche. Ein Gemeinschaftsprojekt f o r s t l i c h e r Versuchsanstalten und Landesforstverwaltungeft. Allgemeine Forst Z e i t s c h r i f t 26_ (21/22): 448-449. A l l e n , G.S, 1960. A method of distinguishing Coastal from Interior Douglas-fir seed. B r i t i s h Columbia Lumberman 4£ (8): 26-30. T^.-_^ r._r._-. 1961. Testing Douglas-fir seed for provenance. Proc. Int. Seed Test. Ass. 26 (3): 388-403. All e n , G.S. and J.N. Owens. 1972. The l i f e history of Douglas-fir. Information Canada, Ottawa. 139 pp. Baker, K.H. 1971. Variation in one-year-old seedlings of Douglas-fir [Pseudotsuga menziesii fMirb.] Franco). B.S.F. thesis, the University of B r i t i s h Columbia, Vancouver. 60 pp. BeckersjB. 1972 Facteurs influencant l a production de graines chez le Douglas. B u l l . Soc. Roy. For. de Belgique 79_ (4): 253-261. Bouvarel, P. 1958. Les repeuplements a r t i f i c i e l s - consequences d'ordre genetique. Journal Forestier Suisse 109 (8/9): 524-535. T T ^ r r r - - . 1969. La selection des semences f o r e s t i e r e s . Ou en sommes-nous en 1969? Revue ForestieYe Franchise 21_: 403-414. Bouvarel, P., R. Tomassone, and 0. S z i k l a i . 1969. Multivariate analyses for Douglas-fir. Mimeographed. 4 pp. On f i l e at the Faculty of Forestry, the University of B r i t i s h Columbia, Vancouver. Buck, J.M., R.S. Adams, J . Cone, M.T. Comkle, W.J. Libby, C.J. Eden, and M.J. Knight. 1970. C a l i f o r n i a tree seed zones. U.S. Department of Agric', C a l i f . Region Forest Service. San Francisco, C a l i f . 5;pp. Chylarecki, H. 1970. Letter dated October 7, 1970, to Dr. 0. S z i k l a i . Faculty of Forestry, the University of B r i t i s h Columbia, Vancouver. Cornet d'Elzius, Ch. 1963. Origine des graines de Douglas. B u l l . Soc. Roy. Forest, de Belgique 70_ (10): 489-493. Davis, P.H., and V.H. Heywood. 1963. Principles of angiosperm taxonomy. Oliver and Boyd, Edinburgh and London. 556 pp. Dawson, I.J.M. 1964. Some notes on Douglas-fir in B r i t a i n . Forestry Commission, Genetics Section. Mimeographed. -95-El-Lakany, M.H., and 0. S z i k l a i . 1971. Intraspecific variation in nuclear c h a r a c t e r i s t i c s of Douglas-fir. Advancing Frontiers of Plant Sciences 28: 363-378. _ . 1972. Further investigations of in t r a s p e c i f i c v a r i a t i o n in DNA contents of Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco). Submitted for publication to Silvae Genetica. 20 pp. Falconer, D.S. 1960. Introduction to quantitative genetics. Oliver and Boyd, Edinburgh and London. 365 pp. F e r r e l , W.K., and E.S. Woodard. 1966. Effects of seed origin on drought resistance of Douglas-fir [Pseudotsuga menziesii [Mirb.] Franco). Ecology 47;(3): 499-503. Flake, R.H., E. von Rudloff, and B.L. Turner. 1969. Quantitative study of c l i n a l v a r i a t i o n in Juniperus virginiana using terpenoid data. Proc. Natl. Acad. S c i . U.S.A. 64 (2): 487-494. Floehr, W. 1958. In Goehre, K. die Douglasie und ihr Holz. Akademie-Verlag, B e r l i n . 595 pp. Flous, F. 1936. Revision du genre Pseudotsuga. Travaux du Laboratoire Forestier de Toulouse. T. 1. .1937. Revision du genre Pseudotsuga. B u l l . Soc. Hist. Nat. Toulouse 7J_ (1/2): 33-164. Fowells, H.A. 1965. S i l v i c s of forest trees of the United States. U.S. Forest Service, Division of Timber Management Research. Department of Agriculture, Washington. 762 pp. Franklin, J.F., and Greathouse, T.E. 1968. Seed or i g i n studies: Noble/California Red F i r species complex. Proceedings, Western Forest Nursery Council 1968 (11-16A). Freese, F. 1964. Linear regression methods for forest research. U.S. Forest Service. Research Paper FPL 17. 136 pp. Galoux, A. 1951. Les principales essences f o r e s t i e r e s de I'Amerique septentrionale temperee. Leur Introduction en Belgique. Travaux de la Station de Recherches de Groenendal-Serie B, No. 13. 141 pp. Garman, E.H. 1970. Pocket guide to the trees and shrubs of B r i t i s h Columbia. 4th Ed. V i c t o r i a . Department of Lands, Forests and Water Resources, B r i t i s h Columbia Forest Service. 131 pp. -96-G r i f f i t h , B.G. 1968. Phenology, growth, and flower and cone production of 154 Douglas-fir trees on the University Research Forest as influenced by climate and f e r t i l i z e r , 1957-1967. Faculty of Forestry, the University of B r i t i s h Columbia, Bulletin No. 6. 70 pp. Guillebaud, W.H. 1956. Douglas-fir provenance experiments in Germany. Library Record No. 25,'Part B: 283-295. B r i t i s h Forestry Commission, A l i c e Holt Lodge, Wrecclesham, Surrey. Mimeographed. Guinier, H. 1948. Variation et heredite chez les essences f o r e s t i e r e s . Consequences s y l v i c o l e s . Journal Forestier Suisse 99_ (4):161-175. Haddock, P.G., and 0. S z i k l a i . 1966. Seed, c o l l e c t i o n zones for Douglas-f i r in Canada, Proceedings of the Sixth World Forestry Conqress, Madrid. Vol. II: 1467-1473. Haddock, P.G., J . Walters, and A. Kozak. 1967. Growth of Coastal and Interior provenances of Douglas-fir [Pseudotsuga menziesii [Mirb.] Franco) at Vancouver and Haney in B r i t i s h Columbia. Faculty of Forestry, the University of B r i t i s h Columbia. Research paper No. 79. 32 pp. Hanover, J.W. 1966. Environmental variation in the monoterpenes of Pinus montioola Dougl. Phytochemistry 5_ (4): 713-717. Hanover, J.W.t 1971. Genetics of terpenes. II. Genetic variances and interrelationships of mohoterpene concentrations in Pinus montioola. Heredity 27_ (2): 237-245. Hanover, J.W.,- and M.M. Furniss. 1966. Monoterpene concentration in Douglas-fir in r e l a t i o n to geographic location and resistance to attack by the Douglas-fir beetle. U.S. Forest Service Research Paper NC-6: 23-28. Irgens-Moller, H. 1966. Of what value i s climatic data in assigning off-source seed to reforestation s i t e ? Western Forest Genetics Association. Proceedings 1965 Olympia, Washington. P. 23-24. Isaac, L.A., and E.J. Dimock II. 1958. S i l v i c a l c h a r a c t e r i s t i c s of Douglas-fir, var. menziesii* U.S. Forest Service, P a c i f i c Northwest Forest and Range Expt. Stat. S i l v i c a l Series No. 9. 18 pp. Jahn, G. 1955. Gegenueberstellung Deutscher und Amerikanischer Douglasien-Provenienzversuche. Allgemeine Forst-und jagdzeitung 126 (4): 68-76. Jabbar, A., and E. Brochmann-Hanssen. 1961. Amino acids in opium. Journ.. Pharm. Sciences 50 (5): 406-408. -97-Koestler, J . 1950. Waldbaud. Paul Parey, B e r l i n und Hamburg. 418 pp. Lacaze, J.-F. 1968. Comparaison de quelques provenances de Douglas dans 1'Arboretum des Barres. Revue Forestiere Francaise 2TJ (1): 28-32. Lacaze, J.-F. et R. Tomassone. 1967. Contribution a l'etude de la v a r i a b i l i t y du Douglas (Pseudotsuga menziesii Mirb.). Annales des Sciences ForestieYes 24 Q): 85-106. Lanier, L. 1966. Les maladies cryptogramiques du Douglas en France. Revue Forestiere Franchise 1_8 (4): 247-265 MacDonald, J . , R.F. Wood, M.V.. Edwards, and J.R. Aldhous. 1957. Exotic forest trees in Great B r i t a i n . Forestry Commission Bu l l e t i n 30. 167 pp. Merendi, A. 1965. (Ought Douglas-fir to be grown in Italy?) I t a l . For. Mont. 20 (6): 255-260, Morris, W.G., R.R. S i l e n , and H, Irgens-Moller. 1957. Consistency of bud bursting in Douglas-fir. Journal of Forestry 55_ (3): 208-210. Nanson, A. 1968. La valeur des tests precoces dans l a selection des arbres f o r e s t i e r s en p a r t i c u l i e r au point de vue de l a croissance. Station de Recherches des Eaux et Forets. Groenendal-Hoeilaart. Ph.D. thesis. 242 pp. Orr-Ewing, A.L. 1965. Inbreeding and single crossinq i n Douglas-fir. Forest Science 11_ (3): 279-290. Otto, H.-J. 1972. Zu den Standortanspruechen der Douglasie -Moeglichkeiten fuer den Anbau im Pleistozaen (Dilivium) Niedersachsens. Forstarchiv 43 (4/5): 62-65. P i n t a r i c , K. 1967. Das Hoehenwachstum verschiedener Herkuenfte der Douglasie (Pseudotsuga taxifolia B r i t t . ) i n den ersten Lebensjahren. Allgemeine Forst-und Jagdzeitung 138_ (9): 198-205. Preston, R.J. 1966. North American Trees. The M.I.T. Press, Massachusetts Institute of Technology, Cambridge, Massachusetts, U.S.A. 395 pp. Rosenstock, A. 1972. Zur Ermittlung der geographischen Variationen der Douglasie. Allgemeine Forst Z e i t s c h r i f t 27 (9/10): 156. Rowe, J.S. 1964. Environmental preconditioning, with special reference to f o r e s t r y . Ecology 45 (2): 399-493. Rudloff, E. von. 1972. Chemosystematic studies i n the genus Pseudotsuga. I. Leaf o i l analysis of the Coastal and Rocky Mountain v a r i e t i e s of the Douglas-fir. Canadian Journal of Botany 50 (5): 1025-1040. -98-Schober, R. 1963a. Experiences with the Douglas-fir in Europe. FAO/Forgen 63/-4/5. pp II1-18. . 1963b. Erfahrungen mit der Douglasie in Europa. Allgemeine Forst Z e i t s c h r i f t 18 (30): 473-474, T8 (31): 487-489, 18 (34): 515-519. Seagraves, P. 1971. C l a s s i f i c a t i o n using discriminant functions. Computing Centre, the University of B r i t i s h Columbia. 4 pp. . 1972. Stepwise discriminant analysis. Computing Centre, the University of B r i t i s h Columbia, Vancouver. 13 pp. S i l e n , R.R. 1962. A study of genetic control of bud bursting in Douglas-fir. Journal of Forestry 60 (7): 472-475. . 1966. A 50-year r a c i a l study, of Douglas-fir in Western Oregon and Washington. (Abstract). Western Forest Genetics Association. Proceedings 1965. Olympia, Washington. P. 6-7. Simak, M. 1967. Seed weight of larch, from difference provenances [Lavix deeidua M i l l . ) . Studia Forestalia Suecica No, 57. 31 pp. Sweet, G.B, 1965. Provenance differences in P a c i f i c Coast Douglas-f i r . Silvae Genetica 1_4 (2): 46-56. S z i k l a i , 0. 1964. Variation and inheritance of some physiological and morphological t r a i t s in Pseudotsuga menziesii lMirb.3 Franco var. menziesii. Ph.D..thesis. The University of B r i t i s h Columbia, Vancouver. 129 pp. .1969. Preliminary notes on the variation in cone and seed morphology of Douglas-fir {Pseudotsuga menziesii [Mirb.] Franco). FO-FTB-69-6/9. T r o l l , C., and Kh. Paffen. 1965. Map No. 5: Seasonal climates of the Earth. In Rodenwaldt, E., and H.J. Jusatz. 1965. World maps of climatology. Springer Verlag, Berlin-Heidelberg-New York. 28 pp.+ appendix. Tusko, F.F. 1963. A study of v a r i a b i l i t y in certain Douglas-fir populations in B r i t i s h Columbia. Ph.D. thesis. The University of B r i t i s h Columbia, Vancouver. 173 pp. Viebig, J . 1967. Anbau der gruenen Douglasie im Forstbezirk Eberbach. Allgemeine Forst Z e i t s c h r i f t 22_ (40): 689-690. Volk, K. 1969. Herkuenfte, Standortsansprueche, waldbauliche Behandlung, Leistung und Verwertung der Douglasie aus Suedbadischer Sicht. Allgemeine Forst Z e i t s c h r i f t 24~(37): 715-717. -99-Vredenburch, C.L.H. Van, and J.G.A. La Bastide. 1969. The influence of meteorological factors on the cone crop of Douglas-fir in the Netherlands. Silvae Genetica ]8_ (5-6): 182-186. Wakel.ey, P.C., and T.E. Bercaw. 1965. L o b l o l l y pine provenance test at age 35. Journal of Forestry 63 (3): 168-174. Western Forest Seed Council. 1966. Tree seed zone map for Washington and Oregon. Meeting of W.F.T.S.C. A p r i l , 1966. Portland, Oregon. 2 pp. + appendix. Wheat, J.G. 1966. The seed movement problem in the Douglas-fir region. Western Forest Genetics Association. Proceedings 1965. Olympia, Washington. P. 3-5. Wilkinson, R.C, J.W. Hanover, J.W. Wright, and R.H. Flake. 1971. Genetic variation in the monoterpene composition of White Spruce. Forest Science 17_ (1): 83-90. Wright, J.W., F.H. Kung, R.A. Read, W.A. Lemmien, and J.N. Bright. 1971. Genetic variation in Rocky Mountain Douglas-fir. Silvae.Genetica 20 [3): 54-60. Yao, C. 1971. Geographic variation in seed weight, some cone scale measurements and seed germination of Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco). M.F. thesis. The University of B r i t i s h Columbia, Vancouver, 88 pp. Zimmermann, H. 1972. Zum Anbau von Douglasie in Hessen. Allgemeine Forst Z e i t s c h r i f t 27 (17): 311-312. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0075390/manifest

Comment

Related Items